Autophagy - Autophagy

EMBO Molecular Medicine

cross-journal focus Autophagy - Autophagy

EDITORS Andrea Leibfried Editor [email protected] | T +

Andrea worked with Jan Lohmann on stem cell maintenance in the plant Arabidopsis before moving to the eld of trafcking. In she obtained her PhD from the Université Pierre et Marie Curie in Paris, for which she studied DE-Cadherin trafcking in Drosophila with Yohanns Bellaiche at the Curie Institute. She then went to Anne Ephrussi’s lab at the EMBL in Heidelberg to work on oocyte polarity and mRNA trafcking in Drosophila. Andrea joined The EMBO Journal in .

Nonia Pariente Senior Editor [email protected] | T +

Nonia joined EMBO Reports in August . She studied biochemistry and molecular biology in Madrid’s Autónoma University, where she also gained her PhD on the generation of new antiviral strategies against RNA viruses. She did a four-year post-doc at UCLA focusing on the development of new strategies for gene therapy.

Céline Carret Editor [email protected] | T +

Céline Carret completed her PhD at the University of Montpellier, France, characterising host immunodominant antigens to ght babesiosis, a parasitic disease caused by a unicellular EMBO Apicomplexan parasite closely related to the malaria agent Plasmodium. She further developed Molecular her post-doctoral career on malaria working at the Wellcome Trust Sanger Institute in Cambridge, Medicine UK and Instituto de Medicina Molecular in Lisbon, Portugal. Céline joined EMBO Molecular Medicine as a Scientic Editor in March .

Maria Polychronidou Editor [email protected] | T +

Maria received her PhD from the University of Heidelberg, where she studied the role of nuclear membrane proteins in development and aging. During her post-doctoral work, she focused on the analysis of tissue-specic regulatory functions of Hox transcription factors using a combination of computational and genome-wide methods.

emboj.embopress.org | embor.embopress.org | embomolmed.embopress.org | msb.embopress.org full articles

The EMBO Journal Article Induction of autophagy supports the bioenergetic demands of quiescent muscle stem cell activation. Tang AH, Rando TA. DOI:10.15252/embj.201488278 | Published 14.10.2014

EMBO Reports Review Getting ready for building: signaling and autophagosome biogenesis. Abada A, Elazar Z. DOI:10.15252/embr.201439076 | Published 15.07.2014 Article PI3P phosphatase activity is required for autophagosome maturation and autolysosome formation. Wu Y, Cheng S, Zhao H, Zou W, Yoshina S, Mitani S, Zhang H, Wang H. DOI:10.15252/embr.201438618| Published 14.08.2014

EMBO Molecular Medicine Articles Lysosomal dysfunction and impaired autophagy underlie the pathogenesis of amyloidogenic light chain-mediated cardiotoxicity. Guan J, Mishra S, Qiu Y, Shi J, Trudeau K, Las G, Liesa M, Shirihai OS, Connors LH, Seldin DC, Falk RH, MacRae CA, Liao R. DOI:10.15252/emmm.201404190 | Published 15.10.2014 Selective clearance of aberrant tau proteins and rescue of neurotoxicity by transcription factor EB. Polito VA, Li H, Martini-Stoica H, Wang B, Yang L, Xu Y, Swartzlander DB, Palmieri M, di Ronza A, Lee VMY, Sardiello M, Ballabio A, Zheng H. DOI:10.15252/emmm.201303671 | Published 28.07.2014

For further reading please see inside back cover

Article

Induction of autophagy supports the bioenergetic demands of quiescent muscle stem cell activation

Ann H Tang1,2 & Thomas A Rando1,2,3,*,†

Abstract autophagy can protect cells against nutrient stress (Mizushima & Komatsu, 2011). Nitrogen-starved yeast, for example, induce auto- The exit of a stem cell out of quiescence into an activated state is phagy to produce amino acids to maintain viability (Onodera & characterized by major metabolic changes associated with Ohsumi, 2005). Similarly, lymphocytes subjected to growth factor increased biosynthesis of proteins and macromolecules. The regu- deprivation that prevents nutrient uptake activate autophagy to lation of this transition is poorly understood. Using muscle stem survive (Lum et al, 2005). Autophagy can also safeguard against cells, or satellite cells (SCs), we found that autophagy, which nutrient deprivation at the organismal level, as studies have shown catabolizes intracellular contents to maintain proteostasis and to that mice rely on autophagy to survive the neonatal starvation produce energy during nutrient deprivation, was induced during period (Kuma et al, 2004; Komatsu et al, 2005). SC activation. Inhibition of autophagy suppressed the increase in More recent studies have demonstrated that autophagy also serves ATP levels and delayed SC activation, both of which could be as an adaptive response to many other stressors including intense partially rescued by exogenous pyruvate as an energy source, exercise, ER stress, infection, hypoxia, and oxidative stress (He & suggesting that autophagy may provide nutrients necessary to Klionsky, 2009; Kroemer et al, 2010; He et al, 2012). For example, meet bioenergetic demands during this critical transition from autophagy responds to oxidative stress by removing damaged quiescence to activation. We found that SIRT1, a known nutrient mitochondria (Wen et al, 2013). Furthermore, autophagy can also sensor, regulates autophagic flux in SC progeny. A deficiency of eliminate intracellular pathogens, such as viruses and bacteria (Yano SIRT1 led to a delay in SC activation that could also be partially & Kurata, 2011). Interestingly, autophagy may also facilitate the rescued by exogenous pyruvate. These studies suggest that auto- immune response to infection, as resting T cells in which autophagy is phagy, regulated by SIRT1, may play an important role during SC inhibited cannot activate (Hubbard et al, 2010). Blocking autophagy activation to meet the high bioenergetic demands of the activation would therefore hinder the ability of a cell to respond to stress. process. Stem cells depend on intact autophagic machinery for the maintenance of states, characteristics, and processes that underlie stem cell Keywords activation; autophagy; quiescence; satellite cell; SIRT1 functions. When autophagy is inhibited, defects in quiescence, differ- Subject Categories Autophagy & Cell Death; Metabolism; Stem Cells entiation, and self-renewal have been reported (Guan et al, 2013; DOI 10.15252/embj.201488278 | Received 19 February 2014 | Revised 29 Phadwal et al, 2013). It has been proposed that long-lived stem cells August 2014 | Accepted 1 September 2014 that rarely divide rely on autophagy to remove damaged proteins and organelles for the maintenance of quiescence (Guan et al, 2013). This is supported by findings that HSCs that accumulate mitochondria and reactive oxygen species fail to maintain quiescence when Introduction blocked for autophagy (Liu et al, 2010; Mortensen et al, 2011). Addi- tionally, autophagy can selectively dispose of proteins and organelles Macroautophagy, hereafter referred to as autophagy, is a homeo- that can inhibit normal differentiation and self-renewal of stem cells static process with dual functions as a cellular quality control mech- (Mortensen et al, 2010, 2011; Mizushima & Komatsu, 2011). anism and a recycling system (Mizushima & Komatsu, 2011; Singh While the autophagic removal of damaged or extraneous cellular & Cuervo, 2011). Double-membraned structures engulf cytoplasmic components has been shown to be necessary for the maintenance components into vesicles called autophagosomes that later fuse with and function of stem cells, the contribution of autophagy to the lysosomes to break down their contents from which amino acids metabolic needs of stem cells has been less well studied. Pathways and other degradation products may be recycled for protein synthe- and molecules that sense and regulate cellular energy status, sis or for the tricarboxylic acid (TCA) cycle to generate energy however, have been well documented to influence stem cell func- (Ravikumar et al, 2010; Yang & Klionsky, 2010). In this way, tion (Rafalski & Brunet, 2011; Folmes et al, 2012). Manipulations of

1 Paul F. Glenn Center for the Biology of Aging, Stanford University School of Medicine, Stanford, CA, USA 2 Department of Neurology and Neurological Sciences, Stanford University School of Medicine, Stanford, CA, USA 3 Neurology Service and Rehabilitation Research and Developmental Center of Excellence, Veterans Affairs Palo Alto Health Care System, Palo Alto, CA, USA *Corresponding author. Tel: +1 650 849 1999; E-mail: [email protected] †This article has been contributed to by US Government employees and their work is in the public domain in the USA

ª 2014 The Authors The EMBO Journal 1 The EMBO Journal Autophagy in satellite cell activation Ann H Tang & Thomas A Rando Ann H Tang & Thomas A Rando Autophagy in satellite cell activation The EMBO Journal

machinery, the significance of the bioenergetic contributions from enormous increase in synthetic activity associated with the activa- uninjured autophagy in stem cell function has yet to be directly tested. tion process. Stem cells activating out of quiescence to generate proliferating progeny that will differentiate or self-renew encounter different bioenergetic requirements at each juncture (Folmes et al, 2012). Results

One crucial bioenergetic hurdle occurs during activation. Quiescent 1.5 stem cells, characterized by their low metabolic state and lower Autophagic flux is induced during SC activation mitochondrial content and activity, must meet a high demand for energy, reducing cofactors, and amino acids needed to support To determine whether autophagy is induced in SCs during the cellular growth during activation (Lunt & Vander Heiden, 2011; Hsu process of muscle regeneration, we used LC3-GFP transgenic mice, & Qu, 2013). It is known that quiescent hematopoietic stem cells in which an integral protein in autophagosome formation, LC3, is (HSCs) generate ATP primarily through glycolysis rather than the tagged with GFP (Mizushima et al, 2004). We examined LC3 expres- more productive oxidative phosphorylation (Mantel et al, 2010; sion in QSCs in uninjured muscle and in ASCs and SC progeny (i.e., 2.5 Simsek et al, 2010; Miharada et al, 2011; Suda et al, 2011). The proliferating cell populations derived from activated SCs) between level of ATP production in a quiescent state may be insufficient 1.5 and 5 days after the induction of activation by muscle injury. during stem cell activation and may activate nutrient sensors that LC3 was not detectable in QSCs but was detectable early during SC

mediate cellular metabolic homeostasis. Identifying those mediators activation and remained elevated during the phase of proliferative a�erInjury Days would enhance our understanding of the metabolic contribution to expansion of SC progeny (Fig 1A). We also analyzed LC3 expression

the regulation of stem cell activation. by immunohistochemical analysis with an LC3B antibody in 5 In this report, we studied the role of autophagy in the activation muscles of Pax7CreER/+; ROSAeYFP/+ mice (Srinivas et al, 2001; Nishijo of muscle stem cells, or ‘satellite cells’ (SCs), from the quiescent et al, 2009), in which tamoxifen treatment marks SCs with an state into the cell cycle. SCs, which are responsible for the regenera- enhanced yellow fluorescent protein (eYFP) reporter, and confirmed tive capacity of skeletal muscle (Brack & Rando, 2012; Yin et al, that SCs expressed LC3 by 1.5 days after injury (Supplementary Fig 2013), exist primarily in the quiescent state until they receive signals S1A). By 5 days after injury, as self-renewal, characterized as + B C to activate in order to begin proliferating. We found that autophagic Pax7 /MyoDÀ, is occurring for many SC progeny that have not already initiated the differentiation process (Wang & Rudnicki, flux is induced as quiescent SCs (QSCs) activate and proceed to ** enter the cell cycle and that inhibition of that flux leads to a delay in 2012), LC3 was no longer detectable in those cells (Fig 1A; Supple-

CQ 100

the activation process. SC activation is associated with a large mentary Fig S1B). These observations suggest that autophagy is - * increase in cellular ATP, an increase that is suppressed when auto- induced in SCs during activation from the quiescent state. 75 phagy is inhibited. Both delayed SC activation and suppressed ATP Because QSCs in situ are small and compact with little cytoplasm increases by inhibition of autophagy could be partially rescued by for the detection of autophagosomes, we confirmed the induction of 50 providing SCs with an exogenous energy source in the form of the autophagy by assessing autophagic flux in QSCs from uninjured 25 nutrient pyruvate. This suggests the possibility that autophagy may LC3-GFP mice and in ASCs and SC progeny from injured LC3-GFP degrade cellular components in order to provide substrates for mice. We isolated these cells to a purity of ~98% by fluorescence-

% Cells Cells with IAF % 0 energy generation during SC activation. Consistent with this hypo- activated cell sorting (FACS) (Cheung et al, 2012; Liu et al, 2013) + CQ thesis, we also show that SIRT1, a key nutrient sensor, modulates (Supplementary Fig S2). We have demonstrated, based on detailed autophagic flux during SC activation and that knocking out sirt1 in transcriptional analyses, that this FACS scheme allows us to purify

QSC ASC SC Progeny Figure 1. Autophagy is induced during SC activation after muscle injury. A Immunostaining of cryosections from tibialis anterior (TA) muscles of LC3-GFP mice to detect LC3 expression. Arrows indicate SCs in resting muscle (‘uninjured’; top ▸ row) and ASCs and SC progeny in injured muscles 1.5 days (second row) and 2.5 days (third row), respectively, after injury. Arrows and dotted circles indicate self- + * renewing (i.e., Pax7 /MyoDÀ) SCs 5 days after injury (bottom row). LC3 is expressed in ASCs and SC progeny but not in quiescent SCs (or in self-renewing SCs). In D -CQ +CQ -CQ +CQ E 6 uninjured fibers, LC3 is expressed at a low, diffuse level that becomes elevated from 2.5 to 5 days after injury. Scale bar: 30 lm. B Autophagic flux in SCs, ASCs, and SC progeny freshly sorted from uninjured and injured lower hindlimbs of LC3-GFP mice. To test for changes in autophagic flux, QSCs, ASCs, and SC progeny were sorted by FACS to a purity of ~98% from uninjured muscles and from 1.5 or 2.5 days after injury, respectively. Cells were plated onto 4 chamber slides and treated in vitro with chloroquine (CQ) or control vehicle for 2 h prior to fixing. GFP punctae, indicative of autophagosomes, accumulate upon CQ LC3B-II treatment and reflect autophagic flux. The punctae are indicated by arrows. Scale bar: 16 lm. C Percentage of SCs with induced autophagic flux (IAF). LC3-GFP punctae were counted from replicate experiments illustrated in (B); those with greater than three 2 Rela�ve punctae are considered to have induced autophagic flux (*P < 0.05; **P < 0.01). D Western blot analysis of SCs from WT mice. FACS-sorted SCs and ASCs from uninjured muscle or from muscle 1.5 days after injury, respectively, were plated and β-Ac�n Autophagic Flux 0 treated with CQ for 2 h in vitro. Western blots of cell lysates were probed with anti-LC3B and anti-b-actin antibodies. QSC ASC E Quantification of autophagic flux. The intensities of bands for LC3B-II in three independent Western blot analyses, of which Fig 3D is a representative, were first QSC ASC normalized to levels of GAPDH, and then ratios of the intensities of +CQ to –CQ conditions were calculated for each time point (*P < 0.05). Figure 1. Source data are available online for this figure.

2 The EMBO Journal ª 2014 The Authors ª 2014 The Authors The EMBO Journal 3 The EMBO Journal Autophagy in satellite cell activation Ann H Tang & Thomas A Rando Ann H Tang & Thomas A Rando Autophagy in satellite cell activation The EMBO Journal

the mTOR pathway, for example, have been reported to perturb SCs results in a phenotype similar to that observed when autophagy A pluripotency, proliferation, differentiation, and self-renewal of stem is inhibited. Together, these data suggest a model in which the cells (Murakami et al, 2004; Chen et al, 2008; Sampath et al, 2008; metabolic demands of SC activation are sensed by SIRT1 which in Zhou et al, 2009; Easley et al, 2010). Since downstream targets of turn activates the autophagic machinery in order to generate nutri- these metabolic pathways include components of the autophagic ents that are essential for the generation of ATP to support that machinery, the significance of the bioenergetic contributions from enormous increase in synthetic activity associated with the activa- uninjured autophagy in stem cell function has yet to be directly tested. tion process. Stem cells activating out of quiescence to generate proliferating progeny that will differentiate or self-renew encounter different bioenergetic requirements at each juncture (Folmes et al, 2012). Results

CQ 100 the activation process. SC activation is associated with a large mentary Fig S1B). These observations suggest that autophagy is - * increase in cellular ATP, an increase that is suppressed when auto- induced in SCs during activation from the quiescent state. 75 phagy is inhibited. Both delayed SC activation and suppressed ATP Because QSCs in situ are small and compact with little cytoplasm increases by inhibition of autophagy could be partially rescued by for the detection of autophagosomes, we confirmed the induction of 50 providing SCs with an exogenous energy source in the form of the autophagy by assessing autophagic flux in QSCs from uninjured 25 nutrient pyruvate. This suggests the possibility that autophagy may LC3-GFP mice and in ASCs and SC progeny from injured LC3-GFP degrade cellular components in order to provide substrates for mice. We isolated these cells to a purity of ~98% by fluorescence-

QSCs and ASCs (Liu et al, 2013). We treated both QSCs and ASCs We found further evidence for induction of autophagy during dramatically increased to > 80% over the following 24 h (Fig 2C). that SCs transfected with atg5/7 siRNAs had lower levels of cyclins in vitro with an inhibitor of autophagy, chloroquine (CQ), for 2 h SC activation by examining fiber-associated SCs in fiber explants Notably, autophagy was induced in ASCs that had not initiated A and E and phospho-Rb but higher levels of p27 relative to control to allow the accumulation of autophagosomes that appear as ex vivo (Brack et al, 2007; Boutet et al, 2012). Single muscle fibers DNA synthesis and hence had not entered the cell cycle; only SCs (Supplementary Fig S6). These data show that inhibition of GFP+ punctae (Mizushima et al, 2010). QSCs showed no detect- were isolated from LC3-GFP mice and allowed to activate in about two-thirds of the ASCs with induced autophagy after 36 h autophagy impacts cell cycle entry prior to S phase for SCs activat- able autophagic flux (only diffuse LC3 expression), whereas more culture over the course of 48 h in the presence of EdU, which were EdU positive (Fig 2C). These data confirm that autophagy is ing out of quiescence. than half of the ASCs showed an induction of autophagic flux we use to identify the final stage of SC activation during which induced early during SC activation, prior to the initiation of DNA within 1.5 days after injury (Fig 1B and C). That proportion they enter the cell cycle. We assessed autophagic flux at 12-h synthesis. Autophagy contributes bioenergetically to the process of increased to more than 80% in SC progeny by 2.5 days after intervals by treating fibers for 2 h with CQ prior to fixation. Auto- SC activation injury (Fig 1C). Moreover, ASCs from wild-type (WT) mice phagic flux was not detected in SCs during the first 12 h in culture Inhibition of autophagy leads to a delay in SC activation showed an increase in the lipidated, membrane-bound form of with CQ treatment (Supplementary Fig S3A). Within the next Since one role of autophagy is to produce energy by catabolizing LC3b, LC3b-II, in response to CQ treatment (Fig 1D and E). These 12 h, as ASCs were just beginning to enter the cell cycle (Supple- To determine the functional significance of the increase in autopha- intracellular contents (Rabinowitz & White, 2010), we reasoned that data demonstrate that autophagic flux increases during SC activa- mentary Fig S3B), an increase in autophagic flux could already be gic flux during SC activation, we tested the effect of inhibiting auto- autophagy may support the cellular energy demands occurring tion in vivo. detected in just over 5% of the SCs (Fig 2A–C). That percentage phagy during ex vivo activation of fiber-associated SCs and in vitro during the activation process. It has been shown that quiescent activation of QSCs. Autophagic flux increases in isolated QSCs HSCs have lower mitochondrial content and activity than fast- just as it does during ex vivo activation of fiber-associated SCs cycling HSCs or hematopoietic progenitors and generate energy DAPI LC3-GFP Pax7/MyoD EdU Merged A CQ (Fig 3A–C; Supplementary Fig S4A and B). Administration of CQ or primarily through glycolysis, whereas increased ATP production 3-methyladenine (3-MA), chemical inhibitors at two different stages and mitochondrial membrane potential are associated with progres- - of the autophagic process (Klionsky et al, 2008; Mizushima et al, sion through the G1 phase of the cell cycle (Schieke et al, 2008; 2010), inhibited SC activation, measured as a reduction of the Folmes et al, 2012). This increase in oxidative metabolism is neces- 24 number of SCs entering the cell cycle, in fiber-associated SCs ex sary to fuel increased biosynthesis for DNA replication as well as vivo and sorted SCs in vitro from WT mice (Fig 3D). In addition, transcriptional and translational processes to support cell growth + we blocked autophagy more specifically with siRNAs against atg5 prior to cell division (Lunt & Vander Heiden, 2011). Recently, we and atg7, essential genes of the autophagic pathway (Kuma et al, reported the identification of an alert phase of the quiescent state,

2004; Komatsu et al, 2005). We confirmed that this led to a reduc- termed GAlert, in which SCs, though still quiescent, have higher - tion in transcript levels for these two genes (Supplementary Fig mitochondrial activity and correlatively larger cellular volume than

36 S5A), a lower autophagic flux (Supplementary Fig S5B), and an QSCs in G0 (Rodgers et al, 2014). We therefore compared mitochon- inhibition of SC activation (Fig 3E). To rule out the possibility that drial activity, ATP content, and cellular volumes in QSCs, ASCs, and + inhibition of autophagy either by pharmacological or siRNA treat- SC progeny. MitoTracker labeling was about 5- and 10-fold higher ment caused irreversible cell cycle arrest, we cultured the SCs for in ASCs and SC progeny, respectively, reflecting a significant an additional 12 h after inhibitor or siRNA treatment and found increase in mitochondrial activity compared with QSCs (Fig 4A and B). that the cumulative measure of cell cycle entry was not signifi- ATP levels, too, increased with SC activation, with ASCs and SC - cantly affected (Supplementary Fig S5C). This indicates that the progeny exhibiting 10 and 100 times greater levels, respectively, Time a�er Isola�on a�er Time (Hours) Isola�on 48 inhibition of SC activation observed at 24 h was not due to a than those in QSCs (Fig 4C). Cellular size, as determined by relative permanent cell cycle arrest but rather to a delay in activation. We cell volume, increased by 1.4 and 2.7 times in ASCs and SC progeny, + also confirmed with anti-active-caspase-3 staining that atg5/7 respectively, compared to that of QSCs (Fig 4D; Supplementary Fig siRNA-transfected SCs were not prone to cell death (Supplementary S7). These data confirm that cellular metabolism increases during Fig S5D). To further confirm the results of delayed activation due activation, with ATP levels increasing in far greater proportion than to pharmacologic or siRNA inhibition of autophagy, we genetically cell volume. B C knocked out atg5 in SCs using a mouse transgenic for a SC-specific, Since ATP content in ASCs is much greater than in QSCs, we ER 100 *** +ve tamoxifen-inducible Cre allele and homozygous for floxed atg5 next tested whether autophagy contributes to this increase. We * EdU EdU-ve alleles (Hara et al, 2006; Nishijo et al, 2009). Because autophagy inhibited autophagy in freshly isolated QSCs with atg5 and atg7 75 * may have an additional role in clearing damaged proteins and siRNAs and found a reduction in the increase in their ATP

with IAF 50 organelles in long-lived quiescent stem cells (Guan et al, 2013), we content during activation (Fig 5A). We therefore conclude that performed an acute knockout of atg5 to observe only the short- autophagy serves to provide bioenergetic resources to SCs under- 25 Cells term effects of inhibiting autophagy during SC activation. We first going activation from the quiescent state. To test whether the % % 0 confirmed that recombination in the atg5 locus had occurred delay in QSC activation by inhibition of autophagy can be attrib- 0 12 24 36 48 24 36 48 Time a�er Isola�on (Hours) (Supplementary Fig S5E). Similar to pharmacologic inhibition of uted to insufficient energy levels, we blocked autophagy in autophagy and by atg5 and atg7 siRNA transfection, acute genetic freshly isolated QSCs and fiber-associated SCs and supplemented Time a�er Isola�on (Hours) deletion of atg5 likewise led to a delay in DNA synthesis in fiber- the media with an exogenous metabolite, sodium pyruvate, as an associated SCs and sorted SCs (Fig 3F). We conclude from these energy source. We found that the delay in activation resulting data that the increase in autophagic flux is essential for normal from the siRNA transfections could be partially rescued by Figure 2. Autophagy is induced in fiber-associated SCs during ex vivo activation. activation kinetics of quiescent SCs. sodium pyruvate (Fig 5B). Moreover, ATP levels in atg5/7 siRNA- A Immunostaining and EdU detection of single fibers from LC3-GFP mice cultured in the presence of EdU for up to 48 h. Parallel cultures were assessed at 12-h Since DNA synthesis is a late component of the process of SC transfected SCs could also be partially rescued by sodium pyru- intervals between 24 and 48 h and were treated with CQ for 2 h prior to fixing. Arrows indicate LC3-GFP punctae. Scale bar: 5 lm. B Higher resolution images of CQ-treated fiber-associated SCs acquired at 63× magnification. Arrows indicate LC3-GFP punctae. Scale bar: 5 lm. activation, we investigated cell cycle proteins governing the progres- vate addition (Supplementary Fig S8). These data suggest that the C Percentage of fiber-associated SCs with induced autophagic flux (IAF). A cell exhibiting IAF was defined as one having greater than three GFP punctae after CQ sion through the G1/S checkpoint. An increase in cyclin A, cyclin E, delay of SC activation that occurs when autophagy is inhibited is + treatment. The percentage of cells with IAF was calculated and categorized according to EdU incorporation status (i.e., EdU or EdUÀ) at each time point (*P < 0.05; and phosphorylated retinoblastoma protein (Rb) and a decrease in due to a failure in the generation of sufficient energetic resources ***P < 0.001). p27 levels are among the changes typically observed during this necessary for the biosynthetic processes required for the progression (Morgan, 1997; Lundberg & Weinberg, 1998). We found activation process.

4 The EMBO Journal ª 2014 The Authors ª 2014 The Authors The EMBO Journal 5 The EMBO Journal Autophagy in satellite cell activation Ann H Tang & Thomas A Rando Ann H Tang & Thomas A Rando Autophagy in satellite cell activation The EMBO Journal

A B A B 100 ** * 75

50 - CQ 25

% Cells with IAF % 0 0 2412 Time in Culture (Hours) C -CQ +CQ -CQ +CQ

+ CQ LC3B-II

GAPDH 0 12 24 0 24 C D Time in Culture (Hours) Time in Culture (Hours)

D Fibers Sorted SCs

E Fibers Sorted SCs Figure 4. ASCs and SC progeny have increased metabolic activity than QSCs. A Increase in functional mitochondria of ASCs and SC progeny versus QSCs. Mononuclear cells from uninjured and injured muscles were stained with antibodies for the sorting of myogenic cells, treated with MitoTracker Deep Red FM, and analyzed by FACS. B Relative MitoTracker incorporation of QSCs, ASCs, and SC progeny. MitoTracker signal intensities obtained from replicate experiments as illustrated in (A) were normalized to levels in QSCs from uninjured muscles (**P < 0.01). C Relative ATP contents of QSCs, ASCs, and SC progeny. QSCs, ASCs, and SC progeny were sorted from uninjured muscles or from muscles 1.5 or 2.5 days after injury, respectively. ATP contents were normalized to levels obtained from QSCs from uninjured muscles (*P < 0.05; **P < 0.01). D Cell volumes in ASCs and SC progeny versus QSCs. Volumes of QSCs, ASCs, and SC progeny were determined based on forward scatter from FACS analyses. Relative volumes were calculated from standard curves from beads of standard sizes (*P < 0.05; **P < 0.01). siRNA tnemtaerT ANRis Treatment

F Fibers Sorted SCs A B

Figure 3. Blocking autophagy inhibits SC activation. A Induced autophagy in in vitro-activated SCs. QSCs were FACS-sorted from uninjured LC3-GFP mice and cultured for 24 h. The cultures were treated with CQ for 2 h at 12-h time points prior to fixing. Arrows indicate LC3-GFP punctae. Scale bar: 16 lm. B Percentage of SCs with IAF. LC3-GFP punctae were counted in SCs from replicate experiments as illustrated in (A); those with greater than three punctae were considered to have induced autophagy (*P < 0.05; ***P < 0.001). Figure 5. Blocking autophagy leads to a bioenergetic deficiency during SC activation. C Western blot analysis of QSCs and in vitro-activated SCs. Sorted QSCs were plated and treated in vitro with CQ for 2 h after 0 or 24 h (at which point they are ASCs) A Decrease in ATP levels of ASCs after inhibition of autophagy. FACS-sorted QSCs were transfected with siRNAs against atg5 and atg7, cultured for 24 h to allow in vitro in culture. The Western blot was probed with anti-LC3B and anti-GAPDH antibodies. activation and at which point they are ASCs, and then collected for ATP assay. Control cultures were transfected with a negative control siRNA (*P < 0.05). D Inhibition of SC activation by pharmacological inhibition of autophagy. Freshly isolated single fibers (left) or FACS-sorted QSCs (right) were cultured for 24 h in the B Partial rescue of delay in activation by addition of exogenous metabolite. Freshly isolated single fibers (left) and FACS-sorted QSCs (right) were transfected with presence of EdU and treated with CQ or 3-MA for 8 h prior to fixing. The percentage of cells positive for EdU was assessed. Control cultures were treated with no siRNAs against atg5 and atg7 and cultured for 24 h with EdU and 20 lM sodium pyruvate. The samples were then assessed for EdU incorporation. Control cultures autophagic inhibitors (*P < 0.05; **P < 0.01). were transfected with a cyclophilin B siRNA (n.s., not significant; *P < 0.05). E Inhibition of SC activation by atg5 and atg7 knockdown. Freshly isolated single fibers (left) or FACS-sorted QSCs (right) were transfected with siRNAs against atg5 and atg7, cultured with EdU for 24 h, and assessed for EdU incorporation. Control cultures were transfected with cyclophilin B siRNA (*P < 0.05; ***P < 0.001). F Inhibition of activation by genetic loss of atg5. Freshly isolated single fibers (left) or FACS-sorted QSCs (right) were isolated from Pax7CreER/+; ATG5fl/fl mice in media SIRT1 induces autophagy during SC activation containing 4-hydroxytamoxifen, cultured in media with EdU and 4-hydroxytamoxifen for 24 h, and assessed for EdU incorporation (n.s., not significant; *P < 0.05). cellular metabolism have increasingly been shown to regulate stem cell maintenance and function (Rafalski & Brunet, 2011; Source data are available online for this figure. How the need for more energy during QSC activation induces an Folmes et al, 2012), we reasoned that molecules or processes autophagic response is not yet understood. Because regulators of known to sense and respond to a cell’s nutritional status may

6 The EMBO Journal ª 2014 The Authors ª 2014 The Authors The EMBO Journal 7 The EMBO Journal Autophagy in satellite cell activation Ann H Tang & Thomas A Rando Ann H Tang & Thomas A Rando Autophagy in satellite cell activation The EMBO Journal

A B A B 100 ** * 75

50 - CQ 25

% Cells with IAF % 0 0 2412 Time in Culture (Hours) C -CQ +CQ -CQ +CQ

+ CQ LC3B-II

GAPDH 0 12 24 0 24 C D Time in Culture (Hours) Time in Culture (Hours)

D Fibers Sorted SCs

F Fibers Sorted SCs A B

/ signal for the induction of autophagy during SC activation. We this, we examined autophagic flux in sirt1À À SC progeny Because SIRT1 is a positive regulator of autophagic flux, we next with and deacetylates ATG5, 7, and 8 in a nutrient-deprivation model hypothesized that SIRT1, which responds to metabolic changes obtained from mice transgenic for a SC-specific, tamoxifen- tested whether loss of sirt1 would phenocopy the inhibition of auto- using murine embryonic fibroblasts. To determine whether a similar and mediates the benefits of caloric restriction (Herranz & inducible Pax7CreER allele and homozygous for floxed sirt1 alleles phagy with respect to QSC activation. We found that FACS-sorted mechanism occurs in SCs, we tested for endogenous protein–protein / Serrano, 2010; Chalkiadaki & Guarente, 2012), may regulate auto- (Li et al, 2007). Administration of 4-hydroxytamoxifen leads to and fiber-associated sirt1À À SCs were delayed in activation interactions between SIRT1 and ATG5 and ATG7 and found that phagic flux in SCs in response to increased energy needs during CRE-mediated recombination that excises exon 4 of sirt1, thus (Fig 7A). We next tested whether this delay was due to a metabolic SIRT1 indeed interacted with ATG7 but not with ATG5 (Fig 8A). / +/ / / activation. In addition to its function in metabolism, SIRT1 has creating sirt1À À cells in the SC lineage. These cells also showed deficiency. We therefore stained sirt1 À and sirt1À À QSCs with Moreover, we found higher levels of acetylated ATG7 in sirt1À À SC previously been shown to deacetylate a number of autophagy reduced autophagic flux by immunocytochemical staining and by MitoTracker and found that the latter had lower mitochondrial progeny than in the sirt1+/+ population (Fig 8B), suggesting that proteins in vitro and to modulate autophagic flux under starvation Western blotting (Fig 6B and C). To test whether loss of sirt1 activity than the former (Supplementary Fig S10). Supplementing SIRT1 deacetylates ATG7. These data also suggest that the regulation conditions (Lee et al, 2008; Hariharan et al, 2010). In order to test affects autophagic flux during SC activation, we transfected the culture media with sodium pyruvate partially rescued the delay of autophagy by SIRT1 occurs at least in part through its interaction / whether SIRT1 could modulate the autophagic response in prolif- siRNAs against sirt1 into SCs from LC3-GFP mice and found that in activation in sirt1À À SC populations (Fig 7B). Taken together, with ATG7. The absence of a physical interaction between SIRT1 and erative SC progeny, we used a small molecular inhibitor of SIRT1, this also decreased autophagic flux (Fig 6D; Supplementary Fig S9). these results demonstrate that SIRT1 is necessary for the normal ATG5 in our studies suggests that our findings of delayed SC activa- EX-527 (Napper et al, 2005), and found that it led to a These data confirm that SIRT1 normally functions to induce auto- induction of autophagy during SC activation, a process that fulfills a tion by conditional deletion of atg5 may be SIRT1 independent. dose-dependent decrease in autophagic flux (Fig 6A). To confirm phagic flux during SC activation and in SC progeny. bioenergetic need during this critical cellular transition. Clearly, In another approach, we explored two metabolic pathways, the however, a considerable degree of autophagy is induced even in the AMPK and mTOR pathways, that have been shown to link SIRT1 absence of SIRT1 (Fig 6C and D), reflecting the fact that there are activity to the regulation of autophagy (Ghosh et al, 2010; Dunlop & A B sirt1+/+ sirt1-/- likely multiple parallel pathways of autophagy inductions and Tee, 2013; Parzych & Klionsky, 2014; Hong et al, 2014; Ou et al, regulation. 2014). When we knocked out sirt1 in SC progeny, AMPKa became hypophosphorylated, whereas the phosphorylation levels of mTOR Mechanism of SIRT1 regulation of autophagy in SCs targets, S6 and 4E-BP1, remained unchanged (Fig 8C). These data suggest that SIRT1 may modulate autophagy in SCs through the - CQ - CQ To investigate how SIRT1 may regulate autophagy, we first examined AMPK pathway and support the hypothesis that autophagy is components of the autophagic machinery as potential targets of induced during SC activation in response to a relative lack of nutri- SIRT1. Lee et al (2008) had previously shown that SIRT1 interacts ent availability, contributing to the evidence that autophagy may

A CQ + + + CQ

0 5 25

EX-527 (μM)

C D sirt1+/+ sirt1-/- 8 ** * 100 - + - + CQ 6 SIRT1 75

SIRT1Δex4 4 50 B Rela�ve LC3B-II 2 25 Autophagic Flux % Cells with IAF % 0 0 β-Actin sirt1+/+ sirt1-/- Control sirt1 siRNA Treatment

Figure 6. SIRT1 mediates autophagic flux. A Reduction of autophagic flux resulting from chemical inhibition of SIRT1. SC progeny obtained from WT mice were incubated with SIRT1 inhibitor EX527, treated with CQ for 2 h, and immunostained with anti-LC3B antibody. Arrows indicate LC3B punctae. Scale bar: 16 lm. B Reduction of autophagic flux resulting from genetic loss of sirt1. SC progeny obtained from WT or Pax7CreER/+; SIRT1fl/fl mice were treated with vehicle control or +/+ / 4-hydroxytamoxifen in vitro to generate sirt1 and sirt1À À cells, respectively. The cells were then treated with CQ or vehicle control, fixed, and immunostained with anti-LC3B antibody. Scale bar: 16 lm. +/+ / C Western blot analysis of the inhibition of autophagic flux resulting from genetic loss of sirt1. Lysates of sirt1 and sirt1À À cells were analyzed by Western blot analysis (left panel). Blots were probed with anti-LC3B and anti-b-actin antibodies. To quantify autophagic flux (right panel), the intensities of bands for LC3B-II from three independent Western blots were first normalized to levels of GAPDH, and then ratios of the intensities of +CQ to –CQ conditions were calculated for each time Figure 7. SIRT1 mediates SC activation. point (*P < 0.05). CreER/+ fl/+ CreER/+ fl/fl / +/ A Delay in activation of SCs with genetic loss of sirt1. Pax7 ; SIRT1 and Pax7 ; SIRT1 mice were treated with tamoxifen to create sirt1À À and sirt1 À D Decrease in autophagic flux in in vitro-activated SCs resulting from sirt1 knockdown. Sorted QSCs from LC3-GFP mice were transfected with siRNAs against sirt1, genotypes in the SC lineage. Single fibers (left) and QSCs (right) sorted by FACS to a purity of ~98% from these animals were cultured for 24 h in the presence of EdU. cultured for 24 h, and treated with CQ. LC3-GFP punctae were counted, and cells with greater than three punctae were characterized as having induced autophagic EdU incorporation was then determined (*P < 0.05). flux (IAF). Control cultures were treated with a negative control siRNA (**P < 0.01). / +/ B Partial rescue of delay in activation by addition of an exogenous metabolite. sirt1À À and sirt1 À fiber-associated (left) and sorted QSCs (right) were obtained as in (A) Source data are available online for this figure. and cultured with EdU and 20 lM sodium pyruvate. EdU incorporation was assessed after 24 h in culture (n.s., not significant; *P < 0.05).

8 The EMBO Journal ª 2014 The Authors ª 2014 The Authors The EMBO Journal 9 The EMBO Journal Autophagy in satellite cell activation Ann H Tang & Thomas A Rando Ann H Tang & Thomas A Rando Autophagy in satellite cell activation The EMBO Journal

A CQ + + + CQ

0 5 25

EX-527 (μM)

C D sirt1+/+ sirt1-/- 8 ** * 100 - + - + CQ 6 SIRT1 75

SIRT1Δex4 4 50 B Rela�ve LC3B-II 2 25 Autophagic Flux % Cells with IAF % 0 0 β-Actin sirt1+/+ sirt1-/- Control sirt1 siRNA Treatment

A WB: SIRT1 C is genetically knocked out in Myf5+ progenitor cells that give rise to unrelated to mitochondrial function, such as an increase in glyco- muscle and brown adipose tissue demonstrated the requirement for lytic flux that is decoupled from oxidative phosphorylation. SIRT1 autophagy in proper brown fat and skeletal muscle development However, our data are consistent with many others showing (Martinez-Lopez et al, 2013). Another recent study using an HSC- increased oxidative activity in proliferating progenitor cells relative Input * specific knockout of atg7 suggests that inhibition of autophagy may to stem cells (Folmes et al, 2012; Hsu & Qu, 2013; Shyh-Chang ATG5 1.00 lead to a loss of quiescence (Mortensen et al, 2011). Other studies et al, 2013). Furthermore, our observations that inhibiting auto- p-AMPKα 0.75 perturbing metabolic pathways that may regulate autophagy show phagy suppresses the increase in ATP content of activating SCs and IP ATG7 defects in the maintenance of quiescence. For example, knocking delays their activation support the hypothesis that autophagy IgG AMPKα 0.50 out Pten or TSC1 in HSCs de-represses the mTOR pathway, a major provides energy sources that are needed during activation from AMPKα negative regulator of autophagy, and causes inappropriate prolifera- quiescence. We hypothesize that activation out of quiescence leads β-ac�n p- 0.25 tion (Yilmaz et al, 2006; Zhang et al, 2006; Gan et al, 2008). A simi- to a state of relative nutrient deprivation as cells face a tremendous 0 lar loss of HSC quiescence is seen in knockouts of FoxO1 or FoxO3 increase in bioenergetic demand for all of the processes of growth, 1.5 n.s. (Miyamoto et al, 2007; Tothova et al, 2007), proteins which also macromolecular synthesis, and organellogenesis. This state of p-S6 function in promoting autophagy (Sengupta et al, 2009; Hariharan nutrient-deprivation stress could be the signal that activates the 1.0 et al, 2010). Because deletions of Pten or FoxO1, 3, and 4 also lead SIRT1 pathway. B ** S6 WB 2.0 p-S6 to a greater propensity for neural stem cells to enter the cell cycle Notably, the fact that inhibition of autophagy delays, rather than 0.5 (Groszer et al, 2006; Paik et al, 2009; Renault et al, 2009), meta- prevents, cell cycle entry suggests that other processes may compen- 1.5 β-ac�n Input SIRT1 bolic regulation may be a general mechanism governing stem cell sate for the bioenergetic insufficiency in autophagy-deficient SCs. 0

Amounts 1.0 quiescence. However, autophagy as a downstream effector of these Possibilities include glycolysis to generate additional ATP and pyru- 1.5 n.s. Input Ac-ATG7 p-4E-BP1 pathways has not been reported. vate and fatty acid oxidation (FAO) to generate NADH, FADH2, and ATG7 0.5 of Using multiple approaches, we show that the inhibition of acetyl CoA (Carracedo et al, 2013; Hsu & Qu, 2013). If QSCs have a IP: Ac-K 1.0 ATG7 Rela�ve 0 autophagy does not lead to a loss of SC quiescence but rather substantial glycolytic reserve, then glycolytic flux may increase in sirt1+/+ sirt1-/- 4E-BP1 delays the activation of these cells out of the quiescent state. response to the shortfall in energy during activation (Nicholls et al, 0.5 p-4E-BP1 Rela�ve AmoutnsRela�ve phosphoproteins of Unlike previous studies which used genetic knockout approaches 2010; Pelletier et al, 2014). FAO, which can produce twice as much β-ac�n 0 and analyses long after the gene deletions, we used pharmaco- ATP as carbohydrate metabolism when normalized to dry mass sirt1+/+ sirt1-/- logical and siRNA approaches to inhibit autophagy acutely. As (Carracedo et al, 2013), has been shown to support HSC mainte- such, our studies were designed to avoid compensatory or nance as well as promote ATP production in response to metaboli-

Figure 8. The regulation of autophagy in SCs by SIRT1. secondary effects associated with chronic autophagy inhibition. cally stressed cancer cells (Zaugg et al, 2011; Carracedo et al, 2012; +/+ / Even in our conditional knockout studies of atg5, we examined Ito et al, 2012). Furthermore, FAO can generate anaplerotic fuels A SIRT1 interacts with ATG7. Anti-ATG5 or anti-ATG7 antibodies were used in immunoprecipitation (IP) reactions with protein lysates from sirt and sirtÀ À SC progeny. Western blots were probed with an anti-SIRT1 antibody. the process of SC activation immediately after gene deletion. It that can feed into the TCA cycle as succinyl-CoA (Velez et al, 2013). +/+ / B Loss of sirt1 leads to hyperacetylation of ATG7. An anti-acetylated lysine antibody was used in IP reactions with protein lysates from sirt and sirtÀ À SC progeny. will be interesting to determine whether the phenotypes of FAO could therefore be another potential source of energy for SC Western blots were probed with an anti-ATG7 antibody (left panel). ATG7 bands from IP reactions on three independent Western blots were normalized to ATG7 input chronic inhibition of autophagy are indeed secondary or tertiary activation. bands prior to calculating relative amounts of protein (right panel) (**P < 0.01). +/+ / effects of the kinds of defects in stem cell activation that we Cells have an intricate system of metabolic pathways and nutri- C Effect of loss of sirt1 on AMPK and mTOR pathways. Protein lysates from sirt and sirtÀ À SC progeny were subjected to Western analysis and probed with antibodies against the phosphorylated and non-phosphorylated forms of AMPKa,S6, and 4E-BP1 proteins (left panel). Bands for phosphoproteins were normalized to demonstrate here. It will also be interesting to examine the ent sensors to ensure that their bioenergetic needs are fulfilled bands for total protein prior to calculating relative amounts of phosphoproteins (right panel) (n.s., not significant; *P < 0.05). effects of inhibition of autophagy on later stages of SC activa- (Shanware et al, 2013). How these pathways trigger an autophagic Source data are available online for this figure. tion, proliferation, and differentiation that we did not examine response is not fully understood. Two pathways that have been here. shown to regulate autophagy in response to nutrient levels are the Stem cells have unique metabolic hallmarks that characterize AMPK and mTOR pathways (Parzych & Klionsky, 2014). mTOR their quiescent, proliferative, and differentiated states (Folmes et al, can inhibit autophagy by disrupting the ULK1–ATG13–FIP200 play a critical role in cellular bioenergetics during times of relative Our findings parallel to those of Hubbard et al (2010) who 2012). For example, quiescent HSCs, which have low mitochondrial complex that initiates autophagy (Ganley et al, 2009; Hosokawa nutrient deprivation. demonstrated that autophagy produces energy required for another content and oxidative activity, primarily generate ATP from glyco- et al, 2009; Jung et al, 2009), while AMPK can activate autophagy cellular process with a high bioenergetic demand, the activation of lysis rather than oxidative phosphorylation (Piccoli et al, 2005; directly through ULK1 or indirectly through the inhibition of the resting T cells in response to antigen recognition. Similar to our Lonergan et al, 2007; Simsek et al, 2010; Miharada et al, 2011). mTOR pathway (Inoki et al, 2003; Egan et al, 2011; Kim et al, Discussion observations during SC activation, they detected an induction of Proliferating cells, on the other hand, have an increased demand for 2011; Dunlop & Tee, 2013). SIRT1, in turn, can negatively regulate autophagy after T-cell stimulation and showed that blocking auto- biogenesis and primarily use oxidative phosphorylation to generate the mTOR pathway and positively regulate AMPK (Lan et al, 2008; Our results demonstrate the importance of autophagy in SC activa- phagy leads to reduced ATP production and hinders T-cell activa- ATP (Hsu & Qu, 2013). As proliferating cells cycle, aerobic ATP Ghosh et al, 2010; Kapahi et al, 2010; Takeda-Watanabe et al, tion. Autophagic flux appears to be too low to be detectable in QSCs, tion, which can be rescued by exogenous pyruvate. Intriguingly, production and oxidative phosphorylation fluctuate, increasing in 2012). Our studies indicate that SIRT1, perhaps in relation to its but increases along with ATP content occur during SC activation. they observed a difference in autophagosomal cargo selection in G1 phase (Schieke et al, 2008). In fact, without sufficient energy role as a nutrient sensor, may be a mediator of the induction of Notably, the increased flux is detected at a time point when the resting and activated T cells, highlighting the importance of and metabolites, proliferating cells will arrest in G1 phase (Holley & autophagic flux during SC activation when nutrient demands are majority of the cells have not yet entered the cell cycle. We show that substrate specificity of autophagy in its quality control and energy Kiernan, 1974; Jones et al, 2005). Stem cells activating out of a likely to be extremely high. Our findings that AMPK is hypo- / inhibiting autophagy leads to a decrease in autophagic flux, a reduc- production modes. Along these lines, the possibility that the role of quiescent state have lower levels of mitochondria and lower oxida- phosphorylated in sirt1À À SC progeny while phospho-S6 and tion in ATP content, and a delay in activation, suggesting a causal substrate-specific autophagy in that study or in our study could tive capacity than proliferating progenitor cells (Hsu & Qu, 2013; phospho-4E-BP1 remain unchanged suggest that SIRT1 signals link between autophagic flux, bioenergetic status, and SC activation. involve the selective degradation of inhibitors of the activation Shyh-Chang et al, 2013). We show that activating SCs have through AMPK rather than the mTOR pathway to regulate auto- That a partial rescue in the delay in activation can be achieved with processes has not been excluded. increased mitochondrial activity with a concomitant increase in phagy in SC progeny. sodium pyruvate confirms a bioenergetic requirement that auto- Whereas several studies have demonstrated the importance of ATP content relative to QSCs. Of course, these observations do not Previous studies have shown that SIRT1 induces autophagy in phagy may fulfill during activation. We also show that SIRT1, a autophagy in the maintenance and function of stem cells (Guan exclude the possibility that the increase in ATP was not a direct starvation or pathological conditions (Lee et al, 2008; Hariharan known nutrient sensor, induces autophagic flux during SC activation et al, 2013; Phadwal et al, 2013), none have specifically addressed consequence of the increase in mitochondrial activity since et al, 2010; Jeong et al, 2013). Lee et al (2008) demonstrated in and that its inhibition also causes a delay in SC activation. stem cell activation from a quiescent state. One study in which atg7 increases in ATP content could result from metabolic processes vitro interaction between SIRT1 and three ATG proteins, leading to

10 The EMBO Journal ª 2014 The Authors ª 2014 The Authors The EMBO Journal 11 The EMBO Journal Autophagy in satellite cell activation Ann H Tang & Thomas A Rando Ann H Tang & Thomas A Rando Autophagy in satellite cell activation The EMBO Journal

ings of Lee et al that SIRT1 can regulate components of the autopha- transcripts was calculated according to the DDCT method (Suda assay was performed, if required. Fibers were washed and gic process (Lee et al, 2008). Unlike the previous studies which Injured and uninjured hindlimb muscles were dissected and et al, 2011). mounted in mounting media. For analysis of GFP punctae or LC3B used nutrient deprivation and induced neurotoxicity as stressors digested into mononuclear cells as previously described (Cheung immunoreactivity in isolated cells, the same procedure was (Lee et al, 2008; Hariharan et al, 2010; Jeong et al, 2013), we show et al, 2012; Liu et al, 2013). Briefly, hindlimb muscles were Determination of autophagic flux performed as for fibers, except that washes were done with PBS. that the induction of autophagy can be mediated, at least in part, by dissected and cut into small pieces with scissors. They were then For immunostaining of muscle sections, TA muscles were fixed in SIRT1 under conditions of normal nutrient availability. digested in Ham’s F10 medium with 10% horse serum (i.e., wash Autophagic flux was determined by the relative accumulation of 0.5% PFA for 5 h at room temperature, dehydrated in 20% sucrose Our studies demonstrating that autophagy contributes to the media) with collagenase II at 500 U/ml at 37°C for 1.5 h. The LC3B-II by Western blot analysis or by the assessment of GFP punc- in PBS, and frozen in Optimal Cutting Temperature mounting metabolic adaptation that SCs undergo during activation add to the mixture was then washed and digested for an additional 30 min at tae in SCs from LC3-GFP mice upon inhibition of autophagy by media (Sakura Finetek). Cryosections were collected, fixed in 2% understanding of the importance of metabolic flexibility in stem cell 37°C in wash media with 100 U/ml collagenase II and 2 U/ml culturing with 20 lM CQ for 2 h (Mizushima et al, 2004, 2010; PFA for 5 min at room temperature, and washed in PBS with 0.3% fate determination. Stem cells adapt their energy usage and produc- dispase. Finally, the mixture was triturated with a 20-gauge needle Klionsky et al, 2008). To analyze autophagic flux by quantifying Triton X-100. The samples were blocked in PBS with 0.3% Triton tion via the manipulation of different metabolic pathways to support and washed. Myogenic cells were then isolated by fluorescence- punctae, myofibers or SCs from LC3-GFP mice were fixed as X-100 and 20% donkey serum for at least 1 h at room temperature the changing bioenergetic needs during quiescence, proliferation, activated cell sorting (FACS) by negative selection with anti- described below, and the punctae were counted. and then washed. Mouse primary antibodies were conjugated to differentiation, or self-renewal (Folmes et al, 2012). The studies CD31-FITC, anti-CD45-FITC, and anti-Sca-1-Pacific Blue antibodies secondary antibodies with the Zenon kit (Life Technologies) and reported here suggest that the activation out of quiescence is an (Biolegend). An anti-VCAM-biotin primary antibody and a MitoTracker staining incubated along with primary antibodies from other species with example of a stem cell function that requires rapid and dramatic streptavidin-PE-cy7 secondary antibody (Biolegend) were used for sections overnight at 4°C. Sections were washed, incubated in changes in the metabolic activity and that the induction of autopha- positive selection. The purity of the sorted cells, as determined by Hindlimb muscles were dissected and digested into mononuclear secondary antibodies for 1 h at room temperature, washed again, gic activity may be a critical component of those metabolic shifts. immunocytochemical staining with anti-Pax7 and anti-MyoD cells and stained with antibodies for satellite cell isolation by FACS and mounted in mounting media. Primary antibodies were as Autophagy may therefore be an additional general mechanism antibodies, was ~98%. as described above. They were then incubated with MitoTracker follows: mouse anti-Pax7 (DSHB), mouse anti-MyoD (Dako), rabbit which confers on stem cells another degree of metabolic flexibility deep red FM (Life Technologies) at 37°C for 30 min. The cells were anti-LC3B (Cell Signaling), rabbit anti-GFP (Life Technologies), to respond to fluctuating bioenergetic demands. SC progeny isolation washed twice prior to analysis on the FACSAria II or III. chick anti-GFP (Aves), and rat anti-Laminin (Sigma-Aldrich). Secondary antibodies used are conjugated to Alexa fluorescent dyes Hindlimb muscles were subjected to needle injury and harvested ATP quantification (Life Technologies). Materials and Methods 2.5 days later. The muscles were digested and sorted as described for SC isolation above. FACS-sorted SC progeny were then plated in ATP content of SCs was determined with the ATP Bioluminescence Protein immunoprecipitation Mouse strains Ham’s F-10 nutrient mixture with 10% horse serum (HS) onto Assay Kit CLS II (Roche) according to manufacturer’s recommenda- 35-mm tissue culture dishes coated with extracellular matrix tions. SC progeny were harvested, washed in PBS, and lysed in RIPA lysis LC3-GFP transgenic mice and ATG5fl/fl mice, in which exon 3 of (Sigma) at a 1:500 dilution. The media was exchanged with Ham’s buffer (50 mM Tris–HCl, pH 7.5; 150 mM NaCl; 5 mM EDTA; 1% atg5 is flanked by loxP sites, were obtained from the Riken Bio- F-10 with 20% fetal bovine serum (FBS) and 2.5 ng/ml fibroblast Determination of cell volume NP-40; 0.5% sodium deoxycholate; 0.1% SDS) with 1× protease Resource Center (Mizushima et al, 2004; Hara et al, 2006). ATG5fl/fl growth factor (FGF) (Peprotech) the next day and maintained for inhibitors (Roche), phosphatase inhibitors, and 5 lM trichostatin A mice were crossed with Pax7CreER mice, in which inducible Cre the subsequent 3–4 weeks to allow the proliferative expansion of The forward scatter (FSC) of SCs was measured with either a (Sigma). Cells were incubated on ice for 30 min, and the lysate was recombinase is knocked into the SC-specific Pax7 locus, to create the SC progeny. The purity of the culture, as determined by FACSAria II or FACSAria III, and the cellular volumes were deter- passed through a 30-gauge needle. After the debris was removed by Pax7CreER/+; ATG5fl/+ control or Pax7CreER/+; and ATG5fl/fl experi- immunocytochemical staining with anti-Pax7 and anti-MyoD mined from a standard curve graphed from the FSC of standard centrifugation, the lysate was pre-cleared with magnetic beads (Cell mental animals. C57BL6, ROSAeYFP/eYFP, and SIRT1fl/fl mice, in antibodies, was > 99%. microspheres from the Flow Cytometry Size Calibration Kit (Life Signaling) at 4°C for at least 3 h. The beads were then removed, and which exon 4 of sirt1 is flanked by loxP sites, were obtained Technologies). the lysate was incubated with antibody overnight at 4°C. Magnetic from The Jackson Laboratory (Srinivas et al, 2001; Li et al, 2007). Single myofiber isolation beads were added to the lysate and incubated at 4°C for 2 h and at Pax7CreER/+; SIRT1fl/+ control or Pax7CreER/+; and SIRT1fl/fl experi- Immunofluorescence room temperature for 15 min. The beads were then washed mental animals were created by crossing SIRT1fl/fl mice with To isolate single myofibers (Rosenblatt et al, 1995), extensor digito- 5 × 10 min with RIPA buffer. Protein was eluted from the beads Pax7CreER mice. Recombination at the loxP sites was induced in rum longus (EDL) muscles were dissected and digested in Ham’s Isolated QSCs or ASCs, proliferating SC progeny, or single fibers with 3× SDS loading buffer, and Western analysis was performed. SCs in control or experimental animals with tamoxifen treatment F-10 media with type II collagenase at 500 units/ml for 75 min at with associated QSCs or ASCs were fixed in 2% paraformaldehyde Antibodies were as follows: rabbit anti-acetylated lysine, rabbit (Nishijo et al, 2009). Animal husbandry, surgical procedures, and 37°C. The fibers were then dissociated and washed in Ham’s F-10 (PFA) for 5 min at room temperature and washed in PBS with anti-ATG7, and rabbit anti-ATG5 (Cell Signaling). drug administration were performed according to the guidelines with 10% HS five times and then plated in Ham’s F-10 with 10% HS 0.1% Triton X-100. They were blocked with PBS with 0.1% Triton established by the Veterinary Medical Unit of the Veterans Affairs and 0.5% chick embryo extract (U.S. Biological). Half the media X-100 and 20% donkey serum for 1 h at room temperature and Statistical analysis Health Care System in Palo Alto. was exchanged with Ham’s F-10 media with 20% FBS on each then incubated with primary antibodies overnight at 4°C. The cells subsequent day in culture. or single fibers were then washed and incubated in Alexa- Quantitative analyses were performed on experiments done at least Mouse procedures conjugated species-specific secondary antibodies and 40,6-diami- in triplicate and expressed as means standard error of mean Æ 5-ethynyl-20-deoxyuridine (EdU) incorporation assay dino-2-phenylindole (DAPI) at room temperature for 1 h. The (SEM). Statistical significance was determined with two-tailed Mice at 8 weeks of age were injected intraperitoneally six times specimens were washed and mounted in mounting media (Electron Student’s t-tests. over the course of 3 weeks with 100 ll tamoxifen (Sigma- EdU was added to cell or fiber cultures at a final concentration of Microscopy Sciences). For analysis of GFP punctae in fiber-associ- Aldrich) that was resuspended at 50 mg/ml in corn oil and etha- 10 lM. Detection was performed with the Click-iT EdU Imaging Kit ated SCs from LC3-GFP mice, fibers were fixed in 2% PFA for Supplementary information for this article is available online:

nol. Mice were first anesthetized with isoflurane/O2 mixture (Life Technologies) according to manufacturer’s protocol. 5 min at room temperature, washed in PBS with 5% horse serum, http://emboj.embopress.org

12 The EMBO Journal ª 2014 The Authors ª 2014 The Authors The EMBO Journal 13 The EMBO Journal Autophagy in satellite cell activation Ann H Tang & Thomas A Rando Ann H Tang & Thomas A Rando Autophagy in satellite cell activation The EMBO Journal

the deacetylation of those proteins. They also showed early perina- prior to injury. To study muscle regeneration in the tibialis ante- Quantitative polymerase chain reaction (qPCR) and permeabilized in 50 lg/ml digitonin in PBS and 5% HS for / tal lethality in sirt1� � animals that could be delayed by injections rior (TA) muscle, 50 ll of 1.2% BaCl2 (w/v in ddH2O) was 5 min at 37°C followed by quenching in 50 mM NH4Cl in PBS and of an exogenous metabolite. In our studies, we show that sirt1 dele- injected into one site in the TA. To generate ASCs, the same The RNeasy Plus Micro Kit (Qiagen) was used for RNA extrac- 5% HS for 5 min at 37°C and washes in PBS and 5% HS. Incuba- tion phenocopies the inhibition of autophagy with respect to a delay volume was injected into multiple areas in each lower limb. All tion from SCs. Reverse transcription was performed with the tions in primary antibodies were performed in 1× permeabilization in SC activation and a decrease in autophagic flux. In addition, we procedures were approved by the Institutional Animal Care and High Capacity cDNA Reverse Transcription Kit (Life Technolo- buffer (eBioscience) and 10% donkey serum overnight at 4°C. The demonstrate a direct interaction between SIRT1 and ATG7 in SC Use Committee. gies), and qPCR was performed with the LightCycler 480 Probe fibers were washed and incubated with secondary antibodies and / progeny and hyperacetylated ATG7 levels in sirt1� � SC progeny Master kit (Roche) and TaqMan probes (Life Technologies) in DAPI in 1× permeabilization buffer and 10% donkey serum at relative to sirt1+/+ SC progeny. We therefore corroborate the find- Satellite cell isolation the LightCycler480 II System (Roche). Relative quantification of room temperature for 1 h. Fibers were washed and EdU Click-IT ings of Lee et al that SIRT1 can regulate components of the autopha- transcripts was calculated according to the DDCT method (Suda assay was performed, if required. Fibers were washed and gic process (Lee et al, 2008). Unlike the previous studies which Injured and uninjured hindlimb muscles were dissected and et al, 2011). mounted in mounting media. For analysis of GFP punctae or LC3B used nutrient deprivation and induced neurotoxicity as stressors digested into mononuclear cells as previously described (Cheung immunoreactivity in isolated cells, the same procedure was (Lee et al, 2008; Hariharan et al, 2010; Jeong et al, 2013), we show et al, 2012; Liu et al, 2013). Briefly, hindlimb muscles were Determination of autophagic flux performed as for fibers, except that washes were done with PBS. that the induction of autophagy can be mediated, at least in part, by dissected and cut into small pieces with scissors. They were then For immunostaining of muscle sections, TA muscles were fixed in SIRT1 under conditions of normal nutrient availability. digested in Ham’s F10 medium with 10% horse serum (i.e., wash Autophagic flux was determined by the relative accumulation of 0.5% PFA for 5 h at room temperature, dehydrated in 20% sucrose Our studies demonstrating that autophagy contributes to the media) with collagenase II at 500 U/ml at 37°C for 1.5 h. The LC3B-II by Western blot analysis or by the assessment of GFP punc- in PBS, and frozen in Optimal Cutting Temperature mounting metabolic adaptation that SCs undergo during activation add to the mixture was then washed and digested for an additional 30 min at tae in SCs from LC3-GFP mice upon inhibition of autophagy by media (Sakura Finetek). Cryosections were collected, fixed in 2% understanding of the importance of metabolic flexibility in stem cell 37°C in wash media with 100 U/ml collagenase II and 2 U/ml culturing with 20 lM CQ for 2 h (Mizushima et al, 2004, 2010; PFA for 5 min at room temperature, and washed in PBS with 0.3% fate determination. Stem cells adapt their energy usage and produc- dispase. Finally, the mixture was triturated with a 20-gauge needle Klionsky et al, 2008). To analyze autophagic flux by quantifying Triton X-100. The samples were blocked in PBS with 0.3% Triton tion via the manipulation of different metabolic pathways to support and washed. Myogenic cells were then isolated by fluorescence- punctae, myofibers or SCs from LC3-GFP mice were fixed as X-100 and 20% donkey serum for at least 1 h at room temperature the changing bioenergetic needs during quiescence, proliferation, activated cell sorting (FACS) by negative selection with anti- described below, and the punctae were counted. and then washed. Mouse primary antibodies were conjugated to differentiation, or self-renewal (Folmes et al, 2012). The studies CD31-FITC, anti-CD45-FITC, and anti-Sca-1-Pacific Blue antibodies secondary antibodies with the Zenon kit (Life Technologies) and reported here suggest that the activation out of quiescence is an (Biolegend). An anti-VCAM-biotin primary antibody and a MitoTracker staining incubated along with primary antibodies from other species with example of a stem cell function that requires rapid and dramatic streptavidin-PE-cy7 secondary antibody (Biolegend) were used for sections overnight at 4°C. Sections were washed, incubated in changes in the metabolic activity and that the induction of autopha- positive selection. The purity of the sorted cells, as determined by Hindlimb muscles were dissected and digested into mononuclear secondary antibodies for 1 h at room temperature, washed again, gic activity may be a critical component of those metabolic shifts. immunocytochemical staining with anti-Pax7 and anti-MyoD cells and stained with antibodies for satellite cell isolation by FACS and mounted in mounting media. Primary antibodies were as Autophagy may therefore be an additional general mechanism antibodies, was ~98%. as described above. They were then incubated with MitoTracker follows: mouse anti-Pax7 (DSHB), mouse anti-MyoD (Dako), rabbit which confers on stem cells another degree of metabolic flexibility deep red FM (Life Technologies) at 37°C for 30 min. The cells were anti-LC3B (Cell Signaling), rabbit anti-GFP (Life Technologies), to respond to fluctuating bioenergetic demands. SC progeny isolation washed twice prior to analysis on the FACSAria II or III. chick anti-GFP (Aves), and rat anti-Laminin (Sigma-Aldrich). Secondary antibodies used are conjugated to Alexa fluorescent dyes Hindlimb muscles were subjected to needle injury and harvested ATP quantification (Life Technologies). Materials and Methods 2.5 days later. The muscles were digested and sorted as described for SC isolation above. FACS-sorted SC progeny were then plated in ATP content of SCs was determined with the ATP Bioluminescence Protein immunoprecipitation Mouse strains Ham’s F-10 nutrient mixture with 10% horse serum (HS) onto Assay Kit CLS II (Roche) according to manufacturer’s recommenda- 35-mm tissue culture dishes coated with extracellular matrix tions. SC progeny were harvested, washed in PBS, and lysed in RIPA lysis LC3-GFP transgenic mice and ATG5fl/fl mice, in which exon 3 of (Sigma) at a 1:500 dilution. The media was exchanged with Ham’s buffer (50 mM Tris–HCl, pH 7.5; 150 mM NaCl; 5 mM EDTA; 1% atg5 is flanked by loxP sites, were obtained from the Riken Bio- F-10 with 20% fetal bovine serum (FBS) and 2.5 ng/ml fibroblast Determination of cell volume NP-40; 0.5% sodium deoxycholate; 0.1% SDS) with 1× protease Resource Center (Mizushima et al, 2004; Hara et al, 2006). ATG5fl/fl growth factor (FGF) (Peprotech) the next day and maintained for inhibitors (Roche), phosphatase inhibitors, and 5 lM trichostatin A mice were crossed with Pax7CreER mice, in which inducible Cre the subsequent 3–4 weeks to allow the proliferative expansion of The forward scatter (FSC) of SCs was measured with either a (Sigma). Cells were incubated on ice for 30 min, and the lysate was recombinase is knocked into the SC-specific Pax7 locus, to create the SC progeny. The purity of the culture, as determined by FACSAria II or FACSAria III, and the cellular volumes were deter- passed through a 30-gauge needle. After the debris was removed by Pax7CreER/+; ATG5fl/+ control or Pax7CreER/+; and ATG5fl/fl experi- immunocytochemical staining with anti-Pax7 and anti-MyoD mined from a standard curve graphed from the FSC of standard centrifugation, the lysate was pre-cleared with magnetic beads (Cell mental animals. C57BL6, ROSAeYFP/eYFP, and SIRT1fl/fl mice, in antibodies, was > 99%. microspheres from the Flow Cytometry Size Calibration Kit (Life Signaling) at 4°C for at least 3 h. The beads were then removed, and which exon 4 of sirt1 is flanked by loxP sites, were obtained Technologies). the lysate was incubated with antibody overnight at 4°C. Magnetic from The Jackson Laboratory (Srinivas et al, 2001; Li et al, 2007). Single myofiber isolation beads were added to the lysate and incubated at 4°C for 2 h and at Pax7CreER/+; SIRT1fl/+ control or Pax7CreER/+; and SIRT1fl/fl experi- Immunofluorescence room temperature for 15 min. The beads were then washed mental animals were created by crossing SIRT1fl/fl mice with To isolate single myofibers (Rosenblatt et al, 1995), extensor digito- 5 × 10 min with RIPA buffer. Protein was eluted from the beads Pax7CreER mice. Recombination at the loxP sites was induced in rum longus (EDL) muscles were dissected and digested in Ham’s Isolated QSCs or ASCs, proliferating SC progeny, or single fibers with 3× SDS loading buffer, and Western analysis was performed. SCs in control or experimental animals with tamoxifen treatment F-10 media with type II collagenase at 500 units/ml for 75 min at with associated QSCs or ASCs were fixed in 2% paraformaldehyde Antibodies were as follows: rabbit anti-acetylated lysine, rabbit (Nishijo et al, 2009). Animal husbandry, surgical procedures, and 37°C. The fibers were then dissociated and washed in Ham’s F-10 (PFA) for 5 min at room temperature and washed in PBS with anti-ATG7, and rabbit anti-ATG5 (Cell Signaling). drug administration were performed according to the guidelines with 10% HS five times and then plated in Ham’s F-10 with 10% HS 0.1% Triton X-100. They were blocked with PBS with 0.1% Triton established by the Veterinary Medical Unit of the Veterans Affairs and 0.5% chick embryo extract (U.S. Biological). Half the media X-100 and 20% donkey serum for 1 h at room temperature and Statistical analysis Health Care System in Palo Alto. was exchanged with Ham’s F-10 media with 20% FBS on each then incubated with primary antibodies overnight at 4°C. The cells subsequent day in culture. or single fibers were then washed and incubated in Alexa- Quantitative analyses were performed on experiments done at least Mouse procedures conjugated species-specific secondary antibodies and 40,6-diami- in triplicate and expressed as means standard error of mean Æ 5-ethynyl-20-deoxyuridine (EdU) incorporation assay dino-2-phenylindole (DAPI) at room temperature for 1 h. The (SEM). Statistical significance was determined with two-tailed Mice at 8 weeks of age were injected intraperitoneally six times specimens were washed and mounted in mounting media (Electron Student’s t-tests. over the course of 3 weeks with 100 ll tamoxifen (Sigma- EdU was added to cell or fiber cultures at a final concentration of Microscopy Sciences). For analysis of GFP punctae in fiber-associ- Aldrich) that was resuspended at 50 mg/ml in corn oil and etha- 10 lM. Detection was performed with the Click-iT EdU Imaging Kit ated SCs from LC3-GFP mice, fibers were fixed in 2% PFA for Supplementary information for this article is available online: nol. Mice were first anesthetized with isoflurane/O2 mixture (Life Technologies) according to manufacturer’s protocol. 5 min at room temperature, washed in PBS with 5% horse serum, http://emboj.embopress.org

Acknowledgements Folmes CD, Dzeja PP, Nelson TJ, Terzic A (2012) Metabolic plasticity in Jeong JK, Moon MH, Lee YJ, Seol JW, Park SY (2013) Autophagy induced by the upregulation of CD34 and loss of pluripotency: a potential strategy for We thank the members of the Rando Lab for critical discussions. We thank Dr. stem cell homeostasis and differentiation. Cell Stem Cell 11: 596 – 606 class III histone deacetylase Sirt1 prevents prion peptide neurotoxicity. reducing oxidative risk in stem cells. Cell Cycle 9: 2008 – 2017 Charles Keller for the Pax7CreER mice and Dr. Noboru Mizushima and the Riken Gan B, Sahin E, Jiang S, Sanchez-Aguilera A, Scott KL, Chin L, Williams DA, Neurobiol Aging 34: 146 – 156 Martinez-Lopez N, Athonvarangkul D, Sahu S, Coletto L, Zong H, Bastie CC, Bioresource Center for permission to use the LC3-GFP and ATG5fl/fl mice. We Kwiatkowski DJ, DePinho RA (2008) mTORC1-dependent and -independent Jones RG, Plas DR, Kubek S, Buzzai M, Mu J, Xu Y, Birnbaum MJ, Thompson Pessin JE, Schwartz GJ, Singh R (2013) Autophagy in Myf5+ progenitors are grateful to Dr. Albert La Spada for providing us with the LC3-GFP and regulation of stem cell renewal, differentiation, and mobilization. Proc CB (2005) AMP-activated protein kinase induces a p53-dependent regulates energy and glucose homeostasis through control of brown fat ATG5fl/fl mice. This work was supported by the Glenn Foundation for Medical Natl Acad Sci U S A 105: 19384 – 19389 metabolic checkpoint. Mol Cell 18: 283 – 293 and skeletal muscle development. EMBO Rep 14: 795 – 803 Research and by grants from the NIH (R01 AG23806, Transformative R01 Ganley IG, du Lam H, Wang J, Ding X, Chen S, Jiang X (2009) Jung CH, Jun CB, Ro SH, Kim YM, Otto NM, Cao J, Kundu M, Kim DH (2009) Miharada K, Karlsson G, Rehn M, Rorby E, Siva K, Cammenga J, Karlsson S AG047820, and P01 AG036695) to TAR. ULK1.ATG13.FIP200 complex mediates mTOR signaling and is essential for ULK-Atg13-FIP200 complexes mediate mTOR signaling to the autophagy (2011) Cripto regulates hematopoietic stem cells as a autophagy. J Biol Chem 284: 12297 – 12305 machinery. Mol Biol Cell 20: 1992 – 2003 hypoxic-niche-related factor through cell surface receptor GRP78. Cell Author contributions Ghosh HS, McBurney M, Robbins PD (2010) SIRT1 negatively regulates the Kapahi P, Chen D, Rogers AN, Katewa SD, Li PW, Thomas EL, Kockel L (2010) Stem Cell 9: 330 – 344 AHT contributed to the conception and design of the project, acquisition, mammalian target of rapamycin. PLoS ONE 5: 0009199 With TOR, less is more: a key role for the conserved nutrient-sensing TOR Miyamoto K, Araki KY, Naka K, Arai F, Takubo K, Yamazaki S, Matsuoka S, analysis, and interpretation of the data, and the writing of the manuscript. Groszer M, Erickson R, Scripture-Adams DD, Dougherty JD, Le Belle J, Zack JA, pathway in aging. Cell Metab 11: 453 – 465 Miyamoto T, Ito K, Ohmura M, Chen C, Hosokawa K, Nakauchi H, TAR contributed to conception and design of the project, the interpretation of Geschwind DH, Liu X, Kornblum HI, Wu H (2006) PTEN negatively Kim J, Kundu M, Viollet B, Guan KL (2011) AMPK and mTOR regulate Nakayama K, Nakayama KI, Harada M, Motoyama N, Suda T, Hirao A the data, and the writing of the manuscript. regulates neural stem cell self-renewal by modulating G0-G1 cell cycle autophagy through direct phosphorylation of Ulk1. Nat Cell Biol 13: (2007) Foxo3a is essential for maintenance of the hematopoietic stem cell entry. Proc Natl Acad Sci U S A 103: 111 – 116 132 – 141 pool. Cell Stem Cell 1: 101 – 112 Conflict of interest Guan JL, Simon AK, Prescott M, Menendez JA, Liu F, Wang F, Wang C, Klionsky DJ, Abeliovich H, Agostinis P, Agrawal DK, Aliev G, Askew DS, Baba Mizushima N, Yamamoto A, Matsui M, Yoshimori T, Ohsumi Y (2004) In vivo The authors declare that they have no conflict of interest. Wolvetang E, Vazquez-Martin A, Zhang J (2013) Autophagy in stem cells. M, Baehrecke EH, Bahr BA, Ballabio A, Bamber BA, Bassham DC, analysis of autophagy in response to nutrient starvation using transgenic Autophagy 9: 830 – 849 Bergamini E, Bi X, Biard-Piechaczyk M, Blum JS, Bredesen DE, Brodsky JL, mice expressing a fluorescent autophagosome marker. Mol Biol Cell 15: Hara T, Nakamura K, Matsui M, Yamamoto A, Nakahara Y, Suzuki-Migishima Brumell JH, Brunk UT et al (2008) Guidelines for the use and 1101 – 1111 References R, Yokoyama M, Mishima K, Saito I, Okano H, Mizushima N (2006) interpretation of assays for monitoring autophagy in higher eukaryotes. Mizushima N, Yoshimori T, Levine B (2010) Methods in mammalian Suppression of basal autophagy in neural cells causes neurodegenerative Autophagy 4: 151 – 175 autophagy research. Cell 140: 313 – 326 Boutet SC, Cheung TH, Quach NL, Liu L, Prescott SL, Edalati A, Iori K, Rando disease in mice. Nature 441: 885 – 889 Komatsu M, Waguri S, Ueno T, Iwata J, Murata S, Tanida I, Ezaki J, Mizushima N, Komatsu M (2011) Autophagy: renovation of cells and tissues. TA (2012) Alternative polyadenylation mediates microRNA regulation of Hariharan N, Maejima Y, Nakae J, Paik J, Depinho RA, Sadoshima J (2010) Mizushima N, Ohsumi Y, Uchiyama Y, Kominami E, Tanaka K, Chiba T Cell 147: 728 – 741 muscle stem cell function. Cell Stem Cell 10: 327 – 336 Deacetylation of FoxO by Sirt1 plays an essential role in mediating (2005) Impairment of starvation-induced and constitutive autophagy in Morgan DO (1997) Cyclin-dependent kinases: engines, clocks, and Brack AS, Conboy MJ, Roy S, Lee M, Kuo CJ, Keller C, Rando TA (2007) starvation-induced autophagy in cardiac myocytes. Circ Res 107: Atg7-deficient mice. J Cell Biol 169: 425 – 434 microprocessors. Annu Rev Cell Dev Biol 13: 261 – 291 Increased Wnt signaling during aging alters muscle stem cell fate and 1470 – 1482 Kroemer G, Marino G, Levine B (2010) Autophagy and the integrated stress Mortensen M, Ferguson DJ, Edelmann M, Kessler B, Morten KJ, Komatsu M, increases fibrosis. Science 317: 807 – 810 He C, Klionsky DJ (2009) Regulation mechanisms and signaling pathways of response. Mol Cell 40: 280 – 293 Simon AK (2010) Loss of autophagy in erythroid cells leads to defective Brack AS, Rando TA (2012) Tissue-specific stem cells: lessons from the skeletal autophagy. Annu Rev Genet 43: 67 – 93 Kuma A, Hatano M, Matsui M, Yamamoto A, Nakaya H, Yoshimori T, Ohsumi removal of mitochondria and severe anemia in vivo. Proc Natl Acad Sci U muscle satellite cell. Cell Stem Cell 10: 504 – 514 He C, Bassik MC, Moresi V, Sun K, Wei Y, Zou Z, An Z, Loh J, Fisher J, Sun Q, Y, Tokuhisa T, Mizushima N (2004) The role of autophagy during the early SA107: 832 – 837 Carracedo A, Weiss D, Leliaert AK, Bhasin M, de Boer VC, Laurent G, Adams Korsmeyer S, Packer M, May HI, Hill JA, Virgin HW, Gilpin C, Xiao G, neonatal starvation period. Nature 432: 1032 – 1036 Mortensen M, Soilleux EJ, Djordjevic G, Tripp R, Lutteropp M, Sadighi-Akha E, AC, Sundvall M, Song SJ, Ito K, Finley LS, Egia A, Libermann T, Bassel-Duby R, Scherer PE, Levine B (2012) Exercise-induced Lan F, Cacicedo JM, Ruderman N, Ido Y (2008) SIRT1 modulation of the Stranks AJ, Glanville J, Knight S, Jacobsen SE, Kranc KR, Simon AK (2011) Gerhart-Hines Z, Puigserver P, Haigis MC, Maratos-Flier E, Richardson AL, BCL2-regulated autophagy is required for muscle glucose homeostasis. acetylation status, cytosolic localization, and activity of LKB1. Possible The autophagy protein Atg7 is essential for hematopoietic stem cell Schafer ZT, Pandolfi PP (2012) A metabolic prosurvival role for PML in Nature 481: 511 – 515 role in AMP-activated protein kinase activation. J Biol Chem 283: maintenance. J Exp Med 208: 455 – 467 breast cancer. J Clin Invest 122: 3088 – 3100 Herranz D, Serrano M (2010) SIRT1: recent lessons from mouse models. Nat 27628 – 27635 Murakami M, Ichisaka T, Maeda M, Oshiro N, Hara K, Edenhofer F, Kiyama H, Carracedo A, Cantley LC, Pandolfi PP (2013) Cancer metabolism: fatty acid Rev Cancer 10: 819 – 823 Lee IH, Cao L, Mostoslavsky R, Lombard DB, Liu J, Bruns NE, Tsokos M, Alt Yonezawa K, Yamanaka S (2004) mTOR is essential for growth and oxidation in the limelight. Nat Rev Cancer 13: 227 – 232 Holley RW, Kiernan JA (1974) Control of the initiation of DNA synthesis in 3T3 FW, Finkel T (2008) A role for the NAD-dependent deacetylase Sirt1 in the proliferation in early mouse embryos and embryonic stem cells. Mol Cell Chalkiadaki A, Guarente L (2012) Sirtuins mediate mammalian cells: low-molecular weight nutrients. Proc Natl Acad Sci U S A 71: regulation of autophagy. Proc Natl Acad Sci U S A 105: 3374 – 3379 Biol 24: 6710 – 6718 metabolic responses to nutrient availability. Nat Rev Endocrinol 8: 2942 – 2945 Li H, Rajendran GK, Liu N, Ware C, Rubin BP, Gu Y (2007) SirT1 modulates Napper AD, Hixon J, McDonagh T, Keavey K, Pons JF, Barker J, Yau WT, Amouzegh 287 – 296 Hong S, Zhao B, Lombard DB, Fingar DC, Inoki K (2014) Cross-talk between the estrogen-insulin-like growth factor-1 signaling for postnatal P, Flegg A, Hamelin E, Thomas RJ, Kates M, Jones S, Navia MA, Saunders JO, Chen C, Liu Y, Liu R, Ikenoue T, Guan KL, Zheng P (2008) TSC-mTOR sirtuin and mammalian target of rapamycin complex 1 (mTORC1) development of mammary gland in mice. Breast Cancer Res 9:R1 DiStefano PS, Curtis R (2005) Discovery of indoles as potent and selective maintains quiescence and function of hematopoietic stem cells by signaling in the regulation of S6 kinase 1 (S6K1) phosphorylation. J Biol Liu F, Lee JY, Wei H, Tanabe O, Engel JD, Morrison SJ, Guan JL (2010) FIP200 inhibitors of the deacetylase SIRT1. J Med Chem 48: 8045 – 8054 repressing mitochondrial biogenesis and reactive oxygen species. J Exp Chem 289: 13132 – 13141 is required for the cell-autonomous maintenance of fetal hematopoietic Nicholls DG, Darley-Usmar VM, Wu M, Jensen PB, Rogers GW, Ferrick DA (2010) Med 205: 2397 – 2408 Hosokawa N, Hara T, Kaizuka T, Kishi C, Takamura A, Miura Y, Iemura S, stem cells. Blood 116: 4806 – 4814 Bioenergetic profile experiment using C2C12 myoblast cells. J Vis Exp 6: Cheung TH, Quach NL, Charville GW, Liu L, Park L, Edalati A, Yoo B, Hoang P, Natsume T, Takehana K, Yamada N, Guan JL, Oshiro N, Mizushima N Liu L, Cheung TH, Charville GW, Hurgo BM, Leavitt T, Shih J, Brunet A, Rando e2511 Rando TA (2012) Maintenance of muscle stem-cell quiescence by (2009) Nutrient-dependent mTORC1 association with the TA (2013) Chromatin modifications as determinants of muscle stem cell Nishijo K, Hosoyama T, Bjornson CR, Schaffer BS, Prajapati SI, Bahadur AN, microRNA-489. Nature 482: 524 – 528 ULK1-Atg13-FIP200 complex required for autophagy. Mol Biol Cell 20: quiescence and chronological aging. Cell Rep 4: 189 – 204 Hansen MS, Blandford MC, McCleish AT, Rubin BP, Epstein JA, Rando TA, Dunlop EA, Tee AR (2013) The kinase triad, AMPK, mTORC1 and ULK1, 1981 – 1991 Lonergan T, Bavister B, Brenner C (2007) Mitochondria in stem cells. Capecchi MR, Keller C (2009) Biomarker system for studying muscle, stem maintains energy and nutrient homoeostasis. Biochem Soc Trans 41: Hsu P, Qu CK (2013) Metabolic plasticity and hematopoietic stem cell Mitochondrion 7: 289 – 296 cells, and cancer in vivo. FASEB J 23: 2681 – 2690 939 – 943 biology. Curr Opin Hematol 20: 289 – 294 Lum JJ, Bauer DE, Kong M, Harris MH, Li C, Lindsten T, Thompson CB (2005) Onodera J, Ohsumi Y (2005) Autophagy is required for maintenance of amino Easley CA, Ben-Yehudah A, Redinger CJ, Oliver SL, Varum ST, Eisinger VM, Hubbard VM, Valdor R, Patel B, Singh R, Cuervo AM, Macian F (2010) Growth factor regulation of autophagy and cell survival in the absence of acid levels and protein synthesis under nitrogen starvation. J Biol Chem Carlisle DL, Donovan PJ, Schatten GP (2010) mTOR-mediated activation of Macroautophagy regulates energy metabolism during effector T cell apoptosis. Cell 120: 237 – 248 280: 31582 – 31586 p70 S6K induces differentiation of pluripotent human embryonic stem activation. J Immunol 185: 7349 – 7357 Lundberg AS, Weinberg RA (1998) Functional inactivation of the Ou X, Lee MR, Huang X, Messina-Graham S, Broxmeyer HE (2014) SIRT1 cells. Cell Reprogram 12: 263 – 273 Inoki K, Zhu T, Guan KL (2003) TSC2 mediates cellular energy response to retinoblastoma protein requires sequential modification by at least two positively regulates autophagy and mitochondria function in embryonic Egan DF, Shackelford DB, Mihaylova MM, Gelino S, Kohnz RA, Mair W, control cell growth and survival. Cell 115: 577 – 590 distinct cyclin-cdk complexes. Mol Cell Biol 18: 753 – 761 stem cells under oxidative stress. Stem Cells 32: 1183 – 1194 Vasquez DS, Joshi A, Gwinn DM, Taylor R, Asara JM, Fitzpatrick J, Dillin A, Ito K, Carracedo A, Weiss D, Arai F, Ala U, Avigan DE, Schafer ZT, Evans RM, Lunt SY, Vander Heiden MG (2011) Aerobic glycolysis: meeting the metabolic Paik JH, Ding Z, Narurkar R, Ramkissoon S, Muller F, Kamoun WS, Chae SS, Viollet B, Kundu M, Hansen M, Shaw RJ (2011) Phosphorylation of ULK1 Suda T, Lee CH, Pandolfi PP (2012) A PML-PPAR-delta pathway for fatty requirements of cell proliferation. Annu Rev Cell Dev Biol 27: 441 – 464 Zheng H, Ying H, Mahoney J, Hiller D, Jiang S, Protopopov A, Wong WH, Chin (hATG1) by AMP-activated protein kinase connects energy sensing to acid oxidation regulates hematopoietic stem cell maintenance. Nat Med Mantel C, Messina-Graham S, Broxmeyer HE (2010) Upregulation of nascent L, Ligon KL, DePinho RA (2009) FoxOs cooperatively regulate diverse mitophagy. Science 331: 456 – 461 18: 1350 – 1358 mitochondrial biogenesis in mouse hematopoietic stem cells parallels pathways governing neural stem cell homeostasis. Cell Stem Cell 5: 540 – 553

14 The EMBO Journal ª 2014 The Authors ª 2014 The Authors The EMBO Journal 15 The EMBO Journal Autophagy in satellite cell activation Ann H Tang & Thomas A Rando Ann H Tang & Thomas A Rando Autophagy in satellite cell activation The EMBO Journal

14 The EMBO Journal ª 2014 The Authors ª 2014 The Authors The EMBO Journal 15 The EMBO Journal Autophagy in satellite cell activation Ann H Tang & Thomas A Rando

Parzych KR, Klionsky DJ (2014) An overview of autophagy: morphology, Simsek T, Kocabas F, Zheng J, Deberardinis RJ, Mahmoud AI, Olson EN, mechanism, and regulation. Antioxid Redox Signal 20: 460 – 473 Schneider JW, Zhang CC, Sadek HA (2010) The distinct metabolic profile of Pelletier M, Billingham LK, Ramaswamy M, Siegel RM (2014) Extracellular hematopoietic stem cells reflects their location in a hypoxic niche. Cell flux analysis to monitor glycolytic rates and mitochondrial oxygen Stem Cell 7: 380 – 390 consumption. Methods Enzymol 542: 125 – 149 Singh R, Cuervo AM (2011) Autophagy in the cellular energetic balance. Cell Phadwal K, Watson AS, Simon AK (2013) Tightrope act: autophagy in stem Metab 13: 495 – 504 cell renewal, differentiation, proliferation, and aging. Cell Mol Life Sci 70: Srinivas S, Watanabe T, Lin CS, William CM, Tanabe Y, Jessell TM, Costantini 89 – 103 F(2001) Cre reporter strains produced by targeted insertion of EYFP and Piccoli C, Ria R, Scrima R, Cela O, D’Aprile A, Boffoli D, Falzetti F, Tabilio A, ECFP into the ROSA26 locus. BMC Dev Biol 1: 27 Capitanio N (2005) Characterization of mitochondrial and Suda T, Takubo K, Semenza GL (2011) Metabolic regulation of hematopoietic extra-mitochondrial oxygen consuming reactions in human hematopoietic stem cells in the hypoxic niche. Cell Stem Cell 9: 298 – 310 stem cells. Novel evidence of the occurrence of NAD(P)H oxidase activity. J Takeda-Watanabe A, Kitada M, Kanasaki K, Koya D (2012) SIRT1 inactivation Biol Chem 280: 26467 – 26476 induces inflammation through the dysregulation of autophagy in human Rabinowitz JD, White E (2010) Autophagy and metabolism. Science 330: THP-1 cells. Biochem Biophys Res Commun 427: 191 – 196 1344 – 1348 Tothova Z, Kollipara R, Huntly BJ, Lee BH, Castrillon DH, Cullen DE, McDowell Rafalski VA, Brunet A (2011) Energy metabolism in adult neural stem cell EP, Lazo-Kallanian S, Williams IR, Sears C, Armstrong SA, Passegue E, fate. Prog Neurobiol 93: 182 – 203 DePinho RA, Gilliland DG (2007) FoxOs are critical mediators of Ravikumar B, Sarkar S, Davies JE, Futter M, Garcia-Arencibia M, hematopoietic stem cell resistance to physiologic oxidative stress. Cell 128: Green-Thompson ZW, Jimenez-Sanchez M, Korolchuk VI, Lichtenberg M, 325 – 339 Luo S, Massey DC, Menzies FM, Moreau K, Narayanan U, Renna M, Siddiqi Velez J, Hail N Jr, Konopleva M, Zeng Z, Kojima K, Samudio I, Andreeff M FH, Underwood BR, Winslow AR, Rubinsztein DC (2010) Regulation of (2013) Mitochondrial uncoupling and the reprograming of intermediary mammalian autophagy in physiology and pathophysiology. Physiol Rev 90: metabolism in leukemia cells. Front Oncol 3: 67 1383 – 1435 Wang YX, Rudnicki MA (2012) Satellite cells, the engines of muscle repair. Nat Renault VM, Rafalski VA, Morgan AA, Salih DA, Brett JO, Webb AE, Villeda Rev Mol Cell Biol 13: 127 – 133 SA, Thekkat PU, Guillerey C, Denko NC, Palmer TD, Butte AJ, Brunet A Wen X, Wu J, Wang F, Liu B, Huang C, Wei Y (2013) Deconvoluting the role of (2009) FoxO3 regulates neural stem cell homeostasis. Cell Stem Cell 5: reactive oxygen species and autophagy in human diseases. Free Radic Biol 527 – 539 Med 65: 402 – 410 Rodgers JT, King KY, Brett JO, Cromie MJ, Charville GW, Maguire KK, Brunson Yang Z, Klionsky DJ (2010) Eaten alive: a history of macroautophagy. Nat Cell C, Mastey N, Liu L, Tsai CR, Goodell MA, Rando TA (2014) mTORC1 Biol 12: 814 – 822 controls the adaptive transition of quiescent stem cells from G0 to G Yano T, Kurata S (2011) Intracellular recognition of pathogens and autophagy (Alert). Nature 510: 393 – 396 as an innate immune host defence. J Biochem 150: 143 – 149 Rosenblatt JD, Lunt AI, Parry DJ, Partridge TA (1995) Culturing satellite cells Yilmaz OH, Valdez R, Theisen BK, Guo W, Ferguson DO, Wu H, Morrison SJ from living single muscle fiber explants. In Vitro Cell Dev Biol Anim 31: (2006) Pten dependence distinguishes haematopoietic stem cells from 773 – 779 leukaemia-initiating cells. Nature 441: 475 – 482 Sampath P, Pritchard DK, Pabon L, Reinecke H, Schwartz SM, Morris DR, Yin H, Price F, Rudnicki MA (2013) Satellite cells and the muscle stem cell Murry CE (2008) A hierarchical network controls protein translation niche. Physiol Rev 93: 23 – 67 during murine embryonic stem cell self-renewal and differentiation. Cell Zaugg K, Yao Y, Reilly PT, Kannan K, Kiarash R, Mason J, Huang P, Sawyer SK, Stem Cell 2: 448 – 460 Fuerth B, Faubert B, Kalliomaki T, Elia A, Luo X, Nadeem V, Bungard D, Schieke SM, McCoy JP Jr, Finkel T (2008) Coordination of mitochondrial Yalavarthi S, Growney JD, Wakeham A, Moolani Y, Silvester J et al (2011) bioenergetics with G1 phase cell cycle progression. Cell Cycle 7: Carnitine palmitoyltransferase 1C promotes cell survival and tumor 1782 – 1787 growth under conditions of metabolic stress. Genes Dev 25: 1041 – 1051 Sengupta A, Molkentin JD, Yutzey KE (2009) FoxO transcription Zhang J, Grindley JC, Yin T, Jayasinghe S, He XC, Ross JT, Haug JS, Rupp D, factors promote autophagy in cardiomyocytes. J Biol Chem 284: Porter-Westpfahl KS, Wiedemann LM, Wu H, Li L (2006) PTEN maintains 28319 – 28331 haematopoietic stem cells and acts in lineage choice and leukaemia Shanware NP, Bray K, Abraham RT (2013) The PI3K, metabolic, and prevention. Nature 441: 518 – 522 autophagy networks: interactive partners in cellular health and disease. Zhou J, Su P, Wang L, Chen J, Zimmermann M, Genbacev O, Afonja O, Horne Annu Rev Pharmacol Toxicol 53: 89 – 106 MC, Tanaka T, Duan E, Fisher SJ, Liao J, Wang F (2009) mTOR supports Shyh-Chang N, Daley GQ, Cantley LC (2013) Stem cell metabolism in tissue long-term self-renewal and suppresses mesoderm and endoderm activities development and aging. Development 140: 2535 – 2547 of human embryonic stem cells. Proc Natl Acad Sci U S A 106: 7840 – 7845

16 The EMBO Journal ª 2014 The Authors Review

Getting ready for building: signaling and autophagosome biogenesis

Adi Abada & Zvulun Elazar*

Abstract autophagosome biogenesis [5,6]. According to the current view, autophagy is a progressive process initiated by the elongation of a Autophagy is the main cellular catabolic process responsible for membrane to form a cup-shaped phagophore into which autophagic degrading organelles and large protein aggregates. It is initiated by cargo is sequestered, a process that initially seemed random but the formation of a unique membrane structure, the phagophore, now is mostly regarded as selective. The autophagic membrane is which engulfs part of the cytoplasm and forms a double-membrane further expanded to its final size and, once sealed, results in a vesicle termed the autophagosome. Fusion of the outer autophag- mature double-membrane autophagosome, the outer membrane of osomal membrane with the lysosome and degradation of the inner which subsequently fuses with the lysosome to create an autolyso- membrane contents complete the process. The extent of autophagy some. The autophagosomal content is then degraded and recycled. must be tightly regulated to avoid destruction of proteins and This review covers recent progress made in understanding the organelles essential for cell survival. Autophagic activity is thus early stages of autophagy. The external signals and environmental regulated by external and internal cues, which initiate the forma- conditions that regulate autophagy—such as growth factors and tion of well-defined autophagy-related protein complexes that nutrient deprivation—and the signal transduction pathways mediate autophagosome formation and selective cargo recruit- involved in such regulation will be discussed, as well as changes in ment into these organelles. Autophagosome formation and the intracellular homeostasis that trigger the autophagic process, such signaling pathways that regulate it have recently attracted as oxidative stress and internal energy levels. Finally, we will substantial attention. In this review, we analyze the different analyze new mechanistic insights into autophagosome biogenesis. signaling pathways that regulate autophagy and discuss recent progress in our understanding of autophagosome biogenesis. Regulation of autophagy Keywords Atgs; autophagosome biogenesis; autophagy; mTOR; signaling DOI 10.15252/embr.201439076 | Received 21 May 2014 | Revised 17 June Eukaryotes have developed signaling networks that control tran- 2014 | Accepted 17 June 2014 | Published online 15 July 2014 scription, translation, and protein modification to adapt to changing EMBO Reports (2014) 15: 839–852 environmental conditions. At times of shortage, cells need to save energy and nutrients by maintaining basal and essential activities. See the Glossary for abbreviations used in this article. As part of the cellular response to such conditions, autophagy—a major cellular catabolic process—is subjected to tight regulation by a network of canonical and unique signaling cascades. It is therefore Introduction important to examine not only cues originating within the cells, but also signaling initiated in response to external changes (Figs 1 and 2). Autophagy is a catabolic process by which proteins and organelles Autophagy is known to be mainly under the control of the key are delivered to the lysosome for degradation. As a self-degrading regulator of cell homeostasis, the Ser/Thr kinase TOR (target of process that is conserved from yeast to man, autophagy is well rapamycin), in yeast, or mTOR in mammals [7]. TOR is found in established as a survival mechanism that maintains cellular homeo- two distinct protein complexes, TORC1 or TORC2 [8,9]. Although stasis under normal growth conditions and enables adaptation both TOR complexes regulate cell metabolism, only TORC1 is under stress [1]. It has also been implicated in disease, develop- directly linked to the regulation of autophagy. ment, and cell differentiation [2–4]. Although beneficial under normal conditions, autophagy can be detrimental in diseases such Extracellular cues as cancer and neurodegenerative disorders. Therefore, cells have developed control mechanisms that tightly regulate their autopha- Amino acid starvation The most extensively characterized inducer gic activity. More than 30 proteins have been identified as related of autophagy is amino acid deprivation (Fig 1A). The absence of to autophagy (Atgs), most of them directly associated with specific amino acids such as leucine and glutamine strongly induces

Department of Biological Chemistry, The Weizmann Institute of Science, Rehovot, Israel *Corresponding author. Tel: +972 8 934 3682; Fax: +972 8 934 4112; E-mail: [email protected]

ª 2014 The Authors EMBO reports Vol 15 | No 8 | 2014 839 EMBO reports From signaling to autophagosome biogenesis Adi Abada & Zvulun Elazar Adi Abada & Zvulun Elazar From signaling to autophagosome biogenesis EMBO reports

Glossary A mTORC2 Akt B P ALFY autophagy-linked FYVE protein NO nitric oxide P PDK1 Ambra1 activating molecule in Beclin1-regulated autophagy NOS nitric oxide synthase T1R1/T1R3 V-ATPase AMPK 5 AMP-activated protein kinase NOX4 NADPH oxidase 4 PI3P 0 IR ATM ataxia-telangiectasia mutated Nrf2 nuclear factor erythroid 2-related factor 2 Lysosome BAR Bin1/amphiphysin/Rvs167 PDK1 phosphoinositide-dependent kinase 1 Bcl2 B-cell lymphoma 2 PERK protein kinase RNA-like endoplasmic reticulum kinase TSC1 PI3KC1 DFCP1 Double FYVE-containing protein 1 pex19 peroxisomal membrane protein SLC7A5– Raptor Rheb GTP RagA/B EGFR epidermal growth factor receptor PI3KC1/3 phosphoinositide 3-kinase complex 1/3 SLC3A5 TSC2 P mTORC1 eIF2a eukaryotic initiation factor 2a PIP2 phosphatidylinositol 4,5-bisphosphate LRS GDP RagC/D ER endoplasmic reticulum PKB protein kinase B ERGIC ER Golgi intermediate compartment PKR protein kinase R Hexokinase FIP200 200-kDa focal adhesion kinase family-interacting protein PtdIns3P/PI3P phosphatidylinositol 3-phosphate P P II FoxO forkhead-box transcription factor class O PTEN phosphatase and tensin homolog Ulk1 FYCO1 FYVE and coiled-coil domain containing 1 Rag Ras-related GTP-binding protein complex FYVE Fab1, YOTB, Vac 1, EEA1 RalA Ras-like protein FoxO GAB2 GRB2-associated binding protein 2 ROS reactive oxygen species GABARAP GABA receptor-associated SH3BP4 SH3-domain binding protein 4 GAP GTPase activating protein SNARE soluble NSF attachment protein receptor protein GARP Golgi-associated retrograde protein SNX18 sorting nexin 18 C Nucleus D GATE16 Golgi-associated ADPase enhancer of 16 kDa STAT3 signal transducer and activator of transcription 3 PI3KC3 TRAF6 GCN general control non-derepressible SUVs small unilamellar vesicles complex mTORC1 GLUD glutamate dehydrogenase TECPR1 tectonin b-propeller repeat containing 1 Beclin1 TRIF GRB2 growth factor receptor-bound protein 2 TIR toll/interleukin-1 receptor MyD88 GSK3 glycogen synthase kinase 3 TLR toll-like receptor GUVs giant unilamellar vesicles TNF tumor necrosis factor Beclin1 eIF2α Nrf2 HOPS homotypic fusion and protein sorting complex TRAF6 tumor necrosis factor receptor-associated factor 6 P GAB2 IKKb inhibitor of nuclear factor jB kinase TRAPPIII transport protein particle PI3KC3 GRB2 IL-1b interleukin 1b TRIF toll/interleukin-1 receptor homology domain- complex Amino acids IFN-c interferon c containing adaptor inducing interferon-b TLR4 IP3R IP3 receptor TSC1/2 tuberous sclerosis protein 1 and 2 P STAT3 P Phosphate JNK c-JUN NH2-terminal kinase Ulk1 UNC-51-like kinase 1 P STAT3 Glucose 6-phosphate LC3 light chain 3 UVRAG UV irradiation resistance-associated gene LKB1 liver kinase B1 VAMP7 vesicle-associated membrane K63 linked ubiquitin EGFR LPS liposaccharide Vps34 vacuolar protein sorting 34 LRS Leucyl-tRNA synthetase WASH Wiskott–Aldrich syndrome protein WASP and SCAR MyD88 myeloid differentiation factor 88 homolog NBR1 neighbor of BRCA1 gene WIPI1/2 WD repeat proteins interacting with phosphoinositides 1/2 Figure 1. Regulation of autophagy by extracellular cues. (A) Amino acids are key regulators of autophagy. When they are in excess, mTORC1 is targeted to the lysosomal membrane, where it is activated by Rheb and inhibits autophagy through phosphorylation of Ulk1 complex subunits. (B) Binding of insulin to its receptor (IR) activates mTOR via the PI3KC1/Akt/TSC pathway, inhibiting autophagy. The expression of autophagy-related proteins is inhibited after the inhibition of FoxO transcription factors by Akt. Glucose 6-phosphate inhibits the activity of hexokinase-II, autophagy, whereas others have a lesser effect. Several mechanisms internalization of essential amino acids by the bidirectional trans- an mTOR activator, inhibiting autophagy. (C) Activation of EGFR by its ligand inhibits autophagy directly by the phosphorylation of Beclin1 or indirectly via GRB2 and GAB2, as have been proposed as part of the amino acid sensing system, all of porter SLC7A5-SLC3A5 [17]. well as via the phosphorylation of STAT3, which releases eIF2a to induce the expression of autophagy-related proteins. (D) TLR4 is activated upon binding of LPS, leading to the which finally mediate autophagic activity through TOR. A decline in Upon starvation, the low amino acid concentration is sensed by recruitment of adaptor proteins to the plasma membrane. As a consequence, TRAF6 is recruited, resulting in the Lys63-linked ubiquitination of Beclin1, allowing it to bind amino acid content is initially sensed at the plasma membrane, and, Rag GTPases on the lysosomal surface [18,19], which then form PI3KC3 and induce autophagy. Nrf2 is activated, up-regulating the expression of p62. in yeast, amino acid transporters were suggested to sense extracellu- inactive heterodimers of RagA/RagB bound to RagC/RagD [19]. See Glossary for definitions and the text for details. lar amino acid concentration to regulate TOR activity [10]. Under amino acid-rich conditions, the Rag complex, together with a However, most studies in yeast induce autophagy by nitrogen star- multi-protein signaling complex known as the Ragulator and the vation, and it has been recently demonstrated that sensing of nitro- vacuolar protein pump v-ATPase, targets mTOR to the lysosome for mTORC1 sensitivity to amino acids [25]. Importantly, mutations in cargo receptor [31]. p62 recruits TRAF6, an E3 ubiquitin ligase that gen differs from that of amino acids [11]. A reduction of its activation [18,20]. It is tempting to speculate that the localization the GAP activity of GATOR-1 are associated with human cancer. is essential for activation of mTOR through Lys63 ubiquitination extracellular amino acids levels was shown to induce autophagy in of active mTOR on the lysosomal membrane allows it to be directly A more direct mechanistic connection of TOR to autophagy is [32]. It would therefore be interesting to determine whether p62 a GCN2-GCN4-dependent manner, whereas how the lack of external regulated by free amino acids produced by protein degradation illustrated by its ability to phosphorylate the Atg1 complex (ULK1 serves as a molecular switch modulating mTOR activity during nitrogen is sensed remains unknown. within the lysosome. Indeed, the v-ATPase was suggested to serve as complex in mammals) [5]. In yeast, TOR binds and phosphorylates changes in growth conditions. If this is indeed the case, it could In mammalian cells, the G protein-coupled receptors (GPCRs) a link between mTOR and the amino acids generated by the lyso- Atg13, detaching it from the Atg1 complex, whereas in mammals, it explain why p62 mediates mTOR activity and autophagic inhibition T1R1/T1R3 have been implicated in the extracellular sensing of some [20]. Active mTORC1 is localized mainly on the lysosomal regulates the constantly assembled Ulk1 complex—Atg13, ULK1, under normal conditions, yet sequesters cargo for lysosomal degra- amino acid availability, which leads to autophagy induction medi- membrane [21,22]; under amino acid deprivation, inactivation of the and FIP200—through direct binding and phosphorylation of Atg13 dation upon autophagic induction. Phosphorylation of Bcl2 by ated by a decrease in mTOR activity [12]. Leucyl-tRNA synthetase Rag complex causes its detachment from raptor (part of mTORC1) and Ulk1 [26–29]. Under unfavorable conditions, the Atg1 complex JNK-1 was recently shown to induce its dissociation from Beclin1, (LRS) was recently shown to sense increases in intracellular leucine and separation of mTOR from the lysosome and its activator Rheb, in yeast is activated by the release of Atg13 from inactive TOR, enabling it to form the PI3KC3 essential for autophagosome forma- levels and to mediate TOR activation by slightly different mecha- resulting in autophagic stimulation [23]. An important connection whereas in mammals, the whole Ulk1 complex is activated by its tion [33,34], thereby providing another important insight into the nisms in yeast and mammalian cells [13,14]. Moreover, leucine also between the Rag complex and tumorigenesis was recently suggested, detachment from mTOR [30]. Ulk1 can be then auto-phosphorylated mechanism by which starvation induces autophagy. regulates mTOR activity through glutamate dehydrogenase (GLUD) as this complex is negatively regulated by the tumor suppressor and phosphorylate Atg13 and FIP200, triggering complex activity in [15]. In addition to its leucine-mediated regulation, mTOR is regu- SH3BP4 [24]. Sabatini and co-workers also identified a protein the initial steps of autophagosome biogenesis [29]. Insulin and glucose starvation At high glucose concentrations, lated by a-ketoglutarate, which is produced during glutamine complex termed GATOR, which is comprised of the subunits GATOR-1 An important link exists between mTORC1, the Rag GTPase autophagy is down-regulated through insulin receptor signaling metabolism [16]. Notably, glutamine plays a key role in the and GATOR-2. GATOR-1 is a GAP for RagA/RagB, mediating complex, and the scaffold protein p62, which is also an autophagic [35]. Binding of insulin to its receptor activates PI3KC1 to generate

840 EMBO reports Vol 15 | No 8 | 2014 ª 2014 The Authors ª 2014 The Authors EMBO reports Vol 15 | No 8 | 2014 841 EMBO reports From signaling to autophagosome biogenesis Adi Abada & Zvulun Elazar Adi Abada & Zvulun Elazar From signaling to autophagosome biogenesis EMBO reports

[51]. It would thus be interesting to further elucidate the molecular Oxidative cues Intracellular pathways lead to the production of switches dictating the regulation of autophagy by the EGFR family ROS, which serve as signaling molecules at low concentrations yet ER NOX4 upon changes in growth conditions. Accordingly, serum starvation are highly hazardous and must be eliminated. The main ROS mole- was recently reported to induce autophagy through GSK3, which cules that participate in autophagic signaling are H2O2 and O2À. ROS phosphorylates and activates the acetyltransferase TIP60, leading to regulate autophagy at various intracellular locations. At the plasma

PI3KC1 acetylation and activation of Ulk1 [52]. membrane, H2O2 directly modifies and inactivates PTEN, which IP3R ROS PTEN inhibits PI3KC1 activity, thus eventually activating mTOR [66]. ROS PERK NO ATM Beclin1 Toll-like receptors Autophagy has been widely implicated in immu- have also been implicated in the activation of JNK, in a process ? nity through its support of immune-cell activity. The toll-like recep- regulated through the interaction of Atg9 with TRAF6 [67] that TRAF6 tor (TLR) family, an essential part of the innate immune system, has induces autophagy by up-regulating the expression of different Atg LKB1 been implicated in the regulation of autophagy, and the mechanism proteins [68–70]. In the ER of cardiomyocytes, ROS levels are inten- ATP/AMP Atg9 eIF2α JNK governing this process was recently elucidated [53] (Fig 1D). Two tionally up-regulated by NOX4 upon glucose deprivation to induce adaptor proteins, MyD88 and TRIF [54], recruit the E3 ligase TRAF6 autophagy through the PERK and eIF2a pathway, thus preventing P to the autophagic regulator Beclin1 via its two TRAF6-binding cell death [71]. TSC1 and TSC2 are targeted to the peroxisomal cGMP IKKβ AMPK domains [55]. Beclin1 is polyubiquitinated with a Lys63-linked membrane by pex19 and pex5, respectively, where they inhibit ubiquitin chain on Lys117, which is located within its BH3 domain. mTOR activity by hydrolyzing GTP-Rheb upon exposure to ROS This induces Beclin1 detachment from Bcl2 to induce autophagy in [72], suggesting that peroxisomes may induce autophagy in Nucleus a process regulated by the deubiquitinating enzyme A20 [55]. response to oxidative stress. Beclin1 is then free to form, together with Atg14, Vps34, and Vps15, In mammalian systems, ROS production at the mitochondria is P P P P TSC1 the PI3KC3 complex essential in the initial stages of autophagosome elevated upon starvation and was found to directly regulate the Beclin1 Vps34 Ulk1 Raptor biogenesis [5]. Moreover, the up-regulation of p62 expression by the activity of Atg4, the priming and delipidating enzyme of Atg8 [73]. TSC2 TLR4 pathway, which is mediated by the transcription factor Nrf2, Upon oxidation, Atg4 is transiently and locally inactivated to stabi- further extends the induction of autophagy in a MyD88- and p38- lize Atg8s in their lipidated active form. Moreover, ROS production dependent manner [56]. Activation of TLR4 thus leads to lysosomal by mitochondria could serve as a signal for their elimination by mTORC1 ROS elimination of invading bacteria, a process termed xenophagy, mitophagy [74]. In this regard, GLUD activity at the mitochondria Mitochondria Peroxisome Rheb mTORC1 P Phosphate which is beyond the scope of the present review [57]. inhibits autophagy, probably through the generation of NADPH, Rheb Cytokines, a large group of immune signaling molecules that are which prevents ROS accumulation [15]. Lysosome Calcium ions secreted to promote differentiation, recruitment, and activation of S-nitrosyl immune cells, have also been implicated in the regulation of auto- Nitric oxide NO is produced in cells by NOS and acts as a signaling Atg4 phagy. The proinflammatory cytokines IL-1b, IFN-c, and TNF were molecule in different immune response pathways and the cardiovas- found to induce autophagy to protect macrophages from bacterial cular system, among other contexts [75]. NO was initially shown to Figure 2. Regulation of autophagy by intracellular cues. infection [58]. In contrast, the cytokines IL-4, IL-13, IL-10, and IL-6 inhibit autophagosome biogenesis in HeLa cells by inducing mTOR Internal cues regulate autophagy on different levels from many intracellular locations. The activity of mTORC1 is regulated at the lysosome and the peroxisome through signal for autophagic inhibition, each via a different signaling path- activation via the AMPK-TSC pathway, through S-nitrosylation of AMPK. Active AMPK indirectly inhibits autophagy by activating the TSC1/2 complex and via inhibition of raptor by phosphorylation, both of which lead to the inhibition of way. IL-10 inhibits autophagy through the Akt signaling pathway, IKKb [76]. In MCF-7 breast cancer cells, however, NO was reported mTORC1. Autophagy is inhibited directly by Ulk1, Vps34, and Beclin1 phosphorylation. ROS molecules activate autophagy at the plasma membrane, the ER, and mitochondria, as well as by up-regulating the expression of autophagy-related proteins. Ca2+ signaling is mediated from its intracellular storages in the mitochondria and the ER. The whereas IL-4 and IL-13 are inhibitory only when autophagy is to lead to the induction of autophagy in an ATM- and mTOR- regulation of autophagy through AMPK induced by NO remains poorly understood. NO regulates mitophagy through cGMP. induced by starvation. IL-6 inhibits autophagy by down-regulating dependent manner [77]. The difference observed between the two the expression of autophagic proteins mediated by STAT3 regulation cell lines may be explained by the lack of LKB1 in HeLa cells, which See Glossary for definitions and the text for details. [58]. is essential in the autophagic regulation of the IKKb signaling pathway by NO. PtdIns3P at the plasma membrane [36], thereby recruiting and acti- pathway, previously implicated only in amino acid starvation [44]. Intracellular cues NO regulates immune responses and was recently shown to be vating both PDK1 and Akt/PKB (Fig 1B). Upon insulin signaling, This issue has not yet been fully resolved, however, as another Autophagy is the main intracellular process responsible for the clear- implicated in xenophagy [78]. NO was shown to nitrosylate cGMP Akt is activated by two pathways, PDK1 and mTORC2, thus mediat- study in mice indicated that long periods of glucose deprivation do ance of defective organelles and protein aggregates caused by aging, following cell stimulation by LPS and IFN-c, forming 8-nitro-cGMP, ing an indirect regulation of the non-autophagic mTOR complex on not inhibit autophagy [45]. cellular malfunction, or both. This is particularly important in long- which can subsequently modify cysteine residues in target proteins. autophagy [37,38]. Akt activation leads to the inhibition of TSC1/2, lived cells such as neurons. The internal state of the cell is monitored Proteins on the bacterial surface of group A Streptococcus (GAS) an mTOR inhibitor, resulting in the inhibition of autophagy [8]. Epidermal growth factor Autophagy plays a crucial metabolic role, to maintain homeostasis under different growth conditions (Fig 2). were modified by S-guanylation after cellular invasion, which Interestingly, loss of the TSC1/2 complex induces constitutive especially when supplies are limited, implying that it is tightly marked bacteria for K63-linked polyubiquitination, inducing their mTOR activity at the plasma membrane due to the enhanced activity linked to growth factors. Recent studies have indeed demonstrated Energy level The energy status of the cell is typically sensed by the engulfment by autophagosomes and lysosomal targeting. It would of RalA/RalB through the exocyst complex [39]. The long-term regu- that the EGFR system inhibits autophagy, either indirectly through ATP/AMP ratio and regulated by AMPK binding [59]. When AMP is be interesting to investigate whether S-guanylation by 8-nitro-cGMP lation of autophagy by Akt includes down-regulation of the expres- GRB2 and GAB2 [46], or directly by phosphorylating Beclin1, in excess, indicating low energy levels, it binds AMPK, leading to serves as a broad modification marking substrates in additional sion of many autophagic proteins through a process mediated by prompting its dimerization to prevent its activity [47] (Fig 1C). phosphorylation and activation by LKB1 [60]. This consequently forms of selective autophagy. the FoxO transcription factor family [40–42]. Alternatively, autophagy can be induced through EGF-dependent activates autophagy via two main signaling pathways: mTOR inhibi- Autophagy was recently directly linked to glucose deprivation in phosphorylation and dimerization of STAT3, releasing its binding of tion through the TSC1/2 complex [61,62] or the phosphorylation of Ca2+ ions Ca2+ is a well-established signaling molecule implicated cardiomyocytes [43]. In low glucose conditions, hexokinase-II—the PKR, the catalytic domain of eIF2a [48,49]. This up-regulates the raptor and binding to 14-3-3 proteins, which also inhibits mTOR in numerous cellular processes, and its cytosolic concentration is first enzyme in the glycolysis pathway—was postulated to directly transcription and translation of the core autophagic proteins LC3 [63]. An alternative mechanism was recently described whereby tightly regulated. The ER and the mitochondria serve as the primary bind and thereby inhibit mTOR. This binding would be inhibited by and Atg5 [50]. Interestingly, expression of the constitutively active upon glucose deprivation, AMPK directly phosphorylates Ulk1 [64], Ca2+ storage organelles. ER stress, for example, leads to the release glucose 6-phosphate, the substrate of hexokinase-II (Fig 1B). EGFR mutant EGFRvIII results in extensive autophagic activity, Vps34, and Beclin1 (members of the PI3KC3 complex), leading to of Ca2+ from internal ER pools into the cytosol to regulate different Furthermore, glucose deprivation in neonatal mice was shown to promoting cell survival in tumor cells especially under stressful PI3KC3 stabilization and activation of autophagy, which ensures cell stages along the autophagy pathway, yet this process remains induce the lysosomal targeting of mTOR through the Rag GTPase conditions, such as uncontrolled Ras signaling and oxidative stress survival [65]. poorly understood at a mechanistic level [79–81].

842 EMBO reports Vol 15 | No 8 | 2014 ª 2014 The Authors ª 2014 The Authors EMBO reports Vol 15 | No 8 | 2014 843 EMBO reports From signaling to autophagosome biogenesis Adi Abada & Zvulun Elazar Adi Abada & Zvulun Elazar From signaling to autophagosome biogenesis EMBO reports

The inositol 1,4,5-trisphosphate receptor, IP3R, is a Ca2+ channel sites [94], and recycling endosomes [95]. This issue has been exten- in the different stages of autophagosome biogenesis to the phago- second UBL conjugation system, that of Atg8 to phosphatidyletha- activated by IP3 binding [82]. It is located on various membranes sively reviewed and is therefore not discussed here [5,6,96]. phore [104]. E3 ubiquitin ligases, such as TRAF6, ubiquitinated nolamine (PE) [117]. The conjugation process is mediated by Atg7 and regulates Ca2+ levels in organelles, and consequently in the protein aggregates, the core autophagic protein Atg5, and the cargo as the E1-conjugating enzyme, and Atg3 as the E2-conjugating cytosol. The N-terminal region of the receptor interacts with Beclin1 Nucleation recruiters p62 and NBR1, have all been shown to bind ALFY enzyme. The mammalian Atg8 is a UBL protein family consisting of through a non-Bcl2-interacting region to regulate autophagy and was The initial step in membrane nucleation for phagophore formation [105,106]. Following protein aggregation in cells, ALFY is exported eight family members grouped into the LC3 and GABARAP subfami- shown to sensitize the receptor to IP3 binding during autophagy is the recruitment of autophagic proteins to a membrane in the cell from the nucleus and targets the protein aggregates to the phago- lies [120]. WIPI2 was recently reported to recruit the Atg12–Atg5– [83]. Interestingly, the IP3R also regulates low Ca2+ levels in the designated by the presence of PtdIns3P. Yeast phagophores are initi- phore via p62. The implication of ALFY in neurodegenerative Atg16 complex to the site of autophagosome formation by directly mitochondria, impairing ATP production and increasing the AMP/ ated at one location termed the pre-autophagosomal structure diseases is consistent with its importance in aggregate clearance interacting with Atg16 [121]. ATP ratio, thereby inducing autophagy in an AMPK-dependent (PAS), whereas in mammals, they are synthesized throughout the [107]. However, the signaling pathways that dictate the cellular The exact mechanism whereby the Atg12–Atg5–Atg16 complex manner [84]. cell. Microscopic analysis of the recruitment order of the autophagy- location of ALFY and its targeting to the membrane remain induces phagophore elongation is still unclear. Using GUVs and related proteins in both yeast and mammalian systems suggested unknown. Importantly, starvation leads to a decrease in ALFY level, purified recombinant proteins, it was initially suggested that the well-defined hierarchies for the order of incorporation of complexes suggesting that this protein is important for the clearance of protein complex participates in vesicle tethering [122]. Atg12 has also been Autophagosome biogenesis into the site of autophagosome formation [97,98]. Regulation of the aggregates yet may be toxic under stressful conditions [108]. It shown to bind Atg3 and carry it to the membrane, promoting Atg8 stability of PI3KC3, which is composed of Vps34, Vps15, Atg14, and would be interesting to determine whether ALFY is implicated in lipidation, which supports its role as an E3 in Atg8 conjugation. According to the current view, autophagosomes originate from a Beclin1, and that of the Ulk1 complex, is essential for the nucleation additional forms of selective autophagy and whether the budding of Interestingly, conjugation of Atg8 to PE is promoted by the Atg12– membrane that elongates until it is finally sealed as a mature double- process in mammalian cells and is regulated by post-translational the phagophore occurs in parallel to ALFY recruitment or sequen- Atg5–Atg16 complex on SUVs but not on GUVs, indicating that membrane autophagosome, which subsequently fuses with the lyso- modifications (Fig 3). Vps34 is a class III PI3K that phosphorylates tially. The order by which proteins are recruited by ALFY and the membrane curvature is a significant factor in the activity of this some, where its content is degraded. The search for the membrane phosphatidylinositol at the designated membrane, generating time point of membrane binding is likely to shed new light on the complex. In agreement with this hypothesis, Atg3 was shown to be origin of the phagophore has been an enticing quest for many years. PtdIns3P [99]. EM analysis utilizing quick-freezing and freeze- early stages of autophagosome biogenesis. targeted to highly curved membranes, where it promotes Atg8–PE The introduction of autophagy-specific molecular tools and sophisti- fracture replica labeling revealed differences in the dispersion of The formation of the PI3KC3 is supported by UVRAG [109] and conjugation [123]. A more recent study in GUVs suggested a slightly cated imaging techniques led the way to the identification of multiple PtdIns3P in yeast and mammalian autophagosomes [100]. In yeast, by Ambra1, a Beclin1-interacting protein [110]. Ambra1 was different scenario, in which Atg12–5–16 initially catalyzes the lipida- cellular membranes as possible sources of the isolation membrane. PtdIns3P was found mostly in the inner membrane leaflets facing recently shown to be a target of mTOR and to be inhibited by its tion of Atg8 to the membrane, which in turn acts to stabilize the The first report in this regard utilized a GFP-tagged FYVE zinc finger the luminal barrier within the double membrane, whereas in phosphorylation at Ser52 under normal growth conditions [111]. association of the Atg12–5–16 complex on the membrane [124]. domain of DFCP1, an ER resident protein that does not participate in mammals, this lipid was mostly localized to the outer autophagoso- Upon autophagic induction, Ambra1 is phosphorylated by Ulk1, Thus, Atg8 is suggested to play a structural role and Atg12–5–16 to autophagy but has high affinity for PtdIns3P on membranes [85]. mal membrane leaflets, suggesting differences in the autophago- which detaches it from dynein on microtubules and targets it to the function as a coat. Both of these in vitro studies need further clarifi- The induction of autophagy apparently leads to the recruitment of some formation process in the different organisms. The site of ER [112]. Ambra1 then binds Ulk1 and TRAF6, promoting the Ulk1 cation. The observed differences might be associated with altera- this artificial reporter protein into an ER subdomain that contains PtdIns3P formation dictates the location of phagophore formation, Lys63-linked polyubiquitination that is essential for creation of the tions in membrane curvature (Fig 4). During the initial stages of autophagic factors such as Atg14 and WIPI and to the formation of as it leads to the recruitment of early autophagosome biogenesis Ulk1 complex [111]. Interestingly, WASH—an endosome-associated autophagosome biogenesis, when the phagophore is still relatively a cup-shaped membrane termed omegasome [85–87]. However, factors, such as the WD-40-repeat-domain containing proteins protein—was shown to compete with Ulk1 ubiquitination and with small and highly curved, the Atg12–Atg5–Atg16 complex might additional membrane sources for phagophore formation have been WIPI1 and WIPI2 [101,102]. In addition, the FYVE-domain contain- Beclin1 binding by Ambra1 [113]. Ambra1 therefore appears to act promote lipidation. As the phagophore continues to grow, its elon- suggested, such as plasma membrane [88,89], mitochondria [90], ing protein ALFY is recruited [103] and was recently defined as an as a novel link between PI3KC3 and the Ulk1 complex, both of gation points exhibit high curvature, whereas the sites already built Golgi [91], ERGIC [92], ER–mitochondria contact sites [93], ER exit adaptor protein able to concentrate several factors that are essential which are essential in the initial steps of autophagosome biogenesis. are less curved and need to be stabilized. At these locations, the Notably, although its activity is crucial, only a limited number of Atg12–Atg5–Atg16 complex might be essential in maintaining Ulk1 effectors have been identified. A recent study in yeast shows membrane structure and stability. that Atg9 is a direct substrate of Atg1 [114], the yeast homolog of Further support for the notion that curvature is important for the JNK TRAF6 AMPK mTORC1 TRAF6 GSK3 Ulk1. As active mTORC1 resides on the lysosomal membrane, the recruitment and activity of the biogenesis machinery comes from inhibited Ulk1 complex could share the same location. To become the association of SNX18 with autophagic induction [125]. SNX18

active, Ulk1 needs to be shuttled from the lysosomal membrane by contains a PX domain that targets it to PIP2 on membranes, as well a mechanism yet to be resolved. A key factor in this process might as SH3 and a BAR domains known to sense and endorse membrane P be Ambra1, owing to its location along microtubules. curvature. It was shown to target Atg16 to perinuclear recycling P P P P Connexins, a family of multispan transmembrane proteins that endosomes and interact directly with LC3 [125], suggesting that Bcl2 Beclin1 Beclin1 Ulk1 Ambra1 TIP60 form plasma membrane gap junctions, were recently suggested to Atg16 mediates LC3 lipidation on highly curved membranes. SNX18 negatively regulate autophagosome formation by direct interaction was further suggested to induce the tubulation of recycling endo- Ambra1 with the PI3KC3 at the plasma membrane. According to the somes that supply membranes for the elongating phagophore. Inter- Ulk1 TRAF6 proposed model, under starvation conditions, Atg14 is incorporated estingly, endosomal tubulation driven by the overexpression of into the plasma membrane, where it releases connexin-induced inhi- TBC1D14, a Rab11 binding protein, was found to be inhibitory for P P P Ac bition by directing these proteins to lysosomal degradation [115]. autophagosome formation, suggesting an antagonist role for SNX18 Bcl2 Beclin1 Vps34 Ulk1 Atg13 Ulk1 [126]. Another pathway thought to target Atg16 to the phagophore Atg14 Vps15 FIP200 Phagophore formation and elongation is through interaction with FIP200, a subunit of the Ulk1 complex, Downstream of the recruitment of WIPI1/2 are two ubiquitin-like as it relieves Atg16 auto-inhibition [127]. The discovery of multiple (UBL) systems specific to the autophagic process. The first is the Atg16 targeting pathways to membranes reflects the vital role of

P Phosphate Ac Acetyl K63 linked ubiquitin conjugation of Atg12, a UBL protein, to Atg5 by the E1 enzyme Atg7 Atg16 in autophagosome biogenesis and strengthens the hypothesis and the E2 enzyme Atg10 [116]. Atg5 binds the N-terminal region of that it plays multiple roles in this process. Atg16 through a non-covalent bond, independently of its interaction The phagophore membrane is subsequently elongated through a

Figure 3. Post-translational modifications regulate PI3KC3 and the Ulk1 complex. with Atg12 [117]. Atg16 creates homodimers, each capable of bind- process that is not fully characterized. A given membrane source Ubiquitination, phosphorylation, and acetylation of Ulk1 and Beclin1 regulate autophagy by promoting or preventing the formation of the Ulk1 complex or PI3KC3. ing an Atg12–Atg5 conjugate resulting in a heterohexamer could continue to elongate until the autophagosome is completed [118,119]. The Atg12–Atg5–Atg16 complex is known to dictate the or, alternatively, small vesicles could fuse with the phagophore to See Glossary for definitions and the text for details. site of autophagosome formation by acting as an E3 ligase in the expand its membrane [128]. In mammals, the VAMP7 SNARE

844 EMBO reports Vol 15 | No 8 | 2014 ª 2014 The Authors ª 2014 The Authors EMBO reports Vol 15 | No 8 | 2014 845 EMBO reports From signaling to autophagosome biogenesis Adi Abada & Zvulun Elazar Adi Abada & Zvulun Elazar From signaling to autophagosome biogenesis EMBO reports

complex—which includes VAMP7, syntaxin 7, syntaxin 8, and Vti1b A LC3I LC3II A Golgi —was the first reported to mediate autophagosome biogenesis [88]. In a parallel yeast study, in which trafficking of ectopically 5 Atg5– expressed Atg9 was studied, the t-SNARE Tlg2 and the v-SNAREs Atg12– 12 Sec22 and Ykt6 were shown to be essential for autophagosome Atg16 biogenesis through mediation of Atg9 trafficking [89]. Notably, Atg9 16 overexpression results in the formation of tubular structures ER suggested to serve as Atg9 reservoirs and as a membrane source for the PAS [129], a finding recapitulated also in mammalian cells [130]. When Atg9 expression was regulated by its endogenous promoter, however, this phenomenon was observed only on small Mitochondria B unilaminar vesicles synthesized de novo from the Golgi membrane [128]. These vesicles were shown to fuse with autophagosomes that contain Atg9 on the outer membrane, which is detached only after autophagosome maturation. Furthermore, relatively few Atg9- B tagged vesicles were shown to fuse with the phagophore, and there- Mitochondria fore, additional membranes sources are required. Atg9 trafficking is regulated by a complex machinery, as became recently apparent ER [131]. TRAPPIII, which is part of the general trafficking machinery, was implicated in Atg9 trafficking under normal growth conditions, but shown to be less necessary under starvation conditions, in which the GARP pathway is essential. In mammalian cells, Atg9 was initially found in the Golgi and is transported to endosomal compartments upon autophagic induction [132], yet it was recently Golgi detected on additional intracellular compartments [130]. It was also shown to be essential in the early stages of autophagosome forma- C tion by transiently interacting with the expanding phagophore [130]. In addition, Atg9 has been recently reported to localize to the plasma membrane, from which it is internalized and fused with 5 Atg16-tagged vesicles, a finding consistent with the suggested Atg9 LC3I LC3II Atg5– 12 involvement of both the plasma membrane and the ER–Golgi system Atg12– in autophagosome formation [95]. Atg16 Adaptor protein/Cargo 16 Atg8 and its mammalian orthologs have been implicated in the elongation of the phagophore [5,133–135]. In yeast, phagophore elongation and autophagosome size are controlled by Atg8 [134,135]. Notably, an in vitro system using liposomes containing Figure 5. Models of autophagosome biogenesis. (A) The current view of autophagosome biogenesis is a continuous process, where an initial membrane for autophagosome formation buds out from an existing organelle different concentrations of PE yielded contradictory results regard- and is further elongated by the fusion of vesicles, some containing Atg9. The Atg12–Atg5–Atg16 complex promotes LC3 lipidation on the highly curved membranes ing the fusogenic activity of Atg8s [89]. LC3, a mammalian homolog while supporting the membrane’s structure as it elongates. Once the autophagosome is sealed, it encapsulates cargo for degradation and the external biogenesis of Atg8, is also involved in phagophore expansion [136]. Interest- machinery is removed. (B) A new model for autophagosome biogenesis. Multiple nucleation membranes bud from several organelles to contribute to the formation of the ingly, the N-terminal region of LC3 and GATE-16 promotes vesicle initial membrane of the autophagosome. Each membrane elongates individually until all are fused to create an autophagosome that encapsulates cargo for lysosomal tethering and fusion in vitro, suggesting their involvement in elonga- degradation. tion of the phagophore membrane [137]. It would therefore be interesting to study the significance of Atg8 in the fusion of Atg9- containing vesicles to the phagophore. curved membrane. If multiple events occur in parallel on different GABARAP homolog LGG-1 and the LC3 homolog LGG-2 act consec- Although typically described as a progressive process whereby sites, it is possible that several biogenesis pathways occur simulta- utively [138]. LGG-2 activity downstream of LGG-1 promotes auto- the edges of a crescent-shaped phagophore elongate continuously to neously in vivo. Therefore, various membrane sources can contrib- phagosome maturation through a pathway that also involves the form autophagosomes, it may be also worth considering an alterna- ute to biogenesis, since within the cell’s third dimension, there are HOPS subunit VPS39. tive view whereby formation of the phagophore membrane is initi- many organelles and membranes in close proximity to the phago- Atg4 plays a dual role by priming Atg8 both for its lipid conjuga- Figure 4. Model of the Atg12–Atg5–Atg16 complex function in ated at several sites, possibly by different membrane sources phore. tion and for its removal from membranes [139]. Interestingly, autophagy. (Fig 5). Different pieces of membranes could then be tied together deconjugation of Atg8 by Atg4 was shown to be important in both (A) The Atg12–Atg5–Atg16 complex is recruited to the phagophore after its by membrane fusion events. This would require many autophagy Sealing and maturation early and late stages of biogenesis, allowing the fusion of auto- initial nucleation. At this stage, the membrane is curved and the complex biogenesis complexes acting on several sites at different stages of The final step of autophagosome biogenesis is the sealing of the phagosomes with lysosomes. There are four members of the Atg4 promotes the lipidation of LC3 with PE. (B) Once the membrane elongates, the complex remains associated with the membrane through LC3 on membranes biogenesis, and attempting to reconstitute such a process in the test phagophore to form a double-membrane vesicle. This rather under- mammalian family, each able to specifically modify different Atg8 with low curvature for their stabilization and continues to promote LC3 tube would be extremely challenging. The available cell-free studied process dictates the final size of the autophagic organelle. family members. In erythrocytes, mammalian Atg4B participates in lipidation at the highly curved edges of the phagophore. (C) As the elongation systems aimed at reconstituting autophagosome formation can Regulation of this process also serves as part of the regulation of the regulation of autophagosome maturation that is necessary for continues, the Atg12–Atg5–Atg16 complex, together with the lipid-conjugated reflect only fragments of the complex process. Accordingly, GUVs fusion with the lysosomes. The mammalian homologs of Atg8 have cell differentiation [140]. The protein FYCO1, previously shown to LC3, forms a coat-like structure that stabilizes the structure of the could represent only events that take place on membranes with rela- been suggested to promote this stage of autophagosome biogenesis. bind LC3 and mediate autophagosome trafficking, was recently phagophore. tively low curvature, whereas small liposomes may mimic highly Consistently, studies in Caenorhabditis elegans showed that the implicated in autophagosome maturation as well [141].

846 EMBO reports Vol 15 | No 8 | 2014 ª 2014 The Authors ª 2014 The Authors EMBO reports Vol 15 | No 8 | 2014 847 EMBO reports From signaling to autophagosome biogenesis Adi Abada & Zvulun Elazar Adi Abada & Zvulun Elazar From signaling to autophagosome biogenesis EMBO reports

proteins. The continuing discovery of connections between well- 10. Hundal HS, Taylor PM (2009) Amino acid transceptors: gate keepers of 28. Hara T, Takamura A, Kishi C, Iemura S, Natsume T, Guan JL, Mizushima Sidebar A: In need of answers established signaling pathways and autophagy-related proteins, as nutrient exchange and regulators of nutrient signaling. Am J Physiol N(2008) FIP200, a ULK-interacting protein, is required for autophago- (i) How are amino acids sensed by TORC1? well as of new regulators of both processes, contributes substantially Endocrinol Metab 296:E603 – E613 some formation in mammalian cells. J Cell Biol 181: 497 – 510 (ii) How are various internal signals integrated to regulate autophagy? to our understanding of the regulation of autophagy in response to 11. Ecker N, Mor A, Journo D, Abeliovich H (2010) Induction of autophagic 29. Jung CH, Jun CB, Ro SH, Kim YM, Otto NM, Cao J, Kundu M, Kim DH (iii) How does the lysosomal location of mTORC1 affect its downstream changes in the extracellular and intracellular environment. flux by amino acid deprivation is distinct from nitrogen starva- (2009) ULK-Atg13-FIP200 complexes mediate mTOR signaling to the signaling on the Ulk1 complex? The mechanism of autophagosome biogenesis is still elusive. Our tion-induced macroautophagy. Autophagy 6: 879 – 890 autophagy machinery. Mol Biol Cell 20: 1992 – 2003 (iv) How do Atg9-containing vesicles contribute to autophagosome knowledge concerning the order of incorporation of the different 12. Wauson EM, Zaganjor E, Lee AY, Guerra ML, Ghosh AB, Bookout AL, 30. Ganley IG, du Lam H, Wang J, Ding X, Chen S, Jiang X (2009) biogenesis? (v) How are the SNARE molecules needed for fusion of the factors into sites of autophagosome formation, as well as on the Chambers CP, Jivan A, McGlynn K, Hutchison MR et al (2012) The G ULK1.ATG13.FIP200 complex mediates mTOR signaling and is essential autophagosome with the lysosome incorporated into the outer functional complexes involved in this process, has increased protein-coupled taste receptor T1R1/T1R3 regulates mTORC1 and auto- for autophagy. J Biol Chem 284: 12297 – 12305 autophagosomal membrane? substantially in recent years. However, the exact details concerning phagy. Mol Cell 47: 851 – 862 31. Duran A, Amanchy R, Linares JF, Joshi J, Abu-Baker S, Porollo A, Hansen (vi) How are the inner and outer membranes of the phagophore sorted the way in which all of these factors act in concert are still missing. 13. Han JM, Jeong SJ, Park MC, Kim G, Kwon NH, Kim HK, Ha SH, Ryu SH, M, Moscat J, Diaz-Meco MT (2011)p62 is a key regulator of nutrient into two sub-domains? The reconstitution of such a step in the test tube will clearly be very Kim S (2012) Leucyl-tRNA synthetase is an intracellular leucine sensor sensing in the mTORC1 pathway. Mol Cell 44: 134 – 146 (vii) How are the cup-shaped phagophores formed and stabilized? (viii) What dictates the final sealing of the autophagosome? challenging, but if successful, it will provide invaluable information for the mTORC1-signaling pathway. Cell 149: 410 – 424 32. Linares JF, Duran A, Yajima T, Pasparakis M, Moscat J, Diaz-Meco MT on this process. There are several open questions remaining in this 14. Bonfils G, Jaquenoud M, Bontron S, Ostrowicz C, Ungermann C, De (2013)K63 polyubiquitination and activation of mTOR by the field (Sidebar A). One of them is how the membrane of the phago- Virgilio C (2012) Leucyl-tRNA synthetase controls TORC1 via the EGO p62-TRAF6 complex in nutrient-activated cells. Mol Cell 51: 283 – 296 phore is sorted into two sub-domains, leading to the formation of complex. Mol Cell 46: 105 – 110 33. Pattingre S, Tassa A, Qu X, Garuti R, Liang XH, Mizushima N, Packer M, The participation of trafficking factors in the sealing and matura- distinct inner and outer membranes. The former membrane will 15. Lorin S, Tol MJ, Bauvy C, Strijland A, Pous C, Verhoeven AJ, Codogno P, Schneider MD, Levine B (2005) Bcl-2 antiapoptotic proteins inhibit tion of autophagosomes is well established. Several Rab proteins are be degraded in the lysosome, whereas the latter fuses with the Meijer AJ (2013) Glutamate dehydrogenase contributes to leucine sens- Beclin 1-dependent autophagy. Cell 122: 927 – 939 known to play a role in autophagosome maturation [142], a process lysosome-limiting membrane to become an integral part of it. ing in the regulation of autophagy. Autophagy 9: 850 – 860 34. Wei Y, Pattingre S, Sinha S, Bassik M, Levine B (2008) JNK1-mediated in which SNARE molecules have also been implicated, as discussed Understanding this process in molecular terms is of broad interest, as 16. Duran RV, Oppliger W, Robitaille AM, Heiserich L, Skendaj R, Gottlieb E, phosphorylation of Bcl-2 regulates starvation-induced autophagy. Mol above. In a recent study, immuno-electron microscopy was used to it may shed light on the mechanism by which the lysosome-limiting Hall MN (2012) Glutaminolysis activates Rag-mTORC1 signaling. Mol Cell 30: 678 – 688 show that syntaxin 13 localizes with LC3 on immature autophago- membrane is protected from degradation. Cell 47: 349 – 358 35. Moruno F, Perez-Jimenez E, Knecht E (2012) Regulation of autophagy somes [143]. Depletion of syntaxin 13 leads to the accumulation of 17. Nicklin P, Bergman P, Zhang B, Triantafellow E, Wang H, Nyfeler B, by glucose in Mammalian cells. Cells 1: 372 – 395 immature autophagosomes, indicating that it participates in the Acknowledgements Yang H, Hild M, Kung C, Wilson C (2009) Bidirectional transport of 36. Manning BD, Tee AR, Logsdon MN, Blenis J, Cantley LC (2002) Identifi- maturation process. In addition, syntaxin 17 was suggested to ZE is the incumbent of the Harold Korda Chair of Biology. We are grateful for amino acids regulates mTOR and autophagy. Cell 136: 521 – 534 cation of the tuberous sclerosis complex-2 tumor suppressor gene promote autophagosomal sealing by trafficking through a still funding from the Israeli Science Foundation (ISF) (Grant Number 535/11), the 18. Sancak Y, Bar-Peled L, Zoncu R, Markhard AL, Nada S, Sabatini DM product tuberin as a target of the phosphoinositide 3-kinase/akt path- unknown mechanism to phagophores at late stages of biogenesis, German-Israeli Foundation (GIF) (Grant Number 1129/157), and the Legacy (2010) Ragulator-Rag complex targets mTORC1 to the lysosomal way. Mol Cell 10: 151 – 162 enabling them to fuse with the lysosome [144]. Recent studies in Heritage Fund (Grant Number 1309/13). surface and is necessary for its activation by amino acids. Cell 141: 37. Sarbassov DD, Guertin DA, Ali SM, Sabatini DM (2005) Phosphorylation Drosophila and mammalian cells suggested that syntaxin 17 binds 290 – 303 and regulation of Akt/PKB by the rictor-mTOR complex. Science 307: the HOPS complex, thereby promoting the fusion between auto- Conflict of interest 19. Sancak Y, Peterson TR, Shaul YD, Lindquist RA, Thoreen CC, Bar-Peled L, 1098 – 1101 phagosomal and lysosomal membranes [145,146]. Syntaxin 17 was The authors declare that they have no conflict of interest. Sabatini DM (2008) The Rag GTPases bind raptor and mediate amino 38. Brozinick JT Jr, Birnbaum MJ (1998) Insulin, but not contraction, acti- also shown to induce, together with Atg14, budding of the source acid signaling to mTORC1. Science 320: 1496 – 1501 vates Akt/PKB in isolated rat skeletal muscle. J Biol Chem 273: membrane at ER contact sites [93]. Additional studies are clearly 20. Zoncu R, Bar-Peled L, Efeyan A, Wang S, Sancak Y, Sabatini DM (2011) 14679 – 14682 needed to clarify the exact role of syntaxin 17 along the autophagic References mTORC1 senses lysosomal amino acids through an inside-out mecha- 39. Martin TD, Chen XW, Kaplan RE, Saltiel AR, Walker CL, Reiner DJ, Der CJ pathway. nism that requires the vacuolar H(+)-ATPase. Science 334: 678 – 683 (2014) Ral and Rheb GTPase activating proteins integrate mTOR and Interestingly, the lysosome-located protein TECPR1 was impli- 1. Boya P, Reggiori F, Codogno P (2013) Emerging regulation and func- 21. Li M, Khambu B, Zhang H, Kang JH, Chen X, Chen D, Vollmer L, Liu PQ, GTPase signaling in aging, autophagy, and tumor cell invasion. Mol Cell cated in fusion between the autophagosome and the lysosome by tions of autophagy. Nat Cell Biol 15: 713 – 720 Vogt A, Yin XM (2013) Suppression of lysosome function induces auto- 53: 209 – 220 binding the Atg12–Atg5 conjugate [147]. Depletion of TECPR1 2. Choi AM, Ryter SW, Levine B (2013) Autophagy in human health and phagy via a feedback down-regulation of MTOR complex 1 (MTORC1) 40. Mammucari C, Milan G, Romanello V, Masiero E, Rudolf R, Del Piccolo resulted in the accumulation of autophagosomes that do not fuse disease. N Engl J Med 368: 651 – 662 activity. J Biol Chem 288: 35769 – 35780 P, Burden SJ, Di Lisi R, Sandri C, Zhao J et al (2007) FoxO3 controls with lysosomes. As the Atg12–Atg5 conjugate is removed from 3. Sato M, Sato K (2013) Dynamic regulation of autophagy and endo- 22. Korolchuk VI, Saiki S, Lichtenberg M, Siddiqi FH, Roberts EA, Imarisio S, autophagy in skeletal muscle in vivo. Cell Metab 6: 458 – 471 autophagosomes prior to their final sealing, these results suggest cytosis for cell remodeling during early development. Traffic 14: Jahreiss L, Sarkar S, Futter M, Menzies FM et al (2011) Lysosomal posi- 41. Zhao J, Brault JJ, Schild A, Cao P, Sandri M, Schiaffino S, Lecker SH, a direct link between autophagosome sealing and removal of 479 – 486 tioning coordinates cellular nutrient responses. Nat Cell Biol 13: 453 – 460 Goldberg AL (2007) FoxO3 coordinately activates protein degradation this complex. It is possible that autophagosomes only reach their 4. Guan JL, Simon AK, Prescott M, Menendez JA, Liu F, Wang F, Wang C, 23. Inoki K, Li Y, Xu T, Guan KL (2003) Rheb GTPase is a direct target of by the autophagic/lysosomal and proteasomal pathways in atrophying final maturation, cued by the removal of all biogenesis machinery, Wolvetang E, Vazquez-Martin A, Zhang J (2013) Autophagy in stem TSC2 GAP activity and regulates mTOR signaling. Genes Dev 17: muscle cells. Cell Metab 6: 472 – 483 immediately before their fusion with the lysosomes. Importantly, cells. Autophagy 9: 830 – 849 1829 – 1834 42. Tan SH, Shui G, Zhou J, Shi Y, Huang J, Xia D, Wenk MR, Shen HM Atg5 binding by TECPR1 was also suggested to promote the initial 5. Weidberg H, Shvets E, Elazar Z (2011) Biogenesis and cargo selectivity 24. Kim YM, Stone M, Hwang TH, Kim YG, Dunlevy JR, Griffin TJ, Kim DH (2014) Critical role of SCD1 in autophagy regulation via lipogenesis stages of autophagosome formation and to target bacteria to of autophagosomes. Annu Rev Biochem 80: 125 – 156 (2012) SH3BP4 is a negative regulator of amino acid-Rag GTPase-m- and lipid rafts-coupled AKT-FOXO1 signaling pathway. Autophagy 10: the phagophore during xenophagy, but not upon starvation [148]. 6. Rubinsztein DC, Shpilka T, Elazar Z (2012) Mechanisms of autophago- TORC1 signaling. Mol Cell 46: 833 – 846 226 – 242 The exact role of TECPR1 in autophagy thus remains to be some biogenesis. Curr Biol 22:R29 – R34 25. Bar-Peled L, Chantranupong L, Cherniack AD, Chen WW, Ottina KA, 43. Roberts DJ, Tan-Sah VP, Ding EY, Smith JM, Miyamoto S (2014) Hexoki- elucidated. 7. Laplante M, Sabatini DM (2012) mTOR signaling in growth control and Grabiner BC, Spear ED, Carter SL, Meyerson M, Sabatini DM (2013)A nase-II positively regulates glucose starvation-induced autophagy disease. Cell 149: 274 – 293 Tumor suppressor complex with GAP activity for the Rag GTPases that through TORC1 inhibition. Mol Cell 53: 521 – 533 8. Sarkar S (2013) Regulation of autophagy by mTOR-dependent and signal amino acid sufficiency to mTORC1. Science 340: 1100 – 1106 44. Efeyan A, Zoncu R, Chang S, Gumper I, Snitkin H, Wolfson RL, Kirak O, Concluding remarks mTOR-independent pathways: autophagy dysfunction in neurodegener- 26. Hosokawa N, Sasaki T, Iemura S, Natsume T, Hara T, Mizushima N Sabatini DD, Sabatini DM (2013) Regulation of mTORC1 by the Rag ative diseases and therapeutic application of autophagy enhancers. (2009) Atg101, a novel mammalian autophagy protein interacting with GTPases is necessary for neonatal autophagy and survival. Nature 493: Autophagy lies at a fundamental junction in cell fate, determining Biochem Soc Trans 41: 1103 – 1130 Atg13. Autophagy 5: 973 – 979 679 – 683 death or survival. The integration between stress-induced pathways 9. Shimobayashi M, Hall MN (2014) Making new contacts: the mTOR 27. Mercer CA, Kaliappan A, Dennis PB (2009) A novel, human Atg13 bind- 45. Naito T, Kuma A, Mizushima N (2013) Differential contribution of insu- and autophagic proteins is rather complex, given that different cues network in metabolism and signalling crosstalk. Nat Rev Mol Cell Biol ing protein, Atg101, interacts with ULK1 and is essential for macro- lin and amino acids to the mTORC1-autophagy pathway in the liver are funneled to activate a relatively small pool of autophagic 15: 155 – 162 autophagy. Autophagy 5: 649 – 662 and muscle. J Biol Chem 288: 21074 – 21081

848 EMBO reports Vol 15 | No 8 | 2014 ª 2014 The Authors ª 2014 The Authors EMBO reports Vol 15 | No 8 | 2014 849 EMBO reports From signaling to autophagosome biogenesis Adi Abada & Zvulun Elazar Adi Abada & Zvulun Elazar From signaling to autophagosome biogenesis EMBO reports

46. Wheeler M, Domin J (2001) Recruitment of the class II phosphoinosi- 65. Kim J, Kim YC, Fang C, Russell RC, Kim JH, Fan W, Liu R, Zhong Q, Guan 81. East DA, Campanella M (2013) Ca2+ in quality control: an unresolved carboxypeptidase Y sorting in Saccharomyces cerevisiae. J Cell Biol 152: tide 3-kinase C2beta to the epidermal growth factor receptor: role of KL (2013) Differential regulation of distinct Vps34 complexes by AMPK riddle critical to autophagy and mitophagy. Autophagy 9: 1710 – 1719 519 – 530 Grb2. Mol Cell Biol 21: 6660 – 6667 in nutrient stress and autophagy. Cell 152: 290 – 303 82. Kiviluoto S, Vervliet T, Ivanova H, Decuypere JP, De Smedt H, Missiaen 100. Cheng J, Fujita A, Yamamoto H, Tatematsu T, Kakuta S, Obara K, 47. Wei Y, Zou Z, Becker N, Anderson M, Sumpter R, Xiao G, Kinch L, 66. Lee SR, Yang KS, Kwon J, Lee C, Jeong W, Rhee SG (2002) Reversible L, Bultynck G, Parys JB (2013) Regulation of inositol 1,4,5-trisphosphate Ohsumi Y, Fujimoto T (2014) Yeast and mammalian autophagosomes Koduru P, Christudass CS, Veltri RW et al (2013) EGFR-mediated Beclin inactivation of the tumor suppressor PTEN by H2O2. J Biol Chem 277: receptors during endoplasmic reticulum stress. Biochim Biophys Acta exhibit distinct phosphatidylinositol 3-phosphate asymmetries. Nat 1 phosphorylation in autophagy suppression, tumor progression, and 20336 – 20342 1833: 1612 – 1624 Commun 5: 3207 tumor chemoresistance. Cell 154: 1269 – 1284 67. Tang HW, Liao HM, Peng WH, Lin HR, Chen CH, Chen GC (2013) Atg9 83. Decuypere JP, Welkenhuyzen K, Luyten T, Ponsaerts R, Dewaele M, 101. Proikas-Cezanne T, Waddell S, Gaugel A, Frickey T, Lupas A, 48. Jutten B, Rouschop KM (2014) EGFR signaling and autophagy depen- interacts with dTRAF2/TRAF6 to regulate oxidative stress-induced JNK Molgó J, Agostinis P, Missiaen L, De Smedt H, Parys JB et al (2011) Ins Nordheim A (2004) WIPI-1 alpha (WIPI49), a member of the novel dence for growth, survival, and therapy resistance. Cell Cycle 13: activation and autophagy induction. Dev Cell 27: 489 – 503 (1,4,5)P3 receptor-mediated Ca2+ signaling and autophagy induction 7-bladed WIPI protein family, is aberrantly expressed in human 42 – 51 68. Adler V, Yin Z, Fuchs SY, Benezra M, Rosario L, Tew KD, Pincus MR, are interrelated. Autophagy 7: 1472 – 1489 cancer and is linked to starvation-induced autophagy. Oncogene 23: 49. Shen S, Niso-Santano M, Adjemian S, Takehara T, Malik SA, Minoux H, Sardana M, Henderson CJ, Wolf CR et al (1999) Regulation of JNK 84. Cárdenas C, Miller RA, Smith I, Bui T, Molgó J, Müller M, Vais H, 9314 – 9325 Souquere S, Mariño G, Lachkar S, Senovilla L et al (2012) Cytoplasmic signaling by GSTp. EMBO J 18: 1321 – 1334 Cheung KH, Yang J, Parker I et al (2010) Essential regulation of cell 102. Polson HE, de Lartigue J, Rigden DJ, Reedijk M, Urbe S, Clague MJ, STAT3 represses autophagy by inhibiting PKR activity. Mol Cell 48: 69. Chen K, Vita JA, Berk BC, Keaney JF Jr (2001) c-Jun N-terminal kinase bioenergetics by constitutive InsP3 receptor Ca2+ transfer to mitochon- Tooze SA (2010) Mammalian Atg18 (WIPI2) localizes to omega- 667 – 680 activation by hydrogen peroxide in endothelial cells involves dria. Cell 142: 270 – 283 some-anchored phagophores and positively regulates LC3 lipidation. 50. Rouschop KM, van den Beucken T, Dubois L, Niessen H, Bussink J, SRC-dependent epidermal growth factor receptor transactivation. J Biol 85. Axe EL, Walker SA, Manifava M, Chandra P, Roderick HL, Habermann A, Autophagy 6: 506 – 522 Savelkouls K, Keulers T, Mujcic H, Landuyt W, Voncken JW et al (2010) Chem 276: 16045 – 16050 Griffiths G, Ktistakis NT (2008) Autophagosome formation from 103. Simonsen A, Birkeland HC, Gillooly DJ, Mizushima N, Kuma A, Yoshi- The unfolded protein response protects human tumor cells during 70. Meng Y, Yong Y, Yang G, Ding H, Fan Z, Tang Y, Luo J, Ke ZJ (2013) membrane compartments enriched in phosphatidylinositol 3-phosphate mori T, Slagsvold T, Brech A, Stenmark H (2004) Alfy, a novel hypoxia through regulation of the autophagy genes MAP1LC3B and Autophagy alleviates neurodegeneration caused by mild impairment of and dynamically connected to the endoplasmic reticulum. J Cell Biol FYVE-domain-containing protein associated with protein granules and ATG5. J Clin Invest 120: 127 – 141 oxidative metabolism. J Neurochem 126: 805 – 818 182: 685 – 701 autophagic membranes. J Cell Sci 117: 4239 – 4251 51. Guo JY, Chen HY, Mathew R, Fan J, Strohecker AM, Karsli-Uzunbas G, 71. Sciarretta S, Zhai P, Shao D, Zablocki D, Nagarajan N, Terada LS, Volpe 86. Hayashi-Nishino M, Fujita N, Noda T, Yamaguchi A, Yoshimori T, 104. Filimonenko M, Isakson P, Finley KD, Anderson M, Jeong H, Melia TJ, Kamphorst JJ, Chen G, Lemons JM, Karantza V et al (2011) Activated M, Sadoshima J (2013) Activation of NADPH oxidase 4 in the Yamamoto A (2009) A subdomain of the endoplasmic reticulum forms Bartlett BJ, Myers KM, Birkeland HC, Lamark T et al (2010) The selec- Ras requires autophagy to maintain oxidative metabolism and tumori- endoplasmic reticulum promotes cardiomyocyte autophagy and a cradle for autophagosome formation. Nat Cell Biol 11: 1433 – 1437 tive macroautophagic degradation of aggregated proteins requires the genesis. Genes Dev 25: 460 – 470 survival during energy stress through the protein kinase RNA-acti- 87. Yla-Anttila P, Vihinen H, Jokitalo E, Eskelinen EL (2009) 3D tomography PI3P-binding protein Alfy. Mol Cell 38: 265 – 279 52. Lin SY, Li TY, Liu Q, Zhang C, Li X, Chen Y, Zhang SM, Lian G, Liu Q, vated-like endoplasmic reticulum kinase/eukaryotic initiation factor 2 reveals connections between the phagophore and endoplasmic reticu- 105. Clausen TH, Lamark T, Isakson P, Finley K, Larsen KB, Brech A, Øvervatn Ruan K et al (2012) GSK3-TIP60-ULK1 signaling pathway links growth alpha/activating transcription factor 4 pathway. Circ Res 113: lum. Autophagy 5: 1180 – 1185 A, Stenmark H, Bjørkøy G, Simonsen A et al (2010)p62/SQSTM1 and factor deprivation to autophagy. Science 336: 477 – 481 1253 – 1264 88. Moreau K, Ravikumar B, Renna M, Puri C, Rubinsztein DC (2011) Auto- ALFY interact to facilitate the formation of p62 bodies/ALIS and their 53. Into T, Inomata M, Takayama E, Takigawa T (2012) Autophagy in regu- 72. Zhang J, Kim J, Alexander A, Cai S, Tripathi DN, Dere R, Tee AR, phagosome precursor maturation requires homotypic fusion. Cell 146: degradation by autophagy. Autophagy 6: 330 – 344 lation of Toll-like receptor signaling. Cell Signal 24: 1150 – 1162 Tait-Mulder J, Di Nardo A, Han JM et al (2013) A tuberous sclerosis 303 – 317 106. Isakson P, Lystad AH, Breen K, Koster G, Stenmark H, Simonsen A 54.O’Neill LA, Bowie AG (2007) The family of five: TIR-domain-containing complex signalling node at the peroxisome regulates mTORC1 and 89. Nair U, Jotwani A, Geng J, Gammoh N, Richerson D, Yen WL, Griffith J, (2013) TRAF6 mediates ubiquitination of KIF23/MKLP1 and is required adaptors in Toll-like receptor signalling. Nat Rev Immunol 7: 353 – 364 autophagy in response to ROS. Nat Cell Biol 15: 1186 – 1196 Nag S, Wang K, Moss T et al (2011) SNARE proteins are required for for midbody ring degradation by selective autophagy. Autophagy 9: 55. Shi CS, Kehrl JH (2010) TRAF6 and A20 regulate lysine 63-linked ubiquiti- 73. Scherz-Shouval R, Shvets E, Fass E, Shorer H, Gil L, Elazar Z (2007) macroautophagy. Cell 146: 290 – 302 1955 – 1964 nation of Beclin-1 to control TLR4-induced autophagy. Sci Signal 3: ra42 Reactive oxygen species are essential for autophagy and specifically 90. Hailey DW, Rambold AS, Satpute-Krishnan P, Mitra K, Sougrat R, Kim 107. Filimonenko M, Stuffers S, Raiborg C, Yamamoto A, Malerød L, Fisher 56. Fujita K, Maeda D, Xiao Q, Srinivasula SM (2011) Nrf2-mediated induc- regulate the activity of Atg4. EMBO J 26: 1749 – 1760 PK, Lippincott-Schwartz J (2010) Mitochondria supply membranes for EM, Isaacs A, Brech A, Stenmark H, Simonsen A (2007) Functional tion of p62 controls Toll-like receptor-4-driven aggresome-like induced 74. Scherz-Shouval R, Elazar Z (2011) Regulation of autophagy by ROS: autophagosome biogenesis during starvation. Cell 141: 656 – 667 multivesicular bodies are required for autophagic clearance of protein structure formation and autophagic degradation. Proc Natl Acad Sci physiology and pathology. Trends Biochem Sci 36: 30 – 38 91. van der Vaart A, Griffith J, Reggiori F (2010) Exit from the Golgi is aggregates associated with neurodegenerative disease. J Cell Biol 179: USA108: 1427 – 1432 75. Thomas DD, Ridnour LA, Isenberg JS, Flores-Santana W, Switzer CH, required for the expansion of the autophagosomal phagophore in yeast 485 – 500 57. Gomes LC, Dikic I (2014) Autophagy in antimicrobial immunity. Mol Cell Donzelli S, Hussain P, Vecoli C, Paolocci N, Ambs S et al (2008) The Saccharomyces cerevisiae. Mol Biol Cell 21: 2270 – 2284 108. Isakson P, Holland P, Simonsen A (2013) The role of ALFY in selective 54: 224 – 233 chemical biology of nitric oxide: implications in cellular signaling. Free 92. Ge L, Melville D, Zhang M, Schekman R (2013) The ER-Golgi intermedi- autophagy. Cell Death Differ 20: 12 – 20 58. Deretic V, Saitoh T, Akira S (2013) Autophagy in infection, inflammation Radic Biol Med 45: 18 – 31 ate compartment is a key membrane source for the LC3 lipidation step 109. Itakura E, Kishi C, Inoue K, Mizushima N (2008) Beclin 1 forms two and immunity. Nat Rev Immunol 13: 722 – 737 76. Sarkar S, Korolchuk VI, Renna M, Imarisio S, Fleming A, Williams A, of autophagosome biogenesis. Elife 2:e00947 distinct phosphatidylinositol 3-kinase complexes with mammalian 59. Hardie DG, Ross FA, Hawley SA (2012) AMPK: a nutrient and energy Garcia-Arencibia M, Rose C, Luo S, Underwood BR et al (2011) 93. Hamasaki M, Furuta N, Matsuda A, Nezu A, Yamamoto A, Fujita N, Atg14 and UVRAG. Mol Biol Cell 19: 5360 – 5372 sensor that maintains energy homeostasis. Nat Rev Mol Cell Biol 13: Complex inhibitory effects of nitric oxide on autophagy. Mol Cell 43: Oomori H, Noda T, Haraguchi T, Hiraoka Y et al (2013) Autophago- 110. Fimia GM, Stoykova A, Romagnoli A, Giunta L, Di Bartolomeo S, 251 – 262 19 – 32 somes form at ER-mitochondria contact sites. Nature 495: 389 – 393 Nardacci R, Corazzari M, Fuoco C, Ucar A, Schwartz P et al (2007) 60. Shaw RJ, Bardeesy N, Manning BD, Lopez L, Kosmatka M, DePinho RA, 77. Tripathi DN, Chowdhury R, Trudel LJ, Tee AR, Slack RS, Walker CL, 94. Graef M, Friedman JR, Graham C, Babu M, Nunnari J (2013) ER exit Ambra1 regulates autophagy and development of the nervous system. Cantley LC (2004) The LKB1 tumor suppressor negatively regulates Wogan GN (2013) Reactive nitrogen species regulate autophagy sites are physical and functional core autophagosome biogenesis Nature 447: 1121 – 1125 mTOR signaling. Cancer Cell 6: 91 – 99 through ATM-AMPK-TSC2-mediated suppression of mTORC1. Proc Natl components. Mol Biol Cell 24: 2918 – 2931 111. Nazio F, Strappazzon F, Antonioli M, Bielli P, Cianfanelli V, Bordi M, 61. Inoki K, Ouyang H, Zhu T, Lindvall C, Wang Y, Zhang X, Yang Q, Acad Sci U S A 110:E2950 – E2957 95. Puri C, Renna M, Bento CF, Moreau K, Rubinsztein DC (2013) Diverse Gretzmeier C, Dengjel J, Piacentini M, Fimia GM et al (2013) mTOR Bennett C, Harada Y, Stankunas K et al (2006) TSC2 integrates Wnt 78. Ito C, Saito Y, Nozawa T, Fujii S, Sawa T, Inoue H, Matsunaga T, Khan autophagosome membrane sources coalesce in recycling endosomes. inhibits autophagy by controlling ULK1 ubiquitylation, self-association and energy signals via a coordinated phosphorylation by AMPK and S, Akashi S, Hashimoto R et al (2013) Endogenous nitrated nucleotide Cell 154: 1285 – 1299 and function through AMBRA1 and TRAF6. Nat Cell Biol 15: 406 – 416 GSK3 to regulate cell growth. Cell 126: 955 – 968 is a key mediator of autophagy and innate defense against bacteria. 96. Lamb CA, Yoshimori T, Tooze SA (2013) The autophagosome: origins 112. Di Bartolomeo S, Corazzari M, Nazio F, Oliverio S, Lisi G, Antonioli M, 62. Inoki K, Zhu T, Guan KL (2003) TSC2 mediates cellular energy response Mol Cell 52: 794 – 804 unknown, biogenesis complex. Nat Rev Mol Cell Biol 14: 759 – 774 Pagliarini V, Matteoni S, Fuoco C, Giunta L et al (2010) The dynamic to control cell growth and survival. Cell 115: 577 – 590 79. Williams A, Sarkar S, Cuddon P, Ttofi EK, Saiki S, Siddiqi FH, Jahreiss L, 97. Koyama-Honda I, Itakura E, Fujiwara TK, Mizushima N (2013) Temporal interaction of AMBRA1 with the dynein motor complex regulates 63. Gwinn DM, Shackelford DB, Egan DF, Mihaylova MM, Mery A, Vasquez Fleming A, Pask D, Goldsmith P et al (2008) Novel targets for Hunting- analysis of recruitment of mammalian ATG proteins to the autophago- mammalian autophagy. J Cell Biol 191: 155 – 168 DS, Turk BE, Shaw RJ (2008) AMPK phosphorylation of raptor mediates ton’s disease in an mTOR-independent autophagy pathway. Nat Chem some formation site. Autophagy 9: 1491 – 1499 113. Xia P, Wang S, Du Y, Zhao Z, Shi L, Sun L, Huang G, Ye B, Li C, Dai Z a metabolic checkpoint. Mol Cell 30: 214 – 226 Biol 4: 295 – 305 98. Suzuki K, Kubota Y, Sekito T, Ohsumi Y (2007) Hierarchy of Atg proteins in et al (2013) WASH inhibits autophagy through suppression of Beclin 1 64. Kim J, Kundu M, Viollet B, Guan KL (2011) AMPK and mTOR regulate 80. Ganley IG, Wong PM, Gammoh N, Jiang X (2011) Distinct autophagoso- pre-autophagosomal structure organization. Genes Cells 12: 209 – 218 ubiquitination. EMBO J 32: 2685 – 2696 autophagy through direct phosphorylation of Ulk1. Nat Cell Biol 13: mal-lysosomal fusion mechanism revealed by thapsigargin-induced 99. Kihara A, Noda T, Ishihara N, Ohsumi Y (2001) Two distinct Vps34 114. Papinski D, Schuschnig M, Reiter W, Wilhelm L, Barnes CA, Maiolica A, 132 – 141 autophagy arrest. Mol Cell 42: 731 – 743 phosphatidylinositol 3-kinase complexes function in autophagy and Hansmann I, Pfaffenwimmer T, Kijanska M, Stoffel I et al (2014) Early

850 EMBO reports Vol 15 | No 8 | 2014 ª 2014 The Authors ª 2014 The Authors EMBO reports Vol 15 | No 8 | 2014 851 EMBO reports From signaling to autophagosome biogenesis Adi Abada & Zvulun Elazar Adi Abada & Zvulun Elazar From signaling to autophagosome biogenesis EMBO reports

850 EMBO reports Vol 15 | No 8 | 2014 ª 2014 The Authors ª 2014 The Authors EMBO reports Vol 15 | No 8 | 2014 851 EMBO reports From signaling to autophagosome biogenesis Adi Abada & Zvulun Elazar

steps in autophagy depend on direct phosphorylation of atg9 by the 131. Shirahama-Noda K, Kira S, Yoshimori T, Noda T (2013) TRAPPIII is atg1 kinase. Mol Cell 53: 471 – 483 responsible for vesicular transport from early endosomes to Golgi, facil- 115. Bejarano E, Yuste A, Patel B, Stout RF Jr, Spray DC, Cuervo AM (2014) itating Atg9 cycling in autophagy. J Cell Sci 126: 4963 – 4973 Connexins modulate autophagosome biogenesis. Nat Cell Biol 16: 132. Young AR, Chan EY, Hu XW, Kochl R, Crawshaw SG, High S, Hailey DW, 401 – 414 Lippincott-Schwartz J, Tooze SA (2006) Starvation and ULK1-dependent 116. Kaiser SE, Mao K, Taherbhoy AM, Yu S, Olszewski JL, Duda DM, Kurinov cycling of mammalian Atg9 between the TGN and endosomes. J Cell Sci I, Deng A, Fenn TD, Klionsky DJ et al (2012) Noncanonical E2 recruit- 119: 3888 – 3900 ment by the autophagy E1 revealed by Atg7-Atg3 and Atg7-Atg10 133. Yang A, Li Y, Pantoom S, Triola G, Wu YW (2013) Semisynthetic lipidat- structures. Nat Struct Mol Biol 19: 1242 – 1249 ed LC3 protein mediates membrane fusion. ChemBioChem 14: 117. Mizushima N, Noda T, Ohsumi Y (1999) Apg16p is required for the 1296 – 1300 function of the Apg12p-Apg5p conjugate in the yeast autophagy path- 134. Nakatogawa H, Ichimura Y, Ohsumi Y (2007) Atg8, a ubiquitin-like way. EMBO J 18: 3888 – 3896 protein required for autophagosome formation, mediates membrane 118. Kuma A, Mizushima N, Ishihara N, Ohsumi Y (2002) Formation of the tethering and hemifusion. Cell 130: 165 – 178 approximately 350-kDa Apg12-Apg5.Apg16 multimeric complex, medi- 135. Xie Z, Nair U, Klionsky DJ (2008) Atg8 controls phagophore expansion ated by Apg16 oligomerization, is essential for autophagy in yeast. during autophagosome formation. Mol Biol Cell 19: 3290 – 3298 J Biol Chem 277: 18619 – 18625 136. Weidberg H, Shvets E, Shpilka T, Shimron F, Shinder V, Elazar Z (2010) 119. Mizushima N, Kuma A, Kobayashi Y, Yamamoto A, Matsubae M, Takao LC3 and GATE-16/GABARAP subfamilies are both essential yet act differ- T, Natsume T, Ohsumi Y, Yoshimori T (2003) Mouse Apg16L, a novel ently in autophagosome biogenesis. EMBO J 29: 1792 – 1802 WD-repeat protein, targets to the autophagic isolation membrane with 137. Weidberg H, Shpilka T, Shvets E, Abada A, Shimron F, Elazar Z (2011) the Apg12-Apg5 conjugate. J Cell Sci 116: 1679 – 1688 LC3 and GATE-16 N termini mediate membrane fusion processes 120. Shpilka T, Weidberg H, Pietrokovski S, Elazar Z (2011) Atg8: an auto- required for autophagosome biogenesis. Dev Cell 20: 444 – 454 phagy-related ubiquitin-like protein family. Genome Biol 12: 226 138. Manil-Segalen M, Lefebvre C, Jenzer C, Trichet M, Boulogne C, Satiat- 121. Dooley HC, Razi M, Polson HE, Girardin SE, Wilson MI, Tooze SA Jeunemaitre B, Legouis R (2014) The C. elegans LC3 acts downstream of (2014) WIPI2 links LC3 conjugation with PI3P, autophagosome GABARAP to degrade autophagosomes by interacting with the HOPS formation, and pathogen clearance by recruiting Atg12–5-16L1. subunit VPS39. Dev Cell 28: 43 – 55 Mol Cell, DOI 10.1016/j.molcel.2014.05.021 139. Nakatogawa H, Ishii J, Asai E, Ohsumi Y (2012) Atg4 recycles inappro- 122. Romanov J, Walczak M, Ibiricu I, Schuchner S, Ogris E, Kraft C, Martens priately lipidated Atg8 to promote autophagosome biogenesis. S(2012) Mechanism and functions of membrane binding by the Autophagy 8: 177 – 186 Atg5-Atg12/Atg16 complex during autophagosome formation. EMBO J 140. Betin VM, Singleton BK, Parsons SF, Anstee DJ, Lane JD (2013) Auto- 31: 4304 – 4317 phagy facilitates organelle clearance during differentiation of human 123. Nath S, Dancourt J, Shteyn V, Puente G, Fong WM, Nag S, Bewersdorf J, erythroblasts: evidence for a role for ATG4 paralogs during autophago- Yamamoto A, Antonny B, Melia TJ (2014) Lipidation of the LC3/GABARAP some maturation. Autophagy 9: 881 – 893 family of autophagy proteins relies on a membrane-curvature-sensing 141. Ma J, Becker C, Reyes C, Underhill DM (2014) Cutting edge: FYCO1 domain in Atg3. Nat Cell Biol 16: 415 – 424 recruitment to dectin-1 phagosomes is accelerated by light chain 3 124. Kaufmann A, Beier V, Franquelim HG, Wollert T (2014) Molecular protein and regulates phagosome maturation and reactive oxygen mechanism of autophagic membrane-scaffold assembly and disassem- production. J Immunol 192: 1356 – 1360 bly. Cell 156: 469 – 481 142. Ao X, Zou L, Wu Y (2014) Regulation of autophagy by the Rab GTPase 125. Knævelsrud H, Søreng K, Raiborg C, Håberg K, Rasmuson F, Brech A, network. Cell Death Differ 21: 348 – 358 Liestøl K, Rusten TE, Stenmark H, Neufeld TP et al (2013) Membrane 143. Lu Y, Zhang Z, Sun D, Sweeney ST, Gao FB (2013) Syntaxin 13,a remodeling by the PX-BAR protein SNX18 promotes autophagosome genetic modifier of mutant CHMP2B in frontotemporal dementia, is formation. J Cell Biol 202: 331 – 349 required for autophagosome maturation. Mol Cell 52: 264 – 271 126. Longatti A, Lamb CA, Razi M, Yoshimura S, Barr FA, Tooze SA (2012) 144. Itakura E, Kishi-Itakura C, Mizushima N (2012) The hairpin-type TBC1D14 regulates autophagosome formation via Rab11- and tail-anchored SNARE syntaxin 17 targets to autophagosomes for fusion ULK1-positive recycling endosomes. J Cell Biol 197: 659 – 675 with endosomes/lysosomes. Cell 151: 1256 – 1269 127. Nishimura T, Kaizuka T, Cadwell K, Sahani MH, Saitoh T, Akira S, Virgin 145. Jiang P, Nishimura T, Sakamaki Y, Itakura E, Hatta T, Natsume T, Mizushi- HW, Mizushima N (2013) FIP200 regulates targeting of Atg16L1 to the ma N (2014) The HOPS complex mediates autophagosome-lysosome isolation membrane. EMBO Rep 14: 284 – 291 fusion through interaction with syntaxin 17. Mol Biol Cell 25: 1327 – 1337 128. Yamamoto H, Kakuta S, Watanabe TM, Kitamura A, Sekito T, 146. Takáts S, Pircs K, Nagy P, Varga Á, Kárpáti M, Hegedus K, Kramer H, Kondo-Kakuta C, Ichikawa R, Kinjo M, Ohsumi Y (2012) Atg9 vesicles Kovács AL, Sass M, Juhász G (2014) Interaction of the HOPS complex are an important membrane source during early steps of autophago- with Syntaxin 17 mediates autophagosome clearance in Drosophila. some formation. J Cell Biol 198: 219 – 233 Mol Biol Cell 25: 1338 – 1354 129. Mari M, Griffith J, Rieter E, Krishnappa L, Klionsky DJ, Reggiori F (2010) 147. Chen D, Fan W, Lu Y, Ding X, Chen S, Zhong Q (2012) A mammalian An Atg9-containing compartment that functions in the early steps of autophagosome maturation mechanism mediated by TECPR1 and the autophagosome biogenesis. J Cell Biol 190: 1005 – 1022 Atg12-Atg5 conjugate. Mol Cell 45: 629 – 641 130. Orsi A, Razi M, Dooley HC, Robinson D, Weston AE, Collinson LM, Tooze 148. Ogawa M, Yoshikawa Y, Kobayashi T, Mimuro H, Fukumatsu M, Kiga K, SA (2012) Dynamic and transient interactions of Atg9 with autophago- Piao Z, Ashida H, Yoshida M, Kakuta S et al (2011) A Tecpr1-dependent somes, but not membrane integration, are required for autophagy. Mol selective autophagy pathway targets bacterial pathogens. Cell Host Biol Cell 23: 1860 – 1873 Microbe 9: 376 – 389

852 EMBO reports Vol 15 | No 8 | 2014 ª 2014 The Authors Scientific Report

PI3P phosphatase activity is required for autophagosome maturation and autolysosome formation

Yanwei Wu1,2, Shiya Cheng2, Hongyu Zhao3, Wei Zou2, Sawako Yoshina4, Shohei Mitani4, Hong Zhang3 & Xiaochen Wang2,*

Abstract formation [2]. The Atg18/Atg2 and Atg1 complexes regulate retrieval of Atg9, which supplies membrane for autophagosome Autophagosome formation is promoted by the PI3 kinase complex formation [2]. Except for Atg8/LC3-II, Atg proteins dissociate from and negatively regulated by myotubularin phosphatases, indicat- sealed autophagosomes, which then fuse directly with vacuoles in ing that regulation of local phosphatidylinositol 3-phosphate yeast, but interact with endocytic compartments in higher eukary- (PtdIns3P) levels is important for this early phase of autophagy. otes before fusing with lysosomes [3]. Consistent with this, essential Here, we show that the Caenorhabditis elegans myotubularin phos- endocytic regulators such as Rab GTPases and ESCRT proteins phatase MTM-3 catalyzes PtdIns3P turnover late in autophagy. play important roles in autophagosome maturation, while Beclin1- MTM-3 acts downstream of the ATG-2/EPG-6 complex and binding proteins UVRAG and Rubicon regulate both endocytic upstream of EPG-5 to promote autophagosome maturation into transport and autophagy [4]. autolysosomes. MTM-3 is recruited to autophagosomes by Phosphatidylinositol 3-phosphate (PtdIns3P) regulates various PtdIns3P, and loss of MTM-3 causes increased autophagic associa- cellular processes by recruiting specific protein effectors to target tion of ATG-18 in a PtdIns3P-dependent manner. Our data reveal membranes. Upon autophagy induction, PtdIns3P is produced at the critical roles of PtdIns3P turnover in autophagosome maturation PAS in yeast or a subdomain of the ER or ER–mitochondria contact and/or autolysosome formation. sites in higher eukaryotes and promotes autophagosome biogenesis by recruiting PtdIns3P-binding effectors such as Atg18/WIPI and Keywords autolysosome; autophagy; Caenorhabditis elegans; MTM-3; DFCP1 [2,5–7]. In addition to the PI3 kinase Vps34, which generates PtdIns3P PtdIns3P, recent studies indicate that myotubularin phosphatases, Subject Categories Autophagy & Cell Death; Membrane & Intracellular which convert PtdIns3P to PI, also play a role in autophagy. There Transport is evidence that both Jumpy/MTMR14 and MTMR3 negatively regu- DOI 10.15252/embr.201438618 | Received 12 February 2014 | Revised 10 July late early events in the mammalian autophagy pathway [8,9]. These 2014 | Accepted 11 July 2014 | Published online 14 August 2014 studies suggest that at the initiation step, the local PtdIns3P level is EMBO Reports (2014) 15: 973–981 tightly controlled by both PI3 kinase and phosphatase to ensure that autophagy occurs at an appropriate level. PtdIn3P regulation in autophagosome initiation has been extensively studied, but little is Introduction known about how it is controlled in autophagosome maturation into autolysosomes. Ymr1, the only myotubularin phosphatase in yeast, In the process of macroautophagy (hereafter called autophagy), promotes dissociation of Atg proteins from autophagosomes and cytosolic materials are sequestered within double-membrane auto- thus positively regulates autophagy progression [10]. It is unclear phagosomes, which mature to fuse with lysosomes where cargoes whether the distinct effect of human and yeast myotubularin phos- are degraded [1]. Autophagosome biogenesis is controlled by four phatases on autophagy is due to intrinsic species-specific differences protein complexes. The Atg1/Atg13 complex and the class III PI3 in autophagy, or because PtdIns3P turnover has opposing effects if kinase Vps34 complex are required for induction and nucleation of it occurs at different stages of autophagy. the isolation membrane or phagophore, while two ubiquitin-like Caenorhabditis elegans contains three active myotubularin phos- conjugation systems (Atg5-Atg12, Atg16, and Atg8-PE) act sequen- phatases (MTM-1, 3, and 6) and two inactive members (5 and 9) tially to elongate the isolation membrane, leading to autophagosome [11]. MTM-1 is important for apoptotic cell clearance, while MTM-6

1 College of Biological Sciences, China Agriculture University, Beijing, China 2 National Institute of Biological Sciences, Beijing, China 3 State Key Laboratory of Biomacromolecules, Institute of Biophysics, Chinese Academy of Sciences, Beijing, China 4 Department of Physiology, School of Medicine and Institute for Integrated Medical Sciences, Tokyo Women’s Medical University, Shinjuku-ku, Tokyo, Japan *Corresponding author. Tel: +86 10 80726688 8535; E-mail: [email protected]

ª 2014 The Authors EMBO reports Vol 15 | No 9 | 2014 973 EMBO reports MTM-3 promotes autolysosome formation Yanwei Wu et al Yanwei Wu et al MTM-3 promotes autolysosome formation EMBO reports

worms (Supplementary Fig S4C–H0). Thus, endocytosis is not obvi- We examined autophagy-related processes in mutants that are defec- ously affected in mtm-3 mutants. tive in mtm-1, 3, 5, 6, and 9 (Supplementary Fig S1A). The C. elegans p62 homolog, SQST-1 (SeQueSTosome-related protein), associates MTM-3 is widely distributed and can be recruited to with various protein aggregates and is removed by autophagy during autophagic structures embryogenesis [16]. SQST-1 is weakly expressed and diffusely local- EFE’ F’ ized in the cytoplasm of wild-type embryos, but forms numerous We generated a GFP::MTM-3 reporter driven by the mtm-3 promoter

aggregates in autophagy mutants [16]. We found that SQST-1 aggre- (Pmtm-3GFP::MTM-3), which efficiently rescued the autophagy gates were not detected in wild type but accumulated significantly in phenotype in mtm-3 mutants (Supplementary Fig S3B and H). GFP:: tm4475, a deletion mutant of mtm-3 that removes all but the first 66 MTM-3 was diffuse in the cytoplasm of almost all cells during

amino acids of the MTM-3 protein (Fig 1A-B0; Supplementary Figs embryogenesis (Fig 2A; Supplementary Fig S3K–L0). In larvae and S1A and S2A and B0; Supplementary Methods). Loss of mtm-1, 5, 6, adults, MTM-3 was found in various cell types including pharyngeal or 9 did not cause accumulation of SQST-1 (Supplementary Fig S1A– cells, vulva muscle cells, intestine and cells in the tail region F; Supplementary Methods). Similarly, aggregates of SEPA-1, an (Supplementary Fig S3M–P). GFP::MTM-3 expression was unaltered GH autophagy substrate that mediates aggregation and autophagic in autophagy mutants, suggesting that MTM-3 itself is not removed degradation of PGL granules (somatic PGL-1- and PGL-3-positive by autophagy (Supplementary Fig S3Q–T). To examine whether granules) in C. elegans embryos, were not observed in late embry- MTM-3 is recruited to autophagosomes or their precursors, we onic stages in wild-type or mtm-1, 5, 6, and 9 mutants, but persisted constructed a GFP reporter of MTM-3(C459S), a catalytically inac-

in mtm-3(tm4475) embryos (Fig 1C and D0; Supplementary Figs tive substrate-trapping form of MTM-3 (Fig 2J). GFP::MTM-3 S1G–K and S2C and D0) [17]. Other autophagy substrates, including (C459S) was expressed at a higher level than the wild-type protein C17E4.2 and C33D9.6, also accumulated ectopically in mtm-3 and formed punctate structures that overlapped well with endoge-

embryos (Supplementary Fig S2E-H0) [18]. Thus, loss of mtm-3, but nous LGG-1 puncta (Fig 2B, D-D00 and I; Supplementary Fig S3J). not other myotubularin phosphatases, affects degradation of various Loss of lgg-1, which disrupts autophagosome biogenesis, but not autophagic substrates. The C. elegans Atg8/LC3 homolog, LGG-1, rab-5 or rab-7, which affect endosome formation, significantly which is essential for autophagosome biogenesis, associates with reduced the number of GFP::MTM-3(C459S) puncta, suggesting that autophagosome membranes and their precursors. In wild type, LGG- MTM-3(C459S) localizes to autophagic structures (Fig 2H). Loss of 1 puncta are mostly seen around the 100–200-cell stage and decrease VPS-34, the PI3 kinase that produces PtdIns3P on autophagic struc-

as embryos develop [16] (Supplementary Fig S1E and E0). In mtm-3 tures, dramatically reduced GFP::MTM-3(C459S) puncta and their mutants, however, we observed significantly increased LGG-1 co-localization with LGG-1 (Fig 2G–I). Moreover, GFP::MTM-3 Figure 1. Loss of mtm-3 impairs autophagy. puncta in late embryonic stages and accumulation of both LGG-1-I (C459S, del PH-G), which lacks the phosphoinositide-binding PH- A-F0 Confocal fluorescent images of wild-type (A, C, E) and mtm-3(tm4475) (B, D, F) embryos at the 1.5-fold stage stained by DAPI (A–F) and anti-SQST-1 (A0,B0), anti-SEPA-1 (C0,D0) or anti-LGG-1 (E0,F0) antibodies. Scale bars: 5 lm. and LGG-1-II (lipid-conjugated form of LGG-1), suggesting a defect in GRAM domain of MTM-3, failed to form puncta (Fig 2C) [20]. These G Western blot analysis of LGG-1-I and LGG-1–II (lipidated) in wild type, epg-6, and mtm-3. autophagy (Fig 1F and G). LGG-1 puncta were normal in mtm-1, 5, data suggest that MTM-3 is recruited to autophagosomes by H mtm-3 and epg-6 mutant L1 larvae have shortened mean and maximum life span in the absence of food. At least 300 animals were scored each day. 6, or 9 mutant embryos (Supplementary Fig S1L–P). In addition to PtdIns3P. Mutation of atg-2, which blocks completion of autophago- Source data are available online for this figure. degrading protein aggregates, autophagy activity is required for the somes, did not affect MTM-3(C459S) puncta or their co-localization

survival of newly hatched L1 larvae in the absence of food [19]. Like with LGG-1 (Fig 2F–F00, H and I). other autophagy-defective mutants, loss of mtm-3 severely affected [21]. To determine where MTM-3 acts in this pathway, we exam- and L; Supplementary Fig S5D–D000) [23]. The PtdIns3P-binding the survival of starved L1 larvae, reducing the mean life span to MTM-3 acts downstream of ATG-9, ATG-18, and ATG-2 in the ined the morphology and distribution of PGL granules (detected protein ATG-18/WIPI1/2 acts early in autophagosome formation 3.5 days from 18.0 days in wild type (Fig 1H). These data further aggrephagy pathway by anti-SEPA-1 antibodies) and LGG-1 puncta in mtm-3, atg, or and loss of its function caused accumulation of PGL granules and suggest that loss of mtm-3 function impairs autophagy. Recombinant epg single mutants and atg/epg;mtm-3 double mutants. In mtm-3 LGG-1 puncta, which were close but did not overlap (Fig 3F and L;

and PtdIns(3,5)P2 in vitro (Fig 2J; Supplementary Methods). protein aggregates (aggrephagy) during C. elegans embryogenesis. and dispersed in the cytoplasm, with 31.6% of PGL granules in epg-8;mtm-3, mtm-3;atg-9, or mtm-3;atg-18 double mutants Expressing wild-type but not catalytically inactive MTM-3 comple- The C. elegans Atg1 kinase complex UNC-51-EPG-1-EPG-9 acts overlapping with LGG-1 (Fig 3A, A0 and L; Supplementary Fig resembled those in epg-8, atg-9, or atg-18 single mutants, suggest- tely rescued the autophagy phenotypes of mtm-3(lf), indicating that first, followed by the PI3 kinase complex EPG-8-VPS-34-BEC-1. S5A–A000). epg-8 encodes a highly divergent functional homolog of ing that epg-8, atg-9, and atg-18 are epistatic to mtm-3 in the MTM-3 acts as a lipid phosphatase to promote autophagy (Fig 2J; ATG-18/WIPI1/2 acts earlier than the ATG-2-EPG-6 complex, Atg14, a component of the Vps34 PI3K complex [21]. epg-8 aggrephagy pathway (Fig 3C, E, G and L; Supplementary Fig

defects in mtm-3 mutants (Supplementary Fig S3D and H), suggest- to regulate autolysosome formation, while CUP-5/TRPML1 is S5B–B000) [22]. Similarly, in atg-9 mutants, the fewer and bigger from omegasomes. epg-8, atg-9, and atg-18 are epistatic to atg-2 ing that human MTMR3 can substitute for MTM-3 in C. elegans required for degradation of autophagic cargo in autolysosomes LGG-1 puncta were mostly separable from PGL granules (Fig 3D and epg-6 in the aggrephagy pathway [23]. atg-2 mutants

974 EMBO reports Vol 15 | No 9 | 2014 ª 2014 The Authors ª 2014 The Authors EMBO reports Vol 15 | No 9 | 2014 975 EMBO reports MTM-3 promotes autolysosome formation Yanwei Wu et al Yanwei Wu et al MTM-3 promotes autolysosome formation EMBO reports

and 9 regulate endocytic transport [12–15]. The cellular functions of autophagy. Overexpression of MTM-3 did not cause LGG-1 accumu- A A’ B B’ MTM-3 and 5 are unclear. Here, we identified C. elegans MTM-3 as lation, consistent with a positive role of it in autophagy (Supplemen- a positive regulator of autophagy. Our data indicate that PtdIns3P tary Fig S3E–I). turnover catalyzed by myotubularin phosphatase plays important mtm-6 and 9 are required for endocytosis by the C. elegans roles in autophagosome maturation and/or autolysosome formation. macrophage-like coelomocytes [12], but in mtm-3(tm4475) mutants, GFP secreted from body wall muscle cells was efficiently endocytosed and transported by coelomocytes, which contained Results and Discussion normal endosomes and lysosomes as in wild type (Supplementary C C’ D D’ Fig S4A and B0). Uptake, trafficking and degradation of the Loss of MTM-3 causes defects in autophagy-related processes C. elegans yolk protein VIT-2 were also normal in mtm-3(tm4475) worms (Supplementary Fig S4C–H0). Thus, endocytosis is not obvi- We examined autophagy-related processes in mutants that are defec- ously affected in mtm-3 mutants. tive in mtm-1, 3, 5, 6, and 9 (Supplementary Fig S1A). The C. elegans p62 homolog, SQST-1 (SeQueSTosome-related protein), associates MTM-3 is widely distributed and can be recruited to with various protein aggregates and is removed by autophagy during autophagic structures embryogenesis [16]. SQST-1 is weakly expressed and diffusely local- EFE’ F’ ized in the cytoplasm of wild-type embryos, but forms numerous We generated a GFP::MTM-3 reporter driven by the mtm-3 promoter aggregates in autophagy mutants [16]. We found that SQST-1 aggre- (Pmtm-3GFP::MTM-3), which efficiently rescued the autophagy gates were not detected in wild type but accumulated significantly in phenotype in mtm-3 mutants (Supplementary Fig S3B and H). GFP:: tm4475, a deletion mutant of mtm-3 that removes all but the first 66 MTM-3 was diffuse in the cytoplasm of almost all cells during amino acids of the MTM-3 protein (Fig 1A-B0; Supplementary Figs embryogenesis (Fig 2A; Supplementary Fig S3K–L0). In larvae and S1A and S2A and B0; Supplementary Methods). Loss of mtm-1, 5, 6, adults, MTM-3 was found in various cell types including pharyngeal or 9 did not cause accumulation of SQST-1 (Supplementary Fig S1A– cells, vulva muscle cells, intestine and cells in the tail region F; Supplementary Methods). Similarly, aggregates of SEPA-1, an (Supplementary Fig S3M–P). GFP::MTM-3 expression was unaltered GH autophagy substrate that mediates aggregation and autophagic in autophagy mutants, suggesting that MTM-3 itself is not removed degradation of PGL granules (somatic PGL-1- and PGL-3-positive by autophagy (Supplementary Fig S3Q–T). To examine whether granules) in C. elegans embryos, were not observed in late embry- MTM-3 is recruited to autophagosomes or their precursors, we onic stages in wild-type or mtm-1, 5, 6, and 9 mutants, but persisted constructed a GFP reporter of MTM-3(C459S), a catalytically inac- in mtm-3(tm4475) embryos (Fig 1C and D0; Supplementary Figs tive substrate-trapping form of MTM-3 (Fig 2J). GFP::MTM-3 S1G–K and S2C and D0) [17]. Other autophagy substrates, including (C459S) was expressed at a higher level than the wild-type protein C17E4.2 and C33D9.6, also accumulated ectopically in mtm-3 and formed punctate structures that overlapped well with endoge- embryos (Supplementary Fig S2E-H0) [18]. Thus, loss of mtm-3, but nous LGG-1 puncta (Fig 2B, D-D00 and I; Supplementary Fig S3J). not other myotubularin phosphatases, affects degradation of various Loss of lgg-1, which disrupts autophagosome biogenesis, but not autophagic substrates. The C. elegans Atg8/LC3 homolog, LGG-1, rab-5 or rab-7, which affect endosome formation, significantly which is essential for autophagosome biogenesis, associates with reduced the number of GFP::MTM-3(C459S) puncta, suggesting that autophagosome membranes and their precursors. In wild type, LGG- MTM-3(C459S) localizes to autophagic structures (Fig 2H). Loss of 1 puncta are mostly seen around the 100–200-cell stage and decrease VPS-34, the PI3 kinase that produces PtdIns3P on autophagic struc- as embryos develop [16] (Supplementary Fig S1E and E0). In mtm-3 tures, dramatically reduced GFP::MTM-3(C459S) puncta and their mutants, however, we observed significantly increased LGG-1 co-localization with LGG-1 (Fig 2G–I). Moreover, GFP::MTM-3 Figure 1. Loss of mtm-3 impairs autophagy. puncta in late embryonic stages and accumulation of both LGG-1-I (C459S, del PH-G), which lacks the phosphoinositide-binding PH- A-F0 Confocal fluorescent images of wild-type (A, C, E) and mtm-3(tm4475) (B, D, F) embryos at the 1.5-fold stage stained by DAPI (A–F) and anti-SQST-1 (A0,B0), anti-SEPA-1 (C0,D0) or anti-LGG-1 (E0,F0) antibodies. Scale bars: 5 lm. and LGG-1-II (lipid-conjugated form of LGG-1), suggesting a defect in GRAM domain of MTM-3, failed to form puncta (Fig 2C) [20]. These G Western blot analysis of LGG-1-I and LGG-1–II (lipidated) in wild type, epg-6, and mtm-3. autophagy (Fig 1F and G). LGG-1 puncta were normal in mtm-1, 5, data suggest that MTM-3 is recruited to autophagosomes by H mtm-3 and epg-6 mutant L1 larvae have shortened mean and maximum life span in the absence of food. At least 300 animals were scored each day. 6, or 9 mutant embryos (Supplementary Fig S1L–P). In addition to PtdIns3P. Mutation of atg-2, which blocks completion of autophago- Source data are available online for this figure. degrading protein aggregates, autophagy activity is required for the somes, did not affect MTM-3(C459S) puncta or their co-localization survival of newly hatched L1 larvae in the absence of food [19]. Like with LGG-1 (Fig 2F–F00, H and I). other autophagy-defective mutants, loss of mtm-3 severely affected [21]. To determine where MTM-3 acts in this pathway, we exam- and L; Supplementary Fig S5D–D000) [23]. The PtdIns3P-binding the survival of starved L1 larvae, reducing the mean life span to MTM-3 acts downstream of ATG-9, ATG-18, and ATG-2 in the ined the morphology and distribution of PGL granules (detected protein ATG-18/WIPI1/2 acts early in autophagosome formation 3.5 days from 18.0 days in wild type (Fig 1H). These data further aggrephagy pathway by anti-SEPA-1 antibodies) and LGG-1 puncta in mtm-3, atg, or and loss of its function caused accumulation of PGL granules and suggest that loss of mtm-3 function impairs autophagy. Recombinant epg single mutants and atg/epg;mtm-3 double mutants. In mtm-3 LGG-1 puncta, which were close but did not overlap (Fig 3F and L;

MTM-3 displayed significant phosphatase activity toward PtdIns3P Autophagy proteins act in a stepwise pathway to remove various (tm4475) embryos, PGL granules and LGG-1 puncta are spherical Supplementary Fig S5F–F000) [23]. PGL granules and LGG-1 puncta and PtdIns(3,5)P2 in vitro (Fig 2J; Supplementary Methods). protein aggregates (aggrephagy) during C. elegans embryogenesis. and dispersed in the cytoplasm, with 31.6% of PGL granules in epg-8;mtm-3, mtm-3;atg-9, or mtm-3;atg-18 double mutants Expressing wild-type but not catalytically inactive MTM-3 comple- The C. elegans Atg1 kinase complex UNC-51-EPG-1-EPG-9 acts overlapping with LGG-1 (Fig 3A, A0 and L; Supplementary Fig resembled those in epg-8, atg-9, or atg-18 single mutants, suggest- tely rescued the autophagy phenotypes of mtm-3(lf), indicating that first, followed by the PI3 kinase complex EPG-8-VPS-34-BEC-1. S5A–A000). epg-8 encodes a highly divergent functional homolog of ing that epg-8, atg-9, and atg-18 are epistatic to mtm-3 in the MTM-3 acts as a lipid phosphatase to promote autophagy (Fig 2J; ATG-18/WIPI1/2 acts earlier than the ATG-2-EPG-6 complex, Atg14, a component of the Vps34 PI3K complex [21]. epg-8 aggrephagy pathway (Fig 3C, E, G and L; Supplementary Fig

Supplementary Fig S3A–C and H). Expression of human MTMR3 which regulates progression from isolation membrane to auto- mutants contained spherical PGL granules and LGG-1 puncta S5C–C000,E–E000 and G–G000). The ATG-2 protein forms a complex driven by the mtm-3 promoter efficiently rescued the autophagy phagosome. EPG-5/mEPG-5 functions downstream of ATG-2-EPG-6 which were mostly separated (Fig 3B and L; Supplementary Fig with EPG-6/WIPI3/4 to regulate formation of autophagosomes defects in mtm-3 mutants (Supplementary Fig S3D and H), suggest- to regulate autolysosome formation, while CUP-5/TRPML1 is S5B–B000) [22]. Similarly, in atg-9 mutants, the fewer and bigger from omegasomes. epg-8, atg-9, and atg-18 are epistatic to atg-2 ing that human MTMR3 can substitute for MTM-3 in C. elegans required for degradation of autophagic cargo in autolysosomes LGG-1 puncta were mostly separable from PGL granules (Fig 3D and epg-6 in the aggrephagy pathway [23]. atg-2 mutants

A BC Figure 2. GFP::MTM-3(C459S) associates with autophagic structures. ◀ A–C Fluorescent images of wild-type embryos (1.5-fold stage) expressing GFP::MTM-3 (A), GFP::MTM-3(C459S) (B), or GFP::MTM-3(C459S, del PH-G) (C). D-G″ Confocal fluorescent images of wild-type (D-D″), mtm-3 (E-E″), atg-2 (F-F″), and vps-34 (G-G″) embryos at the 200-cell stage expressing GFP::MTM-3(C459S) and stained by anti-LGG-1 antibody. Insets show magnified views. Arrows indicate overlapping GFP and LGG-1 puncta. Scale bars: 5 lm. H, I Number of GFP::MTM-3(C459S)-positive puncta and co-localization of GFP::MTM-3(C459S)- and LGG-1-positive puncta in various strains [mtm-3(tm4475), atg-2 (bp576), lgg-1(bp500), vps-34(h797)]. At least 10 (I) and 15 (H) embryos were scored per strain. Data are shown as mean SD. Data from different mutant � backgrounds were compared with wild type. **P < 0.0001; N.S.: not statistically different.

J MTM-3 but not MTM-3(C459S) exhibits phosphatase activity toward PtdIns3P and PtdIns(3,5)P2 in vitro. D D’ D’’ Source data are available online for this figure.

contained numerous PGL granules and LGG-1 puncta which were autolysosome formation was examined using GFP::LGG-1 and enlarged and irregular in shape; over 60% of the PGL granules NUC-1::mCHERRY (Supplementary Fig S7). To corroborate this, co-localized with LGG-1 puncta (Fig 3H and L; Supplementary Fig we examined autophagosomes by transmission electron microscopy

S6A–A000) [23]. The morphology and distribution of PGL granules (TEM). In mtm-3(tm4475) mutant embryos, double-membrane and LGG-1 puncta in mtm-3;atg-2 double mutants resembled that autophagosomes, but not autolysosomes, were readily observed, E E’ E’’ in atg-2 single mutants (Fig 3A0, H, I and L; Supplementary Fig whereas none of these autophagic structures were easily seen in S6B–B000), suggesting that atg-2 is epistatic to mtm-3 in the aggre- wild type (Fig 4J–L). These data indicate that loss of mtm-3 function phagy pathway. epg-5 encodes a metazoan-specific autophagy impairs autolysosome formation. gene, which acts downstream of the ATG-2-EPG-6 complex [23]. We observed very similar patterns of PGL granules and LGG-1 Loss of MTM-3 increases the association of ATG-18 with

puncta in mtm-3, epg-5, and epg-5;mtm-3 embryos (Fig 3A0, J and K; autophagic structures Supplementary Fig S6C–D000). The co-localization of PGL granules F F’ F’’ with LGG-1 puncta was also indistinguishable in mtm-3 (31.6%), As a PtdIns3P phosphatase, MTM-3 may modulate the PtdIns3P epg-5 (25.7%) or epg-5;mtm-3 double mutants (26.0%) (Fig 3L). level on autophagic structures. We examined the localization of Together, these data indicate that MTM-3 acts at a similar step to ATG-18::GFP, the PtdIns3P-binding effector, in wild-type and mtm-3 EPG-5 in the aggrephagy pathway, and this step is downstream mutants [23]. ATG-18::GFP was diffuse in the cytoplasm of wild of EPG-8, ATG-9, ATG-18, and ATG-2. type but formed puncta in mtm-3 mutants (Fig 5A, B and J; Supple- mentary Fig S8E). ATG-18 remained diffuse in other autophagy Autophagosomes in mtm-3 mutants are not fused mutants like atg-2 or epg-5, indicating that the ATG-18::GFP puncta G G’ G’’ with lysosomes seen in mtm-3(lf) are not simply protein aggregates formed in autophagy-defective mutants (Supplementary Fig S8A–D0). The ATG-18:: EPG-5 is suggested to be involved in the formation of functional GFP-positive structures in mtm-3 mutants co-localized well with autolysosomes [21,23]. We examined autolysosome formation in puncta labeled by endogenous LGG-1 and were significantly reduced mtm-3, epg-5, and epg-5;mtm-3 worms. In mtm-3(tm4475) embryos, in number in mtm-3; lgg-1 RNAi worms (Fig 5D, I and J), indicating 78.6% of SQST-1::GFP aggregates did not co-localize with the lyso- that they associated with autophagic structures. Loss of VPS-34 somal reporter NUC-1::mCHERRY [24], suggesting that autophagic greatly reduced the number of ATG-18::GFP puncta, and two muta-

substrates are not within autolysosomes (Fig 4A–A00 and H). In epg-5 tions in ATG-18, FKKG and FTTG, which abolish its binding to H I mutants, 42.5% of SQST-1::GFP puncta were clearly separated from PtdIns(3)P, disrupted ATG-18 puncta formation in mtm-3 embryos NUC-1::mCHERRY, while 46.4% stayed very close to but did not (Fig 5C, E–H and J) [23]. These data suggest that loss of mtm-3 overlap with lysosomes (Fig 4B, H and I). The close association of elevates the PtdIns3P level on autophagic structures, leading to SQST-1::GFP with lysosomes in epg-5 mutants was mostly increased autophagic association of the PtdIns3P-binding effector suppressed in epg-5;mtm-3 double mutants, with 77.7% of SQST-1:: ATG-18. GFP being separated from lysosomes, a phenotype resembling that In summary, we identified MTM-3 as a positive regulator of in mtm-3 single mutants (Fig 4C, H and I). These data suggest that autophagy in C. elegans. MTM-3 is recruited by and regulates MTM-3 acts earlier than EPG-5 in autolysosome formation. Loss of PtdIns3P on autophagic structures to promote autophagosome function of laat-1, which encodes a lysosomal lysine/arginine trans- maturation into autolysosomes (Supplementary Fig S8F). MTM-3 porter, severely affects lysosome function, causing accumulation of may hydrolyze PtdIns3P to trigger release of PtdIns3P-binding effec- autophagic cargo in autolysosomes [25]. CUP-5, the C. elegans func- tors from autophagosomes, thus allowing recruitment of proteins tional ortholog of the mammalian lysosomal channel protein MLN1/ involved in next-step maturation or fusion with lysosomes. In yeast, J TRPML1, regulates lysosome biogenesis and loss of cup-5 impairs myotubularin Ymr1 catalyzes PtdIns3P turnover to release Atg degradation of autolysosomal contents [26–28]. We found that 94% proteins from autophagosomes for fusion with vacuoles [10]. Thus, and 98% of SQST-1::GFP puncta co-localized with lysosomes in PtdIns3P clearance from the autophagosomal surface is probably a laat-1 and cup-5 mutant embryos, respectively, indicating accumula- prerequisite for autolysosome formation and myotubularin phos- tion of autophagy substrates in autolysosomes (Fig 4D, F and H). phatases play evolutionarily conserved roles in this process. It Lysosomal accumulation of SQST-1::GFP in laat-1 and cup-5 remains to be seen if this mechanism applies to mammalian auto- mutants was significantly reduced when mtm-3 was inactivated by phagosomes. RNAi (Fig 4E, G and H), suggesting that loss of mtm-3 impairs Worm and yeast myotubularin phosphatases promote autolyso- autolysosome formation. Similar phenotypes were observed when some formation, while human Jumpy and MTMR3 are suggested to

976 EMBO reports Vol 15 | No 9 | 2014 ª 2014 The Authors ª 2014 The Authors EMBO reports Vol 15 | No 9 | 2014 977 EMBO reports MTM-3 promotes autolysosome formation Yanwei Wu et al Yanwei Wu et al MTM-3 promotes autolysosome formation EMBO reports

J MTM-3 but not MTM-3(C459S) exhibits phosphatase activity toward PtdIns3P and PtdIns(3,5)P2 in vitro. D D’ D’’ Source data are available online for this figure.

A A’ B C A A’ A’’ B

D EF G C DEF

H IJ K G HI

JKL

Figure 3. MTM-3 acts downstream of EPG-8, ATG-9, ATG-18, and ATG-2 in the aggrephagy pathway. A–K Confocal fluorescent images of 200-cell-stage embryos from the indicated strains stained by both anti-SEPA-1 and anti-LGG-1 antibodies. Insets show magnified views. Arrows indicate overlapping SEPA-1 and LGG-1 puncta. The wild type contains no visible SEPA-1 aggregates at this stage and is therefore not shown. Scale bars: 5 lm. Figure 4. Loss of mtm-3 impairs autolysosome formation. L The percentage of SEPA-1 puncta co-localizing with LGG-1 in each strain. At least 10 embryos were scored per strain. Data are shown as mean SD. **P < 0.0001; A–G Confocal fluorescent images of 1.5-fold-stage embryos from the indicated strains expressing both SQST-1::GFP and NUC-1::mCHERRY. Insets show magnified views. Æ N.S.: not statistically different. Arrows indicate overlapping or closely associated GFP and mCHERRY. Scale bars: 5 lm. The wild type contains no visible SQST-1::GFP puncta at this stage and is therefore not shown. Source data are available online for this figure. H, I The percentage of SQST-1::GFP-positive puncta that overlap with (H) or are close to (I) NUC-1::mCHERRY-positive structures in each strain. At least 10 embryos were scored per strain. Data are shown as mean SD. **P < 0.0001; N.S.: not statistically different. Æ J–L TEM analysis of embryos in wild type (I) and mtm-3(tm4475) (J, K). Double-membrane autophagosomes (arrows) were observed in mtm-3 but not wild-type embryos. Scale bars: 500 nm. negatively regulate autophagy at the initiation step (Supplementary Materials and Methods Source data are available online for this figure. Fig S8F). PtdIn3P turnover may therefore have distinct effects on autophagy at different stages (initiation versus maturation). It is Quantification analysis important to determine how specific myotubularin phosphatases are temporally and spatially regulated to control PtdIns3P turnover at Fluorescent confocal images of embryos were collected to determine different areas of the embryo (1) were quantified, and 10 embryos Phosphatase assay different stages and how this affects autophagy. Our observation that (1) the percentage of SQST-1::GFP or GFP::LGG-1 puncta that were scored in each strain. The number of GFP::MTM-3(C459S)- MTM-3 overexpression does not inhibit autophagy suggests that co-localize or are close to NUC-1::mCHERRY-positive structures and and ATG-18::GFP-positive structures in embryos was scored directly MTM-3 phosphatase activity was determined with a Malachite rate-limiting regulatory factors are involved in controlling recruit- (2) the percentage of GFP::MTM-3(C459S)- or SEPA-1-puncta that under the fluorescent microscope. At least 15 embryos were Green Assay kit (Echelon Biosciences Inc., USA) as instructed by ment, release or activity of MTM-3 on autophagic structures. co-localize with LGG-1 puncta. Puncta in whole embryos (2) or five quantified in each strain. the manufacturer. Briefly, 500 ng recombinant HIS-tagged MTM-3

978 EMBO reports Vol 15 | No 9 | 2014 ª 2014 The Authors ª 2014 The Authors EMBO reports Vol 15 | No 9 | 2014 979 EMBO reports MTM-3 promotes autolysosome formation Yanwei Wu et al Yanwei Wu et al MTM-3 promotes autolysosome formation EMBO reports

A A’ B C A A’ A’’ B

D EF G C DEF

H IJ K G HI

JKL

A BCD unpaired t-test. Data were considered statistically different at 11. Xue Y, Fares H, Grant B, Li Z, Rose AM, Clark SG, Skolnik EY (2003) P < 0.05 (indicated in the figure legend). Genetic analysis of the myotubularin family of phosphatases in Caenor- See Supplementary Methods for a list of strains, imaging analy- habditis elegans. J Biol Chem 278: 34380 – 34386 sis, construction of reporters, RNAi and statistical analyses. 12. Dang H, Li Z, Skolnik EY, Fares H (2004) Disease-related myotubularins function in endocytic traffic in Caenorhabditis elegans. Mol Biol Cell 15: Supplementary information for this article is available online: 189 – 196 http://embor.embopress.org 13. Silhankova M, Port F, Harterink M, Basler K, Korswagen HC (2010) Wnt signalling requires MTM-6 and MTM-9 myotubularin lipid-phosphatase EFGH Acknowledgements function in Wnt-producing cells. EMBO J 29: 4094 – 4105 We thank Dr. G. Ou (Tsinghua University) for plasmids and Dr. Isabel Hanson 14. Zou W, Lu Q, Zhao D, Li W, Mapes J, Xie Y, Wang X (2009) Caenorhabd- for editing services. Some strains were provided by the CGC, which is funded itis elegans myotubularin MTM-1 negatively regulates the engulfment by NIH Office of Research Infrastructure Programs (P40OD010440). This work of apoptotic cells. PLoS Genet 5:e1000679 was supported by the National Science Foundation of China (31325015), the 15. Neukomm LJ, Nicot AS, Kinchen JM, Almendinger J, Pinto SM, Zeng S, National Basic Research Program of China (2013CB910100), and an Interna- Doukoumetzidis K, Tronchère H, Payrastre B, Laporte JF et al (2011) The tional Early Career Scientist grant from the Howard Hughes Institute to XW. phosphoinositide phosphatase MTM-1 regulates apoptotic cell corpse I I’ I’’ I’ clearance through CED-5-CED-12 in Caenorhabditis elegans. Development Author contributions 138: 2003 – 2014 YW and XW designed the experiments, analyzed the data, and wrote the 16. Tian Y, Li Z, Hu W, Ren H, Tian E, Zhao Y, Lu Q, Huang X, Yang P, Li X manuscript. YW performed most of the experiments. SC and WZ performed et al (2010) Caenorhabditis elegans screen identifies autophagy genes some genetic experiments, and HZhao performed the EM analysis. SY, SM and specific to multicellular organisms. Cell 141: 1042 – 1055 HZhang provided reagents. 17. Zhang Y, Yan L, Zhou Z, Yang P, Tian E, Zhang K, Zhao Y, Li Z, Song B, Han J et al (2009) SEPA-1 mediates the specific recognition and degra- Conflict of interest dation of P granule components by autophagy in Caenorhabditis The authors declare that they have no conflict of interest. elegans. Cell 136: 308 – 321 J 18. Lin L, Yang P, Huang X, Zhang H, Lu Q (2013) The scaffold protein EPG-7 links cargo-receptor complexes with the autophagic assembly machin- References ery. J Cell Biol 201: 113 – 129 19. Kovacs AL, Zhang H (2010) Role of autophagy in Caenorhabditis elegans. 1. Klionsky DJ (2007) Autophagy: from phenomenology to molecular FEBS Lett 584: 1335 – 1341 understanding in less than a decade. Nat Rev Mol Cell Biol 8: 20. Lorenzo O, Urbe S, Clague MJ (2005) Analysis of phosphoinositide bind- 931 – 937 ing domain properties within the myotubularin-related protein MTMR3. 2. Mizushima N, Yoshimori T, Ohsumi Y (2011) The role of Atg proteins in J Cell Sci 118: 2005 – 2012 autophagosome formation. Annu Rev Cell Dev Biol 27: 107 – 132 21. Lu Q, Wu F, Zhang H (2013) Aggrephagy: lessons from Caenorhabditis 3. Noda T, Fujita N, Yoshimori T (2009) The late stages of autophagy: how elegans. Biochem J 452: 381 – 390 does the end begin? Cell Death Differ 16: 984 – 990 22. Yang P, Zhang H (2010) The coiled-coil domain protein EPG-8 plays an 4. Mehrpour M, Esclatine A, Beau I, Codogno P (2010) Overview of essential role in the autophagy pathway in Caenorhabditis elegans. Auto- Figure 5. Loss of mtm-3 causes increased autophagic association of ATG-18. macroautophagy regulation in mammalian cells. Cell Res 20: phagy 7: 159 – 165 A–H Fluorescent images of embryos at the 4-fold stage in the indicated strains expressing ATG-18::GFP (A–D), ATG-18(FKKG)::GFP (E, F), or ATG-18(FTTG)::GFP (G, H). 748 – 762 23. Lu Q, Yang P, Huang X, Hu W, Guo B, Wu F, Lin L, Kovács AL, Yu L, I-I000 Confocal fluorescent images of 4-fold mtm-3 embryos expressing ATG-18::GFP (I0) and stained by DAPI (I) and anti-LGG-1 (I″) antibody. Insets show magnified 5. Hamasaki M, Furuta N, Matsuda A, Nezu A, Yamamoto A, Fujita N, Zhang H et al (2011) The WD40 repeat PtdIns(3)P-binding protein views. Arrows indicate ATG-18 puncta labeled by anti-LGG-1 antibody. Scale bars: 5 lm. J Number of wild-type or mutated ATG-18::GFP puncta in each strain. At least 15 embryos were scored per strain (mean SD). **P < 0.0001; N.S: not statistically Oomori H, Noda T, Haraguchi T, Hiraoka Y et al (2013) Auto- EPG-6 regulates progression of omegasomes to autophagosomes. Dev � different. phagosomes form at ER-mitochondria contact sites. Nature 495: Cell 21: 343 – 357 Source data is available online for this figure. 389 – 393 24. Guo P, Hu T, Zhang J, Jiang S, Wang X (2010) Sequential action of 6. Dall’Armi C, Devereaux KA, Di Paolo G (2013) The role of lipids in the Caenorhabditis elegans Rab GTPases regulates phagolysosome formation control of autophagy. Curr Biol 23:R33 – R45 during apoptotic cell degradation. Proc Natl Acad Sci USA 107: or MTM-3(C459S) was incubated with 1 mM DiC8PtdIns3P or After washing with PBST (PBS + 0.2% Tween 20), samples were 7. Noda T, Matsunaga K, Taguchi-Atarashi N, Yoshimori T (2010) Regula- 18016 – 18021

DiC8PtdIns(3,5)P2 in 25 ll reaction buffer (50 mM Tris-HCl, incubated with Texas red-conjugated and/or FITC-conjugated tion of membrane biogenesis in autophagy via PI3P dynamics. Semin 25. Liu B, Du H, Rutkowski R, Gartner A, Wang X (2012) LAAT-1 is the lyso-

100 mM NaCl, 2 mM CaCl2, 2 mM DTT) at 20°C overnight. One secondary antibodies (Jackson ImmunoResearch, 112-075-003 and Cell Dev Biol 21: 671 – 676 somal lysine/arginine transporter that maintains amino acid homeosta- hundred microliters Malachite Green Solution was then added to 115-095-003) at a 1:200 dilution for 1 h at room temperature. The 8. Vergne I, Roberts E, Elmaoued RA, Tosch V, Delgado MA, Proi- sis. Science 337: 351 – 354 each reaction and incubated for 20 min at room temperature. Absor- stained samples were washed and mounted in 15% VECTASHIELD kas-Cezanne T, Laporte J, Deretic V (2009) Control of autophagy initi- 26. Sun T, Wang X, Lu Q, Ren H, Zhang H (2011) CUP-5, the Caenorhabditis bance was measured at 620 nm. Control reactions contained no mounting medium with DAPI (VECTOR) and visualized using a ation by phosphoinositide 3-phosphatase Jumpy. EMBO J 28: elegans ortholog of the mammalian lysosomal channel protein MLN1/ MTM-3 protein (blank). Each reaction was performed in triplicate to Zeiss LSM 510 Meta inverted confocal microscope (Carl Zeiss, 2244 – 2258 TRPML1, is required for proteolytic degradation in autolysosomes. Auto- get the mean absorbance. Two independent experiments were Germany). 9. Taguchi-Atarashi N, Hamasaki M, Matsunaga K, Omori H, Ktistakis NT, phagy 7: 1308 – 1315 performed with similar results, and one of them is shown in Fig 2J. Yoshimori T, Noda T (2010) Modulation of local PtdIns3P levels by the 27. Fares H, Greenwald I (2001) Regulation of endocytosis by CUP-5, the Statistical analysis PI phosphatase MTMR3 regulates constitutive autophagy. Traffic 11: Caenorhabditis elegans mucolipin-1 homolog. Nat Genet 28: 64 – 68 Immunostaining 468 – 478 28. Treusch S, Knuth S, Slaugenhaupt SA, Goldin E, Grant BD, Fares H (2004) The standard deviation (SD) was used as the y error bar for bar 10. Cebollero E, van der Vaart A, Zhao M, Rieter E, Klionsky DJ, Helms JB, Caenorhabditis elegans functional orthologue of human protein Mixed-stage embryos were fixed and incubated with anti-LGG-1 charts plotted from the mean value of the data. Data derived from Reggiori F (2012) Phosphatidylinositol-3-phosphate clearance plays h-mucolipin-1 is required for lysosome biogenesis. Proc Natl Acad Sci (1: 1,000) or anti-SEPA-1 (1:10,000) in blocking buffer at 4°C overnight. different genetic backgrounds were compared by Student’s two-way a key role in autophagosome completion. Curr Biol 22: 1545 – 1553 USA 101: 4483 – 4488

980 EMBO reports Vol 15 | No 9 | 2014 ª 2014 The Authors ª 2014 The Authors EMBO reports Vol 15 | No 9 | 2014 981 EMBO reports MTM-3 promotes autolysosome formation Yanwei Wu et al Yanwei Wu et al MTM-3 promotes autolysosome formation EMBO reports

980 EMBO reports Vol 15 | No 9 | 2014 ª 2014 The Authors ª 2014 The Authors EMBO reports Vol 15 | No 9 | 2014 981 EMBO Molecular Medicine Autophagy and AL amyloid cardiomyopathy Jian Guan et al Research Article

to prevent increased ROS generation, decreased ATP production and AL-LC impairs autophagic flux loss of cellular function. Damaged mitochondria are cleared intracellularly by a complex quality control mechanism involving mito- Defective mitochondria are cleared intracellularly by a complex Lysosomal dysfunction and impaired autophagy phagy and the lysosome. Mitophagy refers to a macro-autophagic macro-autophagic response (Codogno, 2014). By Western blot, process that selectively removes mitochondria. Of note, macro- levels of the autophagy marker LC3-II were markedly increased in autophagy has also been implicated in handling proteotoxic events cardiomyocytes exposed to AL-LC (Fig 1D), as well as the number underlie the pathogenesis of amyloidogenic light that cannot be mediated by the proteasome. Herein, utilizing in vitro of autophagosomes in AL-LC-exposed cardiomyocytes overexpress- isolated cardiomyocytes and an in vivo zebrafish model of AL-LC ing GFP-LC3 (Mizushima et al, 2010), (Supplementary Fig S1). chain-mediated cardiotoxicity toxicity, we find that disruption of autophagic flux is the underlying Increased LC3-II levels and number of autophagosomes (detected as mechanism critical for the induction of mitochondrial dysfunction GFP-LC3 punctae) may be indicative of either elevated autophagy Jian Guan1,†, Shikha Mishra1,†, Yiling Qiu1, Jianru Shi1,‡, Kyle Trudeau2, Guy Las2, Marc Liesa2, Orian S and development of AL amyloid cardiomyopathy. induction or defective clearance. To distinguish between induction and clearance of autophagosomes, E64d and Pepstatin A were used 2 3 3 4 1 1,4,* Shirihai , Lawreen H Connors , David C Seldin , Rodney H Falk , Calum A MacRae & Ronglih Liao to inhibit lysosomal enzymes and impede autophagosome clear- Results ance. LC3-II levels accumulated less after lysosome inhibition in AL-LC cardiomyocytes, indicative of a decrease in autophagosome AL-LC triggers mitochondrial dysfunction and ROS production clearance. This decrease was explained both by increased basal Abstract United States and Europe (Merlini et al, 2011), in which widespread LC3-II levels and by decreased LC3-II levels after E64D and Pepstatin tissue infiltration and deposition of amyloid fibrils derived from We have shown that human AL-LC protein provokes excessive ROS A treatments in AL-LC cardiomyocytes when compared to control AL amyloidosis is the consequence of clonal production of amyloi- clonal immunoglobulin light chain (LC) proteins causes multi-organ production and subsequent cellular dysfunction and cell death in (Fig 1E). Furthermore, AL-LC resulted in an increase in p62 accu- dogenic immunoglobulin light chain (LC) proteins, often resulting dysfunction. Greater than 70% of patients with primary LC amyloi- isolated cardiomyocytes (Brenner et al, 2004; Shi et al, 2010); mulation, an established marker of autophagic clearance (Fig 1F). in a rapidly progressive and fatal amyloid cardiomyopathy. Recent dosis present with cardiac involvement (Madan et al, 2010; Falk, however, the source of ROS production has yet to be identified. To We next addressed whether this alteration in macroautophagy was work has found that amyloidogenic LC directly initiate a cardio- 2011), which can progress to debilitating heart failure symptoms determine whether mitochondria contribute to AL-LC-induced ROS also associated with decreased mitophagic clearance. Immunofluo- toxic response underlying the pathogenesis of the cardiomyopathy; and early cardiovascular death (Falk, 2005; Falk, 2011). To date, production, isolated cardiomyocytes were pretreated with the mito- rescent staining of isolated adult cardiomyocytes exposed to AL-LC however, the mechanisms that contribute to this proteotoxicity there are no targeted treatments for amyloid cardiomyopathy (Falk, chondrial-targeted ROS scavenger, Mito-TEMPO, and exposed to showed an increase in p62 levels co-localized with mitochondria, remain unknown. Using human amyloidogenic LC isolated from 2011), owing to a lack of understanding of the basic mechanisms human AL-LC. Using the ROS-sensitive fluorescent dye, DCFDA, we suggesting a perturbation in mitophagy (Fig 1G). Taken together, patients with amyloid cardiomyopathy, we reveal that dysregula- that underlie the pathogenesis of the disease. While amyloid fibril found that increased AL-LC-elicited ROS was abolished with Mito- our data suggest a defect in autophagy flux in cardiomyocytes tion of autophagic flux is critical for mediating amyloidogenic LC deposition within the heart has long been hypothesized to be TEMPO to levels comparable to cells treated with vehicle (Veh) or subjected to human AL-LC protein, with a corresponding inhibition proteotoxicity. Restoration of autophagic flux by pharmacological responsible for disease pathophysiology, there often is dissociation control light chain (Con-LC) proteins isolated from patients with of mitochondrial clearance. intervention using rapamycin protected against amyloidogenic between the degree of amyloid fibril deposition and cardiovascular multiple myeloma (Fig 1A). In addition, AL-LC treatment of cardio- light chain protein-induced pathologies including contractile outcomes. We and others have found that circulating amyloidogenic myocytes was associated with decreased mitochondrial membrane Restoration of autophagic flux attenuates AL-LC-induced cellular dysfunction and cell death at the cellular and organ level and also light chain proteins (AL-LC) directly initiate a potent cardiotoxic potential (depolarization) as determined by TMRE (Fig 1B). One dysfunction and cell death in vitro prolonged survival in an in vivo zebrafish model of amyloid cardio- effect, independent of fibril deposition, and this cardiotoxicity is crit- process that could explain this depolarization is decreased mito- toxicity. Mechanistically, we identify impaired lysosomal function ical to manifestations of amyloid cardiomyopathy, both in vitro and chondrial bioenergetics function associated with decreased ATP To determine whether autophagy dysregulation was causal for to be the major cause of defective autophagy and amyloidogenic in vivo (Liao et al, 2001; Brenner et al, 2004; Migrino et al, 2010, synthesis. In order to confirm this, we measured cellular ATP AL-LC-induced cardiotoxicity, rapamycin, an inhibitor of mTOR LC-induced proteotoxicity. Collectively, these findings detail the 2011; Shi et al, 2010; Sikkink & Ramirez-Alvarado, 2010; Shin et al, levels, as mitochondria are the main contributor to ATP levels in signaling and a potent enhancer of both autophagosome formation downstream molecular mechanisms underlying AL amyloid cardio- 2012). While these findings have changed our understanding of AL cardiomyocytes. Total ATP levels were decreased by AL-LC and not and clearance, was used to restore autophagic flux in cardiomyocytes myopathy and highlight potential targeting of autophagy and lyso- amyloid cardiomyopathy, from one of just passive fibril infiltration by addition of Con-LC (Fig 1C). Thus, these data collectively exposed to AL-LC. Restoration of autophagosome clearance by rapa- somal dysfunction in patients with amyloid cardiomyopathy. to also acknowledging a direct proteotoxicity, the basic mechanisms demonstrate mitochondrial dysfunction and increased mitochon- mycin was confirmed by significant reduction in p62 accumulation in by which this proteotoxicity results in cardiomyopathy remain drial ROS production caused by AL-LC in cardiomyocytes. cardiomyocytes exposed to AL-LC (Fig 2A). Concomitant with Keywords amyloidosis; autophagy; cardiac toxicity; lysosome; mitochondria unknown. Furthermore, an increase in oxidative stress and reactive Subject Categories Cardiovascular System oxygen species (ROS) production is one of the consequences associ- DOI 10.15252/emmm.201404190 | Received 21 April 2014 | Revised 19 September ated with AL-mediated proteotoxicity. However, the source for this Figure 1. AL-LC causes mitochondrial dysfunction and autophagy dysregulation in vitro. ▸ 2014 | Accepted 19 September 2014 | Published online 15 October 2014 increased ROS is unknown. A Using the fluorescent indicator DCFDA, ROS was measured in isolated cardiomyocytes treated with vehicle, Con-LC, AL-LC or AL-LC + Mito-TEMPO. Representative EMBO Mol Med (2014) 6: 1493–1507 A growing body of evidence demonstrates that mitochondrial fluorescent and bright field images are shown in the top panels, and quantitative analysis is summarized in the graph below. AL-LC but not Con-LC increased ROS quality control alterations contribute and can be central to a number and this was blocked by Mito-TEMPO, indicating that ROS is derived from mitochondria. Scale bar = 100 lm. N = 3.*P = 0.020. B Mitochondrial membrane potential was measured using TMRE fluorescent dye in cardiomyocytes following 24-h treatment with vehicle, Con-LC or AL-LC. of human diseases including Alzheimer’s, Parkinson’s, Hunting- Representative TMRE fluorescent and bright field microscopy images are shown in the top panels, and TMRE fluorescence signal was quantified and summarized Introduction ton’s, diabetes and cardiovascular disease (Harris & Rubinsztein, below. AL-LC exposure resulted in loss of mitochondrial membrane potential compared to the other groups. Scale bar = 50 lm. N = 3.*P = 0.031. 2012; Nixon, 2013). The heart is particularly sensitive to perturba- C Quantitative summary of cellular ATP levels in cardiomyocytes following 24-h exposure to vehicle, Con-LC or AL-LC. N = 3.*P = 0.020. AL or light chain amyloidosis (formerly known as primary amyloi- tions of mitochondrial function, given the energetic requirements of D Immunoblot analysis of LC3-II expression in cardiomyocytes following 24-h exposure to vehicle, Con-LC or AL-LC. GAPDH was used as a loading control. Quantitative results are shown in the graph below for comparison. LC3-II expression was significantly increased in the AL-LC group compared to vehicle and Con-LC. N = 3. dosis) is the most commonly diagnosed systemic amyloidosis in the contractile function. Removal of damaged mitochondria is essential 6 *P = 1.9 × 10À . E Immunoblot analysis of AL-LC-induced LC3-II expression in the presence of lysosomal inhibitors E64d and Pepstatin A. GAPDH is used as a loading control. Quantitative results in the graph below show that LC3-II levels in the AL-LC group do not exceed vehicle-treated LC3-II levels following lysosomal inhibition, 6 # 1 Division of Cardiovascular Medicine, Brigham and Women’s Hospital, Harvard Medical School, Boston, MA, USA indicating minimal perturbation to initiation. N = 6.*P = 1.4 × 10À , P = 0.041. 2 Department of Medicine, Boston University School of Medicine, Boston, MA, USA F Immunoblot and quantitative analysis of p62 expression in cardiomyocytes following 24-h exposure to vehicle, Con-LC or AL-LC. Results show a decrease in 3 Amyloidosis Center, Boston University School of Medicine, Boston, MA, USA autophagic flux seen by increased p62 accumulation in the AL-LC group. N = 3.*P = 0.033. 4 Cardiac Amyloidosis Program, Brigham and Women’s Hospital, Harvard Medical School, Boston, MA, USA G Confocal imaging of immunofluorescence staining of p62 (red) in cardiomyocytes following 24-h exposure to vehicle, Con-LC or AL-LC. Mitochondrial co-localization *Corresponding author. Tel: +1 617 525 4864; Fax: +1 617 525 4868; E-mail: [email protected] †These authors contributed equally to this work was visualized using mitochondrial protein COX4 (green) co-staining, and nuclear staining with DAPI (blue). Scale bar = 10 lm. ‡Present address: Division of Cardiovascular Medicine, The Heart Institute, Good Samaritan Hospital, Keck School of Medicine, University of Southern California, Los Angeles, Source data are available online for this figure. CA, USA

ª 2014 The Authors. Published under the terms of the CC BY 4.0 license EMBO Molecular Medicine Vol 6 | No 11 | 2014 1493 1494 EMBO Molecular Medicine Vol 6 | No 11 | 2014 ª 2014 The Authors EMBO Molecular Medicine Autophagy and AL amyloid cardiomyopathy Jian Guan et al Research Article

ª 2014 The Authors. Published under the terms of the CC BY 4.0 license EMBO Molecular Medicine Vol 6 | No 11 | 2014 1493 1494 EMBO Molecular Medicine Vol 6 | No 11 | 2014 ª 2014 The Authors Jian Guan et al Autophagy and AL amyloid cardiomyopathy EMBO Molecular Medicine EMBO Molecular Medicine Autophagy and AL amyloid cardiomyopathy Jian Guan et al

AB decreased p62 levels, rapamycin-treated cardiomyocytes showed evidence for the central role of autophagic dysfunction in the patho- significant attenuation of both mitochondrial dysfunction and intra- genesis of amyloid cardiotoxicity and highlight the use of rapamycin cellular ROS levels (Fig 2B and C), and protection against AL-LC- as a potential therapeutic approach for treatment of this disease. induced cellular contractile dysfunction (Fig 2D) and concomitant impaired intracellular calcium homeostasis (Fig 2E) as well as cell Lysosomal dysfunction directly contributes to AL-LC-triggered death (Fig 2F). Importantly, the concentration of rapamycin used impaired autophagy (10 nM) (Dehay et al, 2010) did not affect downstream substrates of mTOR as measured by S6 kinase activation (Supplementary Fig S2). Our results suggest that AL-LC-induced dysregulation of autophagic In addition, pharmacologic inhibition of autophagy by chloroquine flux may reside at the stage of autophagosome clearance, the final (CQ) reversed rapamycin attenuation of p62 accumulation in cardio- step in the autophagy process in which the lysosome plays a pivotal myocytes exposed to AL-LC (Fig 2G) and abrogated the beneficial role. We sought to examine lysosomal function in response to AL-LC effects of rapamycin on contractile function (Fig 2D), calcium tran- exposure. The number of acidic vesicles (including lysosomes) per sient amplitude (Fig 2F) and cell survival (Fig 2H), consistent with cell, as measured using LysoTracker staining, was markedly rapamycin protecting against AL-LC via improvement of autophagic compromised in isolated cardiomyocytes exposed to AL-LC for 24 h flux. (Fig 4A), concomitant with a loss of lysosomal acidity, assessed by LysoSensor, a pH-sensitive fluorescent probe (Fig 4B) 24 h follow- Rapamycin protects against AL-LC proteotoxicity in vivo ing AL-LC exposure. Associated with loss of lysosomal function, quantitative PCR revealed downregulation of lysosome-related To determine the role of AL-LC-induced impaired autophagy in vivo, genes including cathepsin D, lysosomal-specific vacuolar ATPases we utilized a recently reported zebrafish model of AL-LC cardiotox- (Fig 4C), as well as a transcriptional regulator of lysosomal biogene- icity, characterized by impaired cardiac function and early cardio- sis and function, TFEB (transcription factor EB) at the mRNA vascular death following injection of human AL-LC (Mishra et al, (Fig 4D) and protein (Fig 4E) levels following 24 h of AL-LC expo- C D E 2013). Consistent with our in vitro findings, zebrafish injected with sure. Importantly, decreased TFEB expression was restored to base- AL-LC showed increased LC3-II and p62 levels (Fig 3A–B) compared line levels following rapamycin treatment (Fig 4E). to Con-LC. Electron microscopy of heart tissue revealed increased To determine whether downregulation of TFEB is central to autophagosome number as indicated by the accumulation of AL-LC-induced cardiotoxicity, TFEB was overexpressed in isolated double-membrane vesicle structures with AL-LC exposure (Fig 3C). cardiomyocytes (Supplementary Fig S3A) and in zebrafish (Supple- Autophagic flux was restored in AL-LC-injected zebrafish via treat- mentary Fig S3B). Overexpression of TFEB protected against ment with 10 nM rapamycin (Tobin & Beales, 2008), seen by contractile dysfunction and restored calcium transient amplitude in decreased p62 comparable to control levels (Fig 3D). Peak aortic cardiomyocytes exposed to AL-LC (Fig 4F–G) and prevented AL-LC- flow, an indicator of cardiac function, was decreased in AL-LC- associated cardiac cell death in vivo (Fig 4H) with greatly improved injected fish (Fig 3E) and restored to control levels with rapamycin survival (Fig 4I). treatment. Similarly, AL-LC-triggered cell death in zebrafish hearts To determine the temporal importance of lysosomal and auto- was reduced following rapamycin treatment (Fig 3F and G). phagic dysfunction, we examined the time course of activation of Survival was markedly impaired following injection of human previously established critical components of the AL-LC cardiotoxic AL-LC in zebrafish and was significantly rescued with rapamycin response. We found that lysosomal function was impaired early, treatment (Fig 3H). Rapamycin did not alter survival in Con-LC within 3 h of AL-LC exposure in isolated cardiomyocytes (Fig 5A). animals (Fig 3H). Together, our in vivo data provide further Six hours following AL-LC exposure, autophagic dysfunction was F G

Figure 2. Restoration of autophagic flux with rapamycin attenuates AL-LC-induced cellular dysfunction and cell death in vitro. ▸ A Immunoblot analysis of p62 on cardiomyocytes following 24-h exposure to vehicle, Con-LC or AL-LC in the absence or presence of 10 nM rapamycin. Quantitative results summarized below show decreased p62 expression following rapamycin treatment. N = 3.*P = 0.003, #P = 0.006. B Mitochondrial function is rescued by rapamycin treatment, shown by quantitative analysis of mitochondrial membrane potential using TMRE dye in cardiomyocytes following exposure to vehicle, Con-LC or AL-LC for 24 h in the presence or absence of rapamycin. N = 3.*P = 0.046, #P = 0.005 between indicated groups. 4 # C ROS levels are reduced by treatment with rapamycin, shown using DCFDA in cardiomyocytes exposed to vehicle, Con-LC or AL-LC. N = 3.*P = 2.3 × 10À , P = 0.002 between indicated groups. D Contractile function was measured in cardiomyocytes exposed to AL-LC for 24 h in the absence or presence of rapamycin with or without chloroquine (2.5 lM). 4 # Quantitative analysis was performed by calculating percent cell shortening. N = 3.*P = 3.3 × 10À , P = 0.007. E Calcium transient amplitude, measured in isolated cardiomyocytes, was quantified following exposure to vehicle or AL-LC in the presence or absence of rapamycin with or without chloroquine. Representative tracings are shown, and quantitative analysis in the graph shows a rescue of AL-LC-induced decrease in calcium transient amplitude following rapamycin treatment. N = 3,*P = 0.002, #P = 0.006. F Rapamycin treatment reduces apoptosis in cardiomyocytes exposed to AL-LC. TUNEL staining was performed to quantify cell death in cardiomyocytes following exposure to vehicle, Con-LC or AL-LC with or without rapamycin treatment. Cell death was measured as percent TUNEL-positive nuclei relative to total cell number. N = 3.*P = 0.015, #P = 0.035 between indicated groups. G Verification that rapamycin rescue was autophagy dependent was demonstrated using chloroquine (2.5 lM) administered to AL-LC + rapamycin-treated cardiomyocytes. p62 accumulation was measured using immunoblot analysis, with GAPDH as a loading control. N = 5.*P = 0.023. H Cardiomyocytes were treated with chloroquine in the presence of rapamycin, and immunoblot analysis was performed to probe for active caspase 3 levels, normalized to GAPDH expression. N = 3.*P = 0.007.

Figure 1. Source data are available online for this figure.

ª 2014 The Authors EMBO Molecular Medicine Vol 6 | No 11 | 2014 1495 1496 EMBO Molecular Medicine Vol 6 | No 11 | 2014 ª 2014 The Authors Jian Guan et al Autophagy and AL amyloid cardiomyopathy EMBO Molecular Medicine EMBO Molecular Medicine Autophagy and AL amyloid cardiomyopathy Jian Guan et al

Figure 1. Source data are available online for this figure.

ABC noted by accumulation of GFP-LC3, a result of reduced autophagic of autophagic flux pharmacologically with rapamycin or genetically degradation (Ni et al, 2011) (Fig 5B). Loss of mitochondrial clear- through overexpression of TFEB protects against AL-LC cardiotoxic- ance was followed by decreased mitochondrial membrane potential, ity and may represent a novel therapeutic approach for treatment of as measured by the mitochondrial membrane potential-sensitive amyloid cardiomyopathy. dye TMRE (Fig 5C) at 12 and 24 h following AL-LC exposure, For our experiments, we utilized human Bence-Jones proteins to respectively. Increased ROS was detected by DCFDA (Fig 5D) 24 h study the signaling effects associated with toxic amyloid precursor following AL-LC exposure indicating it to be a late event. The proteins. We found that light chain solubility is equivalent for AL-LC temporal cascade of molecular events triggered by the AL-LC is and Con-LC at the 20 lg/ml concentration used in this study (Sup- summarized in Fig 5E. plementary Fig S5A). Additionally, under non-reducing native conditions, we see that the majority of AL-LC and Con-LC proteins migrate Human AL amyloid cardiomyopathy is associated with impaired to a molecular weight consistent with a dimeric state (Supplemen- autophagy and defective lysosomal function tary Fig S5B), while under reducing conditions, these proteins are found at a molecular weight consistent with monomeric state (Sup- To determine the applicability of our findings to human amyloid plementary Fig S5C). Further investigation is necessary to determine cardiomyopathy, we examined markers of autophagy in heart tissue whether the dimer form of the light chain proteins observed in our samples obtained from patients with AL amyloid cardiomyopathy. study represents a true oligomeric state or merely a state of associa- Electron microscopy of human heart tissue revealed dramatic differ- tion. Light chain has two subtypes, lambda and kappa. For our DE ences in mitochondrial ultrastructure in amyloid cardiomyopathy, experiments presented in this study, the lambda subtype of AL-LC with loss of normal spatial distribution between mitochondria and and kappa subtype of Con-LC proteins were used. Importantly, prior cardiac myofilaments, as well as vacuolization, loss of cristae, swell- work from our group has found no difference in the cardiotoxic ing and enlargement (Supplementary Fig S4). Significant accumula- response for amyloidogenic kappa versus lambda light chain tion of autophagosomes was observed throughout the tissue proteins (Guan et al, 2013; Mishra et al, 2013). These observations (Fig 6A) with increased LC3-II and p62 levels (Fig 6B and C) and a suggest that both kappa and lambda AL-LC proteins exert a similar decrease in TFEB expression (Fig 6D). cardiotoxic response, whereas neither kappa nor lambda Con-LC proteins resulted in cardiomyocyte toxicity or dysfunction, even at tenfold higher concentrations in vitro (100 lg/ml) and in vivo Discussion (1,000 lg/ml) than routinely used concentrations of AL-LC (Supple- mentary Fig S6; Mishra et al, 2013). Prior work has detailed an intrinsic cardiotoxic response to human Our previous studies have indicated that ROS generation is a amyloidogenic light chain proteins that underlies the development of critical factor contributing to AL-LC-induced cellular pathology AL amyloid cardiomyopathy (Liao et al, 2001; Brenner et al, 2004; where phenotypic rescue is seen with antioxidant administration Migrino et al, 2010, 2011; Shi et al, 2010; Shin et al, 2012; Guan (Brenner et al, 2004; Shi et al, 2010). Data presented here expand et al, 2013; Mishra et al, 2013). While stress-activated kinases and upon our previous work by demonstrating the mitochondrial origin ROS generation have been identified as downstream components of of the increased ROS. Mitochondrial dysfunction was found to be the cardiomyocyte response to human AL-LC, the fundamental cellu- closely associated with AL-LC-induced pathology, resulting from lar mechanisms underlying the proteotoxicity remain elusive. Here, impaired autophagic flux. Dysregulated autophagy has been impli- we find that inhibition of autophagic flux and, specifically, lysosomal cated as key in the pathology of a number of human diseases, dysfunction is central to AL-LC cardiotoxicity and the development ranging from neurodegenerative to cardiovascular diseases, and of amyloid cardiomyopathy. We further demonstrate that restoration F GH more recently, protein misfolding diseases including desmin-related

Figure 3. Restoration of autophagic flux via rapamycin attenuates AL-LC-induced cellular dysfunction and cell death in vivo. ▸ A, B Immunoblot analysis of LC3-II and p62 expression in zebrafish lysate, 3 days post-injection of Con-LC or AL-LC (100 lg/ml). GAPDH was used as a loading control. N = 5 (A); N = 4 (B). *P = 0.043, #P = 0.029. C Transmission electron micrographs of cardiac tissue of Con-LC and AL-LC-injected zebrafish 3 days post-injection (5 days post-fertilization). Top panels show changes in mitochondrial morphology and presence of double-membrane structures (autophagosomes) at 6,800× in AL-LC fish compared to Con-LC fish. Bottom panel shows higher magnification of mitochondrial structure and autophagosomes. Scale bar = 500 nm. N = 3 per group. D Zebrafish were treated with 10 nM rapamycin after receiving Con-LC or AL-LC injection. Immunoblot analysis was performed to measure p62 expression and normalized to GAPDH as a loading control. Fifteen fish were homogenized per experiment. Quantitative analysis is shown for comparison below representative blots. N = 3.*P = 0.011. E Peak flow was measured in zebrafish embryos using color Doppler echocardiography. Representative color Doppler peak flow tracings of zebrafish 3 days post- injection with or without treatment with rapamycin. Peak flow quantitation analysis is shown in the panel below. N = 5.*P = 0.001. F Cell death was quantified in zebrafish 3 days post-injection with or without rapamycin treatment. Expression of active caspase 3 was measured using immunoblot. 15 zebrafish were homogenized per experiment. N = 4.*P = 0.029. G Representative confocal images of hearts isolated from zebrafish 3 days post-injection with or without rapamycin treatment. Hearts are stained for TUNEL-positive 4 nuclei (red) and DAPI counterstain. TUNEL-positive nuclei are quantified and graphed in the right panel. Scale bar = 25 lm. N = 4 per group. *P = 6.7 × 10À , #P = 0.001. H Kaplan–Meier analysis of zebrafish survival following injection of Con-LC or AL-LC in the absence or presence of rapamycin. Survival was monitored daily. The survival of AL-LC-injected fish was significantly prolonged in the presence of rapamycin. N = 25 per group. *P = 0.0001.

Figure 2. Source data are available online for this figure.

ª 2014 The Authors EMBO Molecular Medicine Vol 6 | No 11 | 2014 1497 1498 EMBO Molecular Medicine Vol 6 | No 11 | 2014 ª 2014 The Authors Jian Guan et al Autophagy and AL amyloid cardiomyopathy EMBO Molecular Medicine EMBO Molecular Medicine Autophagy and AL amyloid cardiomyopathy Jian Guan et al

Figure 2. Source data are available online for this figure.

A BC cardiomyopathy (Wong & Cuervo, 2010; Bhuiyan et al, 2013). of rapamycin was primarily through improvement of clearance. Autophagy is a dynamic process that starts with the formation of a Lysosomal function plays a critical role in the clearance of auto- double-membrane autophagosome complex that engulfs and then phagosomes (Eskelinen & Saftig, 2009), and dysfunction or defi- degrades cellular waste products, including organelles such as ciency of the lysosome has been implicated in other amyloid-related defective mitochondria. Thus, defects in either formation or clear- diseases, such as Parkinson’s disease (Dehay et al, 2010). In our ance of autophagosomes could result in the observed increased disease model, we observed AL-LC-induced loss of lysosomal acid- levels of LC3-II expression and the formation of double-membrane ity, with associated downregulation of lysosomal genes and ATPases structures induced by AL-LC. Investigation of each step of the auto- required for maintenance of pH, as well as decreased TFEB, a criti- phagic flux following AL-LC exposure was therefore required to cal transcriptional regulator of lysosomes (Settembre et al, 2011; identify the exact point of dysregulation in the autophagy pathway Decressac et al, 2013; Pastore et al, 2013). Furthermore, rapamycin (Mizushima et al, 2010). Through use of lysosomal inhibitors, we administration resulted in a decrease in autophagosome accumula- found that formation of autophagosomes, as measured by LC3-II tion and attenuated dopaminergic neuronal cell death, both of levels in AL-LC-exposed cardiomyocytes, was similar to control which were associated with increased numbers of functional lyso- treated cells, indicating that the primary defect in AL-LC exposure somes (Dehay et al, 2010). In our systems, genetic overexpression is in autophagosome clearance. The autophagic substrate, p62, also of TFEB protected against AL-LC cardiotoxicity in vitro and in vivo known as sequestosome 1 (SQSTM1), is a scaffold protein that and prolonged survival in our zebrafish model. Notably, rapamycin binds ubiquitinated protein leading to the formation of autophago- treatment restored TFEB expression to control levels in AL-LC- somes for subsequent degradation (Kubli & Gustafsson, 2012). We treated cardiomyocytes, further confirming that its mechanism of D EF not only found increased p62 accumulation in our in vitro and action was through targeting lysosomal function. in vivo experimental models in response to AL-LC, but also obser- In summary, the studies presented here show that lysosomal- ved an increase in p62 levels in explanted human hearts from dependent autophagic dysregulation governs the pathogenesis of AL cardiomyopathy patients, further supporting a decrease in auto- AL-LC-induced cellular dysfunction and death. The dysregulation of phagic clearance as the cause for impaired autophagic flux. It is autophagy leads to the accumulation of depolarized mitochondria, noteworthy that the increased p62 protein levels most likely subsequent generation of ROS, and eventual cellular dysfunction resulted from protein accumulation and not increased transcription, and cell death. The therapeutic potential of autophagy-related as we find that p62 mRNA levels are not increased in AL cardiomy- targets was evident following rescue of AL-LC-induced mortality opathy patient hearts compared to control hearts (Supplementary in vivo not only using rapamycin, but also following transient over- Fig S7). expression of TFEB. Our temporal studies suggest that lysosomal Recent studies have reported beneficial effects of rapamycin insufficiency is among the earliest events that occur in response to through autophagy activation in a number of experimental disease AL-LC, and is subsequently followed by dysregulation of autophagy, models (Bove et al, 2011; Cortes et al, 2012; Cai & Yan, 2013). mitochondrial dysfunction, ROS production, and ultimately overt Restoration of autophagic flux by rapamycin in our system was cellular death and dysfunction. In conjunction with the evidence of associated with protection against AL-LC-induced pathology. The profound lysosomal-dependent dysregulation of autophagy in beneficial effects of rapamycin were negated with chloroquine, patients with AL amyloid cardiomyopathy, these studies highlight which neutralizes lysosomal pH thereby inhibiting autophagosome the potential of targeting lysosomal-mediated autophagy as the clearance, supporting the conclusion that the mechanism of action treatment of the AL amyloid cardiomyopathy patients.

G H Figure 4. Lysosomal dysfunction contributes to AL-LC-induced dysregulation of autophagy and consequent cellular dysfunction and death in vitro and in vivo. ▸ A Lysosomal labeling using LysoTracker red in cardiomyocytes following exposure to vehicle, Con-LC or AL-LC is shown (top panels), with corresponding bright field images shown below. Quantitation of fluorescent signal is shown in the graph on the right. Scale bar = 50 lm. N = 6.*P = 0.009. B Alterations in lysosomal pH were measured using LysoSensor in cardiomyocytes treated with vehicle, Con-LC or AL-LC. pH changes were measured by calculating the ratio between green (basic) and red (acidic) LysoSensor signal. Quantitation is shown on the right indicating loss of lysosomal acidity in AL-LC treated cardiomyocytes. Scale bar = 10 lm. N = 8.*P = 0.038. C Quantitative PCR analysis of cardiomyocytes following 24-h exposure to vehicle, Con-LC or AL-LC reveals changes in mRNA encoding the lysosomal gene products cathepsin D and vacuolar ATPase subunits 1 and 2. All three targets were significantly downregulated in the AL-LC group. N = 3.*P = 0.015, #P = 0.002, 4 **P = 1.9 × 10À . D Quantitative PCR analysis of cardiomyocytes following 24-h exposure to vehicle, Con-LC and AL-LC reveals decreased mRNA level of the lysosomal transcriptional factor TFEB. N = 4.*P = 0.003. E Protein expression of TFEB in cardiomyocytes was measured using immunoblot analysis following 24-h exposure to vehicle, Con-LC or AL-LC. AL-LC-induced downregulation of TFEB protein expression was prevented by rapamycin treatment. N = 3.*P = 0.043, #P = 0.032. F TFEB was overexpressed in cardiomyocytes using adenovirus, and adenoviral GFP overexpression was used as a control. Contractile function was measured following exposure to vehicle or AL-LC. In cells overexpressing TFEB, AL-LC-induced decreased cell shortening was rescued compared to GFP-expressing myocytes. N = 5. *P = 0.002. G TFEB was overexpressed in cardiomyocytes using adenovirus. Adeno-GFP was used as a control. Calcium transient amplitude was measured following exposure to either vehicle or AL-LC. In cells overexpressing TFEB, AL-LC-induced decrease in calcium amplitude was rescued compared to control groups. N = 5.*P = 0.002. H Hearts were isolated 2 days post-injection of vehicle, Con-LC or AL-LC from zebrafish overexpressing TFEB, or control mRNA. Hearts were stained for TUNEL-labeled nuclei with a DAPI counterstain. Cell death was calculated as a percent of TUNEL-positive nuclei to total cell number. Scale bar = 20 lm. N = 3–5 per group. *P = 0.003, #P = 0.002. I Kaplan–Meier analysis of survival following injection of Con-LC or AL-LC in zebrafish overexpressing control or TFEB mRNA. Survival was monitored daily. During the Figure 3. time course of transient overexpression of TFEB, AL-LC-induced mortality was rescued significantly. N = 25 per group. *P = 0.0003.

ª 2014 The Authors EMBO Molecular Medicine Vol 6 | No 11 | 2014 1499 1500 EMBO Molecular Medicine Vol 6 | No 11 | 2014 ª 2014 The Authors Jian Guan et al Autophagy and AL amyloid cardiomyopathy EMBO Molecular Medicine EMBO Molecular Medicine Autophagy and AL amyloid cardiomyopathy Jian Guan et al

AB C AB

C D

F G E H

I Figure 5. Temporal analysis of AL-LC-induced toxicity and dysregulation of autophagy. A Lysosomal labeling in cardiomyocytes following exposure to vehicle, Con-LC or AL-LC using LysoTracker red at 3, 6 and 12 h post-AL-LC exposure. Quantitation of fluorescent signal is shown in the graph. N = 3.*P = 0.011, #P = 0.032, **P = 0.029. B Dysregulation of autophagy following AL-LC exposure was monitored using a GFP-LC3 cleavage assay. Degradation of the GFP-LC3 fusion protein was monitored via immunoblotting against GFP antibody at 2, 6, and 12 h post-AL-LC exposure. A significant delay in the degradation of GFP-LC fusion protein was seen by accumulation of GFP-LC fusion protein in cardiomyocytes starting 6 h following exposure to AL-LC. N = 5.*P = 0.003, #P = 0.038. C Mitochondrial membrane potential was measured using TMRE fluorescent dye in cardiomyocytes 6, 12 and 24 h following treatment with vehicle or AL-LC. TMRE fluorescence signal was quantified and summarized in the graph. AL-LC exposure resulted in loss of mitochondrial membrane potential compared to the other groups at 12 and 24 h post-AL-LC exposure. N = 3.*P = 0.020, #P = 0.029. D Using fluorescent indicator DCFDA, ROS was measured in isolated cardiomyocytes treated with vehicle, Con-LC and AL-LC for 3, 12 and 24 h. Quantitative analysis is summarized in the graph. AL-LC but not Con-LC increased ROS only at 24 h following AL-LC exposure. N = 3.*P = 0.004. E Schematic illustration of temporal events involved in AL-LC-induced pathology. Source data are available online for this figure. Figure 4.

ª 2014 The Authors EMBO Molecular Medicine Vol 6 | No 11 | 2014 1501 1502 EMBO Molecular Medicine Vol 6 | No 11 | 2014 ª 2014 The Authors Jian Guan et al Autophagy and AL amyloid cardiomyopathy EMBO Molecular Medicine EMBO Molecular Medicine Autophagy and AL amyloid cardiomyopathy Jian Guan et al

AB C AB

C D

F G E H

AB and breeding of zebrafish were conducted as described previously pre-chilled methanol at 20°C for 30 min and incubated with TUNEL À (Guan et al, 2013; Mishra et al, 2013). reaction mixture (Roche) in a moisture chamber for 1 h at 37°C. Slides were washed with PBS for three times and mounted in anti-fade Chemicals and reagents medium containing DAPI (Vectorlabs). For detection of cell death in vivo, individual hearts were dissected from zebrafish in Tyrode’s General chemicals and reagents were obtained from Sigma unless solution containing 3% BSA. Hearts were transferred to a microwell otherwise specified. Mito-TEMPO was from Santa Cruz Biotechnol- plate and fixed in 4% paraformaldehyde for 20 min. Hearts were ogy. Low-glucose DMEM with phenol red or without phenol red, rinsed in PBS and permeabilized overnight in PBS with 0.1% Tween TMRE, LysoTracker Red, LysoSensor Blue/Yellow, DCFDA and at 4°C. Hearts were washed three times in PBS, and cell death was Laminin were acquired from Invitrogen. Rapamycin was from Cell detected using TUNEL reaction mixture (Roche) in a moisture cham- Signaling Technology. Trypsin and Collagenase were purchased ber for 1 h at 37°C. Hearts were washed with PBS and placed directly from Willington. Antibodies were obtained from Santa Cruz Biotech- into mounting medium containing DAPI. For cardiomyocytes, images nology (TFEB), Cell signaling Technology (total and phosphoryla- were acquired with excitation wavelengths of 405 and 488 nm, ted S6 Kinase, COX4), MBL (LC3), R&D Systems (GAPDH), Sigma and 4–5 pictures were taken from each slide using Axiovision (b-actin), Abcam (active caspase 3) and Abnovo (p62). Secondary fluorescence microscope (Zeiss). For whole fish hearts, images were antibodies for immunohistochemistry (donkey anti-mouse antibody- taken with excitation wavelengths of 405 and 555 nm, and images Alexa Fluor 555 and donkey anti-rabbit antibody-Alexa Fluor 488) were taken using LSM700 confocal microscopy (Zeiss). Percent of were from Invitrogen. Adenovirus purification kit, ATPlite kit and apoptotic cell death was calculated as TUNEL-positive nuclei divided CD TUNEL kit were purchased from Adenopure, Perkin Elmer and by total nuclei. TUNEL-positive nuclei were manually counted, and Roche, respectively. Full-length human TFEB-GFP adenovirus was the total nuclei were counted using ImageJ software (NIH). All of the purchased from Vector BioLabs, and adeno-GFP virus was used as a counting was performed in a blinded fashion. Expression of active control. MOIs of the two adenovirus were adjusted to infect the adult caspase 3 for both cell lysates and zebrafish lysate was determined cardiomyocytes for achieving an equal expression level of GFP as using immunoblotting against active caspase 3. described previously. Experiments were started 24 h following adenovirus infection. Contractile function and intracellular calcium RNA isolation and quantitative PCR measurements were performed 24 h following AL-LC exposure. To measure gene expression in mRNA level, total RNA was isolated Cardiomyocyte isolation and culture using Trizol (Invitrogen) extraction method. Prior to synthesize cDNA, DNAase treatment was performed subsequently to remove As previously described (Jain et al, 2003), rat ventricular cardiomyo- the residual DNA contamination (Turbo DNAase, Ambion). iScriptTM cytes were isolated from adult male Wistar rats using a collagenase- cDNA Synthesis Kit (Bio-Rad) was used for first-strand cDNA Figure 6. Lysosomal dysfunction and autophagic dysregulation in AL amyloid cardiomyopathy. based enzymatic digestion method. Cardiomyocytes were treated synthesis. Quantitative PCR was performed using standard curve A Transmission electron micrographs of cardiac tissue isolated from donor (top panels) or AL cardiomyopathy patients (bottom panels). Disruption in muscle with vehicle (ultrapure water), 20 lg/ml of Con-LC or AL-LC at method using the iCycler PCR (Bio-Rad). The primers are as follows: organization, abnormal mitochondria and double-membrane structures (autophagosomes) are prevalent in heart tissue of AL amyloid cardiomyopathy patients designated time points as described in the Results section. Neonatal for rat cardiomyocytes, TFEB forward primer: CTCGAAGTCGGGG and absent in tissue from healthy donors. Scale bar = 500 nm. rat cardiomyocytes were isolated from 1- to 2-day-old Wistar rats AACTAGG, reverse primer: CTGCAGTCGAGGGAAGACAG; GAPDH B–D Immunoblot analysis for expression of autophagic markers LC3-II (B) and p62 (C) as well as TFEB (D) protein expression in heart tissue isolated from donor or AL cardiomyopathy patients. Immunoblotting revealed increased LC3-II and p62 expression, as well as decreased TFEB protein expression in AL amyloid (Charles River Laboratory #003) as previously described (Guan et al, forward primer: GGTGATGCTGGTGCTGAGTA, reverse primer: cardiomyopathy patients compared to healthy donors. N = 5 (B), N = 6 (C and D). *P = 0.005 (B and C), *P = 0.012 (D). 2013). TTGCTGACAATCTTGAGGGA; cathepsin D (Cts D) forward primer: Source data are available online for this figure. GTGGCTTCATGGGGATGGAC, reverse primer: GGAGCAAGTTA Cell contractility measurement and intracellular GAGTGTGGCA; vacuolar ATPase subunit 1 (VOA1) forward primer: calcium measurements TCTCCACCCATTCAGAGGAC, reverse primer: CCTTCCATGATCAG CAGGAT; vacuolar ATPase subunit 2 (VOA2) forward primer: CAG Materials and Methods control human hearts were purchased from the National Disease Cellular contractile function was measured in cultured adult ventricu- TTCCGAGACCTCAACCA, reverse primer: GTTTAACAGGTGGTGC Research Interchange. Additional information regarding human lar cardiomyocytes using video edge detection, and intracellular GGGA; for fish tissue, TFEB forward primer: GCCACGAGAACGA Human tissues and light chain protein samples is listed in Supplementary Table S2. calcium levels were determined with calcium-sensitive fluorescent dye GATGGAT, reverse primer: GCAGATCCAGACTACCGGGG; and Fura-2, as described previously (Shi et al, 2010). Following treatment EF1a forward primer: CTGGAGGCCAGCTCAAACATGG, EF1a All procedures related to human light chain protein and heart tissues Animal care or addition of light chain protein, cardiomyocytes were perfused with reverse primer: ACTCGTGGTGCATCTCAACAGACT. were reviewed and approved by the Institution Review Board (IRB) 1.2 mmol/l Ca2+ Tyrode’s buffer at 37°C under pacing at 5 Hz. at Boston University School of Medicine and Massachusetts General All animal (rat and zebrafish) procedures were reviewed and Percentage cellular shortening was calculated as the ratio of the differ- Immunohistochemistry Hospital. Bence-Jones proteins, including amyloidogenic LC isolated approved by the Institutional Animal Care and Use Committee at ence between systolic and diastolic cell length over diastolic cell from AL amyloid patients (AL-LC) and non-amyloidogenic LC Harvard Medical School. Rats and zebrafish were housed in Associa- length. Calcium transient amplitude was calculated as the difference in Following treatment, adult cardiomyocytes were washed twice with isolated from non-amyloidosis multiple myeloma patients (Con-LC), tion for Assessment and Accreditation of Laboratory Animal Care the intracellular calcium levels between systolic and diastolic phases. 1× PBS and then fixed/permeabilized with acetone/methanol (1:2) were obtained from urine purification in collaboration with Boston (AAALC)-accredited animal care facilities under a 12-h light–dark 4–6 cells were measured per biological replicate, and three biological solution at 20°C for 20 min. Incubation with 3% BSA solution for À University Amyloidosis Center (Liao et al, 2001; Connors et al, cycle and were fed with laboratory chow. Adult rats for cardiomyo- replicates were performed and grouped for statistical analysis. 1 h at room temperature was performed to minimize non-specific 2007). Immunoblotting was used to determine the purity of LC cyte isolation were purchased from Charles River Laboratory binding. Cells were then incubated with two primary antibodies proteins as described previously (Connors et al, 2007). Additional (male Wistar rats, 180–220 g, catalog #003). Neonatal rats for Cell death assays (anti-p62 [1:100] and COX4 [1:1,000]) at 4°C for 18 h followed by information regarding LC protein is listed in Supplementary Table cardiomyocyte isolation were purchased from Charles River subsequent incubation with secondary antibodies (donkey anti- S1. Explanted hearts of patients with AL amyloid cardiomyopathy Laboratory (Wistar rats, p1-p2, catalog #003). Wild-type zebrafish Cultured cardiomyocytes were washed with PBS and immediat- mouse Alexa Fluor 555 [1:300] and donkey anti-rabbit Alexa Fluor were collected at the Massachusetts General Hospital. Non-disease were purchased from Ekkwill Waterlife Resources (Ruskin). Care ely fixed in 4% paraformaldehyde. Cells were permeabilized in 488 [1:300]) for 1 h at 37°C to detect p62 and COX4, respectively.

ª 2014 The Authors EMBO Molecular Medicine Vol 6 | No 11 | 2014 1503 1504 EMBO Molecular Medicine Vol 6 | No 11 | 2014 ª 2014 The Authors Jian Guan et al Autophagy and AL amyloid cardiomyopathy EMBO Molecular Medicine EMBO Molecular Medicine Autophagy and AL amyloid cardiomyopathy Jian Guan et al

After final washing with PBS, the slides were mounted with VECTA- determination of mitochondrial-derived ROS, cardiomyocytes were Assessment of cardiac function in zebrafish SHIELD mounting media (Vector Lab). Zeiss LSM700 fluorescence pre-treated with Mito-TEMPO (Santa Cruz Biotech), a mitochondrial- The paper explained confocal microscope was used to visualize p62 (Ex/Em: 550/ specific ROS scavenger, at a concentration of 100 nM for 45 min Either Con-LC or AL-LC (100 lg/ml) was introduced into zebrafish 600 nm) and COX4 (Ex/Em: 488/525 nm) with 63× lens. DAPI was prior to experimental manipulation. circulation via venous injection as previously described. Following Problem used to stain for the nuclei. randomization, fish were treated with vehicle or rapamycin [10 nM] Amyloid diseases are a family of protein misfolding diseases that result in fibril deposition in various organs throughout the body. AL Autophagic flux measurement (Tobin & Beales, 2008). 5dpf zebrafish were embedded in 4% low- amyloidosis is the most common systemic amyloid disease and is Immunoblot melting agarose (Invitrogen) made with E3 water. E3 water was characterized by over-production of abnormal light chain (AL-LC) For autophagosome clearance, adult rat cardiomyocytes were added to a level of 2 mm above the agarose. Color Doppler echocar- proteins. These amyloidogenic precursor proteins possess an inherent For rat cardiomyocytes, protein was extracted using cell lysis buffer treated with either Con-LC or AL-LC (20 lg/ml) or vehicle for 24 h. diography was performed using MS700 probe (Vevo2100, Visual- proteotoxicity that contributes to the development of a fatal AL (Cell Signaling) with 1 mM PMSF (Sigma) and then subjected to Autophagy-specific substrate p62 was measured with immunoblot- Sonics) to determine the peak aortic flow velocity at 50 MHz. The amyloid cardiomyopathy. However, the molecular mechanisms underlying AL-LC-associated cardiac proteotoxicity remain unknown. sonication. The protein concentration was determined by Dc ting. For autophagosome generation rate, adult rat cardiomyocytes color Doppler gate was placed on the dorsal edge of ventricle. The protein assay (Bio-Rad). For zebrafish, 15 embryos were suspended were treated with either AL-LC (20 lg/ml) or vehicle for 48 h in the maximal flow velocity of each fish was acquired in a blinded fash- Results directly in 50-ll SDS loading buffer and homogenized using a presence or absence of lysosomal inhibitors (E64d and Pepstatin A ion. Per each animal, the acquisition part starting at embedding was Here, we find that the cell death and cardiomyocyte dysfunction asso- tissue homogenizer (TissueLyser II, Qiagen). Following homogeni- at the concentration of 5 lg/ml) (Hamacher-Brady et al, 2006). kept within 3 min to ensure fish health during time of measure- ciated with AL-LC proteotoxicity are caused by a dysregulation in auto- zation, the samples were centrifuged and total protein homogen- LC3-II levels were then measured with immunoblotting. ment. 6–8 fish were examined for each group. phagy, a cellular housekeeping process critical for maintaining ates were obtained. 30 lg of total protein or 18 ll of fish protein homeostasis. Using both cellular and zebrafish models of AL-LC toxicity, lysate was loaded onto Criterion XT bis-tris precast gels (4–12%) GFP-LC3 cleavage assay Transient TFEB overexpression zebrafish model we show that AL-LC proteins cause a defect in autophagic flux, specifically in the clearance phase, with impaired lysosomal function. Lyso- – (Invitrogen) or PAGEr Gold precast gels (4 20%) (Lonza) for elec- somal dysfunction and autophagic dysregulation were similarly evident trophoresis. Protein was electrotransferred to a PVDF membrane Neonatal cardiomyocytes were infected with GFP-LC3 adenovirus. A pair of primers was designed to amplify fish TFEB mRNA from in cardiac tissue explanted from human patients with AL amyloid-asso- (Millipore) at 30 volts for 16–18 h at 4°C. After blocking in 5% Twenty-four hours following adenoviral infection, neonatal cardio- whole fish mRNA samples: upstream primer: ATTTAGGTGACAC ciated cardiomyopathy. Restoration of autophagic flux through genetic BSA in PBS, proteins of interest were detected by incubation with myocytes were exposed to either vehicle, Con-LC or AL-LC for 2, 6 TATAGAAATGTCGTCACGCATCGGCCT; downstream primer: CCG restoration of lysosomal function or pharmacologic manipulation with appropriate primary antibodies overnight at 4°C. After washing, or 12 h. Following 1× PBS wash, cells were manually harvested CTCGAGTCACTGTATATC. mMESSAGE mMACHINE Kit (Invitro- the small molecule rapamycin protected against AL-LC proteotoxicity and the development of AL amyloid cardiomyopathy. blots were incubated with corresponding secondary antibodies. using a cell lifter. Protein homogenate from harvested lysed cells are gen) was used to synthesize zebrafish TFEB mRNA. Reverse TFEB Odyssey infrared scanner (Li-Cor) was used to determine the infra- then subjected to immunoblotting for GFP. GFP antibody was used mRNA was synthesized for control. The quality/quantity of mRNA Impact red fluorescent signal, and GAPDH was used as a reference gene to detect the presence of GFP-LC3 fusion protein. Degradation of the was determined by Nanodrop spectrometer (Thermo Scientific). Our studies illustrate the cellular defects underlying the pathogenesis for normalization. fusion protein is reduced under conditions when autophagic flux is 20 pg of mRNA was injected into zebrafish embryos at single cell of AL amyloidosis-associated proteotoxicity and highlight the thera- inhibited (Ni et al, 2011). stage. AL-LC was introduced into fish circulation via venous injec- peutic potential of rapamycin in the treatment of AL amyloid cardio- Mitochondrial membrane potential and ATP measurement tion 2 days post-fertilization. Cardiac cell death was determined at myopathy. LysoTracker and LysoSensor staining 4dpf. Fish survival was monitored until day 7 post-fertilization. Following Con-LC, AL-LC (20 lg/ml) or vehicle administration for designated number of hours, cultured cardiomyocytes were incu- Following designated hour treatment with vehicle, Con-LC and Statistical analysis bated with cell permeable, mitochondrial membrane potential- AL-LC (20 lg/ml), cultured cardiomyocytes were incubated with Author contributions sensitive fluorophore TMRE (Invitrogen) at the concentration of cell permeable, lysosomal-specific probe LysoTracker (Invitrogen) All data are shown as mean standard error. Statistical differences JG and SM designed, conducted, analyzed and interpreted the data as well as Æ 10 nM for 30 min. Cardiomyocytes were then washed with warm at a concentration of 100 nM for 30 min. Cardiomyocytes were between mean values for two groups were evaluated by Student’s drafted the manuscript. These two authors contributed to this manuscript PBS 2 times. TMRE fluorescence was acquired with excitation washed twice with warm PBS. Cell images were acquired using t-test using GraphPad Prism software and confirmed using Microsoft equally and their names are listed in alphabetical order. YQ and JS conducted, wavelengths of 555 nm, and 4–5 pictures were taken from each LSM700 confocal microscopy (excitation wavelength at 555 nm) Excel. Individual P-values are denoted within the figure legends. analyzed and interpreted the data. KT, GL, ML and OSS assisted in experimen- dish using LSM700 confocal microscopy (Zeiss). Mean fluorescence and analyzed with ImageJ software. For determination of lysosomal P < 0.05 was considered as significant. tal design and data interpretation related to mitochondrial biology and auto- intensity of individual cardiomyocytes was determined per picture pH, cardiomyocytes were incubated with LysoSensor Blue/Yellow phagy processes. They also made critical suggestions toward the writing of the with ImageJ software (NIH). Cellular ATP level was determined (Invitrogen) at a concentration of 1 lM for 3 min prior to measure- Supplementary information for this article is available online: manuscript. LHC and DCS provided human light chain proteins and critical with ATPlite kit according to the manufacture’s manual. Briefly, ment using LSM710 two-photon confocal microscopy (excitation http://embomolmed.embopress.org suggestions for data interpretation, as well as edited the manuscript. RHF cardiomyocytes were cultured on a 12-well plate. Following 24-h wavelength was 720 nm, and emission wavelengths were collected provided critical suggestions and was consulted for correlation of experimental treatment of LCs, warm PBS was used to wash cells gently. 75 ll of from 400 to 461 nm for the blue emission and 510 to 630 nm for the Acknowledgments data with clinical observations. CAM assisted in experimental design and data PBS was added to each well, followed by 75 ll of cell lysis buffer. yellow emission). Images were analyzed using imageJ soft- We would like to thank Drs. Eva Plovie and Deepa Mishra for assistance with in interpretation as well as supervised the experiments related to zebrafish stud- Plate was subjected to 5-min shaking at 700 r.p.m. to break up the ware. Green fluorescence signal (blue emission) represents basic vivo experiments. We would like to thank Dr. Alaattin Kaya for his help in ies and helped in the editing of the manuscript. RL initiated the project and plasma membrane. 150 ll of ATP luminescent reaction buffer with conditions, and red fluorescence signal (yellow emission) represents native gel electrophoresis. We would like to thank Dr. Federica del Monte for was responsible for the overall experimental design, data interpretation and substrate was added into each well, and the luminescence signal acidic conditions. Pseudocoloring (the ratio of green/red) was done her helpful scientific discussion and her generosity in providing human manuscript writing. was measured with a SpectraMax M5 Microplate Reader (Molecular to represent lysosomal pH. samples. We would like to thank Drs. Judith Gwathmey, Thomas E. Macgillivray, Device). Marc J Semigran and G William Dec Jr for their generosity in providing human Conflict of interest Electron microscopy heart samples. We would like to thank Ms. Gloria Chan from the Boston The authors declare that they have no conflict of interest. ROS measurement University Amyloidosis Center for preparation of human light chain proteins. 1-mm3 cubes of human heart samples from either non-failing We would like to thank Ms. Amy Cui for her technical help. This work was Following Con-LC or AL-LC (20 lg/ml) or vehicle administration control or AL amyloid cardiomyopathy patients, or whole zebra- supported in part by funding from the National Institutes of Health, HL088533, References for designated number of hours, cultured cardiomyocytes were fish embryos that had been injected with Con-LC or AL-LC, were HL086967, HL093148, HL099073 (R.L.), 1RC1DK90696 (D.C.S.), 5R01AG031804 incubated with cell permeable, redox-sensitive fluorophore DCFDA fixed with 2.5% glutaraldehyde overnight at 4° and then embed- (L.H.C.), HL109264 (C.A.M.) as well as the Demarest Llyod Jr. Foundation and the Bhuiyan MS, Pattison JS, Osinska H, James J, Gulick J, McLendon PM, Hill JA, (Invitrogen) at the concentration of 20 lM for 30 min. Cardio- ded in epoxy resins. Ultrathin sections (80 nm) were stained with Cardiac Amyloidosis Program, Brigham and Women’s Hospital (R.H.F., R.L.) and Sadoshima J, Robbins J (2013) Enhanced autophagy ameliorates cardiac myocytes were then washed with warm PBS 2 times. Cell images uranyl acetate/lead citrate and then examined under the Tecnai the Gruss and Wildflower Foundations and the Amyloid Research Fund at proteinopathy. J Clin Invest 123: 5284 – 5297 were acquired using LSM700 confocal microscopy (excitation G² electron microscope (FEI Inc) in Harvard Medical School EM Boston University (L.H.C., D.C.S.). S.M. is supported by National Institution of Bove J, Martinez-Vicente M, Vila M (2011) Fighting neurodegeneration with wavelength at 488 nm) and analyzed with SigmaScan Pro. For core facility. Health T32 postdoctoral fellowship award (T32HL007604). rapamycin: mechanistic insights. Nat Rev Neurosci 12: 437 – 452

ª 2014 The Authors EMBO Molecular Medicine Vol 6 | No 11 | 2014 1505 1506 EMBO Molecular Medicine Vol 6 | No 11 | 2014 ª 2014 The Authors Jian Guan et al Autophagy and AL amyloid cardiomyopathy EMBO Molecular Medicine EMBO Molecular Medicine Autophagy and AL amyloid cardiomyopathy Jian Guan et al

Brenner DA, Jain M, Pimentel DR, Wang B, Connors LH, Skinner M, Apstein CS, response of light chain (AL) amyloidosis diagnosed in patients with Liao R (2004) Human amyloidogenic light chains directly impair previous diagnosis of multiple myeloma. Mayo Clin Proc 85: 232 – 238 cardiomyocyte function through an increase in cellular oxidant stress. Circ Merlini G, Seldin DC, Gertz MA (2011) Amyloidosis: pathogenesis and new Res 94: 1008 – 1010 therapeutic options. J Clin Oncol 29: 1924 – 1933 Cai Z, Yan LJ (2013) Rapamycin, autophagy, and Alzheimer’s disease. J Migrino RQ, Hari P, Gutterman DD, Bright M, Truran S, Schlundt B, Phillips SA Biochem Pharmacol Res 1: 84 – 90 (2010) Systemic and microvascular oxidative stress induced by light chain Codogno P (2014) Shining light on autophagy. Nat Rev Mol Cell Biol 15: amyloidosis. Int J Cardiol 145: 67 – 68 153 Migrino RQ, Truran S, Gutterman DD, Franco DA, Bright M, Schlundt B, Connors LH, Jiang Y, Budnik M, Theberge R, Prokaeva T, Bodi KL, Seldin DC, Timmons M, Motta A, Phillips SA, Hari P (2011) Human microvascular Costello CE, Skinner M (2007) Heterogeneity in primary structure, dysfunction and apoptotic injury induced by AL amyloidosis light chain post-translational modifications, and germline gene usage of nine proteins. Am J Physiol Heart Circ Physiol 301:H2305 – H2312 full-length amyloidogenic kappa1 immunoglobulin light chains. Mishra S, Guan J, Plovie E, Seldin DC, Connors LH, Merlini G, Falk RH, Macrae Biochemistry 46: 14259 – 14271 CA, Liao R (2013) Human amyloidogenic light chain proteins result in Cortes CJ, Qin K, Cook J, Solanki A, Mastrianni JA (2012) Rapamycin delays cardiac dysfunction, cell death, and early mortality in zebrafish. Am J disease onset and prevents PrP plaque deposition in a mouse model of Physiol Heart Circ Physiol 305:H95 – H103 Gerstmann-Straussler-Scheinker disease. J Neurosci 32: 12396 – 12405 Mizushima N, Yoshimori T, Levine B (2010) Methods in mammalian Decressac M, Mattsson B, Weikop P, Lundblad M, Jakobsson J, Bjorklund A autophagy research. Cell 140: 313 – 326 (2013) TFEB-mediated autophagy rescues midbrain dopamine neurons Ni HM, Bockus A, Wozniak AL, Jones K, Weinman S, Yin XM, Ding WX (2011) from alpha-synuclein toxicity. Proc Natl Acad Sci U S A 110:E1817 – E1826 Dissecting the dynamic turnover of GFP-LC3 in the autolysosome. Dehay B, Bove J, Rodriguez-Muela N, Perier C, Recasens A, Boya P, Vila M Autophagy 7: 188 – 204 (2010) Pathogenic lysosomal depletion in Parkinson’s disease. J Neurosci Nixon RA (2013) The role of autophagy in neurodegenerative disease. Nat 30: 12535 – 12544 Med 19: 983 – 997 Eskelinen EL, Saftig P (2009) Autophagy: a lysosomal degradation pathway Pastore N, Blomenkamp K, Annunziata F, Piccolo P, Mithbaokar P, Maria Sepe with a central role in health and disease. Biochim Biophys Acta 1793: R, Vetrini F, Palmer D, Ng P, Polishchuk E et al (2013) Gene transfer of 664 – 673 master autophagy regulator TFEB results in clearance of toxic protein and Falk RH (2005) Diagnosis and management of the cardiac amyloidoses. correction of hepatic disease in alpha-1-anti-trypsin deficiency. EMBO Mol Circulation 112: 2047 – 2060 Med 5: 397 – 412 Falk RH (2011) Cardiac amyloidosis: a treatable disease, often overlooked. Settembre C, Di Malta C, Polito VA, Garcia Arencibia M, Vetrini F, Erdin S, Circulation 124: 1079 – 1085 Erdin SU, Huynh T, Medina D, Colella P et al (2011) TFEB links autophagy Guan J, Mishra S, Shi J, Plovie E, Qiu Y, Cao X, Gianni D, Jiang B, Del Monte F, to lysosomal biogenesis. Science 332: 1429 – 1433 Connors LH et al (2013) Stanniocalcin1 is a key mediator of amyloidogenic Shi J, Guan J, Jiang B, Brenner DA, Del Monte F, Ward JE, Connors LH, Sawyer DB, light chain induced cardiotoxicity. Basic Res Cardiol 108: 378 Semigran MJ, Macgillivray TE et al (2010) Amyloidogenic light chains induce Hamacher-Brady A, Brady NR, Gottlieb RA (2006) Enhancing macroautophagy cardiomyocyte contractile dysfunction and apoptosis via a non-canonical protects against ischemia/reperfusion injury in cardiac myocytes. J Biol p38alpha MAPK pathway. Proc Natl Acad Sci U S A 107: 4188 – 4193 Chem 281: 29776 – 29787 Shin JT, Ward JE, Collins PA, Dai M, Semigran HL, Semigran MJ, Seldin DC Harris H, Rubinsztein DC (2012) Control of autophagy as a therapy for (2012) Overexpression of human amyloidogenic light chains causes heart neurodegenerative disease. Nat Rev Neurol 8: 108 – 117 failure in embryonic zebrafish: a preliminary report. Amyloid 19: 191 – 196 Jain M, Brenner DA, Cui L, Lim CC, Wang B, Pimentel DR, Koh S, Sawyer DB, Sikkink LA, Ramirez-Alvarado M (2010) Cytotoxicity of amyloidogenic Leopold JA, Handy DE et al (2003) Glucose-6-phosphate dehydrogenase immunoglobulin light chains in cell culture. Cell Death Dis 1:e98 modulates cytosolic redox status and contractile phenotype in adult Tobin JL, Beales PL (2008) Restoration of renal function in zebrafish models cardiomyocytes. Circ Res 93:e9 – e16 of ciliopathies. Pediatr Nephrol 23: 2095 – 2099 Kubli DA, Gustafsson AB (2012) Mitochondria and mitophagy: the yin and Wong E, Cuervo AM (2010) Autophagy gone awry in neurodegenerative yang of cell death control. Circ Res 111: 1208 – 1221 diseases. Nat Neurosci 13: 805 – 811 Liao R, Jain M, Teller P, Connors LH, Ngoy S, Skinner M, Falk RH, Apstein CS (2001) Infusion of light chains from patients with cardiac amyloidosis License: This is an open access article under the causes diastolic dysfunction in isolated mouse hearts. Circulation 104: terms of the Creative Commons Attribution 4.0 1594 – 1597 License, which permits use, distribution and reproduc- Madan S, Dispenzieri A, Lacy MQ, Buadi F, Hayman SR, Zeldenrust SR, tion in any medium, provided the original work is Rajkumar SV, Gertz MA, Kumar SK (2010) Clinical features and treatment properly cited.

ª 2014 The Authors EMBO Molecular Medicine Vol 6 | No 11 | 2014 1507 Research Article

Selective clearance of aberrant tau proteins and rescue of neurotoxicity by transcription factor EB

Vinicia A Polito1,†, Hongmei Li1,†, Heidi Martini-Stoica1,2,3,†, Baiping Wang1,4, Li Yang1, Yin Xu1, Daniel B Swartzlander1, Michela Palmieri4,5, Alberto di Ronza4,5, Virginia M-Y Lee6, Marco Sardiello4,5, Andrea Ballabio4,5,7 & Hui Zheng1,3,4,*

Abstract hyperphosphorylated Tau (pTau) protein in diseased brains. Biochemical and genetic studies of the amyloid precursor protein Accumulating evidence implicates impairment of the autophagy- (APP) and presenilins support a causal role of Ab in AD pathogene- lysosome pathway in Alzheimer’s disease (AD). Recently discov- sis (reviewed by (Hardy, 2006)). Accordingly, Ab-based therapies ered, transcription factor EB (TFEB) is a molecule shown to play have been actively pursued. However, the clinical outcomes of these central roles in cellular degradative processes. Here we investigate therapies have been disappointing so far (reviewed in (Mullard, the role of TFEB in AD mouse models. In this study, we 2012)). Of note, these agents may not target the NFT pathology and demonstrate that TFEB effectively reduces neurofibrillary tangle strong evidence supports a direct role of misfolded Tau and NFTs in pathology and rescues behavioral and synaptic deficits and AD and other neurodegenerative diseases (reviewed by (Gendron & neurodegeneration in the rTg4510 mouse model of tauopathy with Petrucelli, 2009; Mandelkow & Mandelkow, 2012)). Tau, encoded no detectable adverse effects when expressed in wild-type mice. by MAPT, is typically localized to axons where it binds and stabi- TFEB specifically targets hyperphosphorylated and misfolded Tau lizes microtubules. Aberrant Tau misfolding due to hyperphos- species present in both soluble and aggregated fractions while phorylation or other alterations leads to its dissociation from leaving normal Tau intact. We provide in vitro evidence that this microtubules followed by aggregation and redistribution to cell effect requires lysosomal activity and we identify phosphatase and bodies and dendrites. Clinically, NFT pathology correlates with tensin homolog (PTEN) as a direct target of TFEB that is required dementia better than amyloid plaques (Giannakopoulos et al, 2003). for TFEB-dependent aberrant Tau clearance. The specificity and Although no Tau mutations have been found in AD, mutations in efficacy of TFEB in mediating the clearance of toxic Tau species the MAPT gene are causal for a subtype of frontotemporal dementia makes it an attractive therapeutic target for treating diseases of (FTD), termed FTD with Parkinsonism linked to chromosome 17 tauopathy including AD. (FTDP-17). These mutations are known to impair Tau structure and promote its fibrillization (Gendron & Petrucelli, 2009; Mandelkow & Keywords Alzheimer’s disease; tauopathy; TFEB; PTEN; autophagy-lysosomal Mandelkow, 2012). Experimentally, overexpression of FTD-associated pathway MAPT mutant genes in transgenic mice results in NFT development Subject Categories Neuroscience and neurodegeneration (Ramsden et al, 2005; Santacruz et al, DOI 10.15252/emmm.201303671 | Received 14 November 2013 | Revised 26 2005), establishing the neurotoxicity conferred by the mutant Tau June 2014 | Accepted 30 June 2014 | Published online 28 July 2014 proteins. As such, there is increasing interest in developing EMBO Mol Med (2014) 6: 1142–1160 Tau-based therapy for treating diseases of tauopathy including AD and FTD (reviewed in (Brunden et al, 2009)). Macroautophagy (herein referred to as autophagy) is a conserved Introduction mechanism that cells utilize to degrade intracellular long-lived proteins and organelles through lysosome-mediated degradation. Alzheimer’s disease (AD) is characterized by the presence of Accumulating evidence has implicated an impaired autophagy extracellular amyloid plaques consisting of b-amyloid peptides and lysosomal pathway (ALP) in neurodegenerative diseases (see (Ab) and intracellular neurofibrillary tangles (NFT) composed of (Harris & Rubinsztein, 2012; Nixon & Yang, 2011) for recent reviews).

1 Huffington Center on Aging, Baylor College of Medicine, Houston, TX, USA 2 Interdepartmental Program of Translational Biology and Molecular Medicine, Baylor College of Medicine, Houston, TX, USA 3 Medical Scientist Training Program, Baylor College of Medicine, Houston, TX, USA 4 Department of Molecular and Human Genetics, Baylor College of Medicine, Houston, TX, USA 5 Dan and Jan Duncan Neurological Research Institute, Texas Children’s Hospital, Houston, TX, USA 6 Department of Pathology and Lab Medicine, University of Pennsylvania School of Medicine, Philadelphia, PA, USA 7 Department of Translational Medical Sciences, Section of Pediatrics, Telethon Institute of Genetics and Medicine, Federico II University, Naples, Italy *Corresponding author. Tel: +1 713 798 1568; Fax: +1 713 798 1610; E-mail: [email protected] †These authors contributed equally to this work

1142 EMBO Molecular Medicine Vol 6 | No 9 | 2014 ª 2014 The Authors. Published under the terms of the CC BY 4.0 license Vinicia Assunta Polito et al Potent targeting of tauopathy by TFEB EMBO Molecular Medicine EMBO Molecular Medicine Potent targeting of tauopathy by TFEB Vinicia Assunta Polito et al

A B Specific to AD, the ALP has been documented to regulate APP turn- also showed that TFEB had no appreciable effect on amyloid deposi- over and Ab metabolism (Mueller-Steiner et al, 2006; Nixon, 2007; tion (Supplementary Fig S2C–F). Pickford et al, 2008; Rohn et al, 2011; Yang et al, 2011) as well as In sharp contrast, using antibodies against phospho-Tau (AT8, Tau protein degradation (Wang et al, 2009, 2010; Kruger et al, S202/T205 and PHF1, S396/S404) or conformation-specific Tau 2012; Schaeffer et al, 2012; Caccamo et al, 2013; Ozcelik et al, (MC1) species, we found that TFEB treatment drastically reduced 2013). A physiological role of autophagy in Tau homeostasis is phospho-Tau (pTau) and NFT-like pathologies in both the cortex demonstrated through pTau accumulation and neurodegeneration (Fig 1A and quantified in 1B) and hippocampus (Supplementary Fig in mice with neuronal deletion of the autophagy gene Atg7 (Inoue S3) of rTg4510 (Tau) mice. Reduction of MC1-positive Tau supports et al, 2012). a role of TFEB in recognizing misfolded Tau. Western blot analysis Autophagy is mediated by a series of intracellular membrane of detergent soluble protein lysates using these antibodies as well as trafficking events and is executed by lysosomal degradation of antibodies against total Tau and unphosphorylated Tau at the S202/ sequestered contents. Thus, efficient autophagy requires heightened T205 sites (Tau1), allowed us to determine that TFEB had no effect lysosomal activity. The Transcription Factor EB (TFEB) was recently on wild-type Tau proteins (Fig 1C and E, WT vs. WT + TFEB); it, discovered as a master regulator of the ALP through coordinated however, drastically reduced the AT8-, PHF1- and MC1-positive Tau expression of autophagy and lysosomal target genes and enhanced species in Tau transgenic mice (Fig 1C and F, Tau vs. Tau + TFEB). lysosomal biogenesis (Sardiello et al, 2009; Settembre et al, 2011). Furthermore, levels of TFEB overexpression inversely correlated TFEB is normally sequestered in the cytoplasm by phosphorylation, with that of PHF1-Tau (Fig 1D and F). Since the Tau1 levels C E an event shown to be mediated by the key negative autophagy regu- remained constant, reduction of AT8-positive Tau most likely indi- lator mTOR (Roczniak-Ferguson et al, 2012; Settembre et al, 2012). cates that TFEB expression leads to the degradation rather than Accordingly, mTOR inhibition is associated with TFEB dephosphor- dephosphorylation of the phospho-Tau (pTau). However and incon- ylation and nuclear translocation, leading to the induction of down- sistent with this view, total Tau levels were not significantly altered stream targets by binding to the coordinated lysosomal expression (Fig 1C and quantified in 1F), suggesting that the pTau pool may and regulation (CLEAR) element (Sardiello et al, 2009; Settembre represent only a small pool of total Tau levels or that TFEB might et al, 2011). Here we investigated the role of TFEB in APP and Tau also act on Tau phosphorylation/dephosphorylation. transgenic mouse models and found that TFEB potently cleared pTau/NFT pathologies and rescued neurodegenerative and behav- TFEB promotes the clearance of aberrant Tau species ioral deficits without overtly affecting Ab pathology or exhibiting F adverse effects in wild-type mice. To further investigate the nature of pTau regulation by TFEB, we prepared detergent-free cytosolic and sarkosyl-insoluble fractions D from Tau mouse brains with or without TFEB injection. Western Results blot analysis revealed that, consistent with the detergent extractable preparations, TFEB efficiently reduced the CP13- and PHF1-positive TFEB differentially targets Ab and pTau/NFT pathologies pTau while leaving Tau1-positive unphosphorylated Tau intact in both fractions (Fig 2A and quantified in B and C). Interestingly, We used an adeno-associated virus (AAV) delivery approach to while levels of total Tau remained similar in the soluble pool assess the potential effect of TFEB in 5xFAD (Oakley et al, 2006) (Fig 2B), they were significantly reduced in the sarkosyl-insoluble and rTg4510 (Ramsden et al, 2005; Santacruz et al, 2005) trans- preparations in TFEB-treated samples (Fig 2C). These results indi- Figure 1. Potent reduction of pTau proteins by AAV-mediated TFEB expression. genic mouse models, which develop progressive Ab and NFT cate that TFEB targets the detergent-insoluble pTau for degradation. A Immunofluorescence staining of cortex of 4-month-old rTg4510 mice either untreated (Tau) or injected with TFEB (Tau + TFEB) at P0 using phospho-Tau neuropathologies, respectively, starting at approximately 2 months This assessment is consistent with the TFEB’s role in the autophagy- antibodies AT8 (S202/T205) and PHF1 (S396/S404) or conformation-specific antibody MC1 and counter-stained with DAPI. AAV-TFEB was injected at P0 and mice of age. The AAV2/9 vector containing mouse TFEB cDNA (AAV- lysosomal pathway and is in agreement with the published reports were analyzed at 4 months. Scale bar: 200 lm. TFEB) or GFP (AAV-GFP) driven by the CMV promoter was that autophagy activators such as rapamycin or trehalose are B Quantification of staining intensities. Student’s t-test was performed to analyze the significance. n = 5 mice/genotype/treatment group. **P = 0.0011, 5 5 ***P = 6.85 × 10À , and ***P = 9.33 × 10À for AT8, PHF1 and MC1, respectively. Each bar represents average s.e.m. injected into the lateral ventricles of both cerebral hemispheres on effective in removing Tau aggregates (Schaeffer et al, 2012; Ozcelik Æ C Western blot analysis of detergent-extracted brain lysates using anti-total Tau, Tau1, AT8 or MC1 antibodies. WT: wild-type; WT + TFEB: wide-type mice injected postnatal day 0 (P0) of 5xFAD or rTg4510 mouse brains and their et al, 2013). with TFEB; Tau: rTg4510 transgenic mice; Tau + TFEB: rTg4510 transgenic mice injected with TFEB. Two exposures were displayed for total Tau: the short exposure wild-type littermate controls. Assessment of mice injected with To probe the cellular mechanisms mediating TFEB reduction of was used for quantification of Tau transgenic samples; the long exposure was used for side-by-side comparisons of the total Tau levels in WT and Tau mice and for AAV-GFP revealed widespread brain expression (Supplementary soluble pTau, we used a doxycycline inducible cell line expressing quantification of WT total Tau. c-tubulin was used as a loading control. Fig S1A), particularly in cortical and hippocampal neurons the largest human Tau isoform with the P301L mutation (T40PL). D Western blot analysis of TFEB expression in 4-month-old WT or Tau mice with (+TFEB) or without AAV-TFEB P0 injection (the same as in A and C but from another (Supplementary Fig S1B). Quantitative real-time PCR (qRT-PCR) Using the same detergent extraction protocol as that used for brain independent batch of experiments). PHF1 and total Tau antibodies were used to blot the Tau transgenic lysates for correlation with TFEB levels. E, F Quantification of relative band intensities in WT (E) or Tau (F) mice. Student’s t-test shows that TFEB protein levels are significantly increased in WT + TFEB vs WT analysis of forebrain samples 1-month post-injection showed lysates, we show that transfection of TFEB results in prominent (**P = 0.0026) and in Tau + TFEB vs Tau (*P = 0.047). AT8, PHF1 and MC1 levels are significantly reduced in Tau + TFEB vs Tau (*P = 0.032, 0.041, 0.040, approximately a twofold to threefold elevation of TFEB expression reductions of total Tau and CP13- and PHF1-positive pTau without respectively). &, significant correlation between TFEB protein levels and PHF-1 was determined using Pearson Product Moment Correlation test (P = 0.0044). n = 3/ compared to uninjected or GFP-injected controls (Supplementary affecting unphosphorylated Tau detected by the Tau1 antibody group/experiment. Each bar represents average s.e.m. Æ Fig S1C). Long-lasting TFEB expression is evidenced by similar (Fig 2D and quantified in 2E). The combined in vitro and in vivo Source data are available online for this figure. levels of TFEB overexpression when analyzed 4-month post-injection data provides a compelling argument that the primary effect of TFEB (see Fig 1D and F). is to target hyperphosphorylated and misfolded Tau proteins for The 5xFAD and rTg4510 mice were analyzed 4 months after degradation. The insignificant changes of total Tau in the soluble TFEB injection. Western blot analysis revealed no appreciable fraction in vivo may be attributed to the low percentage of pTau in Cell autonomous effects of TFEB in pTau reduction mechanism. We performed P0 injections of AAV-GFP or AAV-GFP/ changes of APP levels in TFEB-injected wild-type or 5xFAD mice as the total Tau pool. However, the possibility remains that TFEB may AAV-TFEB in Tau transgenic mice. Immunostaining using the AT8 compared to uninjected controls (Supplementary Fig S2A and B). affect Tau phosphorylation/dephosphorylation at Tau1-independent Having established a potent effect of TFEB in pTau reduction, we next antibody revealed abundant GFP/AT8-double-positive cells in GFP- Immunohistochemical staining and quantification of Ab pathology sites in the soluble pool. examine whether this reduction is mediated by a cell autonomous injected mice. In contrast, GFP-positive cells were devoid of pTau

ª 2014 The Authors EMBO Molecular Medicine Vol 6 | No 9 | 2014 1143 1144 EMBO Molecular Medicine Vol 6 | No 9 | 2014 ª 2014 The Authors Vinicia Assunta Polito et al Potent targeting of tauopathy by TFEB EMBO Molecular Medicine EMBO Molecular Medicine Potent targeting of tauopathy by TFEB Vinicia Assunta Polito et al

A A B

Figure 2. Biochemical analysis of TFEB-mediated reduction of pTau. A Western blot analysis of detergent-free (Soluble) and sarkosyl-insoluble (Insoluble) preparations of total (total Tau), unphosphorylated (Tau1) or CP13- (S202/T205) and PHF1-positive pTau species in 4-month-old Tau or Tau + TFEB mice with P0 injection. c-tubulin was used as a loading control. B, C Quantification of relative band intensities in soluble (B) and insoluble (C) fractions. n = 3 mice/genotype/treatment group. CP13 and PHF1 are significantly reduced in both soluble and insoluble fractions (Student’s t-test, ***P = 0.00072 and **P = 0.004, for B; and **P = 0.007 and **P = 0.009 for C). D Western blot analysis of total Tau, unphosphorylated Tau (Tau1) and CP13- or PHF1-positive pTau levels in response to doxycycline induction (Tau) and/or TFEB- C FLAG transfection (TFEB) in T40PL cell line. Ctrl: untreated; Tau: 0.5 lg/ml doxycycline treated for 48 h; TFEB: TFEB transfected for 48 h; Tau + TFEB: combined TFEB transfection and DOX treatment. FLAG: anti-FLAG antibody blotting for TFEB expression. c-tubulin was used as a loading control. E Quantification of relative band intensities. CP13, PHF1 and total Tau are significantly reduced by TFEB transfection (Student’s t-test, n = 3, **P = 0.0012, 0.0040, and 0.0019, respectively). Each bar represents average s.e.m. Æ Source data are available online for this figure. staining in GFP/TFEB coinjected mice (Fig 3A), suggesting that TFEB rescues neurodegeneration in rTg4510 mice TFEB mediates pTau clearance in a cell autonomous manner. The same result was also obtained when AAV-GFP/TFEB was coinjected In agreement with the published reports, immunostaining using the into 2-month-old Tau mice (Supplementary Fig S4). To corroborate neuronal marker NeuN reveal severe neurodegeneration in rTg4510 the in vivo findings, we transfected EGFP or TFEB-FLAG expression Tau mice (Fig 4A and B, WT vs. Tau). Life-long expression of TFEB vectors in the doxycycline-induced T40PL cell line and performed did not lead to detectable alteration of neuronal structure in wild-type Figure 3. Intracellular clearance of pTau pathology by TFEB. double immunofluorescence staining of GFP or TFEB with mice (WT vs. WT + TFEB), but resulted in grossly expanded hippo- A Immunohistochemical staining of hippocampus of rTg4510 (Tau) mice injected with AAV-GFP (Tau) or AAV-TFEB/AAV-GFP (Tau + TFEB) at P0 and analyzed at PHF1 every 8 h up to 40 h (Fig 3B). The results show that TFEB campal sizes in Tau transgenic background (Tau vs. Tau + TFEB). 4 months. Images are displayed as GFP only (GFP), AT8 only (AT8) or overlay of GFP/AT8/DAPI (Merge/DAPI). Right panels are enlarged images of the bracketed areas. intensities begin to negatively correlate with PHF1 staining staring at Enhanced neuronal survival is confirmed by quantifying neuronal Arrowheads indicate GFP/AT8-double-positive cells only in GFP-injected mice. Scale bar: 50 lm. n = 2 for AAV-GFP and n = 5 for AAV-TFEB/AAV-GFP; 4–5 sections/ the 16 h time point and persist to 24, 32 (not shown) and 40 h, while numbers in the CA1 area of hippocampus using unbiased stereology mouse were examined. GFP intensities show no correlation with PHF1 at all time points (Fig 4C). In fact, a beneficial effect of TFEB is readily appreciable by B Fluorescence images of GFP and immunofluorescence images of anti-FLAG compared with immunofluorescence images of anti-PHF1 in GFP or TFEB transfected and examined (Fig 3B and quantified in Fig 3C). Furthermore, consistent measuring the brain weight, which documented the same brain doxycycline induced T40PL cell line at 8, 16 or 40 h post-transfection. Merge: Overlay of GFP or TFEB-FLAG with PHF1. Insets are higher resolution images documenting that whereas overlapping GFP and PHF1 immunoreactivity can be observed at all the time points, TFEB- and PHF1-double-positive cells can only be with the in vivo results, while transfection of the GFP vector revealed weight in wild-type mice regardless of TFEB expression (Fig 4D, WT detected at 8 h but not later times. Scale bar: 200 lm; in inset: 50 lM. abundant PHF1- and GFP-double-positive cells, most of the TFEB- vs. WT + TFEB), but a significant increase in TFEB-treated Tau mice C Pearson correlation coefficients (Pearson R-values) showing no correlation between GFP and PHF1 (PHF1/GFP) at any time points, but negative correlation between positive cells exhibited reduced PHF1 intensities (insets in Fig 3B). compared to uninjected Tau controls (Tau vs. Tau + TFEB). Improved TFEB and PHF1 (PHF1/TFEB) starting at 16 h and persisting to 40 h (Student’s t-test, n = 4,*P = 0.026 and **P = 0.008 for 16 and 40 h, respectively, comparing R- These results provide strong support that TFEB mediates time-dependent overall neuronal health is evidenced by reduced neuroinflammation values for GFP/PHF1 with TFEB/PHF1). The whole view field of eight confocal projection slices per view field, four view fields per time point per transfection was used for analysis. Bar graph represents average s.e.m. pTau protein clearance through a cell autonomous mechanism. detected by Iba1 (Fig 4E) and GFAP (Fig 4F). Æ

ª 2014 The Authors EMBO Molecular Medicine Vol 6 | No 9 | 2014 1145 1146 EMBO Molecular Medicine Vol 6 | No 9 | 2014 ª 2014 The Authors Vinicia Assunta Polito et al Potent targeting of tauopathy by TFEB EMBO Molecular Medicine EMBO Molecular Medicine Potent targeting of tauopathy by TFEB Vinicia Assunta Polito et al

A A B

A B A

C D

Figure 4. TFEB ameliorates neuronal loss and neuroinflammation. A Immunofluorescence staining of untreated or TFEB-treated (+TFEB) hippocampus of wild-type (WT) and rTg4510 (Tau) mice using the anti-NeuN antibody. Scale bar: 1000 lm. B Enlarged view of the bracketed areas in A. Scale bar: 50 lm. C Unbiased stereological quantification of NeuN-positive neurons in area CA1 of wild-type (WT) or Tau transgenic mice injected with GFP or TFEB. n = 4 sections/ mice, 5 mice/group. Student’s t-test, **P = 0.0063. D Wet brain weight measurement of WT or Tau mice with (+TFEB) or without TFEB injection. n = 14 mice/group. ***P < 0.001 (2 way ANOVA with Bonferroni post Figure 5. Life-long TFEB treatment improves behavioral performance and synaptic function. hoc). Each bar represents average s.e.m. A Morris water maze test of groups of 4-month-old WT or Tau mice without or with TFEB (+TFEB) treatment, showing longer latency (time to find hidden platform) in Æ E, F Representative immunofluorescence images of cortex (CTX) or hippocampus (HPC) of untreated or TFEB-treated (+TFEB) rTg4510 (Tau) mice stained with anti-Iba1 Tau mice compared to WT controls at any given day of training. TFEB has no effect on WT mice, but significantly reduced the latency in Tau mice. n = 14/group. (E) or anti-GFAP (F) antibodies antibody. Scale bar in E: 200 lm; in F: 400 lm. ***P < 0.001 (2 way ANOVA). B Number of platform crossings in non-target quadrants (NTQ) vs. the target quadrant (TQ). Values of the three NTQs were combined and averaged. n.s.: non- significant. *P = 0.047 and ***P = 0.0005 for TQ of WT vs Tau and Tau vs Tau + TFEB, respectively (Student’s t-test). Each bar represents average s.e.m. Æ C Slope of field excitatory postsynaptic potential (fEPSP) in response to theta-burst stimulation delivered to the Schaffer collateral pathway from WT mice (n = 10 TFEB improves cognitive performance and synaptic function without TFEB treatment (Fig 5A and B). P0-injected mice were used recordings from 6 mice), Tau mice (n = 19 recordings from 6 mice) or Tau mice with TFEB injection (Tau + TFEB; n = 8 recordings from 6 mice). Insets: example fEPSP since widespread brain expression can be achieved by this traces taken before (upper) or after (lower) stimulation from WT, Tau or Tau + TFEB mice. Calibration: 1 mV, 5 msec. D Quantification of average fEPSP slope in the last 10 minutes demonstrating severely impaired LTP in Tau mice (WT vs. Tau) and significant improvement by TFEB To investigate whether TFEB-mediated biochemical and morpholo- approach. Mice were trained with blocks of 4 trials at 2 blocks per treatment (Tau vs. Tau + TFEB). Each bar represents average s.e.m. **P < 0.01 (2 way ANOVA with Bonferroni post hoc). Æ gical changes in rTg4510 mice were accompanied by improved func- day over 4 days, followed by memory testing in a probe trial. Wild- E Cytoscape-generated networks representing genes involved in synaptic function. A majority of downregulated genes (blue dots) are found in the synaptic network tional outcome, we performed Morris water maze (MWM) test in type mice with or without TFEB treatment displayed similar latency when comparing transcription profiles of Tau mice with wild-type controls (Tau vs. WT), while a majority of upregulated genes (red dots) are found when comparing 4-month-old rTg4510 Tau mice and wild-type littermates with or during the training phase and platform crossing in the probe trial TFEB-injected with uninjected Tau mice (Tau + TFEB vs. Tau). n = 4 mice/genotype/treatment group.

ª 2014 The Authors EMBO Molecular Medicine Vol 6 | No 9 | 2014 1147 1148 EMBO Molecular Medicine Vol 6 | No 9 | 2014 ª 2014 The Authors Vinicia Assunta Polito et al Potent targeting of tauopathy by TFEB EMBO Molecular Medicine EMBO Molecular Medicine Potent targeting of tauopathy by TFEB Vinicia Assunta Polito et al

A B A

C D

(WT vs. WT + TFEB), indicating that TFEB expression caused no Since ALP can be regulated at both transcriptional and post- A BC adverse effect in learning and memory in wild-type mice. rTg4510 transcriptional levels, we examined whether the autophagy pathway mice performed poorly in both the latency and probe tests. Expres- can be activated by TFEB in general. We assessed the steady-state sion of TFEB significantly improved learning and enhanced memory levels of LC3-II, the best characterized marker of the autophago- retention as evidenced by shorter latency with training and higher some, in TFEB-transfected T40PL cells. As expected, TFEB overex- platform crossing compared to Tau mice without TFEB injection pression increased autophagosome formation, as demonstrated by (Tau vs. Tau + TFEB). immunoblot analysis showing increased levels of LC3-II (Fig 7A). Synaptic plasticity has long been proposed to be a cellular mech- However, this activity is sensitive to the duration of TFEB transfec- anism underlying learning and memory. Given that we observed a tion. Specifically, a prominent induction of LC3-II can only be significant improvement in the hippocampal-dependent behavior detected 24–48 h post-transfection. LC3-II levels returned to base- with TFEB, we recorded field Schaffer collateral long-term potentia- line afterward (Fig 7A). Downregulation of autophagy at later time tion (LTP) in acute hippocampal slices of Tau mice with or without points (72 h) is accompanied by increased LAMP1 levels, indicating TFEB treatment (Fig 5C). Because the uninjected and injected wild- that TFEB could dynamically regulate ALP by coordinating auto- type mice exhibited similar behavioral performance, we collected phagy activation with lysosomal clearance. parallel recordings only on the uninjected group. Compared to wild- To provide evidence that TFEB-mediated ALP is involved in pTau type controls, the slope of field excitatory postsynaptic potentials clearance, we cotransfected T40PL cells with TFEB and a mono- D (fEPSPs) induced by theta-burst stimulation was greatly reduced meric RFP-GFP-tagged LC3 construct, which serves as a reporter for and remained low in Tau mice (Fig 5C, WT vs. Tau). Expression of the autophagy flux (Kimura et al, 2007). As expected, both RFP-only TFEB partially but significantly rescued the LTP defects (Fig 5D, autolysosomes (red arrows) and GFP/RFP double-positive auto- Tau vs. Tau + TFEB). The improvement of neuronal function by phagosomes (yellow arrow) can be detected (Fig 7B upper inset). TFEB is supported by expression microarray analysis of hippo- Compared to vector transfected controls, there was an increase in campal samples. In agreement with the published report (Kopeikina the number of autolysosomes as a function of TFEB expression et al, 2012), we found downregulation of many of the synaptic (Supplementary Fig S7). Immunostaining with PHF1 showed that protein genes in Tau mice compared to wild-type controls when PHF1-positive staining can be detected within autophagosomes mapped on a synaptic gene network obtained by pathway co-expres- recognized by the double membrane LC3 puncta (Fig 6B inset on sion analysis (Tau vs. WT, Fig 5E and Supplementary Fig S5) right) and that cells with significant LC3 puncta (Fig 7B, thick white (Palmieri et al, 2011; The Gene Ontology Consortium, 2012; Song arrows), indicating increased autophagy, had dramatically lower et al, 2013). These downregulated genes were largely restored upon PHF1 levels compared to nearby LC3 puncta-negative cells. These Figure 6. Analysis of TFEB-mediated lysosomal gene activation in vivo. TFEB expression (Tau + TFEB vs. Tau, Fig 5E and Supplementary results suggest that pTau is recruited to the autophagosomes and A Gene set enrichment analysis (GSEA) of transcriptome changes in TFEB-injected vs. uninjected wild-type (WT + TFEB vs. WT) mice. GSEA of genes annotated as participating in the lysosomal function are reported. The upper panel shows the enrichment generated by GSEA of ranked gene expression data (left in upper panel Fig S5). The fact that we did not find significant changes of the that autophagy activation is associated with pTau degradation. and red in middle panel: upregulated; right in upper panel and blue in middle panel: downregulated). Peak value corresponds to the point (along the ranked synaptic pathway genes by comparing wild-type mice with or with- However, transfection of T40PL cells with lyso-tracker-red followed microarray) where the enrichment score (ES) meets its highest value, while the zero value indicates the boundary (along the ranked microarray) between positive fold out TFEB injection (not shown) supports the notion that the by staining with LC3 and PHF1 revealed the presence of PHF1 in change and negative fold change. In the middle panel, vertical blue bars indicate the position of lysosomal genes within the ranked list. Lower panel shows the increased synaptic gene expression in TFEB-treated Tau mice is due both LC3 puncta-positive and -negative lysosomes, indicating both cumulative distribution of lysosomal genes within the ranked lists. The ranking positions that include 50% of lysosomal genes are indicated. The analysis shows that lysosomal genes have a significant global shift toward upregulated genes in TFEB-injected mice compared with uninjected littermates (ES = 0.56, P = 0.0029). n = 4 to the rescue of Tau-triggered synaptic protein reduction rather than autophagy-dependent and independent lysosomal degradation of mice/group. a dominant effect of TFEB. pTau (Supplementary Fig S8). B Representative image of Tau mice coinjected with TFEB-FLAG/GFP and immunostained with anti-FLAG and LAMP1 antibodies. Merge image highlights that nuclear A functional role of TFEB-mediated lysosomal activation in pTau TFEB-FLAG is correlated with higher LAMP1. Scale bar: 10 lm. Activation of autophagy and lysosomal pathways by TFEB clearance is evidenced by coimmunostaining of LAMP1 and pTau in C Western blot analysis of TFEB, LAMP1 and CTSD expression in Tau mice with (+TFEB) or without AAV-TFEB adult injection. TFEB-transfected cells, which reveal that TFEB nuclear staining is D Quantification of relative band intensities of (C). n = 3 and 4 per group. TFEB protein levels are significantly increased in Tg + TFEB vs Tg (***P = 0.00013) along with LAMP1 protein levels (*P = 0.013) and CTSD protein levels (*P = 0.028), Student’s t-test. Each bar represents average s.e.m. Because TFEB has been shown to regulate the ALP through direct associated with higher LAMP1 and lower PHF1- and MC1-positive Æ Source data are available online for this figure. activation of autophagy and lysosomal target genes (Sardiello Tau (Fig 7C). Direct evidence that pTau is processed through ALP is et al, 2009; Settembre et al, 2011), we performed Gene Set documented by showing that treatment of Tau cells with Leupeptin, Enrichment Analysis (GSEA) (Subramanian et al, 2005) of tran- a lysosomal cysteine and serine protease inhibitor, resulted in scriptome changes of lysosomal and autophagy genes in TFEB- increased total Tau and PHF1 Tau under both basal and TFEB- immunoprecipitation (ChIP) experiment using HeLa cells stably We only observed a 19% increase in PTEN protein level upon AAV- injected vs. uninjected hippocampal samples. As expected, we transfected conditions (Fig 7D and E). Thus, the combined results expressing the TFEB-FLAG vector (Sardiello et al, 2009). Anti-FLAG TFEB P0 injection in wild-type mice (Supplementary Fig S9A and B). found global enrichment of lysosomal target genes (Fig 6A). implicate an important role of the lysosome in TFEB-mediated pTau antibody pull-down followed by PCR amplification showed that, A small (14%) but statistically significant protein level increase is Consistent with the transcriptional upregulation of lysosomal clearance. indeed, a CLEAR-containing PTEN promoter fragment was enriched also observed in adult injected mice (Supplementary Fig S9C and D). genes, immunostaining of LAMP1 in AAV-TFEB-injected brains in the anti-TFEB-FLAG immunoprecipitants (Fig 8C). To establish a The relatively small increase of PTEN protein levels resulting from documents that nuclear TFEB expression is associated with higher PTEN is a direct target of TFEB that mediates pTau clearance functional role of the TFEB/CLEAR interaction, we performed lucif- p0 and adult mice TFEB injection make it difficult to ascertain LAMP1 levels (Fig 6B). This is also corroborated by immunoblot- erase reporter assay by cotransfecting a TFEB expression vector whether these changes drive pTau clearance. In order to further ting of LAMP1 and another lysosomal enzyme, CTSD, which Although we failed to detect global autophagy gene induction by with PTEN-luciferase reporters with or without CLEAR motifs in validate the connection between TFEB and PTEN, we took advan- reveal that TFEB injection is associated with mild but significant TFEB, analysis of the microarray data followed by qRT-PCR N2a cells (Teresi et al, 2008). We observed significant activation of tage of the in vitro system where TFEB can be transiently overex- increases of these lysosomal protein levels (Fig 6C and quantified revealed that expression of phosphatase and tensin homolog luciferase activities only in constructs containing one or both CLEAR pressed more than 20-fold and assayed PTEN expression thereafter. in 6D). (PTEN), a lipid phosphatase that antagonizes the phosphatidylinositol- sequences ( 1334 to +1 and 453 to +1), but not in the vector with- Western blot analysis showed that transfection of TFEB directly À À However and unexpectedly, we did not detect appreciable tran- 3-kinase (PI3K)-Akt-mTOR signaling (Kwon et al, 2001, 2003), was out the CLEAR motif ( 203 to +1) (Fig 8D). The increase in lucifer- drives endogenous PTEN protein expression, and is associated with À scriptional upregulation of autophagy pathway molecules as a func- significantly elevated in TFEB-treated WT and Tau mice (Fig 8A). ase activity is TFEB dose-dependent (Fig 8E). Due to PTEN’s elevated LC3-II levels (Fig 8F and G). Immunostaining for TFEB, tion of TFEB expression (Supplementary Fig S6). qRT-PCR analysis Interestingly, examination of the PTEN promoter identified 2 puta- important roles in regulating various cellular mechanisms, its PTEN and pTau reveals that TFEB overexpression is correlated of selected TFEB lysosomal and autophagy targets confirmed the tive CLEAR sequences (Fig 8B), indicating that TFEB may activate expression at both mRNA and protein levels are controlled by intri- with increased PTEN and reduced PHF1-positive Tau (Fig 8H). As microarray data (Supplementary Fig S6B and C). PTEN by binding to the CLEAR motifs. We thus carried out chromatin cate regulatory mechanisms ((Song et al, 2012) for a recent review). expected, enhanced TFEB-dependent PTEN upregulation is associated

ª 2014 The Authors EMBO Molecular Medicine Vol 6 | No 9 | 2014 1149 1150 EMBO Molecular Medicine Vol 6 | No 9 | 2014 ª 2014 The Authors Vinicia Assunta Polito et al Potent targeting of tauopathy by TFEB EMBO Molecular Medicine EMBO Molecular Medicine Potent targeting of tauopathy by TFEB Vinicia Assunta Polito et al

AB A C

C DE

F G H

Figure 8. PTEN is a direct target of TFEB. A qRT-PCR analysis of PTEN levels as a function of TFEB treatment in WT or Tau mice. n = 3 mice/genotype/treatment group in triplicates. *P = 0.027 and **P = 0.01 Figure 7. Analysis of TFEB-mediated autophagy and lysosomal pathway in vitro. for WT vs WT + TFEB and Tau vs Tau + TFEB, respectively (Student’s t-test). A Western blot analysis of Lamp1 and LC3-I and LC3-II levels in T40PL cells transfected with TFEB-FLAG and harvested at times indicated. c-tubulin was used as a B Putative CLEAR sequences in the PTEN promoter and their positions. TSS: translation start site. 203 to +1, 453 to +1 and 1334 to +1: PTEN promoter fragments À À À loading control. Ctrl: untransfected. linked to the luciferase reporter. B Fluorescent images of T40PL cells transfected with TFEB and RFP-GFP-LC3. Thick white arrows indicate cells with significant LC3 puncta and lower PHF1 compared C ChIP analysis of TFEB binding. HPRT and APRT: negative controls; MCOLN1: positive control. Open bar: IgG control; filled bar: TFEB-FLAG IP. The experiment was done with nearby cells. Images on the upper and right are higher magnification views of the bracketed area of the LC3-GFP and LC3-RFP or Merge panels, respectively. two times each in triplicates. RFP-only autolysosomes are marked by red arrows; GFP/RFP double-positive puncta representing an autophagosome is shown in yellow (upper inset). Right insets D Normalized luciferase activity by transfecting 300 ng of each PTEN-luciferase reporter. Open bar: vector transfection control; filled bar: TFEB transfected. Student’s highlight a cell with colocalization of GFP/RFP-positive autophagosome with PHF1. Scale bar: 10 lm; in inset: 2.5 lm. t-test, n = 4;*P = 0.01; **P = 0.005 C Immunofluorescence images of T40PL cells transfected with TFEB and stained with anti-LAMP1/MC1 (top) or PHF1 (bottom). Arrow marks cells with nuclear TFEB E Dose-dependent luciferase activities in response to increasing concentrations of TFEB using the ( 1344 to +1) PTEN-luciferase reporter. The experiment was done À and its correlation with higher LAMP1 and lower MC1 (top) or PHF1 (bottom) stainings. Arrowhead indicates nearby cells with cytoplasmic TFEB and opposite LAMP1 three times each in triplicates. and MC1/PHF1 patterns. Scale bar: 10 lm. F Western blot analysis of PTEN and LC3 expression in T40PL cells transfected with either pCMV or TFEB. D Western blot analysis of T40PL cells transfected with either pCMV or TFEB, allowed to recover for 24 h, and then treated with 0.5 lg/ml doxycycline and either DMSO G Quantization of blots in F and normalized for loading to c-tubulin. PTEN and LC3-II are significantly increased in cells transfected with TFEB with *P = 0.032 and or 50-lM Leupeptin for 48 h prior to lysis. Black line denotes cropped lanes from a single immunoblot. 0.036, respectively (Student’s t-test, n = 3). E Quantization of the autophagy markers LC3-I, LC3-II (pCMV vs pCMV+Leupeptin P = 0.032, TFEB vs TFEB+Leupeptin P = 0.014), and p62 (pCMV vs pCMV+Leupeptin H Representative immunofluorescent images of T40PL cells transfected with TFEB and triple staining for TFEB (FLAG), PTEN and PHF1. Arrow marks cell with higher P = 0.019, TFEB vs TFEB+Leupeptin P = 0.024), and Tau species markers Tau1, PHF1 (pCMV vs pCMV+Leupeptin P = 0.00082, TFEB vs TFEB+Leupeptin P = 0.022), TFEB and PTEN is correlated with lower PHF1. Arrowhead indicates nearby cell with opposite TFEB/PTEN/PHF1 patterns. Scale bar: 10 lm. Each bar represents and total Tau normalized for loading to c-tubulin from (D). P-values determined by t-test, n = 4. (*P < 0.05) Each bar represents average s.e.m. average s.e.m. Æ Æ Source data are available online for this figure. Source data are available online for this figure.

ª 2014 The Authors EMBO Molecular Medicine Vol 6 | No 9 | 2014 1151 1152 EMBO Molecular Medicine Vol 6 | No 9 | 2014 ª 2014 The Authors Vinicia Assunta Polito et al Potent targeting of tauopathy by TFEB EMBO Molecular Medicine EMBO Molecular Medicine Potent targeting of tauopathy by TFEB Vinicia Assunta Polito et al

AB A C

C DE

F G H

with reduced pAkt (S473), pP70SK6 (T389) and pULK1 (S757), but TFEB primarily triggers the degradation of the misfolded Tau species A B not their total protein levels (Supplementary Fig S10). including but not limited to hyperphosphorylated Tau. However, it Previous reports have implicated a role of PTEN in pTau reduc- is also possible that TFEB-mediated clearance recognizes only the tion (Kerr et al, 2006; Zhang et al, 2006a,b). To determine whether Tau species marked by hyperphosphorylation and that the clearance PTEN expression leads to reduced pTau levels in our system, we of MC1-positive species is the result of hyperphosphorylation. Our transfected either an empty vector (Ctrl), a wild-type PTEN expres- results that TFEB and pTau stainings are mutually exclusive support sion vector (WT) or the phosphatase (C124S) or lipid phosphatase a neuronal-intrinsic role of TFEB in pTau reduction. However, it (G129E) deficient mutants in T40PL cells and measured total and remains possible that other cell types, such as astrocytes and micro- phosphorylated Tau levels upon doxycycline induction (Wang et al, glia, may also participate in TFEB-dependent pTau clearance in a 2007). Biochemical (Fig 9A and B) and immunohistochemical non-cell-autonomous manner. In this regard, it is interesting to note (Fig 9C) analyses showed that only the wild-type PTEN and not the that TFEB has been reported to mediate not only lysosomal biogene- phosphatase-defective mutants were associated with reduced PHF1 sis but also lysosomal exocytosis (Medina et al, 2011). It is difficult Tau, demonstrating a role for PTEN in pTau reduction that requires to probe the cellular mechanisms using the current system as the C its lipid phosphatase activity. To ascertain whether PTEN is required AAV vector infects both neurons and non-neuronal cells. Genetic D in mediating TFEB-dependent pTau reduction, we cotransfected expression of TFEB in these distinct cell types is better suited to either an EGFP or TFEB expressing vector with a PTEN shRNA address this question. construct or a scrambled control into Tau expressing cells (Fig 9D). TFEB has been reported to regulate autophagy through transcrip- Expression of the PTEN shRNA markedly blunted TFEB-mediated tional activation of autophagy genes (Settembre et al, 2011). Our PHF1-positive pTau and total Tau reduction and LC3-II elevation gene expression analysis failed to detect significant upregulation of (Fig 9E), demonstrating that PTEN is critical in mediating TFEB- autophagy targets in TFEB-treated samples. This could be attributed dependent pTau clearance. to the modest TFEB expression or the tissue-specific regulation of autophagy gene induction by TFEB. Regardless, our cell culture studies support a role of TFEB in the dynamic regulation of ALP by Discussion coordinating autophagy activation with lysosomal clearance and our analysis implicates this regulation, in particular lysosomal activity, In this report, we investigated the role of TFEB on AD neuropathol- in TFEB-mediated pTau clearance. Importantly, we identify PTEN as ogy by directly injecting AAV-TFEB into 5xFAD and rTg4510 trans- a bona fide target of TFEB that may mediate the TFEB effect on genic mouse brains, which affords widespread and persistent TFEB pTau reduction. Since PTEN could activate autophagy by antagoniz- expression in vivo. We demonstrate that TFEB exerts a potent activ- ing the PI3K-Akt-mTOR signaling (Kwon et al, 2001, 2003), it is ity in attenuating NFT pathology without overtly affecting Ab deposi- reasonable to hypothesize that the TFEB-PTEN-PI3K-Akt-mTOR E tion or displaying adverse reactions in wild-type mice. The reduction signaling pathway is involved in the pTau clearance through ALP. of pTau/NFTs by TFEB is accompanied by improved neuronal This notion, combined with the published reports showing that survival and function. In light of the substantial evidence supporting TFEB is a substrate of mTOR (Roczniak-Ferguson et al, 2012; a role of ALP in APP turnover and Ab catabolism (Mueller-Steiner Settembre et al, 2012), raises an interesting possibility for a TFEB- et al, 2006; Nixon, 2007; Pickford et al, 2008; Rohn et al, 2011; Yang PTEN-Akt-mTOR-TFEB feedback regulatory loop whereby TFEB et al, 2011), our finding that TFEB has no appreciable effect on Ab induces autophagy through upregulation of PTEN and inhibition of pathology is somewhat unexpected. It is important to note, however, Akt and mTOR, which in turn results in further TFEB activation. this negative data should be interpreted with caution. Our finding is Nevertheless, enhanced lysosomal activity by TFEB may also play limited to the system we employed using the 5xFAD mice and needs crucial roles in pTau/NFT degradation by effectively clearing LC3II- to be further validated in other APP/Ab mouse models and by other positive autophagosomes and/or degrading pTau independent of approaches such as genetic manipulation. It remains possible that macroautophagy. Further studies are needed to delineate the contri- the TFEB expression attainable in our system may not be sufficient bution of autophagy-dependent vs. autophagy-independent lyso- to impact Ab pathology or that APP/Ab may be subject to TFEB somal clearance of pTau and to what extent PTEN-mediates ALP Figure 9. TFEB mediates pTau reduction in a PTEN-dependent manner. independent ALP regulation. Nevertheless, the potent clearance of downstream of TFEB. In addition, it is important to point out that in A Western blot analysis of total Tau and PHF1-positive pTau levels as a function of wild-type or the C124S or G129E mutant PTEN expression. Empty vector (Ctrl)- or pTau/NFTs using the same TFEB injection scheme argues that Ab vitro cell cultures were used for mechanistic studies. The nature of TFEB-transfected cells were used as negative and positive controls, respectively. c-tubulin was used as a loading control. and pTau/NFT pathologies are subject to distinct TFEB regulations. pTau expressed in the acutely induced cells versus in vivo is likely B Quantization of blots in A and normalized for loading to c-tubulin. TFEB and PTEN transfection significantly reduced pTau as probed by PHF1 (P = 0.0095 and Biochemical analysis of rTg4510 mouse brain samples and distinct, especially with regards to its aggregation status. Therefore, 0.0092, respectively), as well as total Tau (P = 0.0052 and 0.034, respectively), n = 3, Student’s t-test. C Representative immunofluorescence imaging showing that cells with wild-type PTEN (WT), but not the C124S or G129E mutants, were correlated with reduced PHF1- T40PL cell lysates revealed that TFEB enhances clearance of only the relevance of the signaling pathways identified here to tauopathies positive Tau. Scale bar: 10 lm. the hyperphosphorylated and misfolded Tau proteins. The fact that in vivo requires further investigation and validation. D Western blot analysis of total Tau, PHF1-positive pTau and LC3 levels in response to TFEB or TFEB and PTEN shRNA expression. Transfection of a scramble PTEN the Tau1-positive unphosphorylated Tau remained constant in all The mechanism mediating the macroautophagy-independent shRNA was used as negative controls. c-tubulin was used as a loading control. preparations strongly supports the notion that TFEB leads to degra- lysosomal degradation of pTau remains to be investigated. One E Quantization of blots in D and normalized for loading to c-tubulin. Comparing with cells transfected with scramble+pEGFP, scramble+TFEB transfection reduces dation rather than dephosphorylation of the aberrant Tau species. possibility is that clearance occurs through chaperone-mediated PHF1 and total Tau intensity (P = 0.038 and 0.035, respectively) and increases PTEN and LC3II (P = 0.04 and 0.0076, respectively) (white slashed bars). shRNA significantly reduce PTEN when transfected with either pEGFP or TFEB (P = 0.0089 and 0.043, respectively), but TFEB’s effect in reducing pTau drops from 60% Nevertheless, although it is clear that TFEB targets the detergent- autophagy (CMA). In this regard, Tau has been reported to contain reduction in the presence of scramble sequence to 30% reduction when combined with shRNA of PTEN. All statistics here are Student’s t-test, n = 3. Each bar insoluble pTau for degradation, we cannot exclude the possibility two CMA targeting motifs obligatory for hsc70 binding and represents average s.e.m. Æ that TFEB may promote pTau dephosphorylation or lower the rate LAMP2A-mediated lysosomal degradation (Wang et al, 2009). Data information: *P < 0.05, **P < 0.01, exact value described above for each analysis. of Tau phosphorylation in the soluble pool. Since TFEB recognizes However, only proteolytically processed Tau fragments were shown Source data are available online for this figure. Tau phosphorylated at multiple sites as well as misfolded Tau to be subject to the CMA pathway (Wang et al, 2009). Since we did marked by MC1 immunoreactivity, we favor a mechanism in which not detect appreciable levels of Tau fragments, the role of CMA in

ª 2014 The Authors EMBO Molecular Medicine Vol 6 | No 9 | 2014 1153 1154 EMBO Molecular Medicine Vol 6 | No 9 | 2014 ª 2014 The Authors Vinicia Assunta Polito et al Potent targeting of tauopathy by TFEB EMBO Molecular Medicine EMBO Molecular Medicine Potent targeting of tauopathy by TFEB Vinicia Assunta Polito et al

our system is not clear. Besides autophagy, PTEN may mediate Tau purchased from commercial sources as follows: AT8 (Pierce) and TIGEM AAV Vector Core Facility as previously described (Settembre SuperScript III reverse transcriptase reagents (Invitrogen). Quanti- homeostasis through additional activities. In particular, the downre- AT100 (Innogenetics), total Tau (DAKO), Tau1 (Millipore), PTEN et al, 2011). For P0 injection, each mouse was injected into the tative PCR was done using Perfecta SYBR Green Fastmix (Quanta gulation of pAkt by PTEN may directly affect Tau phosphorylation (Cell Signaling), LAMP1 (Millipore), LC3 (Novus Biological and lateral ventricles of both cerebral hemispheres with 4.2 × 109 total Biosciences) utilizing the ABI Prism 7000 detection system status independent of mTOR (Zhang et al, 2006a,b). Akt may also Sigma), Akt (Cell Signaling) and pAkt (Cell Signaling), P70S6K (Cell viral particles per side. The AAV-GFP and/or AAV-TFEB-injected (Applied Biosystems). For expression studies, the qRT-PCR results participate in ubiquitin-proteasome degradation of pTau species by Signaling) and pP70S6K (Cell Signaling), ULK1 (Cell Signaling) and wild-type, 5xFAD and rTg4510 mice were euthanized at 1 or were normalized against an internal control (GAPDH). Primers interacting with the HSP90/CHIP ubiquitin ligase complex (Dickey pULK1 (Cell Signaling), Y188 (Epitomics), 6E10 (Covance), NeuN 4 months after the injection. Adult rTg4510 Tau mice at 2 months were designed with Primer Express Version 2.0 software (Applied et al, 2007, 2008). Further, since PTEN has been shown to be essen- (Chemicon), GFAP (DAKO), Iba1 (Waco), GAPDH (Sigma), FLAG of age were injected into the cortex and hippocampus Biosystems) using sequence data from NCBI. GAPDH primers were tial for neural development and synaptic plasticity (Kwon et al, (Biolegend), human TFEB (Cell Signaling), mouse and human TFEB (AP: + 1.5 mm, LAT: + 1.5 mm, DV: + 1.5 mm and AP: 2 mm, used as an internal control for each specific gene amplification. À 2006; Zhou et al, 2009; Sperow et al, 2011; Takeuchi et al, 2013), (Abcam and Abmart), and c-tubulin (Sigma). LAT: 1.5 mm, DV: + 1.75 mm) of both cerebral hemispheres The relative levels of expression were quantified and analyzed by À upregulation of PTEN may directly contribute to the improved according to the stereotaxic atlas of Franklin and Paxinos (Franklin using ABI PRISM Sequence Detection System 7000 software. The neuronal survival and synaptic function in TFEB-treated Tau mice. Animals & Paxinos, 2001) using the same total viral concentrations and real-time value for each sample was averaged and compared using Our studies implicate mTOR as an integral component of the analyzed also at 4 months. the comparative CT method. The relative amount of target RNA TFEB-PTEN-Akt-mTOR autophagy pathway, which is compatible All procedures involving mice were approved by the Institutional was calculated relative to the expression of endogenous reference with a recent report documenting that TFEB rescued a-synuclein Animal Care and Use Committee of the Baylor College of Medicine. Microarray analysis and relative to a calibrator which was the mean CT of control toxicity through downregulation of mTOR (Decressac et al, 2013). The 5xFAD APP (Oakley et al, 2006) and rTg4510 Tau (Ramsden samples. However, a potent role of TFEB in lysosomal biogenesis distin- et al, 2005; Santacruz et al, 2005) transgenic mouse lines were Mice were euthanized at 4 months of age, and hippocampi were guishes the effects of TFEB from autophagy activators mechanisti- obtained from the Jackson Laboratories and produced by crossing the dissected and frozen in liquid nitrogen immediately. Total RNA was Western blotting cally and functionally. Of particular significance, while autophagy APP transgenic mice with C57BL/6J mice, and the transactivator line isolated using RNeasy Lipid Tissue Mini Kit (Qiagen) and tested for leads to decreased endogenous Tau and degradation of pTau/NFT CaMKIIa-tTA (on 129S6 background) with the Tau responder line (on quality assurance on the Aglient 2100 Bioanalyzer. For gene expres- For Westerns without fractionation, cells or forebrain tissues were aggregates, as demonstrated by treating Tau mouse models with the FVB background), respectively. The littermate wild-type mice were sion analysis, we used a Mouse Genome 430 2.0 Array from Affyme- lysed by RIPA buffer (TBS with 1% NP-40, 1% sodium deoxycholic autophagy inducer rapamycin or trehalose (Schaeffer et al, 2012; used as controls. Both males and females were used. Mice within each trix that provides coverage of the transcribed mouse genome in a acid, 0.1% sodium dodecylsulfate, and protease phosphatase inhibi- Caccamo et al, 2013; Ozcelik et al, 2013), TFEB targets both soluble genotype were randomly assigned for GFP or TFEB injections. single array (over 39,000 transcripts). Scanning was done using tor cocktails (Roche)). Cell lysates were sonicated for 6 pulses at and insoluble pTau species with no effect on endogenous Tau. Affymetrix GeneChip Scanner 3000. Affymetrix labeling, hybridiza- 50% duty cycle, incubated at 4°C for 30 min and centrifuged at Because soluble Tau is known to be neurotoxic (Santacruz et al, Cell culture, transfection and luciferase assay tion, staining, washing, scanning and statistical analysis were done 20,000 × g for 15 min. Supernatants were used for SDS-PAGE, 2005), targeting only the aggregated Tau by enhancing autophagy is by the Microarray Core at Baylor College of Medicine (http:// transferred to PVDF membranes and detected using the ECL method expected to offer limited therapeutic benefit. Furthermore, auto- The double-stable Tet-On inducible Tau-expression cell line, T40PL, www.bcm.edu/mcfweb/). Accession number GSE53480. (Pierce). Protein levels were quantified using ImageJ (National Insti- phagy activation without enhanced lysosomal function will likely that expresses full-length human Tau with the P301L mutation Gene set enrichment analysis (GSEA) was performed as previ- tute of Health). In repeated experiments, we used infrared dye lead to the accumulation of autophagic cargoes detrimental to (T40PL) in HEK293 cells under the regulation of a tightly controlled ously described (Subramanian et al, 2005). The cumulative distri- conjugated 2nd antibodies (IRDye 680RD anti-mouse IgG, IRDye neuronal health. Indeed, extensive autophagic vacuoles can be tetracycline-inducible promoter system (Invitrogen) and maintained in bution function was constructed by performing 1,000 random 680RD anti-rat IgG and IRDye 800CW anti-rabbit IgG) in two-color observed in human AD brains and in transgenic mice expressing DMEM supplemented with 10% Tet-system-approved FBS (Clontech), gene set membership assignments. A nominal P-value of < 0.01 combination and LI-COR Odyssey Imaging System for Western blot mutant P301L Tau (Lin et al, 2003; Nixon et al, 2005). TFEB in this 100 units/ml penicillin G, 100 lg/ml streptomycin, 400 lg/ml G418 and an FDR of < 10% were used to assess the significance of the analysis and quantification. We changed from ECL as the detection regard dynamically regulates the ALP by coordinating autophagy (Sigma) + 0.20 lg/ml Puromycin (Clontech) at 37°Cwith5%CO2. enrichment score (ES). Gene Ontology (GO) analyses were method for Western blot to avoid overexposure issues; these near- induction with enhanced lysosomal clearance. As such, targeting Cells were transfected with 3xFLAG-tagged human TFEB plasmid, performed with the web tool DAVID (da Huang et al, 2009) using infrared fluorescent 2nd antibodies result in a much large dynamic TFEB-mediated cellular clearance may offer more superior therapeu- PTEN expression vector or PTEN shRNA using X-tremeGENE 9 DNA default parameters. Redundant terms were manually removed detection range than ECL. tic benefits than autophagy activators (reviewed in (Cuervo, 2011)). transfection reagent (Roche) Lipofectamine 3000 (GE lifesciences) from the resulting lists. The synaptic gene network was obtained For subcellular fractionation, cortex dissected from cerebral hemi- In summary, we demonstrate that TFEB targets only the aberrant according to the company’s suggested protocol. Doxycycline (DOX) by performing pathway co-expression analyses as previously spheres of AAV-TFEB P0-injected mice were weighed, and homoge- hyperphosphorylated and misfolded Tau while leaving the normal (Clontech) was added to the cell medium 24 h after TFEB transfection described (Palmieri et al, 2011; Song et al, 2013) using genes nized by Dounce homogenizer in five volumes of TBS with 10% Tau intact; it is highly efficacious in ameliorating pTau/NFT pathol- to a final concentration of 500 ng/ml. Cells were collected in ice-cold annotated as ‘Synapse’ in the Gene Ontology database (The Gene sucrose, protease inhibitor and phosphatase inhibitor cocktails. The ogy, neurodegeneration and behavioral deficits in rTg4510 mice while Tris-buffered saline (TBS, 50 mM Tris-HCl, pH 7.4, 150 mM NaCl) at Ontology Consortium, 2012). Briefly, ‘Synapse’ genes were used mixture was centrifuged at 800 × g for 5 min. Pellets were resuspended exhibiting no adverse effects on wild-type mice. These features make various times after DOX induction as indicated. to analyze a vast set of transcriptional profiles available at the in the same volume of the above buffer and centrifuged again at the TFEB an attractive therapeutic target for AD and other diseases of N2a cells grown in 12-well plates were cotransfected with a Gene Expression Omnibus (GEO) database (Barrett et al, 2013). same speed. Supernatants from the two steps were pooled together and tauopathy. However, the current work represents a proof-of-concept TFEB-FLAG expression vector and PTEN promoter ( 1334 to +1, Multiple cellular conditions and tissues are represented in this 1 ml of the solution from each mouse was subjected to ultracentrifuga- À study. A TFEB-based therapy likely requires the identification of 453 to +1 or 203 to +1)-luciferase reporters (Teresi et al, 2008), database. To ensure data homogeneity, the analysis was focused tion at 100,000 × g for 1 h. 700 ll of supernatants from ultracentrifuga- À À specific small molecule TFEB activators. Furthermore, it is important together with Renilla luciferase vector using X-tremeGENE 9 DNA on experiments that used the Affymetrix platform Mouse Genome tion were collected as the cytosol fractions. Pellets from the to note that, as a master regulator of lysosomal activity, expression Transfection Reagent (Roche). Twenty-four hours after transfection, 430 2.0 Array. For each ‘Synapse’ gene pair, a pairwise co-expression ultracentrifugation were resuspended in 700 ll of 1% sarkosyl in TBS levels and duration of TFEB need to be properly controlled. Indeed, cells were lysed with Passive Lysis Buffer (Promega). The Dual- score was calculated as their cumulative occurrence in the top with protease inhibitor and phosphatase inhibitor cocktails, sonicated, dysregulation of members of the microphthalmia family of transcrip- Luciferase Reporter Assay System (Promega) was used to determine 3% of correlated genes across all investigated experiments incubated in shaking for 30 min and ultracentrifuged again at tion factors including TFEB have been shown to cause renal carcino- the firefly and Renilla luciferase activities according to the manufac- (Palmieri et al, 2011; Song et al, 2013). The expression correla- 100,000 × g for 1 h. Pellets from the second ultracentrifugation step mas (Haq & Fisher, 2011). Therefore, rigorous studies are needed to turer’s instructions. Measurements were performed with a BD lumi- tion data were then analyzed with Cytoscape (Lopes et al, 2010) were resuspended in 100 ll of 2x SDS-PAGE sample loading buffer, evaluate the safety profiles of potential TFEB activators. nometer, and firefly luciferase values were normalized to Renilla to draw a visual representation of expression relationships among sonicated, and boiled for 5 min, and used as the sarkosyl-insoluble frac- luciferase values. In all experiments, the internal control plasmid genes. Downregulated and upregulated ‘Synapse’ genes in Tau tions. All the above procedures prior to boiling in SDS-sample loading was used to compensate variable transfection efficiencies. All assays and TFEB-injected Tau mice were highlighted in the Cytoscape- buffer were carried out either on ice or in 4°C cold room. Either 5 ll of Materials and Methods were repeated three times with each in triplicates. generated network by using different color codes. cytosol fractions, (1/140 of total) or 10 ll of insoluble fractions (1/10 of each sample) was loaded per gel for SDS-PAGE and the following Antibodies In vivo gene delivery RNA extraction, reverse transcription and qRT-PCR Western blot analysis. All the cytosol fractions have protein concentration 25 1.8 mg/ml by protein concentration assay using Biorad Æ CP13, PHF1 and MC1 antibodies were kind gifts from P. Davies The AAV2/9-CMV-GFP (AAV-GFP), AAV2/9-CMV-TFEB (AAV- Total RNA was extracted from forebrain samples with TRIzol protein dye reagent (Biorad). No protein concentration assay was done (Albert Einstein College of Medicine). Other antibodies were TFEB) or AAV2/9-CMV-TFEB-3xFLAG vectors were produced by the reagent (Invitrogen). Reverse transcription was performed using for the insoluble fractions.

ª 2014 The Authors EMBO Molecular Medicine Vol 6 | No 9 | 2014 1155 1156 EMBO Molecular Medicine Vol 6 | No 9 | 2014 ª 2014 The Authors Vinicia Assunta Polito et al Potent targeting of tauopathy by TFEB EMBO Molecular Medicine EMBO Molecular Medicine Potent targeting of tauopathy by TFEB Vinicia Assunta Polito et al

Histology and Immunofluorescence to the manufacturer’s protocol. Each ChIP experiment required Acknowledgments The paper explained 107 cells. We are grateful to Dr. P. Davies (Albert Einstein School of Medicine) for the gift Cells grown on cover slips were fixed by 4% paraformaldehyde at Oligonucleotide sequences are as follows: Problem of CP13, MC1 and PHF1 antibodies, Dr. C. Eng (Cleveland Clinic Foundation) for room temperature for 10 min after removal of culture medium, HPRT-F: GCCACAGGTAGTGCAAGGTCTT; HPRT-R: TTCATGGC Alzheimer’s disease has two pathological hallmarks: b-amyloid providing the pGL3-PTEN reporter constructs and Dr. X. Jiang (Memorial Sloan washed in TBS for 5 min, followed by the same staining procedure GGCCGTAAAC plaques and neurofibrillary tangles (NFTs), the latter are composed of Kettering Cancer Center) for the PTEN expression and shRNA constructs. We as brain sections. Mice brains were collected after PBS perfusion; APRT-F: GCCTTGACTCGCACTTTTGT; APRT-R: TAGGCGCCATC misfolded Tau protein. In the past, the field has devoted intensive are indebted to C. Spencer and the Baylor College of Medicine IDDRC Adminis- effort on -amyloid based therapies and these have yielded disap- this procedure was followed by an over-night fixation of the brains GATTTTAAG b trative, Genomic and RNA Profiling and Mouse Neurobehavioral cores pointing results so far. One of the concerns is that these agents may (HD024064) for the generous resources and support. We thank A. Cole and N. with 4% paraformaldehyde in PBS. Then the brains were dehy- MCOLN1-F: AGGGGCTCTGGGCTACC; MCOLN1-R: GCCCGCCG not target the NFT pathology and strong evidence has implicated a drated in 30% sucrose in PBS. Immunofluorescence analyses were CTGTCACTG neurotoxic function of NFTs in Alzheimer’s disease and other neurode- Aithmitti for expert technical assistance and members of the Zheng laboratory performed on 30-lm frozen sections. The slices were incubated for PTEN-F: ATGTGGCGGGACTCTTTATG; PTEN-R: ACAGCGGCT- generative diseases. Therefore, there is increasing interest in develop- for constructive discussions. This work was supported by grants from NIH 1 h with TBS blocking solution before incubation overnight with the CAACTCTCAAAC ing Tau-based therapy for treating Alzheimer’s disease. (AG020670, AG032051 and NS076117 to HZ) and the Belfer Neurodegeneration primary antibodies. After washing, the sections were incubated for Consortium. 1 h with secondary antibodies conjugated either with Alexafluor Morris water maze (MWM) assay Results In the current study, we investigated the role of Transcription Factor 488 or Alexafluor 647 (Invitrogen). Images were taken on a confocal Author contributions EB (TFEB), a molecule shown to play central roles in the autophagy microscope (Leica SPE). The number of positive cells was quantified MWM tests were performed in 4-month-old male mice. We used 54 and lysosomal pathway, in the rTg4510 Tau transgenic mouse models. VAP, HL, HM-S, BW, LY, YX, DW, MP, AdR performed experiments and analyzed by using ImageJ software analysis. The number of autophagosomes animals in total for behavior test, at 10–14 mice/group. The genotypes We found that adeno-associated virus-mediated TFEB expression has the data, VAP, HL and HM-S and HZ designed the study, MS performed per cell was quantified based on puncta number in the green chan- and treatment groups were blinded to the experimenter. Briefly, mice no untoward side effects on wild-type mice. In contrast, TFEB is bioinformatics analysis, VML and AB provided the Tau inducible cell line nel. The number of total autophagosomes and autolysosomes per were trained to locate a square platform (9 × 9 cm) hidden in Quadrant highly efficacious in reducing neurofibrillary tangle pathology and and TFEB vectors/viruses, respectively, as well as overall guidance, HZ wrote rescuing behavioral and synaptic deficits and neurodegeneration in cell was quantified based on puncta number in the red channel. The 3 (1–1.5 cm below water surface) of a circular water pool (1.38 m the manuscript together with VAP. All authors made comments on the the Tau mouse model. TFEB specifically targets the misfolded Tau number of autolysosomes per cell was calculated as total number diameter, 21°C) within 60. In the learning sessions, each mouse was species while leaving normal Tau intact. We provide evidence that manuscript. minus number of autophagosomes. trained for 8 training blocks in four consecutive days. Each day, mice this effect requires lysosomal activity and we identify phosphatase would swim for 4 successive trials as one training block before being and tensin homolog (PTEN) as a direct target of TFEB that may be Conflict of interest Quantitative colocalization analysis returned to its holding cage. At the end of the 8th training block on day required for TFEB-dependent aberrant Tau clearance. The authors declare that they have no conflict of interest. 4, mice were subjected to the probe test during which the platform was Impact DOX (1,000 lg/ml) was added to culture medium when T40PL removed and each mouse would swim for 60 s to search for the The specificity and efficacy of TFEB in mediating the clearance of toxic cells were plated on cover slips in 24-well plates. 3xFLAG-TFEB platform. For all the trials, mice were allowed to climb to the platform Tau species and rescuing neurodegeneration is remarkable. Our find- References and pEGFP (control) plasmids were transfected using Lipofecta- and rest for 10 s before they were submitted to the next trial or trans- ings provide proof-of-concept that small molecule TFEB activators mine 3000 (GE lifesciences) 24 h after DOX induction. Cells on ferred back to holding cages. The time for mice to locate the hidden may serve as effective therapy for treating diseases of tauopathy Barrett T, Wilhite SE, Ledoux P, Evangelista C, Kim IF, Tomashevsky M, coverslips were fixed at 0, 8, 16, 24, 32 and 40 h time points for platform during learning, as well as the number of platform crossings including Alzheimer’s disease. Marshall KA, Phillippy KH, Sherman PM, Holko M et al (2013) NCBI GEO: immunofluorescent staining. Cells transfected with 3xFLAG-TFEB during the probe tests, was recorded by Ethovision tracking system archive for functional genomics data sets–update. Nucleic Acids Res 41: were stained by antibodies anti-FLAG (rat mAb) and PHF1 (mouse (Noldus Information Technologies). D991 – D995 mAb) antibodies, while cells transfected with pEGFP were stained dissector method (West & Gundersen, 1990) on Zeiss microscope Brunden KR, Trojanowski JQ, Lee VM (2009) Advances in tau-focused drug by PHF1 only. To minimize the bleed-through effect in immunoflu- Electrophysiology with motorized stage. Data acquisition and analysis was performed discovery for Alzheimer’s disease and related tauopathies. Nat Rev Drug orescent image collection, Alexa-647-anti-mouse IgG (Invitrogen) using software StereoInvestigator by MBF Bioscience. The CA1 Discov 8: 783 – 793 was chosen to be paired with Alexa-488-anti-rat IgG as the second- Field recordings of Schaffer collateral LTP was performed as region of counting was outlined under 5× objective lens on every Caccamo A, Magri A, Medina DX, Wisely EV, Lopez-Aranda MF, Silva AJ, Oddo ary antibody for PHF1. Confocal images were collected on a Leica described (Peethumnongsin et al, 2010). Briefly, brains were 10th section of 35-lm-thick cryostat sections from 0.24 to 1.36 mm S(2013) mTor regulates tau phosphorylation and degradation: TCS SPE microscope; data analysis was carried out using Fiji- isolated from 4-month-old mice and cut into 400-lM slices (Leica). lateral from the midline on one hemisphere of the mouse brain, implications for alzheimer’s disease and other tauopathies. Aging Cell 12: ImageJ (Schindelin et al, 2012). Specifically, for images collected Hippocampal slices were incubated for 1 h at room temperature and according to the stereotaxic coordinates adopted from a mouse brain 370 – 380 using 10× objective lens, 8 confocal projection slices per view then transferred to a heated recording chamber filled with recording atlas (Franklin & Paxinos, 2001). NeuN-positive neurons were Cruz-Orive LM (1999) Precision of Cavalieri sections and slices with local 2 field, 4 view fields per time point per transfection were analyzed ACSF (125 mM NaCl, 2.5 mM KCl, 1.25 mM NaH2PO4, 25 mM counted in counting frame area of 625 lm with sampling grid area errors. J Microsc 193: 182 – 198 2 by plugin ‘coloc2’, in which the threshold of images were chosen NaHCO3, 1 mM MgCl2, 2 mM CaCl2, and 10 mM glucose, saturated of 26,338 lm under 40× objective lens. Volume of the region was Cuervo AM (2011) Cell biology. Autophagy’s top chef. Science 332: 1392 – 1393 automatically for objective analysis. Pearson Product Moment with 95% O2 and 5% CO2) maintained at 32°C. Stimulation of determined by Cavalieri’s method (Cruz-Orive, 1999). Decressac M, Mattsson B, Weikop P, Lundblad M, Jakobsson J, Bjorklund A Correlation (also called Pearson correlation coefficient or Pearson Schaffer collaterals from the CA3 region was performed with bipolar (2013) TFEB-mediated autophagy rescues midbrain dopamine neurons R-value) above the threshold for each time point and transfection electrodes, while borosilicate glass capillary pipettes filled with Statistics from alpha-synuclein toxicity. Proc Natl Acad Sci USA 110:E1817 – E1826 group were averaged and represented. recording ACSF (resistances of 2–3.5 MΩ) were used to record field Dickey CA, Kamal A, Lundgren K, Klosak N, Bailey RM, Dunmore J, Ash P, excitatory postsynaptic potentials (fEPSPs) in the CA1 region. All data are presented as average s.e.m. Power analysis was Shoraka S, Zlatkovic J, Eckman CB et al (2007) The high-affinity Æ CLEAR sequence analysis and chromatin Signals were amplified using a MultiClamp 700 B amplifier (Axon), performed using confident interval of a = 0.05. N = 3–14 mice/ HSP90-CHIP complex recognizes and selectively degrades phosphorylated immunoprecipitation (ChIP) digitized using a Digidata 1322A (Axon) with a 2 kHz low pass filter group were used for each experiment as specified. Outliers were tau client proteins. J Clin Investig 117: 648 – 658 and a 3 Hz high pass filter and then captured and stored using identified using Grubbs’ method with a = 0.05. Pairwise compari- Dickey CA, Koren J, Zhang YJ, Xu YF, Jinwal UK, Birnbaum MJ, Monks B, Sun Human and mouse gene promoters were retrieved from the Ensem- Clampex 9 software (Axon) for offline data analysis. The genotypes sons were analyzed using a two-tailed Student’s t-test, while a two- M, Cheng JQ, Patterson C et al (2008) Akt and CHIP coregulate tau ble database (http://www.ensembl.org) and searched with the and treatment groups were blinded to the experimenter. way ANOVA with Bonferroni post hoc was used for multiple degradation through coordinated interactions. Proc Natl Acad Sci USA 105: CLEAR PWM using the PatSer tool with default parameters as comparisons. Correlation between TFEB protein levels and PHF-1 3622 – 3627 described (Sardiello et al, 2009; Settembre et al, 2011). Stereology was determined using Pearson Product Moment Correlation test. Franklin KBJ, Paxinos G (2001) The Mouse Brain in Stereotaxic Coordinates, The ChIP analysis was performed as previously described (Palmieri P-values less than or equal to 0.05 were considered statistically 2nd edn. San Diego: Academic Press et al, 2011). Briefly, formaldehyde-fixed nuclei were isolated from The number of CA1 neurons was assessed on NeuN-DAB stained significant. *P < 0.05; **P < 0.01; *** P < 0.001. Gendron TF, Petrucelli L (2009) The role of tau in neurodegeneration. Mol HeLa transfectants carrying a TFEB-3xFLAG transgene or a control sections from wild-type and rTg4510 Tau mice with either AAV-GFP Neurodegener 4: 13 HeLa cell line without any tagged transgene. ChIP was performed or AAV-TFEB P0 injections (3 in each genotype and treatment group) Supplementary information for this article is available online: Giannakopoulos P, Herrmann FR, Bussiere T, Bouras C, Kovari E, Perl DP, using the ANTI-FLAG M2 Affinity Gel (Sigma) or IgG according at 4 months of age. Unbiased stereology was performed using optical http://embomolmed.embopress.org Morrison JH, Gold G, Hof PR (2003) Tangle and neuron numbers, but not

ª 2014 The Authors EMBO Molecular Medicine Vol 6 | No 9 | 2014 1157 1158 EMBO Molecular Medicine Vol 6 | No 9 | 2014 ª 2014 The Authors Vinicia Assunta Polito et al Potent targeting of tauopathy by TFEB EMBO Molecular Medicine EMBO Molecular Medicine Potent targeting of tauopathy by TFEB Vinicia Assunta Polito et al

Histology and Immunofluorescence to the manufacturer’s protocol. Each ChIP experiment required Acknowledgments The paper explained 107 cells. We are grateful to Dr. P. Davies (Albert Einstein School of Medicine) for the gift of CP13, MC1 and PHF1 antibodies, Dr. C. Eng (Cleveland Clinic Foundation) for Cells grown on cover slips were fixed by 4% paraformaldehyde at Oligonucleotide sequences are as follows: Problem room temperature for 10 min after removal of culture medium, HPRT-F: GCCACAGGTAGTGCAAGGTCTT; HPRT-R: TTCATGGC Alzheimer’s disease has two pathological hallmarks: b-amyloid providing the pGL3-PTEN reporter constructs and Dr. X. Jiang (Memorial Sloan washed in TBS for 5 min, followed by the same staining procedure GGCCGTAAAC plaques and neurofibrillary tangles (NFTs), the latter are composed of Kettering Cancer Center) for the PTEN expression and shRNA constructs. We as brain sections. Mice brains were collected after PBS perfusion; APRT-F: GCCTTGACTCGCACTTTTGT; APRT-R: TAGGCGCCATC misfolded Tau protein. In the past, the field has devoted intensive are indebted to C. Spencer and the Baylor College of Medicine IDDRC Adminis- this procedure was followed by an over-night fixation of the brains GATTTTAAG effort on b-amyloid based therapies and these have yielded disap- trative, Genomic and RNA Profiling and Mouse Neurobehavioral cores pointing results so far. One of the concerns is that these agents may (HD024064) for the generous resources and support. We thank A. Cole and N. with 4% paraformaldehyde in PBS. Then the brains were dehy- MCOLN1-F: AGGGGCTCTGGGCTACC; MCOLN1-R: GCCCGCCG not target the NFT pathology and strong evidence has implicated a drated in 30% sucrose in PBS. Immunofluorescence analyses were CTGTCACTG neurotoxic function of NFTs in Alzheimer’s disease and other neurode- Aithmitti for expert technical assistance and members of the Zheng laboratory performed on 30-lm frozen sections. The slices were incubated for PTEN-F: ATGTGGCGGGACTCTTTATG; PTEN-R: ACAGCGGCT- generative diseases. Therefore, there is increasing interest in develop- for constructive discussions. This work was supported by grants from NIH 1 h with TBS blocking solution before incubation overnight with the CAACTCTCAAAC ing Tau-based therapy for treating Alzheimer’s disease. (AG020670, AG032051 and NS076117 to HZ) and the Belfer Neurodegeneration primary antibodies. After washing, the sections were incubated for Consortium. 1 h with secondary antibodies conjugated either with Alexafluor Morris water maze (MWM) assay Results In the current study, we investigated the role of Transcription Factor 488 or Alexafluor 647 (Invitrogen). Images were taken on a confocal Author contributions EB (TFEB), a molecule shown to play central roles in the autophagy microscope (Leica SPE). The number of positive cells was quantified MWM tests were performed in 4-month-old male mice. We used 54 and lysosomal pathway, in the rTg4510 Tau transgenic mouse models. VAP, HL, HM-S, BW, LY, YX, DW, MP, AdR performed experiments and analyzed by using ImageJ software analysis. The number of autophagosomes animals in total for behavior test, at 10–14 mice/group. The genotypes We found that adeno-associated virus-mediated TFEB expression has the data, VAP, HL and HM-S and HZ designed the study, MS performed per cell was quantified based on puncta number in the green chan- and treatment groups were blinded to the experimenter. Briefly, mice no untoward side effects on wild-type mice. In contrast, TFEB is bioinformatics analysis, VML and AB provided the Tau inducible cell line nel. The number of total autophagosomes and autolysosomes per were trained to locate a square platform (9 × 9 cm) hidden in Quadrant highly efficacious in reducing neurofibrillary tangle pathology and and TFEB vectors/viruses, respectively, as well as overall guidance, HZ wrote rescuing behavioral and synaptic deficits and neurodegeneration in cell was quantified based on puncta number in the red channel. The 3 (1–1.5 cm below water surface) of a circular water pool (1.38 m the manuscript together with VAP. All authors made comments on the the Tau mouse model. TFEB specifically targets the misfolded Tau number of autolysosomes per cell was calculated as total number diameter, 21°C) within 60. In the learning sessions, each mouse was species while leaving normal Tau intact. We provide evidence that manuscript. minus number of autophagosomes. trained for 8 training blocks in four consecutive days. Each day, mice this effect requires lysosomal activity and we identify phosphatase would swim for 4 successive trials as one training block before being and tensin homolog (PTEN) as a direct target of TFEB that may be Conflict of interest Quantitative colocalization analysis returned to its holding cage. At the end of the 8th training block on day required for TFEB-dependent aberrant Tau clearance. The authors declare that they have no conflict of interest. 4, mice were subjected to the probe test during which the platform was Impact DOX (1,000 lg/ml) was added to culture medium when T40PL removed and each mouse would swim for 60 s to search for the The specificity and efficacy of TFEB in mediating the clearance of toxic cells were plated on cover slips in 24-well plates. 3xFLAG-TFEB platform. For all the trials, mice were allowed to climb to the platform Tau species and rescuing neurodegeneration is remarkable. Our find- References and pEGFP (control) plasmids were transfected using Lipofecta- and rest for 10 s before they were submitted to the next trial or trans- ings provide proof-of-concept that small molecule TFEB activators mine 3000 (GE lifesciences) 24 h after DOX induction. Cells on ferred back to holding cages. The time for mice to locate the hidden may serve as effective therapy for treating diseases of tauopathy Barrett T, Wilhite SE, Ledoux P, Evangelista C, Kim IF, Tomashevsky M, coverslips were fixed at 0, 8, 16, 24, 32 and 40 h time points for platform during learning, as well as the number of platform crossings including Alzheimer’s disease. Marshall KA, Phillippy KH, Sherman PM, Holko M et al (2013) NCBI GEO: immunofluorescent staining. Cells transfected with 3xFLAG-TFEB during the probe tests, was recorded by Ethovision tracking system archive for functional genomics data sets–update. Nucleic Acids Res 41: were stained by antibodies anti-FLAG (rat mAb) and PHF1 (mouse (Noldus Information Technologies). D991 – D995 mAb) antibodies, while cells transfected with pEGFP were stained dissector method (West & Gundersen, 1990) on Zeiss microscope Brunden KR, Trojanowski JQ, Lee VM (2009) Advances in tau-focused drug by PHF1 only. To minimize the bleed-through effect in immunoflu- Electrophysiology with motorized stage. Data acquisition and analysis was performed discovery for Alzheimer’s disease and related tauopathies. Nat Rev Drug orescent image collection, Alexa-647-anti-mouse IgG (Invitrogen) using software StereoInvestigator by MBF Bioscience. The CA1 Discov 8: 783 – 793 was chosen to be paired with Alexa-488-anti-rat IgG as the second- Field recordings of Schaffer collateral LTP was performed as region of counting was outlined under 5× objective lens on every Caccamo A, Magri A, Medina DX, Wisely EV, Lopez-Aranda MF, Silva AJ, Oddo ary antibody for PHF1. Confocal images were collected on a Leica described (Peethumnongsin et al, 2010). Briefly, brains were 10th section of 35-lm-thick cryostat sections from 0.24 to 1.36 mm S(2013) mTor regulates tau phosphorylation and degradation: TCS SPE microscope; data analysis was carried out using Fiji- isolated from 4-month-old mice and cut into 400-lM slices (Leica). lateral from the midline on one hemisphere of the mouse brain, implications for alzheimer’s disease and other tauopathies. Aging Cell 12: ImageJ (Schindelin et al, 2012). Specifically, for images collected Hippocampal slices were incubated for 1 h at room temperature and according to the stereotaxic coordinates adopted from a mouse brain 370 – 380 using 10× objective lens, 8 confocal projection slices per view then transferred to a heated recording chamber filled with recording atlas (Franklin & Paxinos, 2001). NeuN-positive neurons were Cruz-Orive LM (1999) Precision of Cavalieri sections and slices with local 2 field, 4 view fields per time point per transfection were analyzed ACSF (125 mM NaCl, 2.5 mM KCl, 1.25 mM NaH2PO4, 25 mM counted in counting frame area of 625 lm with sampling grid area errors. J Microsc 193: 182 – 198 2 by plugin ‘coloc2’, in which the threshold of images were chosen NaHCO3, 1 mM MgCl2, 2 mM CaCl2, and 10 mM glucose, saturated of 26,338 lm under 40× objective lens. Volume of the region was Cuervo AM (2011) Cell biology. Autophagy’s top chef. Science 332: 1392 – 1393 automatically for objective analysis. Pearson Product Moment with 95% O2 and 5% CO2) maintained at 32°C. Stimulation of determined by Cavalieri’s method (Cruz-Orive, 1999). Decressac M, Mattsson B, Weikop P, Lundblad M, Jakobsson J, Bjorklund A Correlation (also called Pearson correlation coefficient or Pearson Schaffer collaterals from the CA3 region was performed with bipolar (2013) TFEB-mediated autophagy rescues midbrain dopamine neurons R-value) above the threshold for each time point and transfection electrodes, while borosilicate glass capillary pipettes filled with Statistics from alpha-synuclein toxicity. Proc Natl Acad Sci USA 110:E1817 – E1826 group were averaged and represented. recording ACSF (resistances of 2–3.5 MΩ) were used to record field Dickey CA, Kamal A, Lundgren K, Klosak N, Bailey RM, Dunmore J, Ash P, excitatory postsynaptic potentials (fEPSPs) in the CA1 region. All data are presented as average s.e.m. Power analysis was Shoraka S, Zlatkovic J, Eckman CB et al (2007) The high-affinity Æ CLEAR sequence analysis and chromatin Signals were amplified using a MultiClamp 700 B amplifier (Axon), performed using confident interval of a = 0.05. N = 3–14 mice/ HSP90-CHIP complex recognizes and selectively degrades phosphorylated immunoprecipitation (ChIP) digitized using a Digidata 1322A (Axon) with a 2 kHz low pass filter group were used for each experiment as specified. Outliers were tau client proteins. J Clin Investig 117: 648 – 658 and a 3 Hz high pass filter and then captured and stored using identified using Grubbs’ method with a = 0.05. Pairwise compari- Dickey CA, Koren J, Zhang YJ, Xu YF, Jinwal UK, Birnbaum MJ, Monks B, Sun Human and mouse gene promoters were retrieved from the Ensem- Clampex 9 software (Axon) for offline data analysis. The genotypes sons were analyzed using a two-tailed Student’s t-test, while a two- M, Cheng JQ, Patterson C et al (2008) Akt and CHIP coregulate tau ble database (http://www.ensembl.org) and searched with the and treatment groups were blinded to the experimenter. way ANOVA with Bonferroni post hoc was used for multiple degradation through coordinated interactions. Proc Natl Acad Sci USA 105: CLEAR PWM using the PatSer tool with default parameters as comparisons. Correlation between TFEB protein levels and PHF-1 3622 – 3627 described (Sardiello et al, 2009; Settembre et al, 2011). Stereology was determined using Pearson Product Moment Correlation test. Franklin KBJ, Paxinos G (2001) The Mouse Brain in Stereotaxic Coordinates, The ChIP analysis was performed as previously described (Palmieri P-values less than or equal to 0.05 were considered statistically 2nd edn. San Diego: Academic Press et al, 2011). Briefly, formaldehyde-fixed nuclei were isolated from The number of CA1 neurons was assessed on NeuN-DAB stained significant. *P < 0.05; **P < 0.01; *** P < 0.001. Gendron TF, Petrucelli L (2009) The role of tau in neurodegeneration. Mol HeLa transfectants carrying a TFEB-3xFLAG transgene or a control sections from wild-type and rTg4510 Tau mice with either AAV-GFP Neurodegener 4: 13 HeLa cell line without any tagged transgene. ChIP was performed or AAV-TFEB P0 injections (3 in each genotype and treatment group) Supplementary information for this article is available online: Giannakopoulos P, Herrmann FR, Bussiere T, Bouras C, Kovari E, Perl DP, using the ANTI-FLAG M2 Affinity Gel (Sigma) or IgG according at 4 months of age. Unbiased stereology was performed using optical http://embomolmed.embopress.org Morrison JH, Gold G, Hof PR (2003) Tangle and neuron numbers, but not

amyloid load, predict cognitive status in Alzheimer’s disease. Neurology 60: immuno-electron microscopy study. J Neuropathol Exp Neurol 64: Song MS, Salmena L, Pandolfi PP (2012) The functions and regulation of the Wang Y, Martinez-Vicente M, Kruger U, Kaushik S, Wong E, Mandelkow EM, 1495 – 1500 113 – 122 PTEN tumour suppressor. Nat Rev Mol Cell Biol 13: 283 – 296 Cuervo AM, Mandelkow E (2010) Synergy and antagonism of Haq R, Fisher DE (2011) Biology and clinical relevance of the micropthalmia Nixon RA, Yang DS (2011) Autophagy failure in Alzheimer’s disease–locating Song W, Wang F, Savini M, Ake A, di Ronza A, Sardiello M, Segatori L (2013) macroautophagy and chaperone-mediated autophagy in a cell model of family of transcription factors in human cancer. J Clin Oncol 29: the primary defect. Neurobiol Dis 43: 38 – 45 TFEB regulates lysosomal proteostasis. Hum Mol Genet 22: 1994 – 2009 pathological tau aggregation. Autophagy 6: 182 – 183 3474 – 3482 Oakley H, Cole SL, Logan S, Maus E, Shao P, Craft J, Guillozet-Bongaarts A, Sperow M, Berry RB, Bayazitov IT, Zhu G, Baker SJ, Zakharenko SS (2011) West MJ, Gundersen HJ (1990) Unbiased stereological estimation of the Hardy J (2006) A hundred years of Alzheimer’s disease research. Neuron 52: Ohno M, Disterhoft J, Van Eldik L et al (2006) Intraneuronal beta-amyloid Phosphatase and tensin homologue (PTEN) regulates synaptic plasticity number of neurons in the human hippocampus. J Comp Neurol 296: 3 – 13 aggregates, neurodegeneration, and neuron loss in transgenic mice with independently of its effect on neuronal morphology and migration. 1 – 22 Harris H, Rubinsztein DC (2012) Control of autophagy as a therapy for five familial Alzheimer’s disease mutations: potential factors in amyloid J Physiol 590: 777 – 792 Yang DS, Stavrides P, Mohan PS, Kaushik S, Kumar A, Ohno M, Schmidt SD, neurodegenerative disease. Nat Rev Neurol 8: 108 – 117 plaque formation. J Neurosci 26: 10129 – 10140 Subramanian A, Tamayo P, Mootha VK, Mukherjee S, Ebert BL, Gillette MA, Wesson D, Bandyopadhyay U, Jiang Y et al (2011) Reversal of autophagy da Huang W, Sherman BT, Lempicki RA (2009) Systematic and integrative Ozcelik S, Fraser G, Castets P, Schaeffer V, Skachokova Z, Breu K, Clavaguera Paulovich A, Pomeroy SL, Golub TR, Lander ES et al (2005) Gene set dysfunction in the TgCRND8 mouse model of Alzheimer’s disease analysis of large gene lists using DAVID bioinformatics resources. Nat F, Sinnreich M, Kappos L, Goedert M et al (2013) Rapamycin Attenuates enrichment analysis: a knowledge-based approach for interpreting ameliorates amyloid pathologies and memory deficits. Brain 134: Protoc 4: 44 – 57 the Progression of Tau Pathology in P301S Tau Transgenic Mice. PLoS ONE genome-wide expression profiles. Proc Natl Acad Sci USA 102: 258 – 277 Inoue K, Rispoli J, Kaphzan H, Klann E, Chen EI, Kim J, Komatsu M, Abeliovich 8:e62459 15545 – 15550 Zhang X, Li F, Bulloj A, Zhang YW, Tong G, Zhang Z, Liao FF, Xu H (2006a) A(2012) Macroautophagy deficiency mediates age-dependent Palmieri M, Impey S, Kang H, di Ronza A, Pelz C, Sardiello M, Ballabio A Takeuchi K, Gertner MJ, Zhou J, Parada LF, Bennett MV, Zukin RS (2013) Tumor-suppressor PTEN affects tau phosphorylation, aggregation, and neurodegeneration through a phospho-tau pathway. Mol Neurodegener (2011) Characterization of the CLEAR network reveals an integrated Dysregulation of synaptic plasticity precedes appearance of morphological binding to microtubules. FASEB J 20: 1272 – 1274 7: 48 control of cellular clearance pathways. Hum Mol Genet 20: 3852 – 3866 defects in a Pten conditional knockout mouse model of autism. Proc Natl Zhang X, Zhang YW, Liu S, Bulloj A, Tong GG, Zhang Z, Liao FF, Xu H (2006b) Kerr F, Rickle A, Nayeem N, Brandner S, Cowburn RF, Lovestone S (2006) Peethumnongsin E, Yang L, Kallhoff-Munoz V, Hu L, Takashima A, Pautler Acad Sci USA 110: 4738 – 4743 Tumor suppressor PTEN affects tau phosphorylation: deficiency in the PTEN, a negative regulator of PI3 kinase signalling, alters tau RG, Zheng H (2010) Convergence of presenilin- and tau-mediated Teresi RE, Planchon SM, Waite KA, Eng C (2008) Regulation of the PTEN phosphatase activity of PTEN increases aggregation of an FTDP-17 mutant phosphorylation in cells by mechanisms independent of GSK-3. FEBS Lett pathways on axonal trafficking and neuronal function. J Neurosci 30: promoter by statins and SREBP. Hum Mol Genet 17: 919 – 928 Tau. Mol Neurodegener 1: 7 580: 3121 – 3128 13409 – 13418 The Gene Ontology Consortium (2012) Gene Ontology annotations and Zhou J, Blundell J, Ogawa S, Kwon CH, Zhang W, Sinton C, Powell CM, Parada Kimura S, Noda T, Yoshimori T (2007) Dissection of the autophagosome Pickford F, Masliah E, Britschgi M, Lucin K, Narasimhan R, Jaeger PA, Small S, resources. Nucleic Acids Res 41:D530 – D535 LF (2009) Pharmacological inhibition of mTORC1 suppresses anatomical, maturation process by a novel reporter protein, tandem Spencer B, Rockenstein E, Levine B et al (2008) The autophagy-related Wang X, Trotman LC, Koppie T, Alimonti A, Chen Z, Gao Z, Wang J, cellular, and behavioral abnormalities in neural-specific Pten knock-out fluorescent-tagged LC3. Autophagy 3: 452 – 460 protein beclin 1 shows reduced expression in early Alzheimer disease and Erdjument-Bromage H, Tempst P, Cordon-Cardo C et al (2007) mice. J Neurosci 29: 1773 – 1783 Kopeikina KJ, Hyman BT, Spires-Jones TL (2012) Soluble forms of tau are toxic regulates amyloid beta accumulation in mice. J Clin Invest 118: NEDD4-1 is a proto-oncogenic ubiquitin ligase for PTEN. Cell 128: in Alzheimer’s disease. Transl Neurosci 3: 223 – 233 2190 – 2199 129 – 139 License: This is an open access article under the Kruger U, Wang Y, Kumar S, Mandelkow EM (2012) Autophagic degradation Ramsden M, Kotilinek L, Forster C, Paulson J, McGowan E, SantaCruz K, Wang Y, Martinez-Vicente M, Kruger U, Kaushik S, Wong E, Mandelkow EM, terms of the Creative Commons Attribution 4.0 of tau in primary neurons and its enhancement by trehalose. Neurobiol Guimaraes A, Yue M, Lewis J, Carlson G et al (2005) Age-dependent Cuervo AM, Mandelkow E (2009) Tau fragmentation, aggregation and License, which permits use, distribution and reproduc- Aging 33: 2291 – 2305 neurofibrillary tangle formation, neuron loss, and memory impairment in clearance: the dual role of lysosomal processing. Hum Mol Genet 18: tion in any medium, provided the original work is Kwon CH, Luikart BW, Powell CM, Zhou J, Matheny SA, Zhang W, Li Y, Baker a mouse model of human tauopathy (P301L). J Neurosci 25: 10637 – 10647 4153 – 4170 properly cited. SJ, Parada LF (2006) Pten regulates neuronal arborization and social Roczniak-Ferguson A, Petit CS, Froehlich F, Qian S, Ky J, Angarola B, Walther interaction in mice. Neuron 50: 377 – 388 TC, Ferguson SM (2012) The transcription factor TFEB links mTORC1 KwonCH,ZhuX,ZhangJ,BakerSJ(2003) mTor is required for hypertrophy of signaling to transcriptional control of lysosome homeostasis. Sci Signal 5: Pten-deficient neuronal soma in vivo. Proc Natl Acad Sci USA 100: 12923 – 12928 ra42 Kwon CH, Zhu X, Zhang J, Knoop LL, Tharp R, Smeyne RJ, Eberhart CG, Burger Rohn TT, Wirawan E, Brown RJ, Harris JR, Masliah E, Vandenabeele P (2011) PC, Baker SJ (2001) Pten regulates neuronal soma size: a mouse model of Depletion of Beclin-1 due to proteolytic cleavage by caspases in the Lhermitte-Duclos disease. Nat Genet 29: 404 – 411 Alzheimer’s disease brain. Neurobiol Dis 43: 68 – 78 Lin WL, Lewis J, Yen SH, Hutton M, Dickson DW (2003) Ultrastructural Santacruz K, Lewis J, Spires T, Paulson J, Kotilinek L, Ingelsson M, Guimaraes neuronal pathology in transgenic mice expressing mutant (P301L) human A, DeTure M, Ramsden M, McGowan E et al (2005) Tau suppression in a tau. J Neurocytol 32: 1091 – 1105 neurodegenerative mouse model improves memory function. Science 309: Lopes CT, Franz M, Kazi F, Donaldson SL, Morris Q, Bader GD (2010) 476 – 481 Cytoscape Web: an interactive web-based network browser. Bioinformatics Sardiello M, Palmieri M, di Ronza A, Medina DL, Valenza M, Gennarino VA, Di 26: 2347 – 2348 Malta C, Donaudy F, Embrione V, Polishchuk RS et al (2009) A gene Mandelkow EM, Mandelkow E (2012) Biochemistry and cell biology of tau network regulating lysosomal biogenesis and function. Science 325: protein in neurofibrillary degeneration. Cold Spring Harb Perspect Med 2: 473 – 477 a006247 Schaeffer V, Lavenir I, Ozcelik S, Tolnay M, Winkler DT, Goedert M (2012) Medina DL, Fraldi A, Bouche V, Annunziata F, Mansueto G, Spampanato C, Puri Stimulation of autophagy reduces neurodegeneration in a mouse model C, Pignata A, Martina JA, Sardiello M et al (2011) Transcriptional activation of human tauopathy. Brain 135: 2169 – 2177 of lysosomal exocytosis promotes cellular clearance. Dev Cell 21: 421 – 430 Schindelin J, Arganda-Carreras I, Frise E, Kaynig V, Longair M, Pietzsch T, Mueller-Steiner S, Zhou Y, Arai H, Roberson ED, Sun B, Chen J, Wang X, Yu G, Preibisch S, Rueden C, Saalfeld S, Schmid B et al (2012) Fiji: an Esposito L, Mucke L et al (2006) Antiamyloidogenic and neuroprotective open-source platform for biological-image analysis. Nat Methods 9: functions of cathepsin B: implications for Alzheimer’s disease. Neuron 51: 676 – 682 703 – 714 Settembre C, Di Malta C, Polito VA, Garcia Arencibia M, Vetrini F, Erdin S, Mullard A (2012) Sting of Alzheimer’s failures offset by upcoming prevention Erdin SU, Huynh T, Medina D, Colella P et al (2011) TFEB links autophagy trials. Nat Rev Drug Discov 11: 657 – 660 to lysosomal biogenesis. Science 332: 1429 – 1433 Nixon RA (2007) Autophagy, amyloidogenesis and Alzheimer disease. J Cell Sci Settembre C, Zoncu R, Medina DL, Vetrini F, Erdin S, Huynh T, Ferron M, 120: 4081 – 4091 Karsenty G, Vellard MC, Facchinetti V et al (2012) A lysosome-to-nucleus Nixon RA, Wegiel J, Kumar A, Yu WH, Peterhoff C, Cataldo A, Cuervo AM signalling mechanism senses and regulates the lysosome via mTOR and (2005) Extensive involvement of autophagy in Alzheimer disease: an TFEB. EMBO J 31: 1095 – 1108

ª 2014 The Authors EMBO Molecular Medicine Vol 6 | No 9 | 2014 1159 1160 EMBO Molecular Medicine Vol 6 | No 9 | 2014 ª 2014 The Authors Vinicia Assunta Polito et al Potent targeting of tauopathy by TFEB EMBO Molecular Medicine EMBO Molecular Medicine Potent targeting of tauopathy by TFEB Vinicia Assunta Polito et al

ª 2014 The Authors EMBO Molecular Medicine Vol 6 | No 9 | 2014 1159 1160 EMBO Molecular Medicine Vol 6 | No 9 | 2014 ª 2014 The Authors

The EMBO Journal Listeria phospholipases subvert ULK1 translocates EMBO Molecular Medicine host autophagic defenses by to mitochondria and Review stalling pre-autophagosomal phosphorylates FUNDC1 to Reviews A new pathway for mitochondrial structures. regulate mitophagy. TORC —a new player in genome quality control: mitochondrial- Tattoli I, Sorbara MT, Yang C, Tooze SA, Wu W, Tian W, Hu Z, Chen G, Huang L, stability. derived vesicles. Philpott DJ, Girardin SE. Li W, Zhang X, Xue P, Zhou C, Liu L, Zhu Weisman R, Cohen A, Gasser SM. Sugiura A, McLelland GL, Fon EA, McBride Y, Zhang X, Li L, Zhang L, Sui S, Zhao B, DOI: ./emboj.. DOI:./emmm. HM. Feng D. Published .. Published .. DOI: ./embj. DOI: ./embr. Published .. Mfn modulates the UPR and Published .. Mitochondrial response to mitochondrial function via nutrient availability and its role TBC1D5 and the AP2 complex Articles repression of PERK. in metabolic disease. regulate ATG9 trafcking and Parkin-independent mitophagy Muñoz JP, Ivanova S, Sánchez- Gao AW, Cantó C, Houtkooper RH. Wandelmer J, Martínez-Cristóbal P, initiation of autophagy. requires Drp and maintains the DOI: ./emmm. Noguera E, Sancho A, Díaz-Ramos A, Popovic D, Dikic I. integrity of mammalian heart and Published .. Hernández-Alvarez MI, Sebastián D, brain. DOI: ./embr. Mauvezin C, Palacín M, Zorzano A. |Published .. Autophagic control of cell Kageyama Y, Hoshijima M, Seo K, Bedja DOI: ./emboj.. ‘stemness’. D, Sysa-Shah P, Andrabi SA, Chen W, Höke Published .. Structural determinants A, Dawsn VL, Dawson TM, Gabrielson K, Pan H, Cai N, Li M, Liu GH, Izpisua in GABARAP required for Kass DA, Iijima M, Sesaki H. Belmonte JC. XIAP inhibits autophagy via XIAP- the selective binding and DOI: ./embj. DOI: ./emmm. Mdm -p signalling. recruitment of ALFY to LC3B- Published .. Published .. Huang X, Wu Z, Mei Y, Wu M. positive structures. Articles USP regulates mitophagy by DOI: ./emboj.. Lystad AH, Ichimura Y, Takagi K, Yang removing K-linked ubiquitin Published .. Y, Pankiv S, Kanegae Y, Kageyama S, Defects in GABA metabolism conjugates from parkin. Suzuki M, Saito I, Mizushima T, Komatsu affect selective autophagy Autolysosomal β-catenin M, Simonsen A. pathways and are alleviated by Durcan TM, Tang MY, Pérusse JR, Dashti degradation regulates Wnt- EA, Aguileta MA, McLelland GL, Gros P, DOI: ./embr. mTOR inhibition. autophagy-p crosstalk. Published .. Shaler TA, Faubert D, Coulombe B, Fon EA. Lakhani R, Vogel KR, Till A, Liu J, Burnett Petherick KJ, Williams AC, Lane JD, DOI: ./embj. SF, Gibson KM, Subramani S. Ordóñez-Morán P, Huelsken J, Collard TJ, Cytosolic cleaved PINK1 Published .. Smartt HJ, Batson J, Malik K, Paraskeva C, represses Parkin translocation DOI: ./emmm. Published .. Parkin and PINK Function in a Greenhough A. to mitochondria and mitophagy. Vesicular Trafcking Pathway DOI: ./emboj.. Fedorowicz MA, de Vries-Schneider RL, Tyrosine kinase inhibition Regulating Mitochondrial Quality Published .. Rüb C, Becker D, Huang Y, Zhou C, Alessi increases functional parkin- Control. Wolken DM, Voos W, Liu Y, Przedborski S. Beclin- interaction and emboj.embopress.org DOI:./embr. McLelland GL, Soubannier V, Chen CX, enhances amyloid clearance and Published .. McBride HM, Fon EA. cognitive performance. EMBO Reports DOI: ./embj. Loss of iron triggers PINK1/ Lonskaya I, Hebron ML, Desforges NM, Published .. Articles Parkin-independent mitophagy. Franjie A, Moussa CE. DOI: ./emmm. Autophagy-related gene Atg5 Allen GF, Toth R, James J, Ganley IG. Autophagy proteins control Published .. goblet cell function by is essential for astrocyte DOI: ./embor.. potentiating reactive oxygen differentiation in the developing Published .. Embelin inhibits endothelial species production. mouse cortex. mitochondrial respiration and Autophagy in Myf5+ progenitors Wang S, Li B, Qiao H, Lv X, Liang Q, Shi Z, impairs neoangiogenesis during Patel KK, Miyoshi H, Beatty WL, Head RD, regulates energy and glucose Malvin NP, Cadwell K, Guan JL, Saitoh T, Xia W, Ji F, Jiao J. tumor growth and wound healing. homeostasis through control of Akira S, Seglen PO, Dinauer MC, Virgin DOI:./embr. Coutelle O, Hornig-Do HT, Witt A, Andree brown fat and skeletal muscle HW, Stappenbeck TS. Published .. M, Schiffmann LM, Piekarek M, Brinkmann development. DOI: ./emboj.. K, Seeger JM, Liwschitz M, Miwa S, Hallek Published .. Hrr25 kinase promotes selective Martinez-Lopez N, Athonvarangkul D, M, Krönke M, Trifunovic A, Eming SA, autophagy by phosphorylating Sahu S, Coletto L, Zong H, Bastie CC, Wiesner RJ, Hacker UT, Kashkar H. WASH inhibits autophagy the cargo receptor Atg19. Pessin JE, Schwartz GJ, Singh R. DOI: ./emmm. through suppression of Beclin Pfaffenwimmer T, Reiter W, Brach T, DOI: ./embor.. Published .. ubiquitination. Nogellova V, Papinski, Schuschnig M, Published .. Xia P, Wang S, Du Y, Zhao Z, Shi L, Sun L, Abert C, Ammerer G, Martens S, Kraft C. embomolmed.embopress.org Casein kinase 2 is essential for Huang G, Ye B, Li C, Dai Z, Hou N, Cheng X, DOI:./embr. mitophagy. Sun Q, Li L, Yang X, Fan Z. Published .. Molecular Systems Biology DOI: ./emboj.. Kanki T, Kurihara Y, Jin X, Goda T, Ono Published .. Genome-wide screen identies Y, Aihara M, Hirota Y, Saigusa T, Aoki Y, Article Uchiumi T, Kang D. signaling pathways that Screen for mitochondrial DNA Autophagy sequesters damaged regulate autophagy during DOI: ./embor.. copy number maintenance genes |Published .. lysosomes to control lysosomal Caenorhabditis elegans reveals essential role for ATP biogenesis and kidney injury. development. synthase. embor.embopress.org Maejima I, Takahashi A, Omori H, Kimura Guo B, Huang X, Zhang P, Qi L, Liang Q, Fukuoh A, Cannino G, Gerards M, Buckley T, Takabatake Y, Saitoh T, Yamamoto A, Zhang X, Huang J, Fang B, Hou W, Han S, Kazancioglu S, Scialo F, Lihavainen E, Hamasaki M, Noda T, Isaka Y, Yoshimori T. J, Zhang H. Ribeiro A, Dufour E, Jacobs HT. DOI: ./emboj.. DOI:./embr. DOI: ./msb. Published .. Published . . Published ..

msb.embopress.org EMBO Molecular Medicine

embor.embopress.org | emboj.embopress.org | embomolmed.embopress.org | msb.embopress.org

58403 / MITM027320

Autophagy -   Autophagy

Autophagy - Autophagy