Interplay Between Autophagy and the Primary Cilium: Role in Mechanical

Interplay between autophagy and the primary cilium : Role in mechanical stress integration Idil Orhon

To cite this version:

Idil Orhon. Interplay between autophagy and the primary cilium : Role in mechanical stress integration. Cellular Biology. Université Pierre et Marie Curie - Paris VI, 2014. English. NNT : 2014PA066672. tel-01242554

HAL Id: tel-01242554 https://tel.archives-ouvertes.fr/tel-01242554 Submitted on 13 Dec 2015

HAL is a multi-disciplinary open access L’archive ouverte pluridisciplinaire HAL, est archive for the deposit and dissemination of sci- destinée au dépôt et à la diffusion de documents entific research documents, whether they are pub- scientifiques de niveau recherche, publiés ou non, lished or not. The documents may come from émanant des établissements d’enseignement et de teaching and research institutions in France or recherche français ou étrangers, des laboratoires abroad, or from public or private research centers. publics ou privés. Université Pierre et Marie Curie (Paris 6)

Ecole doctorale Physiologie et Physiopathologie

2014

DOCTORAL THESIS for University Pierre et Marie Curie Paris 6

Public Defense : 11 December 2014 by

Idil ORHON

entitled

INTERPLAY BETWEEN AUTOPHAGY AND THE PRIMARY CILIUM:

ROLE IN MECHANICAL STRESS INTEGRATION

Members of the Jury:

Professor Joëlle SOBCZAK-THEPOT President

Doctor Alexandre BENMERAH Reporter

Professor Fulvio REGGIORI Reporter

Doctor Vincent GALY Examinator

Doctor Sophie PATTINGRE Examinator

Doctor Isabelle BEAU Thesis Director

Doctor Patrice CODOGNO Thesis Director

Institut Necker-Enfants Malades (INEM)

INSERM UMR 1151 -CNRS UMR 8253 Université Paris Descartes-Sorbonne Paris Cité

Laboratoire de l’homéostasie cellulaire et signalisation en physiopathologie hépatique et rénale

ACKNOWLEDGMENTS

Firstly, I would like to thank the members of the jury, to Alexandre Benmerah and Fulvio Reggiori for accepting to evaluate this manuscript, to Joelle Sobczak for accepting to be the president of the jury and to Sophie Pattingre and Vincent Galy for accepting to be examiner of this work.

Foremost, I would like to express my gratitude to my thesis supervisors Patrice Codogno and Isabelle Beau for giving me the opportunity to work and learn from them. I am sincerely grateful to Patrice Codogno for his expertise, experience and confidence in me throughout these years. Without his extensive knowledge on autophagy but also on “the outside world”, I certainly could not evolve as a researcher during these years. I am deeply thankful for Isabelle Beau and her presence in my PhD journey starting from day one whether from the next door or in distance. She set an inspiring example for a woman in science with her passion of bench work and guidance.

I am greatly thankful to Ana-Maria Cuervo for giving me the opportunity to be part of an inspiring collaboration that changed my perspective on many different levels. Working in her lab was a condensed learning experience with a very dynamic and kind group of people. That leads to a sincere thanks and gratitude to Olatz Pampliega, an amazing collaborator and scientist whom I feel lucky to have worked with, learnt from and hope to work with again. Without them this work could not have happened.

I would also like to acknowledge Gerard Friedlander, for his support and contribution to this work.

A very special thanks goes to the “primary” Nicolas Dupont. During the last year, his presence changed the dynamism of our lab and this project. Without our brainstorms, his motivation and his passion for pubmed search, I wouldn’t see the finish line. I would especially like to thank him for his input and time on this manuscript.

I would like express my thanks to my chocolate loving group of colleagues with whom I feel very lucky to work with daily: Ahmeeeeeeed Hamai who was ready to answer all of my questions at any time with a smile, Joelle Botti and Chantal Bauvy for their kindness and

support and Etienne Morel who was a support even before joining the lab, Anna Chiara Nascimbeni and Maryam Mehrpour.

I’d like to thank sincerely to the Chatenay-Malabry team with whom I started this journey: Lina Mouna for her friendship and the Turkish/Syrian coffee time, Francoise Magnin for her daily kindness, Mathieu Bertrand, Marion Lussignol, Audrey Esclatine, Marie-Francoise Bernet-Camard, Anne Greffard and Raymonde Amsellem.

I want to thank everybody from Co-lander team “across the hall”, with a shout-out to Christine and Valérie for their management skills and not leaving us in chaos. I would also like to thank to Marie Courbebaisse and Dominique Prié for their insight on the kidney.

I must acknowledge the platforms that helped me: Claudine from Trans-Prot of IPSIT, Nicolas and Raphaelle from IMAGINE.

I must thank Philippe Cardot for his support, Chiara Maiuri, Ezgi Tasdemir and Devrim Gozuacik for their supervision during my ignorant years and for teaching me how to work on the bench, which I love.

I would also like to thank Rosalind, Henrietta, Simone, Charles, Pablo, Samed, Halit Ziya and Yoda for the inspiration.

Last but not least I have to thank my family and friends who make this journey worth sharing. I am grateful for the love and support of my brother, mother and father. Without the patience of my mom and dad, I could not have pursued an education in science so far. I am also thinking of my dear grandmother who passed away at the beginning of this craziness and who I wished could have seen the end of it all. I am very lucky and grateful to have too many friends to name here who shared this experience with me. However, I cannot fail to acknowledge these wonderful people for their love, jokes and support: Seray, Yasemin, Burge, Idil, Irem and especially Isil who supported my daily neurosis towards the end. I love them dearly.

TABLE OF CONTENT

Abbreviations ………………………………………………………………………………….7

Table of Illustrations…………………………………………………………………………..9

Summary……………………………………………………………………………………...10

Résumé………………………………………………………………………………………..11

I. Introduction…………………………………………………………………………..12

1. Cilium: Structure and Signaling……………………………………………………...13 1.1. Structure ……………………………………………………………...13 1.1.1. Motile Cilia and Flagella ……………………………………..13 1.1.2. Primary Cilia…………………………………………………..14 1.1.3. Ciliogenesis and Ciliary Subdomains………………………..15 1.1.4. Intraflagellar Transport Mechanism………………………....18 1.2. Ciliary Signaling………………………………………………………20 1.2.1. Hedgehog Signaling…………………………………………...20 1.2.2. PDGF Pathway………………………………………………..23 1.2.3. Wnt Pathway…………………………………………………..23 1.2.4. Calcium Signaling……………………………………………..24 1.2.5. Fluid flow and Fluid Shear Stress………………………..…..25 2. Autophagy…………………………………………………………………………….30 2.1. Role of Autophagy………………………………………………...….30 2.2. Different Types of Autophagy…………………………………..…...30 2.2.1. Macroautophagy………………………………………………30 2.2.2. Chaperone-mediated Autophagy ……………………………..31 2.2.3. Microautophagy…………………………………………... ….34 2.3. Autophagosome Machinery and Biogenesis …………………………34 2.3.1. Initiation……………………………………………………….35 2.3.2. Elongation…………………………………………………..…40 2.3.3. Maturation……………………………………………………..41 2.3.4. The origin of the Autophagosome…………………………….42 2.4. Regulation of Autophagy …………………………………………….46 2.4.1. Diverse Regulatory Pathways…………………………………46

2.4.2. Post-translational Modifications of Autophagic Machinery…..50 2.4.3. Transcriptional and Epigenetic Regulations of Autophagy…...53 2.5. Physiological Role of Autophagy……………………………………..55 2.5.1. Selective Autophagy…………………………………………..56 2.5.2. Autophagy during pre- and post-natal state…………………...58 2.5.3. Autophagy in the liver………………………………………...58 2.5.4. Autophagy in the brain………………………………………..59 2.5.5. Autophagy in the muscle……………………………………...59 2.5.6. Autophagy in the intestine…………………………………….60 2.5.7. Autophagy in the pancreas……………………………………60 2.5.8. Autophagy in the lung…………………………………………61 2.5.9. Autophagy in innate and adaptive immunity………………….61 2.5.10. Pathophysiological Role of Autophagy……………………….63 2.6. The Role of Autophagy in Kidney Physiology and Physiopathology...67 2.6.1. Physiology of the Kidney: An Introduction…………………...67 2.6.2. Autophagy in the Kidney: Health and Disease………………..71 II. AIMS of the study…………………………………………………………………….77 III. Materials and Methods………………………………………………………………..80 IV. Results …………………...…………………………………………………………...86 1. Objective1:Study the functional interaction between autophagy and primary cilium..87 1.1.Results…………………………………………………………………….89 1.2. Discussion 1……………………………………………………………..90 2. Objective 2: Study the role of autophagy in integration of mechanical stress via primary cilium………………………………………………………………………………….94 2.1. Results……………………………………………………………………96 2.2. Discussion 2……………………………………………………………..106 V. General Discussion and Perspectives………………………………………………..111

BIBLIOGRAPHY………………………………………………………………………………….…119

ANNEXE…………………………………………………………………………………………..…139

Abbreviations

GFP: green fluorescent protein 3-MA: 3-methyladenine Gli: glioma-associated oncogene family 4EBP-1: 4E binding protein-1 GPCRs: G-protein-coupled receptors ADPKD: autosomal dominant polycystic kidney disease HDAC6: histone deacetylase-6 aIFT : anterograde IFT Hh: Hedgehog

AKI: acute kidney injury IFN: interferon

AMPK: AMP-activated protein kinase IFT : Intraflagellar transport

APC: adenomatosis polyposis complex IKK: IκB kinase

AQP: aquaporin IPTG: Isopropyl β-D-1-thgalactopyranoside

ATG: autophagy related JNK1: Jun NH2-terminal kinase 1

ATP: adenosine triphosphate KEC: Kidney epithelial cell

BB : Basal body MAPK: mitogen-activated protein kinase

BBS: Bardet–Biedl syndrome MC: motile cilium

Bcl-2: B cell lymphomas MDCK: Madin-Darby canine kidney

CKD: chronic kidney disease MHC: major histocompatibility complex

CMA: chaperone-mediated autophagy mTORC1: mammalian target of rapamycin complex 1 COPD, chronic obstructive pulmonary disease; NF-κB: nuclear factor-κB CQ: Chloroquine NHE1: Na+/H+ exchanger 1 CTS: Ciliary targeting sequence OFD1: oral-facial-digital syndrome type 1 DN: diabetic nephopathy PAS: pre-autophagosomal structure EndoMT: endothelial-to-mesenchymal transition PC: Primary cilia

ER: endoplasmic reticulum PC1: polycystin 1

ERGIC: ER-Golgi intermediate compartment PC2: polycystin 2

FGF: fibroblast growth factor PCP: planar cell polarity

FOXO: conserved forkhead box O PDGF: platelet-derived growth factor FSS: Fluid Shear Stress

PDGFRα: platelet-derived growth factor receptor alpha

PE: phospholipid phosphatidylethanolamine

PI3K: class III phosphatidylinositol 3-kinase

PI3P: phosphatidylinositol 3-phosphate

PKD: polycystic kidney disease

Ptc1: patched 1 rIFT : retrograde IFT

ROS: reactive oxygen species

S6K: ribosomal protein S6 kinase-1

Shh: sonic hedgehog

Smo: smoothened

Sufu: suppressor of fused protein

TFEB: transcription factor EB

TGF-β: transforming growth factor-β

TGN: trans-Golgi networks

UUO: Unilateral ureteral obstruction

Vangl2: Van Gogh-like 2 protein

Table of Illustrations

Figure 1: Cilia Structure

Figure 2: Classification of cilia

Figure 3: Primary Ciliogenesis models

Figure 4: Intraflagellar Transport

Figure 5: Hedgehog signaling

Figure 6: Mechano-sensing by the cell

Figure 7: The different types of autophagy

Figure 8: The Autophagic pathway

Figure 9: Autophagosome machinery

Figure 10: The Origin of the Autophagosome

Figure 11: mTOR dependent regulation of Autophagy

Figure 12: Physiological Roles of Autophagy

Figure 13: Selective autophagy

Figure 14: Autophagy and Pathophysiology

Figure 15: Physiology of the nephron

Figure 16: Polycystic Kidney Disease and Fluid Flow in the Kidney

Figure 17: Flow stimulates the autophagic flux in MDCK cells

Figure 18: Inhibition of ciliogenesis impairs flow-induced autophagy

Figure 19: Inhibition of autophagy impairs primary cilium cell size regulation

Figure 20: Autophagic machinery (ATG16L1) recruited to the basal body upon fluid flow

Figure 21: Signaling regulation of primary cilium-dependent autophagy activation during flow

Table 1: Pathways regulated by mechanical stress via Primary Cilia

Table 2: Mutations in Autophagy Genes and associated human diseases

Table 3: Structure and signaling from the cilium and relationship with the autophagic pathway

Table 4: Cilia-related proteins degraded by autophagy

Summary

Motile and primary cilia are microtubule-based structures located at the cell surface of many cell types. Cilia govern cellular functions ranging from motility to integration of mechanical and chemical signaling from the environment. In this work we investigate the potential crosstalk between the primary cilium and macroautophagy. Macroautophagy or self-eating is a lysosomal degradative pathway that allows cells to adapt to various stress situations. The rational for the study was based on the survey of the literature showing that many stress situations that trigger primary cilium signaling also stimulates autophagy (serum starvation, calcium mobilization, cell cycle arrest). In the first part of the study we showed that inhibition of ciliogenesis severely impairs serum-induced autophagy in mouse embryo fibroblasts, kidney epithelial cells and neurons. We also showed that in response to serum deprivation many Autophagy-related proteins (Atg proteins) involved in autophagosome formation are co-localized with cilium subdomains (axoneme and basal body). Notably the protein Atg16L1 is co-transported to the basal body with the ciliary protein IFT20. The localization of Atg16L1 to the basal body as well as serum-induced autophagy were severely impaired by inhibiting the Hedgehog signaling pathway either genetic or pharmacological approaches. We also showed that invalidation of ATG genes induced an increase in primary cilium length in basal condition. Cilia were functional in ATG-deficient cells because of the presence of a ciliary pocket and the activation of the Hedgehog signaling pathway. Finally we identified IFT20 as a substrate for autophagy. Thus autophagy is required to regulate the level of IFT20 and consequently that of the length of the primary cilium. In the second part of the work we investigate the role of the cross-talk between autophagy and the primary cilium in regulating the size of kidney epithelial cells. Previous studies have shown that the primary cilium plays a central role in regulating cell size and cell volume. This regulation is important to keep the physiological functions of tubular renal cells by maintaining the planar polarity in kidney tubule. By applying a liquid flow of 1 dyn/cm2 to MDCK or mouse kidney epithelial cells to mimic physiological conditions, we show that the flow induces autophagy and reduction of the cell volume. In absence of cilium we observed that autophagy is not induced and that the cell size/volume is not responsive to the mechanical stress. Finally we showed that ablation of autophagy led also to an impairment of flow-dependent regulation of cell size/volume in ciliated kidney epithelial cells. In conclusion primary cilium-dependent autophagy plays a major role in controlling the epithelial kidney cell size/volume during mechanical stress induced by fluid flow.

Key words: Macroautophagy, IFT, ATG, primary cilium, fluid flow, cell size/volume, kidney epithelial cells

Résumé

Les cils primaires et motiles sont des structures microtubulaires présentent à la surface de nombreux types cellulaires. Les structures ciliées contrôlent de nombreuses fonctions allant de la motilité cellulaire à l’intégration par la cellule de stimuli chimiques et mécaniques. Au cours de cette thèse, nous avons étudié le dialogue entre le cil primaire et l’autophagie, un processus d’autodigestion qui permet à la cellule de s’adapter à des situations de stress. L’hypothèse de ce dialogue reposait sur l’analyse de la littérature montrant que de nombreux médiateurs (calcium, carence en sérum, arrêt du cycle cellulaire) stimulent à la fois l’activité ciliaire et l’autophagie. Dans un premier temps de notre étude nous avons montré que l’inhibition de la ciliogenèse altère l’induction de l’autophagie en réponse à la carence en sérum dans des fibroblastes d’embryon de souris, des cellules épithéliales rénales et des lignées de neurone. Nous avons aussi montré que la carence en sérum induisait une redistribution de nombreuses protéines Atg (Autophagy-related), protéines impliquées dans la biogenèse de l’autophagosome, au niveau du cil primaire (soit au niveau du corps basal soit au niveau de l’axonème). Particulièrement la protéine Atg16L1 est co-transportée vésiculairement au corps basal avec la protéine ciliaire IFT20. L’inhibition génétique ou pharmacologique de la voie de signalisation Hedgehog inhibe à la fois le transport de la protéine Atg16L1 au corps basal et l’induction de l’autophagie en absence de sérum. Nous avons aussi montré que l’invalidation de gènes ATG est associée à une ciliogenèse accrue. Dans ces conditions nous avons conclu sur des bases morphologiques et biochimiques que ces cils primaires sont fonctionnels. La protéine IFT20 s ‘accumule dans les cellules déficientes en autophagie et est dégradée par autophagie dans les cellules sauvage en présence de sérum. Ces résultats montre que l’autophagie basale (autophagie observée en présence de sérum) est un mécanisme qui contribue au contrôle de la croissance du cil primaire. Dans une deuxième partie du travail nous avons étudié l’importance de l’autophagie dans la réponse cellulaire à stress mécanique. Le contrôle de la taille et du volume des cellules épithéliales rénales est un élément important pour maintenir la polarité planaire des cellules tubulaires. Cette propriété est dépendante du cil primaire. Au cours de l’application d’un flux de liquide (1 dyn/cm2) concomitamment à la réduction du volume et de la taille cellulaire nous avons observé une stimulation de l’autophagie. Cette réponse autophagique dépend du cil primaire. L’invalidation de l’autophagie dans des cellules épithéliales ciliées abolit le contrôle du volume et de la taille cellulaire dans les cellules épithéliales rénales. L’ensemble de ces résultats montre le dialogue qui existe en l’autophagie et le cil primaire et l’importance de ce dialogue dans l’intégration par la cellule du stress mécanique.

Mots clés : Macroautophagie, IFT, ATG, cil primaire, stress mécanique, volume et taille cellulaire, cellules épithéliales rénales

INTRODUCTION

I. INTRODUCTION

1. Cilium: Structure and Signaling

1.1. Structure

Cilia are microtubule-based organelles that are present at the surface of many vertebrate cells (Satir and Christensen 2007; Satir, Pedersen et al. 2010). Conventionally, cilia are classified as motile and non-motile according to their microtubule pattern and motility in the organism.

1.1.1. Motile Cilia and Flagella

Motile cilium (MC) is formed by a microtubule based axoneme of 9 outer microtubule doublets and 2 single central microtubules extending from a basal body. Two dynein motors and a radial spoke are attached to the doublets along the axoneme, giving the motility to the organelle and thus the ability to generate fluid and beat-like movement (Figure 1). Motile cilia are found on the surface of the epithelial cells lining the airways and reproductive tracts, as well as on epithelial cells of the ependyma and in the brain and on the embryonic node. Motility of the cilium in these epithelial cells is responsible for mucociliary clearance and ependymal flow. The 9+2 structure of MC is identical to the flagella, which are found in sperm and single-celled eukaryotes (i.e. Paramecium) and that serve traditionally in cell locomotion (Pedersen and Rosenbaum 2008). MC and flagella differ from each other based on their pattern of movement and the number per cell. However, new studies on the structure and function of MC and flagella in different tissues or species show that they can have sensory roles such as MC in the tracheal epithelial cells which have bitter taste receptors or mammalian sperm which possess pH-dependent Ca2+ channel (Takeda and Narita 2011) (See Figure 2 for an overview of cilia classification in relation to motility).

1.1.2. Primary Cilia

Primary cilium (PC), a solitary non-motile cilium, consists of an axoneme of 9 outer doublet microtubules extending from a basal body (BB) that is derived from the older (mother) centriole of the centrosome, and which lacks the 2 single central microtubules involved in motility (Satir and Christensen 2007) (Satir, Pedersen et al. 2010). PC projects from surface of the plasma membrane of a wide range of polarized cells including stem, epithelial, endothelial, connective-tissue muscle cells and neurons (Figure 1).

Figure 1: Cilia Structure

(A) Scanning electron microscopy (SEM) of primary cilia from mouse embryonic fibroblasts. (B) Electron microscopy of PC. (C) Electron microscopy (EM) image on the bottom and schematic representation of the 9 microtubule doublets of PC on the top. (D) SEM image of motile cilia from mouse tracheal epithelium. (E) EM image of motile cilia. (F) EM image on the bottom and schematic representation of 9+2 microtubule doublets of the motile cilia on the top. (Images from Pampliega et al. 2013, Satir et al. 2010 and Sorokin 1968) (G) Schematic representation of the cilium (Orhon et al. 2014 review in press)

Figure 2: Classification of cilia

Schematic diagram showing the classification of motile “9+2”, motile “9+0”, and non-motile “9+0” cilia. The diagram shows the coordinated synchronous motion of motile “9+2” cilia, the rotary motion of motile “9+0” nodal monocilium, and non-motile “9+0” primary monocilium (modified from (Ferkol and Leigh 2012))

1.1.3. Ciliogenesis and Ciliary Subdomains

Ciliogenesis occurs in quiescent cells when the cells enter the G0 phase. When the cells decide to re-enter in a cell cycle, cilium resorption occurs followed by the S phase. Ciliogenesis and cell cycle is regulated by post-translational modifications (Avidor-Reiss and Gopalakrishnan). Sorokin proposed two models of primary ciliogenesis upon TEM observations from epithelial cells and fibroblasts (Sorokin 1962; Sorokin 1968); an ‘extracellular’ pathway and an ‘intracellular’ pathway. In the polarized cells, the ‘extracellular’ model the centrioles move directly to the plasma membrane and the mother centriole serve as a docking center for the growth of the axoneme thus forming the basal body

of the cilium (Benmerah 2012). This model is observed in the ciliated cells of lung or kidney tissues. Alternatively, in non-polarized cells, as in the case for fibroblasts, an intracellular ciliary vesicle grows around the mother centriole together with the axoneme. This vesicle will later fuse with the plasma membrane, which will result in the protrusion of PC to the extracellular space. The BB is formed of a 9 triplet microtubules (Figure 3). The TEM images show that PC can remain invaginated to the plasma membrane which resembles to the flagellar pocket. The ciliary pocket is continuous with the plasma membrane, but functionally different. This small but significant domain of the PC is established after fusion of the plasma membrane with the intracellular ciliary vesicle that grows on top of the BB (Ghossoub, Molla- Herman et al. 2011). The origin and function of the ciliary pocket is not clearly known but it is currently a topic of intense research (Benmerah 2012). The ciliary pocket is characterized by the presence of clathrin coated vesicles resulting from the vesicular trafficking that originates and ends in this region. Acting as a specific endocytic membrane, the ciliary pocket has been described to regulate the transforming growth factor-β (TGF-β) pathway (Molla- Herman, Ghossoub et al. 2010; Clement, Ajbro et al. 2013). The presence or absence of the ciliary pocket depends on the cell type and can be explained by different kinetics during ciliogenesis; for example, ciliary pocket is common in fibroblasts but has rarely been observed in epithelial cells. This rare observation of the pocket in epithelial cells can be explained by a short-lived intracellular pathway of the ciliary pocket docking to the PM (Ghossoub, Molla-Herman et al. 2011). PC is not isolated from the cytoplasm but is highly compartmentalized; it is surrounded by a specialized region of the plasma membrane that holds a different concentration of channels and receptors and the axoneme is separated from the rest of the intracellular compartments by the “transition zone”. The transition zone consists of the “ciliary necklace” and transition fibers and is evolutionarily conserved. It is characterized by the ciliary necklace formed by Y-shaped fibers that connect the microtubule doublets to the ciliary membrane at the base of the cilium distal to the BB. Moreover, this zone is mostly characterized by TEM data (Reiter, Blacque et al. 2012). This ciliary subdomain is highly organized with a molecular structure resembling that of the nuclear pore and serves as a border control where different, precursors and intraflagellar transport (IFT) proteins, along with proteins of different signaling pathways are concentrated and regulated for entry to the axoneme (Kee, Dishinger et al. 2012). The composition of the transition zone and subdomains is not fully known, however several studies suggest that distal appendage proteins play an important role (Avidor-Reiss and Gopalakrishnan 2013). During ciliogenesis, the mother centriole anchors to the associated membrane; the ciliary vesicle membrane or the

plasma membrane. The mother centriole can be distinguished from the daughter centriole by the presence of subdistal and distal appendages. Distal appendages are necessart for ciliogenesis. OFD1 (oral-facial-digital syndrome1) protein is associated with the distal ends of centrioler microtubules and is necessary for centriole docking via the distal appendages which forms the transition fibers at the base of the cilium (Reiter, Blacque et al. 2012; Singla, Romaguera-Ros et al. 2010).

Figure 3: Primary Ciliogenesis models

(A)The “extracellular pathway” which takes place primarily in polarized cells where the mother centriole of the centrosome (green) directly docks to the plasma membrane forming the basal body (BB) and the primary cilium (PC, membrane in red) grows directly in the extracellular milieu. (B) The “intracellular pathway” which takes place primarily in fibroblasts during which a primary ciliary vesicle interacts with the distal appendages of the mother centriole, the axoneme grows then within this vesicle, which elongates thanks to the fusion with incoming secondary vesicles then fusing with the plasma membrane and forming the ciliary pocket (CP).(CP, blue). (B’) Ciliogenesis model proposed for polarized cells with a faster fusion event with the plasma membrane and/or complete extrusion of the cilium. (C) CP acts as a specific endocytic membrane domain (CCP, CCV clathrin-coated pits and vesicles respectively, CM is ciliary membrane). (D) PC in RPE1, human retinal pigment epithelial cells. (E-H) IF images of RPE1 PC. (Ghossoub, Molla-Herman et al. 2011)

1.1.4. Intraflagellar Transport Mechanism

A bi-directional movement in the Chlamydomonas flagella was fist observed by Rosenbaum lab and identified as the intraflagellar transport (IFT) (Kozminski, Johnson et al. 1993). IFT complexes are conserved in ciliated organisms and have different functions. The IFT mechanism plays an essential role in the assembly and maintenance of the cilia or flagella and during the transport and trafficking of ciliary components such as cargo along the axoneme (Rosenbaum and Witman 2002; Sung and Leroux 2013). The bi-directional trafficking along the polarized microtubules of the axoneme is maintained through the action of two motor protein families: kinesins and dyneins (Figure 4) (Pedersen and Rosenbaum 2008). Anterograde IFT (aIFT) trafficking with base-to-tip cargo direction, depends on the heterotrimeric kinesin-2 motor, also known as KIF3 motor complex. This complex consists of two kinesin-2 family proteins KIF3A, KIF3B and KIF-associated protein 3 (KAP3). Retrograde IFT (rIFT) with tip-to-base cargo direction depends on the dynein motor that consists of cytoplasmic dynein 2 heavy chain 1 (DYNC2H1) and cytoplasmic dynein 2 light intermediate chain 1 (DYNC2LI1)) (Pedersen, Veland et al. 2008). The kinesin and dynein motors are conserved in all ciliated organisms with the exception of Plasmodium falciparum (Sung and Leroux 2013). Two large complexes composed of more than 20 proteins called IFT A and IFT B are essential for the building and trafficking along the cilium, the first for retrograde and the latter for anterograde trafficking. Mutations in the two IFT subunits have different consequences; IFT B protein defects lead to absent or shortened cilia whereas defects in IFT A proteins lead to a bulged cilia (Goetz and Anderson 2010). The docking and assembly mechanism of IFT particles are not well characterized. Several IFT proteins are identified over the years in several organisms and a number of these proteins possess specialized roles in ciliogenesis, antero- or retrograde trafficking and transport of ciliary components to the cilia or flagella. Some subunits of IFT complexes, such as IFT20 and DYF- 11/MIP-T3, which are part of the IFT B complex are found in other cellular compartments from where they facilitate transport of vesicles with specific cargo to the BB and axoneme (Pedersen, Veland et al. 2008). Continuous shuttling of IFT20 protein between the Golgi and the PC is required for ciliogenesis and cilia dependent signalizations (Follit, Tuft et al. 2006; Keady, Le et al. 2011). OFD1 protein plays an important role in association with IFT assembly and is required for the recruitment of IFT88 protein (Singla, Romaguera-Ros et al. 2010). Maintenance of the cilia also depends on the BBsome, a complex formed by proteins encoded by genes mutated in Bardet–Biedl syndrome (BBS). Ciliary targeting sequence

(CTS) is recognized by the BBsome and serves target the proteins with CTS to the cilia in a BBSome dependent manner. Several proteins involved in the vesicular trafficking such as the small GTPase Rab8 or Rab11 are evolutionarily conserved in the vesicular trafficking and also in cilia assembly via IFT and BBsome (Sung and Leroux 2013). The mechanisms of the transport and trafficking of proteins involved in various cilia-dependent signaling pathways are not known in details (Goetz and Anderson 2010). However various proteins are observed to interact with IFT. For example the Chlamydomonas orthologue of the calcium channel PKD2 is IFT-dependent (Huang, Diener et al. 2007).

Figure 4: Intraflagellar Transport

Cargo is transported from the base to the tip of the cilium along the microtubule axoneme by the kinesin-2 motor (which consists of KIF3A, KIF3B and KAP3) together with the intraflagellar transport A (IFT A) complexe regulating the anterograde IFT. The retrograde IFT of the cargo, from the tip to the base of the cilium, is mediated by the dynein motor (which consists of cytoplasmic dynein 2 heavy chain 1 (DYNC2H1) and cytoplasmic dynein 2 light intermediate chain 1 (DYNC2LI1)) and by IFT B complexe. Certain basal body proteins also influence ciliary trafficking. Among these are components of the BBSome. (Goetz and Anderson 2010)

1.2. Ciliary Signaling

The major function of PC is sensing the extracellular medium which is mediated by different signaling pathway receptors, ion channels, transporter proteins and their downstream molecules located to the axoneme or to the basal body. PC responds to different sensory modalities such as mechanical stimulus (bending of the cilia following a shear stress) or chemosensation (specific ligand, growth factor, hormone or morphogen recognition). A variety of signaling pathways are coordinated through this organelle during development, tissue homeostasis, cell migration, cellular differentiation, cell cycle and apoptosis. Examples of such pathways are Hedgehog (Hh) pathway, PDGF pathway; Wnt pathway and Ca2+ signaling cascade (Eggenschwiler and Anderson 2007; Satir, Pedersen et al. 2010).

1.2.1. Hedgehog Signaling

Hedgehog (Hh) signaling is one of the pathways that orchestrate development in animals, along with other signaling pathways such as Wnt, Notch, TGF- β and fibroblast growth factor (FGF). The name of the pathway originates from the short and “spiked” phenotype of the cuticle of the Hh mutant Drosophila larvae. Hh pathway is researched in detail in Drosophila and the mammalian pathway remains to be fully discovered in adult cells (Varjosalo and Taipale 2008). The Hh pathway is involved in regulation of cell proliferation, cell fate determination, in the epithelial-to-mesenchymal transition and in maintenance of homeostasis in adult cells (Simpson, Kerr et al. 2009; Briscoe and Therond 2013). The cellular trafficking of these pathways includes coordinated uptake, secretion and localizations of necessary molecules throughout the cell. A cell’s fate is determined according to the concentration gradient of the Hh proteins, which are diffusible lipid-modified morphogens. Primary cilium is a key coordinator of the mammalian Hh signaling pathway as its major signaling components are localized to the cilium, and IFT proteins are essential for trafficking of Hh molecules (Goetz and Anderson 2010). A recent study showed that in the ciliated olfactory sensory neurons of Drosophila, the Hh pathway is dependent on the presence of PC (Kuzhandaivel, Schultz, et al. 2014). This observation suggests that there may be two pathways for Hh mediation within the same organism; a PC-dependent and PC-independent pathway in non-ciliated cells.

There are three secreted Hh ligand member family proteins: Sonic hedgehog (Shh), Indian hedgehog, and desert hedgehog (Varjosalo and Taipale 2008; Ingham, Nakano et al. 2011). Shh is the most widespread ligand, whose auto-catalytically processed lipidated form is secreted into the extracellular matrix. It is expressed in the midline tissue and in the zone of polarizing activity of the limb bud during vertebrate embryogenesis. During organogenesis it is expressed during the development of most epithelial tissues (Varjosalo and Taipale 2008). Hh signaling depends on the balance between activator and repressor forms of the glioma- associated oncogene family (Gli) of zinc-finger transcription factors. The basic signaling pathway involves a 12-transmembrane domain receptor protein Patched 1 (Ptc1 or Ptch1), the 7-transmembrane domain protein Smoothened (Smo), the suppressor of fused protein (Sufu), and the transcription factors Gli members (Figure 5). Hh signaling is suppressed in the absence of Hh ligand through the physiological inhibitory effect of Ptc1 on Smo. When the pathway is activated by Hh ligand binding to Ptc1, Smo is released from Ptc1 and re-localizes to the PC axoneme where it inhibits Sufu activity. Sufu controls the processing and translocation of Gli family transcription factors from the cilium to the nucleus (Goetz and Anderson 2010).

The Gli protein family includes Gli1, Gli2, and Gli3. They are highly related but their biochemical properties differ from one another along with their functions in vivo. This family of zinc-finger transcription factors balances between activator and repressor forms. Activated Gli forms directly induce transcription of Hh target genes including genes encoding Ptc1 and Gli family members. The activity of Gli1 depends on Gli2, and Gli1 and Gli2 act primarily as transcriptional activators. Gli3 exists in two-forms: a non-processed, full-length form that can function as transcriptional activator and a truncated amino-terminal fragment that acts as transcriptional repressor (Robbins, Fei et al. 2012; Aberger and Ruiz 2014). Also, it has been recently described that the activity of Kif7 at the cilium creates a specialized compartment at the top of the cilia where the activity of the Gli proteins is regulated (He, Subramanian et al. 2014). The bifunctional Gli proteins have roles in embryogenesis and adult homeostasis as their target genes regulate cell proliferation (cyclin and Myc genes), cell death (Bcl-2), differentiation (genes encoding Forkhead family transcription factors), and stem cell renewal (Bmi-1). Target genes regulated by the Hh pathway depend on tissue and cell types (Jiang and Hui 2008; Briscoe and Therond 2013). Hh is a key morphogen for body patterning during development of various organs such as pancreas, kidney, lung, foregut, nervous system and limb. Target genes regulated by the Hh pathway may vary among tissues and cell types.

Misregulation of such diversified signaling pathway leads to various unfavorable developmental and pathological conditions including birth defects and tumorigenesis (Keady, Le et al. 2011).

Figure 5: Hedgehog signaling

(A) In the absence of ligand, Ptc1 (Ptch) localizes to the PC and prevents the entry of Smo into the axoneme. Low levels of Sufu-Gli protein complexes are localized to the cilium which promotes the repressor forms of Gli (GliR), inhibiting the transcription of target genes. (B) After activation of the pathway, Smo moves to the ciliary membrane promoting Sufu-Gli complexes translocation, dissociation and Gli (Gli2) accumulation at tip of the cilium. Activated Gli proteins (GliA) are transported out of the cilium by the dynein motor and intraflagellar transport (IFT) particles for the transcriptional activation of the target genes. (Hui and Angers 2011)

1.2.2. PDGF Pathway

Platelet-derived growth factor receptor alpha (PDGFRα), a widely expressed protein, localizes specifically to the PC in quiescent fibroblasts and induces activation of Akt, Mek1/2, Erk1/2, and Rsk pathways at the axoneme and in the basal body of the cilium (Satir and Christensen 2007). The PDGF pathway also induces cilia-dependent activation of MAPK and regulates proliferation, cell survival, and migration during embryogenesis and tissue homeostasis. Primary cilium-dependent PDGFRα activation regulates the ubiquitous plasma membrane Na+/H+ exchanger NHE1, which controls proliferation, cell survival, and migration (Schneider, Stock et al. 2009; Clement, Mally et al. 2013). Defects in this pathway can lead to different human pathologies including gastrointestinal stromal tumors, lung tumors, and ovarian carcinoma (Schneider, Clement et al. 2005).

1.2.3. Wnt Pathway

Canonical and non-canonical Wnt pathways are crucial during development (Kestler and Kuhl 2008; Angers and Moon 2009). The canonical Wnt pathway is regulated by β- catenin and controls proliferation, apoptosis, and cell fate determination via activation of target genes. Activation of Wnt target genes depends on ligand binding to the Frizzled receptor at the plasma membrane and subsequent β-catenin translocation into the nucleus. The stability of β-catenin depends on its degradation by the adenomatosis polyposis complex complex (APC) at the basal body. Recent studies have established the association of Wnt pathways with ciliary components such as kinesin motors. Transport of APC to its various cellular locations via kinesins plays a vital role in microtubule stabilization, activation of protein kinases, and cell polarization. These findings shed light on the contribution of the interaction of Wnt pathway with the PC in the regulation of cytoskeletal architecture (May- Simera and Kelley 2012).

The non-canonical Wnt pathway is controlled by a membrane protein called Van Gogh-like 2 (Vangl2) and regulates cytoskeletal changes, cell adhesion, migration, and polarity (May-Simera and Kelley 2012). Vangl2 functions in asymmetric positioning of motile cilia and in centrosome positioning, which is important for cytoskeletal rearrangement. Mutations of different ciliary proteins, such as Kif3a, IFT88, OFD1 (the protein mutated in oral-facial-digital syndrome type 1) result in dysregulation of planar cell polarity (PCP) and

are associated with hyper- or de-activation of the Wnt pathway depending of the tissue, timing, and localization of these proteins (Corbit, Shyer et al. 2008). PCP is critical for the development of various organs. Apart from the apical/ basal polarity that defines cellular orientation, epithelial cells can also be polarized within the plane of the tissue, orthogonal to the apical/basal axis which characterizes PCP (Carroll and Das 2011). Vertebrate tissues such as postnatal kidney tubule epithelia show PCP components such as oriented cell division which is characterized by the cell division parallel to the axis of kidney tubule elongation and in the plane of the epithelium. (Fedeles and Gallagher 2012).Components of the PCP pathway, such as Vangl2 and inversin (also called nephrocystin 2), localize at the PC, specifically at the basal body and the inversin compartment which is localized at the proximal segment of the cilium beyond the transition zone, during development (Shiba, Yamaoka et al. 2009). Inversin seems to serve as a molecular switch between canonical and non-canonical Wnt pathways by regulating Dishevelled protein degradation at the base of the cilium by the APC (May-Simera and Kelley 2012).

1.2.4. Calcium Signaling

The Ca2+-specific cation channel PC2, which forms an ion channel complex with the G-protein coupled receptor PC1 localized in the axoneme, regulates calcium signaling (Patel and Honore 2010; Patel 2014). PC1 and PC2 are necessary for the calcium response and nitric-oxide release in endothelial cells in response to blood flow (Nauli, Kawanabe et al. 2008). Extracellular flow-induced calcium influx leads to Ca2+ release from the endoplasmic reticulum via ryanodine and inositol triphosphate receptors, resulting in AMP release (Winyard and Jenkins 2011). Kidney epithelial cells respond to urinary flow in the lumen of renal tubules by bending of the PC. PC acts as a mechanosensory antenna in the tubule lining epithelial cells upon fluid shear stress generated by the fluid flow. This stimuli increases the membrane permeability to Ca2+. It has also been shown that different renal cells have additional cilia-independent mechanosensitive Ca2+ responses (Liu, Xu et al. 2003; Rodat- Despoix, Hao et al. 2013). The flow dependent calcium signaling is addressed further in the manuscript. Recently, Delling et al. elegantly showed that the PC is a specialized organelle with respect to intracellular ions by using a ratiometric Ca2+ sensor for individual cilia and patch-clamp methods (Delling, DeCaen et al. 2013). The study demonstrated that ciliary calcium channels, specifically the heteromeric PKD1L1-PKD2L1 channel, maintain a high Ca2+ concentration in the cilium with respect to cytoplasm in resting state, thus regulating a steady diffusion into the cell. The rapid calcium dilution to cytoplasm does not activate a Ca2+ 24

release but positively regulates ciliary Hh trafficking via IFT25, which contains a unique Ca2+ binding site (Delling, DeCaen et al. 2013).

1.2.5. Fluid flow and Fluid Shear Stress

The mechanosensory role of PC is evolutionarily important for various organs. PC can sense different fluid movements such as urine in the kidney tubules, blood in the vascular system, bile in the hepatic biliary system, lacunocanalicular fluid in the bone and cartilage, digestive fluid in the pancreatic duct, nodal flow in the Hensin’s node, and cerebral spinal movement in the nervous system. The transduction of mechanical stimuli by the PC depends on different receptors and channels such as calcium channels polycystin 1 (PC1) and polycystin 2 (PC2) on the axoneme depending on the tissue and thus can induce various intracellular signalizations explained above. See Table 1 for an overview of cilia-dependent mechanical stress in different tissues. There are other mechanosensory cellular compartments in different tissues including microvilli, actin filaments in the kidney, glycocalyx, a membrane-bound filament structure found or caveolae, an invagination of the plasma membrane rich in lipids, receptors and channels including K+ channels, Cl- channels, and Ca2+ channels on the surface of endothelial cells. Tyrosine kinase receptor (TKR), G-protein- coupled receptors (GPCRs) and G-protein itself have been shown to be activated by fluid shear stress. Adhesion proteins (PECAM-1, and VE-cadherin) at sites of cell–cell and cell– matrix attachment are subjected to tension under shear stress. Cytoskeleton is responsible for cell shape regulation and can respond to mechanical forces that deform cells (Ando and Yamamoto 2013) (Figure 6).

Cellular Pathways regulated by mechanical stress via Primary Cilia Mechanosensitive cells Pathways regulated upon mechanical stress References Praetorius and Spring 2003 and Calcium signaling Nauli et al. 2003 Kidney tubular cells Cytoskeletal re-arrangement Essig, Terzi et al. 2001 Polarity Kotsis, Nitschke et al. 2008 mTOR signaling Boehlke, Kotsis et al. 2010 Calcium signaling Nauli, Kawanabe et al. 2008 Nitric Oxide production Cytoskeletal re-arrangement Endothelial cells Hierck, Van der Heiden et al. 2008 Inflammatory response via KFL-2 Endothelial-to-mesenchymal transition Egorova, Khedoe et al. 2011 Goetz et al. 2014; Kallakuri et al. Vascular development and integrity 2014 Bone metabolism via PGE2 Osteoblasts and Osteocytes Malone, Anderson et al. 2007 Bone resorption ECM secretion McGlashan, Haycraft et al. 2007 Chondrocytes ATP-induced Calcium signaling Wann, Zuo et al. 2012 Mesencymal stem cells Osteogenic differentiation Hoey, Tormey et al. 2012 Calcium signaling Gradilone, Masyuk et al. 2007 Cholangiocyte epithelial cells cAMP response Masyuk, Masyuk et al. 2006 Nodal flow via motile cilia Nodal cells McGrath, Somlo et al. 2003 Calcium signaling

Table 1: Pathways regulated by mechanical stress via Primary Cilia

Figure 6 : Mechano-sensing by the cell

Candidates for shear stress sensors on the plasma membranes are presented. Various types of ion channels including K+ channels, Cl- channels, and Ca2+ channels have been shown to be shear stress responsive.TKR, GPCRs, and G-protein itself have been shown to be activated by shear stress. Caveola are membrane microdomains containing a variety of receptors, ion channels, and signalling molecules are involved in shear-stress sensing. Adhesion proteins (PECAM-1, and VE-cadherin) at sites of cell–cell and cell–matrix attachment are subjected to tension under shear stress are such adhesion proteins. Cytoskeleton is responsible for cell shape regulation and can respond to mechanical forces that deform cells. Glycocalyx unfolds under flow conditions, and this conformational change affects the local concentrations and transport of ions, amino acids, and growth factors, or triggers signal transduction. Primary cilia is known to detect flow. Well known flow-responsive calcium channels on the cilium are PC1 and PC2. Shear stress induces an influx of extracellular Ca2+ via PC2. (Ando and Yamamoto 2013)

Endothelial cells cover the inner surface of blood vessels. They are under fluid shear stress (FSS) which is the drag force induced by the viscous blood acting on the vessel wall in parallel to the blood flow (Hierck, Van der Heiden et al. 2008). Endothelial cells are also subjected to cyclic stretch forces perpendicular to the shear forces caused by hydrostatic and blood pressures (Resnick, Yahav et al. 2003). In the vascular system fluid shear stress can vary from 10-40 dyn/cm2 in large arteries such as the aorta with a laminar pulsatile flow and to 1-6 dyn/cm2 on the walls of the veins. FSS is responsible for the homeostasis of the circulatory system and various phenotypes including angiogenesis, blood coagulation or vascular remodeling (Ando and Yamamoto 2013). The impairment of flow sensing can lead to various pathologies including hypertension and atherosclerosis (Davies, Mundel et al. 1995). The ability of endothelial cells to sense the mechanical stimuli depend on various proteins such as ion channels, G-protein receptors, adhesion molecules expressed on the plasma membrane or on supporting structures such as cytoskeleton and lipid bilayer 27

membrane or on microdomains such as caveolae and primary cilium (Davies, Mundel et al. 1995). The presence of PC in endothelial cells depends on the blood flow pattern and exposed shear stress levels in the circulatory system. For example, most ciliated endothelial cells are present in inner curvature of aorta or various arteries where the flow can be disturbed and observed to be oscillatory whereas the endothelial cells on the areas exposed to high shear stress such as on the outer curvature of the aorta are deciliated (Egorova, Van der Heiden et al. (2012). PC is not the sole mechano-sensor in endothelial cells however bending of the cilium is vital for the influx of extracellular calcium and cytoskeletal re-arrangement by appearance of stress actin fibers with re-orientation to the shear axis and also for the regulation of zinc-finger transcription factor Kruppel-like factor-2 (KLF-2) (Hierck, Van der Heiden et al. 2008). The cytosolic Ca2+ increase is followed by nitric oxide production which is important for vascular contractility. Endothelial cells mutated for PKD1 or Tg737 (IFT88) failed to induce this response (Nauli, Kawanabe et al. 2008). The regulation of KLF-2 expression by PC is important in atherosclerosis as during low oscillatory, “athero-prone”, flow in ciliated areas of the circulatory system, KLF-2 expression is suppressed and an inflammatory response is induced which may result in plaque formation (Hierck, Van der Heiden et al. 2008). In areas exposed to high shear stress, KLF-2 expression is induced and cells are observed to lose their cilia which can also correspond to endocardial cushions in the embryo which undergo TGF-β mediated endothelial-to-mesenchymal transition (EndoMT) (Egorova, Khedoe et al. 2011). Two recent studies showed another role of PC in flow sensing during vascular development and integrity in zebrafish. These studies showed that PC was involved during early embryonic stages, during which a low blood flow is established. PC is observed to mediate vascular morphogenesis via PC2-mediated calcium signaling upon flow induced bending (Goetz, Steed et al. 2014). It is also shown that the presence of PC is essential for vascular integrity via the Hh pathway since the cilia impairment increased the risk of intracranial hemorrhage (Kallakuri, Yu et al. 2014).

Lacuno-canalicular fluid in the bone and cartilage creates a dynamic flow that is sensed by the primary cilium of osteocytes in the bone matrix and osteoblasts. The nature of the mechanical stress in the bone, as well as in the vascular system, is still under investigation (Fritton and Weinbaum 2009). PC is required for osteogenesis and bone resorption in response to mechanical stress (Weinbaum, Duan et al. 2011). Osteoblasts up-regulate osteopontin expression, a bone matrix protein along with cytokine prostaglandin E2 (PGE2) regulating bone metabolism in response to flow in a cilia dependent manner. The bone

resorption regulation of osteocytes in the bone is a calcium independent homeostatic response regulated by the PC (Malone, Anderson et al. 2007). As for the cartilage cells, it was previously shown a role of PC in the chondrocyte differentiation and hypertrophy (McGlashan, Haycraft et al. 2007). A mechano-sensory role of chondrocytes is also in question during the mechanical physiological joint loading. Compressive loading in chondrocytes regulates extracellular matrix secretion and the ATP induced Ca2+ signaling via purine receptors in a cilia dependent manner (Wann, Zuo et al. 2012). Mechanical loading induced by fluid shear stress stimulates mesenchymal stem cell differentiation into bone forming osteoblasts in a PC dependent manner via osteogenic gene expression regulation (Hoey, Tormey et al. 2012).

The role of PC in the bile duct is important as the liver cysts observed in ADPKD patients originate from the epithelial cells lining the bile duct, cholangiocytes. It is reported that luminal bile flow induces Ca2+ signaling response and cAMP decrease in ciliary dependent manner via PC1 and PC2 in vitro (Masyuk, Masyuk et al. 2006). The osmotic stress caused by luminal flow regulates absorption of solutes and water secretion by cholangiocytes via TRPV4 channel on the PC leading to the secretion of bicarbonate thus ductal bile formation (Gradilone, Masyuk et al. 2007).

It is known that primary cilia, known as “nodal” cilia is vital for left-right body axis determination during embryonic development. McGrath et al. showed that ciliary PC2 is responsible for the mechansosensation of the nodal flow created by the motile cilia across the node and creating an intracellular Ca2+ signaling (McGrath, Somlo et al. 2003).

The role of primary cilia and its mechanosensory role in the kidney will be detailed later in the manuscript (see 2.6.1.Physiology of the kidney: an introduction).

In the case of motile ciliated tissues, including ependymal cilia, it is shown that the organelle can have a role in flow direction rather than flow sensing. It is reported that motile cilia of the ependymal cells are important for directional regulation of cerebrospinal fluid flow and neuronal migration (Sawamoto, Wichterle et al. 2006).

2. Autophagy

2.1. Role of Autophagy

The word “autophagy” is derived from the Greek words “auto” and “phagy” meaning “self-eating”. The term “autophagy” is first introduced in 1963 by the biochemist Christian de Duve who received a Nobel Prize in Physiology or Medicine in 1974 for his discovery and work on lysosome (Klionsky 2007). Autophagy is an evolutionarily conserved mechanism for eukaryotes and refers to catabolic processes and turnover of cellular contents such as intracellular misfolded or long-lived proteins, damaged organelles and invading microorganisms. It is characterized by the formation of double-membrane vesicles, “autophagosomes”, which sequester cytoplasmic material. However, autophagy is more than just a degradative process, it is a recycling system that responds to various intra- and extracellular stresses by generating metabolites as sources of energy or building blocks for synthesis of new macro-molecules such as amino acids, free fatty acids and sugars (Yang and Klionsky 2010; Boya, Reggiori et al. 2013). The leading role of autophagy is maintaining the energy homeostasis. Stimuli that activate autophagy include for example, nutrient deprivation and other nutritional challenges as well as growth factor withdrawal, oxidative stress, protein accumulation, infection. Autophagy plays an essential role in multitude of physiological processes ranging from adaptation to starvation, cell differentiation and development, as well as tumor suppression, innate and adaptive immunity, lifespan extension, and cell death. Autophagy mechanism is deregulated in various human pathologies, for example cancer, neurodegeneration, obesity, type-2 diabetes, heart and liver diseases, infectious and immune diseases and even physiologic stages such as aging (Rubinsztein, Codogno et al. 2012; Choi, Ryter et al. 2013).

2.2. Different Types of Autophagy

There are three major types of autophagy described in mammals: macroautophagy, chaperone-mediated autophagy (CMA) microautophagy (Figure 7) (Boya, Reggiori et al. 2013).

2.2.1. Macroautophagy

Macroautophagy, usually referred to as autophagy, is the process of bulk degradation of cytoplasmic material. It is characterized by de novo formation of the isolation membrane, called phagophore, which sequesters cytoplasmic content or organelles and sealing of the double membrane vesicle “autophagosome” that will later fuse with lysosomes. This process is regulated by a group of proteins described as autophagy-related (Atg) proteins that are conserved in eukaryotic cells (Rubinsztein, Shpilka et al. 2012). The discovery of Atg genes in yeast was revolutionary for the field and lead to our understanding of the molecular mechanism of autophagy (Tsukada and Ohsumi 1993; Thumm, Egner et al. 1994; Harding, Morano et al. 1995). The formation of autophagosomes as well as the membrane sources of this double-membrane structures are still not 100% characterized. Autophagy is a bulk process, but, in many cases, it can be highly selective of cellular structures, including peroxisomes (pexophagy), mitochondria (mitophagy), endoplasmic reticulum (reticulophagy), ribosomes (ribophagy), midbody, micro-nuclei and part of the nuclei (nucleophagy), lipid droplets (lipophagy), aggregate-prone proteins (aggrephagy), secretory granules (zymophagy), damaged lysosomes (lysophagy), and microorganisms (xenophagy) (Rogov, Dotsch et al. 2014).The molecular aspect of macroautophagy will be focused further in detail in the manuscript.

2.2.2. Chaperone-mediated Autophagy

Chaperone-mediated autophagy (CMA) is a selective form of autophagy where the substrates bind to a cytosolic chaperone, a heat shock-cognate protein of 70 kDa (Hsc70), and then to a lysosomal protein, the lysosomal associated membrane protein type 2A (LAMP‐2A). The recognition of the CMA substrates by Hsc70 requires a motif related to the pentapeptide KFERQ (Dice 1990) after which they are directly targeted to lysosomal surface and interact with the cytoplasmic tail of LAMP-2A. This is followed by the unfolding of the substrate

protein in order to translocate across the lysosomal membrane and to be degraded (Cuervo and Wong 2014). The substrates of CMA are soluble proteins but not organelles, because they are not engulfed by the lysosome but rather transported across the lysosomal membrane. This translocation of substrates requires a multi-protein complex formed by the association of LAMP-2A monomers at the lysosomal membrane. The stability of the LAMP-2A is maintained via Hsp90 at the luminal membrane. Following substrate translocation, LAMP-2A complex disassembles for others to bind. This process can be regulated in a GTP-dependent manner through GTP-bound EIF1α- GFAP (Glial fibrillary acidic protein) interaction during the disassembly of the multimeric complex (Bandyopadhyay, Sridhar et al. 2010).

CMA contributes to quality control of proteins and exists at basal levels however it is maximally induced by stress conditions similar to macroautophagy, such as starvation, oxidative stress or exposure to various toxic compounds (Massey, Zhang et al. 2006; Cuervo and Wong 2014). However in contrast to rapid activation of macroautophagy upon starvation, CMA is gradually activated upon 10 hours of starvation and persists up to 3 days which contributes to the selective degradation of non-essential proteins to generate building blocks of essential proteins’ synthesis (Massey, Kaushik et al. 2006). Both the perturbation and increase of CMA activity can have roles in various pathologies. The reduction of CMA contributes primarily to neurodegenerative diseases such as Parkinson’s and Alzheimer’s diseases; whereas in certain cancer cases increased CMA is observed, contributing to rapid cell growth by sustaining enhanced glycolysis. Another important role of CMA is defined during aging as its activity is significantly decreased over time along with total protein degradation rate reduction in all organisms (Cuervo and Wong 2014). Recent data showed a previously unknown role for CMA via a liver specific conditional LAMP-2A knock-out mouse model. CMA impaired mice showed altered hepatic carbohydrate and lipid metabolism in aging mice suggesting that CMA loss has a negative role on the energetic balance in aging organisms (Schneider, Suh et al. 2014).

Figure 7 : The different types of autophagy

(A)Macroautophagy is characterized by the sequestration of cytosolic material in double- membrane vesicles called autophagosomes. (B) During chaperone-mediated autophagy, proteins with KFERQ motifs are recognized by the Hsc70 chaperone. Later they associates with the integral lysosome membrane protein LAMP-2A. (C) Microautophagy entails the recruitment of targeted components in proximity with the lysosomal membrane, which subsequently invaginates and pinches off. (Boya, Reggiori et al. 2013).

2.2.3. Microautophagy

In the case of microautophagy, the cytoplasmic content is directly engulfed by the lysosome by the invagination of the lysosomal membrane itself. This invagination of lysosomes regulate the lysosomal size by consumption of the membrane formed during macroautophagy (Li, Li et al. 2012). The accompanying membrane structures can function in closure of the cargo in an Atg protein dependent manner. The vesicles containing the autophagic substrates pinch off into the lysosomal lumen to be degraded by lysosomal hydrolases (Cuervo and Wong 2014). Both macro- and micro-autophagy can mediate the degradation of various organelles including peroxisomes, mitochondrion, lipid droplets and mucleus in yeast, however the elimination of various organelles by microautophagy in mammals remain to be investigated (Okamoto 2014). Microautophagy can engulf multivesicular bodies, therefore, it can participate to renewal of cell membrane proteins in endosomes. It has also been shown to participate in glycogen delivery into the lysosome serving as a delivery route (Li, Li et al. 2012).

2.3. Autophagosome Machinery and Biogenesis

Macroautophagy, referred as autophagy from now on, depends on a tight and complex molecular regulation that involves more than 30 Atg proteins that were firstly identified in yeast. Autophagy related proteins involved in the formation and maturation of autophagosomes are almost entirely conserved in all eukaryotes (Mizushima, Yoshimori et al. 2011). Autophagy is characterized by a dynamic membrane re-organization which starts with the formation of a cup-shaped double-membraned sac called the isolation membrane (also referred to as phagophore) in the cytoplasm which expands in size and later seals as the double-membraned autophagosome. Autophagosome size can differ from 0.3-0.9 µm in yeast (Baba, Osumi et al. 1997) and 0.5-1.5 µm in mammals (Mizushima, Ohsumi et al. 2002). Autophagosomes later fuse with lysosomes (with vacuoles in yeast), forming the autolysosomes in which the intracellular content and the inner membrane of the autophagosome are degraded by lysosomal hydrolases. Formation of autophagosome and autolysosomes are highly dynamic. The two processes together can take 7-9 minutes in yeast (Geng, Baba et al. 2008) and autophagosome formation can take 5-10 minutes in mammals (Fujita, Hayashi-Nishino et al. 2008).

There are three important stages in autophagic machinery: 1. Initiation and nucleation (formation of the phagophore); 2.Vesicle elongation (growth and closure of autophagosome); 3.Maturation into autolysosomes. These stages depend on different core Atg protein units: Atg1/ULK complex, the class III phosphatidylinositol 3-kinase (PI3K) (Beclin-1- PtdIns3KC3-Atg14L) complex, the Atg2-Atg18/ WIPI complex, the Atg12-Atg5-Atg16L1 conjugation system, the Atg8/LC3 conjugation system, and Atg9 vesicles. For the relevance of the manuscript, the main focus will be on mammalian autophagic machinery (Mizushima, Yoshimori et al. 2011) (Figure 8).

2.3.1. Initiation

ULK complex

ULK1, mammalian orthologue of yeast Atg1, is considered to have the major role in autophagy (Chan, Kir et al. 2007). ULK complex is the most upstream module of autophagic machinery. The multimeric complex includes unc-51-like kinase 1 (ULK1), focal adhesion kinase family interacting protein of 200kD (FIP200), Atg13 and Atg101 (Itakura and Mizushima 2010). The ULK1-mAtg13-FIP200-Atg101 complex is constitutively formed in the cytosol independent of nutrient conditions. Under nutrient rich conditions ULK1 and Atg13 is phosphorylated and the complex is inactivated by mammalian target of rapamycin complex 1 (mTORC1). Upon autophagy induction, this stable complex translocate to the autophagosome assembly site at the ER membrane, where ULK1 is no longer phosphorylated by mTORC1 (Itakura and Mizushima 2010) (Figure 9). ULK1, a Ser/Thr kinase is known to be required for autophagy induction and recruitment of the ULK complex on the phagophore membrane in mammals (Hara, Takamura et al. 2008). ULK1 is known to phosphorylate Atg13 and FIP200 as well as other proteins such as AMBRA1 or a non autophagic zipper- interacting protein kinase ZIPK protein however the relevance of the substrates of this kinase remains to be further elucidated (Mizushima, Yoshimori et al. 2011). Atg101 interacts with and stabilizes Atg13. ULK1 and ULK2 are mammalian orthologues closely related to yeast Atg1 protein and are shown to be functionally redundant in vivo and to form complexes with Atg13 and FIP200 in a compensatory manner in response to nutrient deprivation (McAlpine, Williamson et al. 2013).

Figure 8: The Autophagic pathway

There are two important stages in autophagy pathway: (i) autophagosome formation, which includes the initiation/nucleation (formation of the phagophore) and elongation/closure, and (ii) autophagosome fusion with the lysosome. Autophagy-related (ATG) proteins involved in the formation of autophagosome and a non-exhaustive list of the protein families involved in the maturation/fusion step are shown in italics.

The class III PI3K complex

There are at least three class III phosphatidylinositol kinase (PI3K) complexes in mammalian cells which have different functions in the cell. The major class III PI3K complex that is required for the autophagosome formation include a PI3 kinase the vacuolar protein sorting 34 (Vps34), Vps15, Beclin1, Atg14 (L) (also called Barkor) and activating molecule in Beclin1-regulated autophagy (AMBRA1) (Itakura and Mizushima 2010).

This Beclin1-Vps34-Vps15-Atg14L complex, downstream of the ULK complex, produces phosphatidylinositol 3-phosphate (PI3P) by phosphorylating phosphatidylinositol (PI) at 3-position hydroxyl group of inositol ring. PI3P production is induced on the autophagosome assembly site on the ER, which generates a platform for other PI3P-binding autophagic proteins for the expansion of the isolation membrane (Shibutani and Yoshimori 2014). PI3P production is necessary for the recruitment of proteins with PI3P binding domains such as FYVE or PX. For example DFCP1 or the WD-repeat protein interacting with phosphoinositides (WIPI1/2, homologues of Atg18) are recruited via PI3P (Itakura and Mizushima 2010; Matsunaga, Morita et al. 2010). Atg14L is essential for the phagophore formation and PI3K activity of the complex (Itakura and Mizushima 2010). AMBRA1 is phosphorylated by ULK1 upon autophagy induction which leads to the release of the Beclin1- Vps34-Vps15-AMBRA1 core complex from the dynein motor on the microtubules to the ER (Di Bartolomeo, Corazzari et al. 2010) (Figure 9).

The complex consisting of Beclin1-Vps34-Vps15-Rubicon and UV irradiation resistance associated gene (UVRAG, Vps38) has a role in endosomal transport (Itakura, Kishi et al. 2008) and negatively regulates autophagosome-lysosome fusion (Zhong, Wang et al. 2009). Atg14L and UVRAG compete with each other to bind to Beclin1 in the class III PI3K complex. UVRAG can bind to Bax-interacting factor 1 (Bif-1) which can increase the PI3K activity and have a positive role in initiation of autophagosomes. However it is shown that UVRAG can also have a negative role in the maturation step upon interaction with RUN domain and cysteine-rich domain containing Beclin1 interacting protein Rubicon (Matsunaga, Saitoh et al. 2009).

Another PI3K complex which includes Bcl-2 negatively regulates autophagosome formation via the Bcl-2 - Beclin1 interaction (Pattingre, Tassa et al. 2005). The dissociation of Bcl-2 from Beclin1 positively regulates autophagy (Mizushima, Yoshimori et al. 2011).

Atg9 Vesicles

Atg9 (mAtg9 or Atg9L1) is the only multi-pass transmembrane protein in the core autophagic machinery. Atg9 is found in small vesicles and tubular structure in the cell similar to its homolog in yeast. Under nutrient-rich conditions Atg9 is observed to be associated with the trans-Golgi networks (TGN) and late endosomes. Upon amino-acid starvation Atg9

positive vesicles are redistributed from the TGN. Atg9 cycles in the cell in a bidirectional manner between TGN and autophagy-related vesicles and this shuttling is ULK1 dependent (Young, Chan et al. 2006). The peripheral population of Atg9 co-localizes partially with LC3 and shows a dynamic interaction between DFCP1 positive phagophores and autophagosomes while it does not get incorporated into the autophagosomal membranes (Figure 9). The retrieval of the Atg9 from the autophagosomes depends on WIPI2 (Orsi, Razi et al. 2012). The specific role of mammalian Atg9 vesicles during autophagy remains to be determined however the existing data suggest that Atg9 vesicles may serve as membrane sources for autophagosome biogenesis and that Atg9 may play a role in phagophore expansion. The role of Atg9 as a source for autophagosome biogenesis will be addressed further (in chapter 2.3.4).

Figure 9: Autophagosome machinery

There are three important stages in autophagic machinery: 1. Initiation and nucleation (formation of the phagophore); 2. Vesicle elongation (growth and closure of autophagosome); 3. Maturation into autolysosomes. These stages depend on different core Atg proteins which are recruited to the autophagosome assembly site: Atg1/ULK complex, the class III phosphatidylinositol 3-kinase (PI3K) (Beclin-1-PtdIns3KC3-Atg14L) complex, the WIPI complex, the Atg12-Atg5-Atg16L1 conjugation system, the Atg8/LC3 conjugation system, and Atg9 vesicles. (Mizushima and Komatsu 2011)

2.3.2. Elongation

Atg12 Conjugation System

Autophagosome elongation constitutes of two “ubiquitin-like” conjugation systems. One is the Atg12-Atg5-Atg16L1 complex. During the formation of this complex Atg12 covalently conjugates with the lysine residue of Atg5 via the C terminal glycine (Mizushima, Noda et al. 1998). Atg7 and Atg10 catalyze this conjugation via their E1-like and E2-like enzyme function respectively, by activating Atg12 and transferring it to Atg10 for Atg5 binding. Atg12-Atg5 complex then binds to Atg16L1 and forms the 2:2:2 stoichiometric complexes via the homo-dimerization of Atg16L1 (Fujioka, Noda et al. 2010). Atg16L2, another isoform of the mammalian Atg16L1, can also bind to the Atg12 conjugation complex; however it seems that this conjugation is not essential for autophagy pathway (Ishibashi, Fujita et al. 2011). This Atg12-Atg5-Atg16L1 complex does not have a de-conjugating enzyme therefore it is formed constitutively in the cell independent of the nutrient conditions. During autophagy induction, the multimeric complexes are associated to the outer surface of the isolation membrane via Atg16L1 and are dissociated from the membrane after the completion of the autophagosome formation (Mizushima, Yamamoto et al. 2001). Therefore the proteins of the complex serve greatly as early markers of the autophagosome formation (Figure 9). It is recently shown that Atg16L1 directly binds to WIPI2b upon starvation, and this interaction is required for LC3 conjugation (Dooley, Razi et al. 2014). Atg12-Atg5- Atg16L1 system was thought to function like an E3-like enzyme for the other ubiquitin-like system of the autophagic machinery, the Atg8/LC3 conjugation system (Hanada, Noda et al. 2007). Recently, structural studies revealed that Atg12 conjugation system interacts with Atg3 and induces a re-arrangement of catalytic site of Atg3 which in turn enhances the conjugase activity of the protein and promoting the LC3-PE reaction (Sakoh-Nakatogawa, et al. 2013).

LC3 conjugation system

Yeast Atg8 protein has many mammalian orthologues including LC3/A/B/C, GABARAP, and GABARAPL1/2/3. LC3B, the most abundant of the LC3 family protein, exists in the cell in two forms, a cytosolic non-lipidated form (LC3-I) and a lipidated form LC3-II which is conjugated to a major membrane phospholipid phosphatidylethanolamine

(PE) (Kabeya, Mizushima et al. 2000). LC3B is translated as precursors (pro-LC3) and the C- terminal peptide is later processed by a cysteine protease Atg4 forming the cytosolic LC3-I. This cleavage leaves the C-terminal glycine exposed for conjugation to the amino group of PE via the E1-like enzyme Atg7 and E2-like enzyme Atg3 (Ichimura, Kirisako et al. 2000). As mentioned above, LC3-PE conjugation requires the Atg12 conjugation system promoting the activity of Atg3, therefore the site of LC3 lipidation is determined by the Atg12-Atg5- Atg16L1 complex and the protein-lipid conjugation takes place on the isolation membrane (Hanada, Noda et al. 2007). LC3-PE, also referred to as LC3-II, remains associated with the autophagosomes until their fusion with lysosomes thus serving as a good marker for autophagic activity. Different LC3 proteins can have different functions in the cell and autophagosome formation (Weidberg, Shvets et al. 2010). It also known that the decrease of Atg8 expression lead to smaller autophagosomes suggesting a role for Atg8/LC3 on autophagosome size therefore on the expansion of the membrane (Xie, Nair et al. 2008).

2.3.3. Maturation

The maturation of mammalian autophagosomes is a multi-step process involving the fusion of various vesicles. Autophagosomes fuse with endosomes forming “amphisomes” and then fuse with lysosomes forming “autolysosomes” (Eskelinen 2005). The acidification of the autophagosomes begins during the maturation step (Dunn 1990; Punnonen, Autio et al. 1992). Following the maturation of amphisomes and autolysosomes, the breakdown and degradation of the autophagosome inner membrane and its cargo depend on the lysosomal hydrolases, whereas the recycling of the macromolecules occurs through the permeases (Yang and Klionsky 2010).

The maturation process depends on RabGTPases, SNAREs, ESCRT proteins, the HOPS complex, some of the lysosomal membrane glycoproteins LAMP-2 variants and the Beclin1-Vps34-Vps15-UVRAG complex which up-regulates the maturation and the endocytic trafficking (Figure 8) (Rusten and Simonsen 2008; Saftig, Beertsen et al. 2008; Manil-Segalen, Culetto et al. 2014; Tooze, Abada et al. 2014). A small GTPase protein Rab7 is required for the fusion of autophagosomes with lysosomes (Jager, Bucci et al. 2004). UVRAG protein can also interact with class C Vps complex, which includes Vps11, Vps16,

Vps18 and Vps33, and this interaction positively regulates the GTPase activity of Rab7 and in turn promotes the maturation of autophagosomes and endocytic trafficking. This interaction is independent of the UVRAG-Beclin1 interaction (Liang, Lee et al. 2008). The class C Vps complex can either exist as the HOPS complex which includes two additional subunits Vps39 and Vps41, or it can also exist as the core vacuole/endosome tethering (CORVET) complex containing Vps3 and Vps8 (Liang, Lee et al. 2008). HOPS complex and Rab7 are essential for endosomal fusion with lysosomes, whereas only Rab7 is found to be necessary for autophagosome-lysosome fusion (Ganley, Wong et al. 2011). UVRAG protein can also bind to Rubicon and in this case Beclin1-Vps34-Vps15-UVRAG-Rubicon complex can inhibit the maturation of autophagosomes (Matsunaga, Saitoh et al. 2009). A SNARE molecule Syntaxin 17 is found on the outer membrane of mature autophagosomes but not on the phagophore membrane. It is hypothesized that Syntaxin 17 serves to prevent an immature fusion of phagophore with the lysosome (Itakura, Kishi-Itakura et al. 2012). Endosomal sorting complex required for transport (ESCRT) proteins are essential for endosomal trafficking and biogenesis and their mutations are involved in various neurodegenerative diseases (Metcalf and Isaacs 2010). ESCRT III complex is shown to have a role in maturation of amphisomes and autolysosomes (Lee, Beigneux et al. 2007). Finally, LAMP-1 and LAMP-2 proteins are also involved in maturation of autophagosomes, however the knock-down of these proteins can result in different phenotypes depending on the cell type and tissue (Eskelinen 2005).

2.3.4. The Origin of the Autophagosome

The increasing number of discoveries and studies on Atg proteins and their trafficking in the cell are revealing the mystery behind the origin(s) of the autophagosomal membrane. For yeast, a specific autophagosome formation site is identified and shown by live-cell imaging: the pre-autophagosomal structure (PAS) where almost all of the Atg proteins assemble and generate autophagosomes (Suzuki, Kirisako et al. 2001). In the mammalian cell there are a number of strong candidates for autophagosome formation site supporting the hypothesis of multiple-membrane source: the ER, mitochondria, the Golgi apparatus, recycling endosomes and the plasma membrane (Shibutani and Yoshimori 2014) (Figure 10).

Endoplasmic Reticulum

The endoplasmic reticulum (ER) is one of the best candidates as the origin of autophagosome formation site and source as the proteins involved in the nucleation such as ULK1 and PI3P effectors are localized to the ER during starvation (Axe, Walker et al. 2008). DFCP1 which binds to PI3P via its FYVE domain is localized to the ER. It is known that upon starvation the autophagosome membranes are emerged from the PI3P-binding FYVE-containing protein 1 (DFCP1) positive cup shaped, ring like structures which are partially LC3 positive (Itakura and Mizushima 2010). These structures are called omegasomes due to the Ω shape. Two 3D tomography studies confirmed the presence of the isolation membrane on the ER via the “ER cradle model”, 1 µm diameter ring like structure sandwiched between two cisternae and physically connected to the ER via a narrow tube-like structure (Hayashi-Nishino, Fujita et al. 2009; Yla-Anttila, Vihinen et al. 2009) (Figure 10). According to this model early autophagic machinery, including Atg14L, DFCP1 Atg5 and ULK1, is located to the omegasome which serve as a cradle for phagophore formation (Shibutani and Yoshimori 2014).

ER-Mitochondria contact sites

Mitochondrion was also a candidate as a site for autophagosome formation following the study of Hailey and colleagues, which showed that LC3 positive membranes co-localized with an outer mitochondrial membrane protein upon starvation. This study proposed that the sharp membrane curvature at the connection site between the mitochondrion and the isolation membrane prevented the transfer of transmembrane proteins from entering autophagosomes (Hailey, Rambold et al. 2010). Following starvation the proteins of the PI3K complex, Atg14L, Beclin1, Vps34 and Vps15, are observed to accumulate at the ER-mitochondria contact sites however Atg5 accumulation at the contact sites is very dynamic (Hailey, Rambold et al. 2010; Hamasaki, Furuta et al. 2013). Syntaxin 17 is also observed to be necessary for the recruitment of Atg14L to the ER-mitochondria contact sites during early autophagosome formation (Hamasaki, Furuta et al. 2013) (Figure 10).

Figure 10: The Origin of the Autophagosome

(A) Membrane sources of autophagosomal membrane. (Shibutani and Yoshimori 2014) (B) 3D models of electron tomography of ER presenting the ER-cradle model of autophagosome formation site. The ER, IM (isolation membrane) and VTC (vesicular tubular cluster) -like compartments are shown in blue or red, yellow and white, respectively. White arrowheads in b indicate the continuity of VTC with the ER. (Hayashi-Nishino, Fujita et al. 2009)

Golgi apparatus

The involvement of Golgi apparatus in autophagosome biogenesis is described mostly in yeast (Geng and Klionsky 2010; van der Vaart and Reggiori 2010). In mammalian cells, LC3 positive structures are formed and bud off from the TGN upon starvation (Guo, Chang et al. 2012). Golgi is often associated with the mammalian Atg9 vesicles since under nutrient- rich conditions Atg9 is localized at the TGN along with late endosomes and is redistributed from the TGN upon autophagy induction (Young, Chan et al. 2006). The ER-Golgi intermediate compartment (ERGIC) which is a very dynamic site for membrane trafficking is

also another candidate as an autophagosome formation site identified by the group of Schekman. Upon starvation, ERGIC recruits Atg14L and initiates phagophore formation and triggers LC3 lipidation in vitro(Ge, Melville et al. 2013).

Plasma Membrane

It was recently shown that the plasma membrane can also contribute to the formation of autophagosomes. The group of David Rubinsztein showed the direct contribution of the plasma membrane for the formation of the phagophore via the interaction of Atg16L1 with the heavy chain of clathrin. Atg16L1 positive vesicles are internalized via the clathrin- dependent endocytosis and mature into early autophagic structures that are LC3 negative, both under basal or autophagy-induced conditions (Ravikumar, Moreau et al. 2010). The maturation of the early Atg16L1 positive-LC3 negative vesicles into the phagophore is shown to require the homotypic fusion of Atg16L1 vesicles via SNARE protein VAMP7 (Tetanus Neurotoxin-insensitive Vesicle-Associated Membrane Protein) and its partner SNAREs (Syntaxin 7, Syntaxin 8 and Vti1b). This step is also important for the vesicle size and is increased with autophagic stimulation (Moreau, Ravikumar et al. 2011).

Recycling Endosomes

Atg9 is observed to be localized to TGN along with endosomes such as recycling endosomes and it can also be found incorporated in autophagosomes (Orsi, Razi et al. 2012). It is also shown that Atg9 and ULK1 positive recycling endosomes are important for autophagosome formation. The redistribution of Atg9 in the cell upon starvation suggests that Atg9 vesicles recruit membrane sources for the increased autophagosome formation (Longatti and Tooze 2010; Longatti, Lamb et al. 2012). The possible contribution of recycling endosomes to autophagosome biogenesis is also described by the recent findings of Atg9 localization to the plasma membrane (Puri, Renna et al. 2013). It is shown that Atg9 is internalized via clathrin-coated vesicles, via a different route from the Atg16L1-associated endocytic vesicles. These Atg9 vesicles associate with Atg16L1 in Rab11 positive recycling endosomes prior to autophagosome formation and increase upon starvation induction (Puri, Renna et al. 2013). However solely the Atg9 vesicles (300-900 nm diameter vesicles) are too small to contribute to lipid intake in the autophagosome membrane, supporting the hypothesis

of multiple organelle sources for phagophore formation (Shibutani and Yoshimori 2014). The distinct role of mammalian Atg9 and Atg16L1 positive recycling endosomesin autophagosome biogenesis is to be determined.

2.4. Regulation of Autophagy

2.4.1. Diverse Regulatory Pathways

Autophagy can be constitutively active in the cell however autophagic activity depends on integration of different inputs and stress stimuli such as various nutrient deprivations including amino acids or growth factor starvation along with stimuli like organelle damage, infection, hypoxia or mechanical stress. These signals are integrated and regulated by diverse cellular pathways.

mTOR-dependent Regulation of Autophagy

One of the best characterized regulatory pathways for autophagy is the mammalian target of TOR (mTOR) pathway. mTOR integrates multiple signaling pathways that are sensitive to the availability of amino acids, ATP, growth factors, and the levels of reactive oxygen species and negatively regulates autophagy (Singh and Cuervo 2011). Besides autophagy, mTOR pathway has diverse functions in the cell such as initiation of mRNA translation, cell growth and proliferation, ribosome biogenesis, transcription, cytoskeletal reorganization and long-term potentiation (Ravikumar, Sarkar et al. 2010). Downstream targets of the mTOR complex includes for example ribosomal protein S6 kinase-1 (S6K), translation initiation factor 4E binding protein-1 (4EBP-1), and eEF2 kinase which are involved in protein synthesis regulation. There are two mTOR complexes; a rapamacyn sensitive mTOR complex1 (mTORC1) and rapamycin insensitive mTORC2. mTORC1, the main regulator of autophagy, includes the catalytic subunit raptor, G protein β -subunit-like protein (GβL) and proline-rich Akt substrate of 40 kDa (PRAS40) (Meijer, Lorin et al. 2014). Rapamycin, a potent autophagy activator, inhibits the kinase activity of mTORC1. In the presence of nutrients, mTORC1 is responsible for the phosphorylation of ULK1 on Serine 757 and its subsequent sequestration therefore inhibiting autophagy. Activation of mTORC1

by amino acids occurs via the translocation of the complex to the lysosomal membranes where Rag GTPases reside. Insulin activates the complex via G protein Rheb (Meijer and Codogno 2011). Inhibition of mTORC1 by rapamycin is mediated by the reduced phosphorylation of S6K and 4EBP1 and releases ULK1 to phosphorylate FIP200 in the ULK1-Atg13-FIP200 complex to be recruited to the phagophore formation site on the ER (Jung, Jun et al. 2009).

AMP-activated protein kinase (AMPK) function as an energy sensor and can sense the changes of AMP/ATP ratio in the cell. During starvation, decreased ATP production stimulates AMP-activated protein kinase (AMPK). AMPK activity is facilitated by upstream kinases such as LKB1 and Ca2+/calmodulin dependent protein kinase β. AMPK inhibits mTORC1 activity by phosphorylation of downstream effectors TSC2 (tuberin) and Raptor resulting in autophagy activation. mTOR is also responsible for the phosphorylation of the transcription factor EB (TFEB). When translocated to the nucleus TFEB activates the synthesis of Atg proteins and regulates lysosomal biogenesis (Settembre and Ballabio 2011). The different classes of PI3K have distinct effects on autophagy.

PI3K class I produce the membrane lipid PIP3 in the presence of growth factor and insulin which induces Akt activation and inhibits autophagy. Akt activation is responsible for the phosphorylation of TSC1/TSC2 complex which regulates cell growth and survival and inhibits autophagy. Beclin1 is shown to be phosphorylated by Akt and protein kinase B which promotes its interaction with vimentin intermediate filament and autophagy inhibition (Wang, Wei et al. 2012). PI3K pathway can be inhibited by the tumor suppressor phosphoinositide phosphatase (PTEN) activity which in turn induces autophagy. The class III PI3K stimulates autophagy via PI3P production. Well known autophagy inhibitors 3-Methyladenine (3-MA) and wortmannin are PI3K inhibitors (Ravikumar, Sarkar et al. 2010).

Figure 11: mTOR dependent regulation of Autophagy

mTOR is part of the mTORC1 complex. AMPK stimulates ULK1 and autophagy by coordinated inhibition of mTOR through phosphorylation of TSC2 and Raptor, and phosphorylation of ULK1. mTOR, activated by amino acids and insulin, phosphorylates ULK1 at a different site from the site of AMPK inhibiting ULK1 activity and autophagy.

Other kinases can modulate autophagy as well. IκB kinase (IKK) can activate JNK1 and AMPK inhibiting mTOR and inducing autophagy in an NF-κB independent manner (Criollo, Senovilla et al. 2009). A serine threonine kinase, eukaryotic translation initiation factor 2α (eIF2α) kinase can regulate stress-induced autophagy upon virus infection or starvation. Mitogen-activated protein kinase/ extracellular signal-regulated kinase (MAPK/ERK) pathway can regulate autophagy in different states. This pathway is usually activated in cancers and sustained ERK activation inhibits mTOR complex and enhance Beclin1 activity. P38 MAPK can also stimulate autophagy via mTOR inhibition. Stress- activated signaling molecule, Jun NH2-terminal kinase 1 (JNK1) can modulate autophagy in response to starvation via Bcl-2-Beclin1 complex dissociation by Bcl-2 phosphorylation (Wei, Pattingre et al. 2008). Beclin1 phosphorylation by death-associated protein kinase (DAPK) can stimulate the dissociation of Bcl-2-Beclin1 complex from the PI3K complex I as well (Zalckvar, Berissi et al. 2009).

mTOR-independent regulation of Autophagy

During energy deprivation, AMPK can also directly phosphorylate ULK1 on different sites, independent of mTORC1. Egan et al. identified Serine 467 and Serine 555 as essential phosphorylation sites of ULK1 for autophagic activity while Kim et al. showed that Serine 317 or Serine 777 phosphorylation is necessary for the induction of autophagosome formation by AMPK in glucose deprivation (Egan, Shackelford et al. 2011; Kim, Kundu et al. 2011) (See Figure 11 for an overview of this pathway). In the absence of glucose AMPK can phosphorylate Beclin1 and regulate Vps34 complex via Atg14L (Kim, Kim et al. 2013).

Phosphoinositol signaling pathway is a mediator of autophagy. Intracellular inositol 1,4,5-trisphosphate (IP3) receptors (IP3R) found on the ER bind IP3 and release Ca2+ into the cytoplasm. Pharmacological agents such as mood-stabilizing drugs like lithium reduce intracellular inositol levels and induce autophagy in an mTOR independent. IP3R is also shown to be a regulator of the Beclin1 complex (Ravikumar, Sarkar et al. 2010). An upstream pathway regulating IP3 levels is the cAMP-Epac-PLC-ε pathway. Ca2+ release from the ER upon IP3R binding negatively regulates autophagy in a cyclic pathway. Pharmacological agents like thapsigargin or ionomycin which induce intracellular calcium levels also inhibit autophagy. Elevated intracellular Ca2+ induce Ca2+ -dependent cysteine proteases called calpains which increases adenyl cyclase activation and therefore cAMP production. Elevated cAMP negatively regulates autophagy creating a cyclic regulation (Gordon, Holen et al. 1993; Williams, Sarkar et al. 2008)

Reactive oxygen species (ROS) can modulate autophagy as well. ROS are produced by incomplete reduction of oxygen and are highly reactive molecules that cause oxidative stress. They are however found to have important signaling properties as well for example

H2O2 can activate NF-κB transcription factor. Starvation or hypoxia increase ROS levels and increase of ROS stimulates autophagy, and specifically mitophagy as a survival mechanism by favoring LC3 lipidation via Atg4 oxidization or by increasing Beclin1 (Scherz-Shouval, Shvets et al. 2007; Chen, McMillan-Ward et al. 2008; Azad, Chen et al. 2009).

2.4.2. Post-translational Modifications of Autophagic Machinery

Some of the post-translational regulations during autophagy induction were mentioned earlier. During autophagosome biogenesis, Atg12–Atg5 and the LC3–PE conjugates require ubiquitin-like reactions and mTOR pathway-dependent regulation of autophagy requires the phosphorylation of ULK1. These modifications are well known in the literature, however new information is surfacing on post-translational modifications such as phosphorylation, ubiquitination and acetylation of Atg machinery (McEwan and Dikic 2011; Boya, Reggiori et al. 2013).

Phosphorylation

ULK1, an essential autophagic kinase can regulate autophagy by phosphorylating various proteins. As mentioned above, ULK1 is phosphorylated by mTOR and AMPK on different sites. Raptor subunit of mTORC1 is phosphorylated by ULK1 which reduces substrate recognition by mTORC1, contributing to a negative feedback loop to inhibit mTORC1 in presence of nutrients. ULK1 can also phosphorylate and negatively regulates AMPK. AMBRA1 is also phosphorylated by ULK, which in turn activates the Atg14L- Beclin1-PI3K class III complex formation. Beclin1 can also be regulated by phosphorylation by Akt or by DAPK which leads to the dissociation of Beclin1 from Bcl-XL to induce autophagy. In response to glucose deprivation, Beclin1 can also be phosphorylated by AMPK (Wirth, Joachim et al. 2013).

Other autophagic proteins are phosphorylated by different kinases. LC3 isoforms can be phosphorylated by cAMP dependent protein kinase A (PKA) or protein kinase C (PKC). This modification reduces the autophagosome formation in an unknown mechanism (Cherra, Kulich et al. 2010; Jiang, Cheng et al. 2010). In S. cerevisiae PKA phosphorylates Atg13 and inhibits its association with the pre-autophagosomal structure (PAS), however this mechanism is not known in mammalian cells (Stephan, Yeh et al. 2009). However apart from ULK1, mTORC1 also phosphorylates mAtg13. The phosphorylation of Beclin1 on Thr119 within its BH3 domain,by DAPK also regulates autophagy induction by its dissociation from Bcl-2- Beclin1complex (Zalckvar, Berissi et al. 2009).

Phosphorylation of other autophagic proteins during selective autophagy increases their substrate binding. p62, phosphorylated by casein kinase 2 interacts with ubiquitinated

proteins and optineurin, phosphorylated by TANK-binding kinase 1 binds to LC3 (Pankiv, Clausen et al. 2007; Geisler, Holmstrom et al. 2010).

Ubiquitylation

Ubiquitin (Ub) is a small 8kDa, evolutionarily conserved protein expressed in all eukaryotic cells. It contains seven lysine residues and can be conjugated via the action of E1 (Ub-activating enzyme), E2 (Ub-conjugating enzyme) and E3 (Ub ligase, which dictates substrate specificity) enzymes. There are several types of ubiquitination: fixation of one Ub protein on the lysine of the substrate, “monoubiquitylation”, fixation of multiple Ub chains, “polyubiquitylation” and fixation of one Ub on several lysines, “multi-monoubiquitylation”(Robinson and Ardley 2004). Ubiquitylation of the cargo proteins plays an important role during selective autophagy for targeting the cargo to the autophagy receptors for the degradation (McEwan and Dikic 2011). Adaptor proteins such as p62/SQSTM1, NBR1, NPD52 or optineurin can bind LC3 via a LIR motif and ubiquitylated cargo via their Ub binding domains and allowing the inclusion of the target protein aggregates into autophagosomes. As for Atg proteins themselves, there is no evidence of ubiquitylated Atg proteins yet, although autophagy itself has a ubiquitin-like system during autophagosome formation. Atg7 is found to function as an E1-like enzyme and activate LC3 and Atg12 while Atg10 functions as a E2-like enzyme conjugating Atg12 to Atg5 and LC3 is conjugated to PE by Atg3. The formation of an Atg12–Atg3 conjugate through the action of Atg7 and the autocatalytic activity of E2-like Atg3 has been described in mammalian cells (Radoshevich, Murrow et al. 2010). This Atg12 conjugate has a role during cell death and mitochondrial expansion (Radoshevich and Debnath 2010).

Acetylation

Another important post-translational modification is acetylation; the addition of an acetyl group into lysine or N-terminus of target proteins which leads to the regulation of protein function, gene transcription and receptor trafficking. Best known regulator of acetylation in autophagy is histone deacetylase 6 (HDAC6), which is responsible for degradation of protein aggregates via selective autophagy, termed as “aggrephagy” (Deribe, Pawson et al. 2010). HDAC6 is also involved in autophagosome maturation through F-actin

network and in mitochondria degradation by autophagy (mitophagy) as well (Lee, Koga et al. 2010). Other deacetylases and acetyltransferases regulate autophagic machinery as well (Hamai and Codogno 2012). It is shown in vitro that acetyltransferase p300 can acetylate Atg5, Atg7, Atg8 and Atg12 proteins, and p300 can directly interact with Atg7 and this acetylation inhibits autophagy under nutrient rich conditions while p300 inhibition increases autophagic flux (Lee and Finkel 2009). Upon starvation the NAD-dependent deacetylase Sirt1 removes acetyl groups from the Atg7, Atg5, Atg12, and Atg8, which allows autophagy induction. Sirt1 knockout mice phenotype resembles that of Atg5 knockout mice suggesting an important role of Sirt1-dependent deacetylation for autophagy at basal levels and in response to starvation (Lee, Cao et al. 2008). Autophagy is also promoted by acetylation of mutant Huntingtin protein and its subsequent clearance whereas a mutant version of Huntingtin that cannot be acetylated accumulates and leads to neurodegeneration (Jeong, Then et al. 2009). Acetylations of tubulin is also necessary during autophagy initiation upon starvation for the recruitment and transport of various Atg proteins and autophagosomal binding proteins by the microtubule network and for the induction of autophagy followed by the activation of kinesin and JNK1 and Bcl-2-Beclin1 complex release (Geeraert, Ratier et al. 2010). HDAC6 mediate the acetylation of tubulin and autophagosome trafficking (Iwata, Riley et al. 2005).

Another acetyltransferase, TIP60 is a histone acetyltransferase and identified as a positive regulator of autophagy in crosstalk between phosphorylation. In response to growth factor or serum withdrawal, TIP60 is activated by phosphorylation on two residues by glycogen synthase kinase-3 (GSK3) and then acetylates ULK1 which increases its kinase activity thus stimulating autophagy induction (Lin, Li et al. 2012). ULK1 activity depends on two different post-translational modification, phosphorylation and acetylation in response to amino acid or growth factor factor deprivation.

Additionally, HLA-B associated transcription 3 (BAT3) is shown to regulate autophagy through the regulation of intracellular localization of acetyltransferase p300 during basal autophagy and in response to starvation. P300 can acetylate Atg7 and p53 and negatively regulate autophagy (Sebti, Prebois et al. 2014).

2.4.3. Transcriptional and Epigenetic Regulations of Autophagy

Autophagy is a cytoplasmic process and mostly regulated by cytoplasmic modifications under different stimuli. However, with more and more knowledge on various transcriptional and epigenetic regulations of autophagy suggest that nucleus is also a key player for autophagy (Fullgrabe, Klionsky et al. 2014).

The evolutionarily conserved forkhead box O (FOXO) family of transcription factors regulates autophagy. FOXO proteins are regulated by phosphorylation-dependent nuclear- cytoplasmic shuttling. FOXO1 and FOXO3 can induce the expression of various core autophagic machinery genes and induce autophagy in a transcription-dependent manner (Sengupta, Molkentin et al. 2009; Zhou, Liao et al. 2012).

Other transcriptional activators have been shown to regulate autophagy. A transcriptional activator of E2F family of transcription factors, E2F1 is a positive autophagy regulator via its target gene BNIP3 which is responsible for the dissociation of Bcl2-Beclin1 association (Polager, Ofir et al. 2008). NF-κB transcriptional factor is involved in this regulation due to its competition for the BNIP3 promoter. However NF-κB is also an inducer of autophagic genes independent of its relation with BNIP3 promoter (Shaw, Yurkova et al. 2008).

A well-known tumor suppressor p53 shuttles constantly through the nuclear pore complex and has a well characterized role as a transcription factor. Cytoplasmic p53 supresses autophagy whereas nuclear p53 activates several autophagy related genes and even regulates the upstream regulatory pathways such as mTOR, AMPK and PI3K pathway related genes (Tasdemir, Maiuri et al. 2008). Other p53 family members like p63 and p73 can also positively induce autophagy related genes (Kenzelmann Broz, Spano Mello et al. 2013).

The transcription factor EB (TFEB) is a key modulator of lysosomal biogenesis and upregulates lysosomal genes. TFEB is also an important positive regulator of autophagy. In response to autophagic stimuli such as starvation, cytosolic TFEB translocates to the nucleus following phosphorylation and induces various autophagic genes (Settembre, De Cegli et al. 2013). ZKSCAN3 (zinc-finger protein with KRAB and SCAN domains 3) has the counter effect of TFEB on autophagy. This DNA-binding protein functions as a transcriptional repressor of autophagy genes and is negatively regulated upon starvation (Moresi, Carrer et al. 2012).

Several other transcription factors regulating autophagy are identified mostly involved in transactivation of autophagy related genes which suggests that a complex transcriptional network in the nucleus regulates the level of autophagic flux in the cell. microRNAs (miRNAs) constitute another regulator family of autophagy in the nucleus. miRNAs are ~22 nucleotide long non-coding RNAs that regulate gene expression and silence specific mRNAs. Selective autophagy can regulate miRNA biogenesis whereas there are exhaustive number of miRNAs that target different ATG genes (Fullgrabe, Klionsky et al. 2014).

Histone acetylation and methylation is shown to modulate autophagy (Rohde and Cardenas 2003; Artal-Martinez de Narvajas, Gomez et al. 2013). In yeast, TOR inhibition is associated with down-regulation of H3 Lys56 acetylation (Chen, Fan et al. 2012). This mechanism and its function remain to be discovered.

2.5. Physiological Role of Autophagy

Autophagy occurs at a basal rate in the cells as a protective mechanism in order to maintain energy homeostasis. Over the years, various autophagy deficient murine models were generated for genes expressing core autophagic proteins such as Beclin1, Atg5 or Atg7. These important studies revealed significant physiological roles for this basal or constitutive autophagy in different organs, tissues or stages of development (Mehrpour, Esclatine et al. 2010) (Figure 12). Autophagy can also be modulated upon integration of different inputs and stress stimuli such as various nutrient deprivations including amino acids or growth factor starvation along with stimuli like organelle damage, infection, hypoxia or mechanical stress. Constitutive and induced autophagy depends on different regulation mechanisms in the cell and can have different physiological consequences. Basal autophagy regulates mainly cellular and tissue homeostasis via renewal of proteins and organelles, while induced autophagy is mainly responsible to maintain energy for the stressed cell and regulate cellular metabolism.

Figure 12: Physiological Roles of Autophagy (Mizushima and Komatsu 2011)

2.5.1. Selective Autophagy

There is another layer of complexity to autophagic activity: selective autophagy. Selective autophagy seem to be more relevant at basal levels and contributes to the the maintenance of tissue homeostasis. Selective identification and targeting the excessive or damaged intracellular content occurs by autophagy adaptor proteins p62/SQSTM1, NBR1, NDP52, optineurin, BNIP3L which recognize and target their substrates to the autophagy machinery (Okamoto 2014; Rogov, Dotsch et al. 2014; Stolz, Ernst et al. 2014) (Figure 13). These receptor proteins identify their ubiquitylated cargo and target them to autophagosomes via their specific interaction with LC3 protein. This interaction is ensured by the LC3- interacting region (LIR) motif with the consensus sequence [W/F/Y]xx[L/I/V] (Birgisdottir, Lamark et al. 2013). Sequestosome 1 (SQSTM1) or p62 contains several domains including a LIR domain and Ubiquitin-associated (UBA) domain which can recognize mono- ubiquitinated proteins and poly-ubiquitinated chains. Damaged proteins are recognized by p62 and form aggregates which will be anchored to autophagosomes by direct interaction with LC3 on the isolation membrane for subsequent degradation (Pankiv, Clausen et al. 2007). As mentioned earlier the inhibition of autophagy can result in accumulation of protein aggregates and damaged organelles contributing to tissue homeostasis as in the case for liver specific Atg7 knock-out mice which developed liver dysfunction (Komatsu, Waguri et al. 2007).

Autophagy receptors may have specialized role as NDP52 binds to ubiquitylated cytosolic Salmonella for autophagic degradation upon infection. Removal of damaged mitochondria by mitophagy is mediated by an E3 ligase, Parkin which is recruited upon loss of mitochondrial membrane potential. p62 is recruited following ubiquitylation of voltage- dependent anion channel 1 (VDAC1) by Parkin and mediates the clearance of damaged mitochondria (Geisler, Holmstrom et al. 2010). This process may not require VDAC for p62 recruitment, but p62 is responsible for mitochondrial clustering prior to mitophagy (Narendra, Kane et al. 2010). The recruitment of Parkin depends on the presence of phosphatase and tensin homolog-induced putative kinase 1 (PINK1) which can bind to Beclin1 independent of its kinase activity (Michiorri, Gelmetti et al. 2010). The complex process of mitochondrial clearance suggest that it is cell type- and context- specific process, with multiple ubiquitylation targets on the mitochondrial membrane (McEwan and Dikic 2011). Autophagy- linked FYVE protein (ALFY) also interacts with p62 and with Atg5 directly and PI3P serving as scaffold protein for autophagy induction (Birgisdottir, Lamark et al. 2013). Selective autophagy can also be induced by extracellular stimuli such as pathogens. p62 intervenes for 56

pathogen degradation via xenophagy (Dupont and Lafont 2009). As mentioned earlier, histone de-acetylase 6 (HDAC6) is involved in selective autophagy. HDAC6 mediate the degradation of aggregate containing inclusion bodies, aggrephagy, and is involved in mitophagy as well (Deribe, Pawson et al. 2010; Lee, Koga et al. 2010).

Figure 13: Selective autophagy

Recognition and degradation by selective autophagy depends on the type of cargo and various adaptor proteins. Question marks indicate unidentified receptor proteins for the specific type of selective autophagy (Rogov, Dotsch et al. 2014).

2.5.2. Autophagy during pre- and post-natal state

During the pre-implantation development, autophagy has a role of eliminating maternal proteins in the fertilized oocyte, which is vital for the development of the embryo (Tsukamoto, Kuma et al. 2008). Two very important mouse models Atg5 and Atg7 KO mice revealed the vital role of autophagy in neonates. Autophagy deficient mice die within one day following the birth. After birth neonates are in a severe starvation state due to sudden termination of nutrient supply through the placenta, during this time basal autophagy is highly up-regulated for amino-acid generation as a direct energy source until it is supplied by milk. Autophagy up-regulation is quiet rapid after 30 minutes of birth and is steady for the next 12 hours (Kuma, Hatano et al. 2004; Komatsu, Waguri et al. 2005). Autophagy has an important role in glucose metabolism in a selective manner by glycogen degradation. This occurs in postnatal period which demands high hepatic glucose production. Glucagon secretion following postnatal hypoglycemia induces cAMP levels promoting autophagy (Kotoulas, Kalamidas et al. 2004).

2.5.3. Autophagy in the liver

Liver, a metabolically active organ, demands high energy supply in the organism. The first role of autophagy in hepatic cells was observed by Christian de Duve himself as glucagon and starvation induced autophagy in liver cells (Deter, Baudhuin et al. 1967). Further observation was stated as amino acids inhibited autophagy in the rat liver (Mortimore and Schworer 1977). It is proposed that the physiological role of autophagy in the liver is to maintain energetic balance and a quality control by breakdown of intracellular energy stores, by control of key regulatory enzymes and by mitochondrial homeostasis preservation (Schneider and Cuervo 2013). The conditional Atg7 KO deficient, unstressed mice liver showed an accumulation of organelles and a disturbed turnover of proteins in hepatocytes leading to compromised mitochondrial function (Komatsu, Waguri et al. 2005). Hepatic autophagy is up-regulated rapidly in 4 to 8 hours of starvation for breakdown of energy stores such as glycogen (glycophagy) and lipids (lipophagy) suggesting that basal autophagy serve to maintain high energetic demands of hepatocytes (Kotoulas, Kalamidas et al. 2006; Singh, Kaushik et al. 2009). Liver autophagy has also an important metabolic role in hepatic

gluconeogenesis. Amino acids produced by autophagy are converted to glucose in the liver which maintains correct blood glucose levels (Ezaki, Matsumoto et al. 2011).

2.5.4. Autophagy in the brain

The role of basal autophagy in neurons was firstly reported in neural specific Atg5 KO mice which developed neurodegenerative symptoms and ubiquitin-positive inclusion bodies in their neurons without the expression of any mutant proteins responsible for neurodegenerative diseases (Hara, Nakamura et al. 2006). The turnover of intracellular proteins by autophagy is important for neuronal cells which require constitutive nutrient supply. Autophagic response in the hypothalamic neurons, specifically lipophagy generates free fatty acids that increase the agouti-related peptide (AgRP) expression which then regulates neuropeptide levels. This response is responsible for the increase of food intake. Autophagy has a role in central nervous system as the conditional loss of Atg7 lead to behavioral defects as well as a massive neuronal loss and accumulation of poly-ubiquitinated proteins. (Komatsu, Waguri et al. 2006). The physiological role of autophagy in neuronal homeostasis extends to its role during pathogenesis of various neurodegenerative diseases such as Huntington’s disease.

2.5.5. Autophagy in the muscle

Skeletal muscle also metabolically active tissue that covers 40% of the body mass is a source for amino acids, especially during catabolic processes, which is used by the liver for glucose production (Lecker, Goldberg et al. 2006; Masiero, Agatea et al. 2009). In the muscle both autophagy and ubiquitin proteasome systems have important functions (Lecker, Goldberg et al. 2006). Muscle specific Atg7 deficient mice showed muscle atrophy, weakness caused by an age-dependent impairment of force along with an accumulation of abnormal mitochondria, sarcoplasmic reticulum distention, sarcomere disorganization and aberrant concentric membranous structures. This phenotype was very similar for muscle specific deletion of Atg5 (Raben, Hill et al. 2008). The inhibitions of basal autophagy lead to an increase of a compensatory proteasome activity and apoptosis leading to muscle loss. These results conclude that the physiological role of autophagy in the skeletal muscle is to control

muscle mass and protein quality control to maintain myofiber integrity (Masiero, Agatea et al. 2009). Recent finding by conditional knock-out of Atg7 in skeletal muscle during physical exercise revealed that autophagy is an adaptive response to exercise to preserve mitochondrial function by controlling proper organelle turnover during damaging contraction (LoVerso, Carnio et al. 2014).

The phenotype of disorganized sarcomere structures and mitochondrial aggregation in skeletal muscles with defective autophagy was observed in cardiac specific Atg5 deficient mice as well (Nakai, Yamaguchi et al. 2007). Autophagy inhibition leads to cardiomyopathy, left ventricular dilatation, contractile dysfunction and an accumulation of poly-ubiquitinated proteins and ER stress in cardiomyocytes (Tanaka, Guhde et al. 2000; Nakai, Yamaguchi et al. 2007). These observations reveal the role of autophagy as a cytoprotective mechanism for maintaining cardiomyocyte function as well as an adaptive response to stress such as pressure overload or ischemia/reperfusion (Nishida, Kyoi et al. 2009).

2.5.6. Autophagy in the intestine

The cytoprotective role of autophagy extends to intestine as well. ATGL1 is identified as a risk gene for the development of Crohn’s disease as the loss of the Atg16L1 or Atg5 in vivo or in vitro leads to abnormal exocytosis, degenerating mitochondria, loss of granule secretion as well as defective innate immune response specifically for Paneth cells, which are specialized intestinal epithelial cells (Cadwell, Liu et al. 2008; Adolph, Tomczak et al. 2013). This phenotype was very similar to patients of Crohn’s disease. Xenophagy is especially important for the intestinal epithelia as it functions as a barrier against dietary and microbial antigens (Randall-Demllo, Chieppa et al. 2013). Autophagy is also responsible for the breakdown of triglycerides for lipid supply in enterocytes upon lipid uptake by food consumption in order to avoid massive accumulation of lipid droplets in the cell (Khaldoun, Emond-Boisjoly et al. 2013).

2.5.7. Autophagy in the pancreas

Basal autophagy has a similar cytoprotective role in the pancreas as in the heart and liver. Autophagy deficiency in the pancreatic β cells leads to accumulation of organelles and

p62, dysfunction of islets and impaired glucose tolerance along with decreased β cell mass and insulin content (Ebato, Uchida et al. 2008; Jung, Chung et al. 2008). The low level of constitutive autophagy in the pancreas is important for the maintenance of functional islets by degradation of potentially toxic cytoplasmic material such as damaged mitochondria which is important during the pathogenesis of insulin resistance (Ebato, Uchida et al. 2008; Fujitani, Ueno et al. 2010).

2.5.8. Autophagy in the lung

Constitutive autophagy is also responsible for the lung homeostasis, specifically bronchial epithelial cells. The inhibition of autophagy in the lung epithelia causes hyper responsiveness to cholinergic stimulus and p62 accumulation (Inoue, Kubo et al. 2011). Autophagy also participates to the pathogenesis of lung diseases such as cystic fibrosis (Mizushima and Komatsu 2011).

Autophagy is also shown to function in different lineages of differentiation including adipocytes, erythrocytes, T cells and B cells, mainly for the elimination of mitochondria (Mizushima and Levine 2010). The role of autophagy in the kidney physiology will be addressed further in detail.

2.5.9. Autophagy in innate and adaptive immunity

Autophagy, as a cellular defense mechanism, responds to infection by degrading intracellular pathogens such as bacteria or virus. Xenophagy, an intracellular process that traps pathogens into autophagosomes, functions as a strict border control for outsiders (Levine 2005). Autophagy can be induced upon viral infection and contribute to innate immunity by selective targeting of microorganisms. Various bacteria following infection via endocytosis (Streptococcus pyogenes), phagocytosis (Mycobacterium tuberculosis) are targeted to autophagosomes via bacterial receptors such as CD46 or interferon (IFN)-γ receptor, or via recognition of ubiquitinated bacterial proteins by autophagy receptors such as p62, NDP52 or optineurin (Vergne, Singh et al. 2006; Dupont and Lafont 2009; Joubert, Meiffren et al. 2009; Watson, Manzanillo et al. 2012). Certain bacteria can escape autophagic degradation by escaping p62 recognition. Viral recognition is regulated by pattern recognition receptors

(PRRs) such as toll-like receptors (TLRs) which can trigger autophagy (Richetta and Faure 2013). Following viral infection in plasmacytoid dendritic cells, autophagy positively promotes antiviral IFN-I synthesis (Lee, Lund et al. 2007). It is also shown that autophagy is required for IFN-γ response against murine norovirus infection. In contrast, different viruses can inhibit or hijack autophagy to interfere with the antiviral innate immunity. For example by targeting a NF-κB subunit for autophagic degradation and inhibiting the inflammatory cytokine synthesis, by inducing protein aggregation or by down regulation of inflammasome response (Richetta and Faure 2013).

Autophagy participates to the adaptive immunity as well by delivering of antigens to major histocompatibility complex (MHC) class II compartments in MHC class II positive cells such as dendritic cells, B lymphocytes and epithelial cells. (Paludan, Schmid et al. 2005). Autophagy contributes to antigen presentation by MHC class I in macrophages to CD8+ T cells cells upon viral infection (English, Chemali et al. 2009). The contribution of autophagy to the adaptive immunity extends to its role during development, survival and proliferation of B and T cells as Atg5-/- mice showed inefficient B and T cell development and proliferation (Pua, Dzhagalov et al. 2007; Miller, Zhao et al. 2008). Recently, it is also been shown that CMA contribute to the adaptive immunity by maintaining T cell activation via selective degradation of negative regulators. This CMA-mediated T cell activation regulation declines with age upon loss of LAMP2-A (Valdor, Mocholi et al. 2014).

2.5.10. Pathophysiological role of autophagy

As described above, constitutive autophagy contributes to homeostasis of various tissues in the organism and the alteration of this homeostatic control leads to altered protein quality control, organelle turnover, DNA damage and disrupted metabolic homeostasis as a result. Various systemic or organ specific diseases are caused by autophagy impairment such as cancer, metabolic dysfunctions, vascular instability, neurodegeneration, cardiomyopathies, non-alcoholic fatty liver disease and autophagy (Ravikumar, Sarkar et al. 2010; Schneider and Cuervo 2014) (Figure 14).

Mutations of autophagic genes are observed to be involved during the development of different diseases such as Parkinson’s, Paget disease, amyotrophis lateral sclerosis (ALS), Crohn’s, Danon’s disease, asthma or systemic lupus erythematosus. These genetic variations and their pathological consequences in the organisms are reviewed by Schneider and Cuervo (Schneider and Cuervo 2014) (Figure 14). Neurodegenerative diseases and the impairment of autophagy is review by Ravikumar and colleagues (Ravikumar, Sarkar et al. 2010). Different mutations and polymorphisms associated with the development of several human diseases are listed in Table 2 in detail.

Impaired CMA, caused by defective substrate targeting and translocation across the lysosomal membrane can result in pathologies as well. The crosstalk between CMA and Huntington’s disease, Parkinson’s disease and taupathies are extensively studied and reviewed (Cuervo and Wong 2014).

Autophagy is known to play an important role in various stages of tumorigenesis. It can function as a tumor suppressor or tumor-promoting. Accordingly, autophagy is regulated positively by tumor suppressors and negatively by oncogenes. The cyto-protective role of constitutive autophagy can prevent cancer initation by limiting ROS production, DNA damage and genomic instability. However, autophagy can be beneficial for tumor progression as it serves to overcome metabolic stress or cytotoxicity induced by chemotherapy (Lorin, Hamai et al. 2013). Autophagy can also have another important role during tumor development by promoting self-renewal and maintenance of cancer stem cells and the targeting of autophagy/lysosomal pathway has been revealed to be very promoter to induce the death of this population (Gong, Bauvy et al. 2013; Yue, Hamai et al. 2013).

TABLE 2. MUTATIONS AND POLYMORPHISMS IN AUTOPHAGY GENES LINKED TO HUMAN DISEASES AUTOPHAGOSOME FORMATION ATG2B, ATG9 Frameshift mutations in gastric and colorectal cancers with microsatellite instability Genetic polymorphisms associated with increased risk of asthma and increased risk of ATG5 systemic lupus erythematosus ATG7 Genetic polymorphism (V471A) associated with earlier onset of Huntington’s disease Genetic polymorphism (T300A) associated with increased risk of Crohn’s disease,impaired intestinal dendritic cell antigen sampling and processing and more ATG16L1 aggressive clinical course, increased risk of colorectal cancer, and increased susceptibility to H. pylori infection, and increased risk of COPD Monoallelic deletion associated with risk of breast, ovarian, and prostate cancer (and BECN1 decreased expression associated with poor prognosis of multiple cancers) EI24/PIG8 Mutations and deletions associated with early onset breast cancers Frameshift mutation associated with autosomal recessive form of hereditary spastic TECPR2 paraparesis WDR45/ Heterozygous mutations associated with static encephalopathy of childhood with WIPI4 neurodegeneration in adulthood AUTOPHAGOSOME MATURATION AND DEGRADATION Mutations that impair autophagosome maturation are associated with frontotemporal CHMP2B dementia and amyotropic lateral sclerosis Mutation that impairs autophagolysosomal degradation of micronuclei causes autosomal CHMP4B dominant posterior polar cataract Mutations that impair autophagosome movement are associated with motor neuron Dynein disease EPG5 Autosomal recessive mutations cause the multisystems disorder, Vici Syndrome Mutations that impair autophagolysosomal fusion associated with distal hereditary motor HspB8 neuropathy type II LAMP2 X-linked deletion associated with Danon’s cardiomyopathy UVRAG Deletion mutation associated with colorectal cancer Mutations that impair autophagosome maturation cause a multisystem disease consisting VCP/p97 of inclusion body myopathy, Paget’s disease of the bone, and frontotemporal dementia ZFYVE26 (spastizin) Autosomal recessive mutations cause hereditary spastic paraparesis type 15 SELECTIVE AUTOPHAGY Mutations associated with autosomal recessive or sporadic early-onset Parkinsons’s PARK2/P disease and with colon, lung, arkin and brain cancers ; genetic polymorphisms associated with risk of M. leprae, S. typhi and S. paratyphi infection PARK6/PI Mutations associated with autosomal recessive or sporadic early-onset Parkinson’s NK1 disease SQSTM1/ Mutations associated with Paget’s disease of the bone, amyotrophic lateral sclerosis, And p62 frontotemporal lobar degeneration SMURF1 Genetic polymorphism associated with increased risk of ulcerative colitis

Table 2: Mutations in Autophagy Genes and associated human diseases

(Modified from Levine, Packer and Codogno. Development of autophagy inducers in Clinical Medicine. J Clin Invest, in press)

Beclin1 is a two tumor suppressor autophagy regulator. BECN1, a haplo-insufficient tumor suppressor gene, is mono-allelically deleted in 40–75% of sporadic breast, ovarian and prostate cancers. The pro-apoptotic tumor suppressor p53 is mutated in more than 50% human cancers and has been correlated with metastasis and poor prognosis (Lorin, Hamai et al. 2013). P53 has been demonstrated to have a dual role in autophagy depending on its subcellular localization. In the nucleus, p53 functions as a pro-autophagic factor in a transcription-dependent or independent manner. The cytoplasmic pool of p53 has been described to inhibit autophagy (Tasdemir, Maiuri et al. 2008). Importantly, the autophagy cargo receptors (as described above) are themselves degraded during selective autophagy. Autophagy inhibition can lead to their accumulation these cargo receptors and in aberrant regulation of their downstream pathways promoting tumorigenesis. This has been illustrated by studies demonstrating tumor-promoting functions for the autophagy cargo receptor p62/sequestosome 1 (SQSTM1). On the other hand, the RAS oncogene, a member of a small GTPase family, is shown to induce autophagy. It regulates cell growth and survival which are increased significantly in cancer progression. Autophagy induced by the active oncogenic RAS, can have two consequences: it can limit oncogenesis via Beclin1 upregulation resulting in cell death or in contrast, it can promote oncogenic transformation via improved cell metabolism. Indeed, Ras activated autophagy can result in senescene induction which limits the proliferation of tumor cells. However, Ras-driven tumors are described to be addicted to autophagy for maintenance of cancer cell homeostasis and elimination of damaged mitochondria (Guo, Chen et al. 2011). As autophagy is reported to have a pro-sruvival role in oncogene-driven tumors, it is an attractive therapeutic target (Mathew and White 2011). There are currently several clinical trials using the lysomotropic agent autophagy inhibitor hydroxychloroquine. Chloroquine is reported to reduce tumor growth however this agent presents limitations on the level of target specificity and shown to improve tumor milieu. So other autophagy targeting agents are being investigated (Mathew and White 2011).

In the objective of this thesis, the relation between autophagy and pathophysiology will be detailed mainly for kidney diseases.

Figure 14: Autophagy and Pathophysiology

Autophagy-related diseases are presented as organ-specific and systemic human diseases. Red dots indicate primary autophagy defects and green dots indicate autophagy changes secondary or reactive to disease. Mutations (m), polymorphisms (p) and haplo-insufficiencies (h) of the indicated autophagy-related protein in the respective diseases are indicated in blue. (Schneider and Cuervo 2014)

2.6. Autophagy in Kidney Physiology and Pathophysiology

2.6.1. Physiology of the Kidney: An Introduction

Mammalian kidney is a highly specialized and architecturally precise organ which coordinates the development of specific cell types with different functions in the kidney network. The special architecture of the kidney is developed to tightly regulate body fluid and filtration of urine. Kidney is evolved to adapt to conserve water, excrete waste, and maintain electrolyte homeostasis within a variety of challenging environments. It is a model organ to study epithelial-mesenchymal interactions, epithelial cell polarization, and branching morphogenesis (Dressler 2006). It plays a fundamental role in maintaining body salt and fluid balance and blood pressure homeostasis through urine filtration across the nephrons

A nephron is the functional unit that filters blood, reabsorbs the filtered electrolytes, solutes and fluid, and excretes wastes and excessive electrolytes and water. Each nephron consists of a renal corpuscle including the glomerulus, which contains a network of capillaries and podocytes, and a tubule unit including proximal tubule, loop of Henle, distal tubule, connecting tubule, and perhaps the collecting duct (Figure 15). As mentioned earlier in the manuscript, cells lining the nephron are continuously exposed to laminar fluid flow and fluid shear stress (FSS) during filtration. This FSS corresponds to 1 dyn/cm2 or smaller in physiological conditions. Flow sensing is maintained by primary cilia and microvilli on the epithelial cells via various channels and signaling pathways described in the first chapter of the manuscript (Weinbaum, Duan et al. 2010).

The glomerulus is exclusively responsible for filtering blood. The tubules of the nephron are specialized for reabsorbing 99% of glomerularly filtered electrolytes and water. Proximal tubules are responsible for reabsorbing approximately 65% of filtered amino acids, solutes, low molecular weight proteins, 80% of the filtered bicarbonates and all filtered glucose. The tubular segments of the nephron where the polarized epithelial cells function to filtrate various molecules via different mechanisms are highly functionally heterogenic in the expression and distribution of biological enzymes, aminopeptidases, sodium and hydrogen exchangers or antiporters such as sodium and hydrogen exchanger 3 (NHE3), Na+-K+- ATPase, Na+/HCO3, organic solutes and glucose, different G protein-coupled receptors (GPCRs) (Zhuo and Li 2013).

Figure 15: Physiology of the nephron

Each nephron is composed of a glomerulus which includes podocytes and a network of capillaries, and a tubule unit including proximal tubule, loop of Henle, distal tubule, connecting tubule, and the collecting duct. The tubules are specialized for reabsorbing filtered electrolytes and water, with different channel concentrations depending on the region (proximal or distal). (Dressler 2006)

Role of Primary Cilia in the Transduction of Fluid Flow in the Kidney

In the kidney, renal epithelial cells are under laminar urinary flow which applies a fluid shear stress (FSS) on the epithelial cells lining the tubules in the nephrons. Fluid shear stress created by laminar flow can be calculated via the equation: τ = 6µQ/bh2 where “Q” is flow rate, µ is medium viscosity, and “b” and “h” are channel width and height, respectively (Weinbaum, Duan et al. 2010). Kidney FSS is around or smaller than 1 dyn/cm2 and is responsible for several phenotypes in the cells such as cytoskeletal re-arrangement (Essig, Terzi et al. 2001), cellular differentiation and signalizations during filtration such as Na+- K+ transport or Ca2+ signaling (Gonzalez, Essig et al. 2013). In the nephron, the “brush border” formed of densely packed actin filament based microvilli is firstly identified as a mechanical sensor, responsible for glomerulotubular balance via Na+ transport upon microvilli bending in the proximal tubules (Du, Duan et al. 2004). In 1997, Schwartz et al. proposed a mechanosensory role of primary cilia (PC) in kidney epithelial cells upon bending induced by FSS (Schwartz, Leonard et al. 1997). This role was developed further by the study of Praetorius and Spring in MDCK cells which are derived from the collecting duct epithelia of canine kidney as they showed the PC-dependent induction of calcium signaling upon fluid

flow in vitro (Praetorius and Spring 2003). Bending of the PC is followed by the influx of calcium (Praetorius and Spring 2001). The identification of two mutated genes in the common ciliopathy polycystic kidney disease (PKD), PKD1 and PKD2 encoding the ciliary calcium channels PC1 and PC2 respectively, further highlighted the possible relevance of PC in the kidney epithelia. The extracellular domain of PC1 functions as a sensor of flow and transduce the mechanical stimuli to the associated calcium permeable cation channel PC2 which induces a calcium influx in the cilium and the subsequent Ca2+ induced Ca2+ release in the cell. The disruption of this PC1/PC2 dependent flow mediated calcium signaling leads to cystic phenotype in vivo (Nauli, Alenghat et al. 2003). PC2 can also function as a mechanosensor channel in the PC in endothelial cells. Other possible mechanical receptor-like channels are localized to the cilium such as TRPV4, a Ca2+ permeable ion channel which interacts with PC2 and mediates flow-induced calcium influx as well as osmolality. However the TRPV4 mutated mice do not develop cystic kidney (Wu, Gao et al. 2007). As mentioned earlier in the manuscript, PC has a role in planar cell polarity (PCP) during development. However, under static conditions ciliated renal epithelial cells show random orientation of mitotic spindles whereas under fluid flow MDCK cells show polarization in vitro suggesting that PC translates flow-induced anterior posterior orientation of kidney cells in vivo (Kotsis, Nitschke et al. 2008).

Primary Cilia and Polycystic Kidney Disease

The polycystic kidney models are important for the study of the role of PC in the kidney, under urine flow. PKD is a genetic disorder that causes formation of fluid-filled cysts in the renal tubules (Chapin and Caplan 2010). Autosomal dominant polycystic kidney disease (ADPKD), the most common form of PKD, is due to mutations in two genes: PKD1 (around 85% of cases) and PKD2 (around 15% of cases) that encode PC1 and PC2 ciliary proteins, respectively (Takiar and Caplan 2011). The polycystins have different roles in signaling, cell-to-cell and cell-to-matrix interactions, and mechanosensation (Chapin and Caplan 2010). ADPKD leads to kidney enlargement with increasing numbers and sizes of cysts over time due to abnormal renal epithelial cell growth and impaired fluid transport resulting in end-stage renal disease in 50% of patients (Figure 16). The role of primary cilia is complex in PKD patients and in rodent models. It has been shown that mutations in various ciliary proteins, including IFT88 (also known as Tg737, Polaris, or Orpk), IFT20, IFT140,

and a subunit of kinesin-II Kif3-α, interfere with cilia structure and can result in cyst formation. Mutations of PC1 and PC2, however, result in functional defects in cilium- dependent signaling: defective mechanosensory properties decrease intracellular Ca2+ and reduce intracellular cAMP clearance (Cowley 2008). Mice with null alleles of Tg737 die during mid-gestation. Mice with a hypomorphic mutation in the gene encoding IFT88 (Tg737) show a very similar phenotype to autosomal recessive PKD with cystic kidneys, hepatic biliary disease, large cysts in the collecting duct, and inability to concentrate urine and have short and bulged form of PC in their kidney epithelia (Pazour, Dickert et al. 2000; Cowley 2008). The connection between mechano-sensory role of cilia and the development of PKD is still under investigation. It was previously reported that mTOR signaling is hyper- activated in PKD, and rapamycin was identified as an effective therapeutic agent against cystogenesis in rat and mice PKD models; however, rapamycin (everolimus) was not a successful treatment in clinical trials with PKD patients (Huber, Edelstein et al. 2012; Ravichandran and Edelstein 2014; Wang and Choi 2014). The regulation of the mTOR pathway is crucial during cystic development. Epithelial cells lining the cysts expressing mutated PC1 protein have increased levels of cell proliferation and abnormally activated mTOR signaling due to a disrupted interaction between PC1 and tuberin, a product of TSC2 gene, that controls mTOR activity (Shillingford, Murcia et al. 2006). Boehlke et al. (Boehlke, Kotsis et al. 2010) showed that in kidney epithelial cells, the AMP-activated protein kinase (AMPK) pathway controls cell volume in an mTOR dependent manner upon Lkb1 recruitment to the axoneme under fluid flow .This regulation was demonstrated to independent of calcium signaling via PC2 in the cilia axoneme. In PKD models the renal epithelial cells lining the cysts are larger and often have impaired cilia (Grantham, Geiser et al. 1987; Brown and Murcia 2003) (Figure 16), suggesting that PC has an important regulatory role for cell size regulation in the kidney tubules. This role remains to be discovered in detail and will be addressed further by our study.

Figure 16: Polycystic Kidney Disease and Fluid Flow in the Kidney

(A)Cyst formation at the level of the cell, nephron, and kidney. Defects in the genes encoding PC1 or PC2 lead to aberrant gene transcription, cell proliferation, and ion secretion, which in turn result in the formation of fluid-filled cysts resulting in the kidney dysfunction. (Chapin and Caplan 2010) (B) Cells lining the cysts are observed to be larger. (Grantham, Geiser et al. 1987) (C) A proposed model of cilia dependent cell size regulation in kidney cells under flow. (Wiczer, Kalender et al. 2010)

2.6.2. Autophagy in the Kidney: Health and Disease

Autophagy is involved in various pathologies in different organs, the kidney is no exception. There is growing information on the role of autophagy in various kidney pathologies such as acute kidney injury (AKI), diabetic nephropathy (DN) and polycystic kidney disease (PKD). Depending on the pathology autophagy can be up- or down-regulated suggesting a complex role of autophagy in this highly specialized organ.

Podocytes are identified to be very vulnerable for renal pathologies and during kidney aging as they cannot be self-renewed. The glomerular podocytes are observed to have a high basal autophagy especially in late developmental stages and mature podocytes. Podocyte specific Atg5 KO mice developed proteinuria, impaired organelle turnover in podocytes and increased compensatory proteasomal machinery which declined with age. Along with the data that autophagy was not required for glomerular development, it can be concluded that basal autophagic activity is mainly important in differenciated podocytes to avoid degeneration and maintain homeostasis (Hartleben, Godel et al. 2010). In kidney tubules, basal autophagy has also a homeostatic role as the autophagy deficient tubules develop deformed mitochondria and aggregate inclusions leading to tubular cell hypertrophy and degeneration over time and finally kidney dysfuntion and (Kimura, Takabatake et al. 2011; Liu, Hartleben et al. 2012).

Acute Kidney Injury

Acute Kidney Injury (AKI) is defined by rapid loss of kidney function and occur due to ischemia/reperfusion injury during kidney transplantation, sepsis or nephrotoxins. AKI can be mortal in a large number of hospitalized patients by 40-80% or can develop into chronic kidney disease (Huber, Edelstein et al. 2012; Wang and Choi 2013). It is characterized by the decline in glomerular filtration, accumulation of nitrogenious waste (urea and nitrogen), perturbation of extracellular fluid volume and electrolyte and cause renal tubular cell injury and death followed by kidney dysfunction and dysfunction of the graph in transplant patients. There is an increasing number of studies on the autophagic activity induction during AKI in vitro, in animal models and biopsies from patients which show an increased number of autophagosomes in tubule cells accompanied by increase in autophagic flux and increase in the expression of autophagic proteins (Huber, Edelstein et al. 2012). Continuous supply of oxygen is vital for organs and cells and it is known that autophagy respond to low oxygen levels in response to hypoxia or ischemic conditions as a cytoprotective mechanism to maintain ATP production. The role of autophagy in hypoxic heart and brain tissue conditions is already known. For the heart, following myocardial infarctus low oxygen or increased oxidative stress activates the transcription factor hypoxia inducible factor 1α (HIF-1α) which induces glycolytic pathway, inhibiting the TCA cycle and fatty acid oxidation which promotes ATP preservation. HIF-1α can also activate BNIP3 and leading to the dissociation of Bcl-2- Beclin1 complex or binding of Rheb and inhibiting mTOR activity (Zhang and Semenza

2008). Following ischemic cerebral stroke accumulation of autophagosomes, autolysosomes and Beclin1 expression is observed along with CMA markers (Nishida, Kyoi et al. 2009), (Dohi, Tanaka et al. 2012).

In the rat kidney tubules, ischemia-reperfusion induced LC3 lipidation and Beclin 1 expression (Wu, Hsiao et al. 2009). In human and mouse kidney epithelial cells in vitro, hypoxia induced LC3 and LAMP-2 positive cells suggesting an increase of autophagic activity (Suzuki, Isaka et al. 2008). In cultured human epithelial cells, ROS induced cell death could be inhibited by addition of autophagy inhibitor 3-MA and silencing of Atg7. These findings are in parallel with the effect of autophagy inhibition on hypoxic cells. Autophagy inhibition by different methods such as 3-methyladenine (3-MA) treatment, silencing of Beclin1 or leads to the inhibition of hypoxia induced cell death by autophagy inhibition. In vivo inhibition of autophagy by systemic 3-MA or chloroquine induced renal ischemia/reperfusion injury in mice concluding that autophagy activation function as a cytoprotective mechanism in face of kidney injury (Jiang, Liu et al. 2010). Nephrotoxins such as cisplatin, cyclosporine and environmental toxins contribute to AKI as well. In nephrotoxic models of AKI in vitro, an increase in autophagosomes and autolysosome in proximal tubule cells prior to apoptotic response is observed (Inoue, Kuwana et al. 2010). Although the data on AKI models suggest a cytoprotective role for autophagy during ischemic injury, the exact mechanism by which autophagy is induced and the balance between autophagy induction and tubular cell death remains to be elucidated (Huber, Edelstein et al. 2012).

Diabetic Nephropathy

Diabetic nephropathy (DN) is a complication resulting in type I and II diabetic patients and can cause end stage renal disease (ESRD). DN phenotype includes glomerular membrane thickening, mesengial expansion, podocyte loss, proteinuria. Injuries to podocytes can have important consequences in the development in DN. Interestingly in DN podocytes there is high mTORC1 and reduced AMPK activity. Systemic treatment of type I and II diabetic mice with rapamycin attenuated DN development while podocyte specific inhibition of mTORC1 results in a DN phenotype in mouse kidney (Lee, Feliers et al. 2007; Inoki, Mori et al. 2011). Podocyte specific Atg5 KO mice also develop a DN-like phenotype with loss of podocytes and glomerulosclerosis in aged mice (Hartleben, Godel et al. 2010). In diabetic rats, the number of autophagosomes is decreased in kidney tubule cells. This observation is in line

with the role of impaired autophagy and subsequent excessive mitochondrial ROS induced oxidative stress during the development of DN (Godel, Hartleben et al. 2011). Autophagy inducers are good candidates for treatment of DN by reducing oxidative stress as observed by the autophagy inducer SIRT1 (Wu, Zhang et al. 2012).

Renal fibrosis

Renal fibrosis is a hallmark of progressive chronic kidney disease (CKD) and ESRD resulting in tubular degeneration and atrophy which ends with kidney failure (Grabias and Konstantopoulos 2013). Unilateral ureteral obstruction (UUO), a progressive renal fibrosis model in rodents, induces TGF-β1 secretion, ROS production and the resulting mitochondrial damage (Xu, Ruan et al. 2013). UUO induced LC3-II lipidation and increase in autophagic vacuoles rapidly which decreased over time before the cells die with apoptosis. They also have a decrease and then an induction of mTOR and AKT activity over time following UUO. Inhibition of autophagy by 3-MA in UUO models increased tubular cell death ratio in the obstructed kidney (Kim, Nam et al. 2011). These observations suggest that upon obstruction, autophagy is activated as a cytoprotective mechanism following mitochondrial damage however fails to do so. Interestingly, in the contralateral kidney which compensates the renal function and cell growth, an induction of autophagic activity is observed and maintained over time, which shows an increase of the activity for constitutive autophagy in the compensatory kidney (Kim, Nam et al. 2011).

Polycystic Kidney Disease

The genetic background of the PKD is described earlier in the manuscript as the development depends on the ciliary proteins (Figure 6). ADPKD results in kidney enlargement and dysfunction with increasing numbers and sizes of cysts over time due to abnormal renal epithelial cell growth and disturbed fluid transport along the tubules. It can lead to end-stage renal disease in 50% of patients and can present its symptoms with a late onset in adult patients. This can be explained by transient obstructive or ischemic injuries to the tubule epithelial cells. PC1 and PC2 are critical for cellular repair and controlled cell growth and division of tubule cells after kidney injury, thus explaining the injury sensitivity of ADPKD kidneys in rodent models (Bastos, Piontek et al. 2009; Prasad, McDaid et al. 2009).

There is an increasing interest in the role of autophagy during PKD development, however a clear molecular connection between autophagy and PKD has not yet been defined (Wang and Choi 2013). In congenital PKD mouse models, a blockage of autophagic flux is observed in early stages along with an increase of Beclin1 expression in late stages of the disease; these features suggest a suppressed autophagic activity (Ravichandran and Edelstein 2014). The role of this reduction in cyst formation may be connected to the decreased senescence and induction of cell proliferation and apoptosis, which are previously defined features in cells lining the cysts. Several hypotheses are being discussed for the role of autophagy in PKD in light of different effects of pharmacological agents used in PKD models and trials with autophagy modulators such as the mTORC1 inhibitor rapamycin and other autophagy inducers. It was previously reported that mTOR signaling is activated in cystic epithelia of PKD patients and mouse models, and rapamycin was identified as an effective therapeutic agent against cystogenesis in rodent PKD models; however, rapamycin was not a successful treatment in clinical trials with PKD patients while another mTOR inhibitor everolimus decreased kidney volume but did not reverse kidney dysfunction (Walz, Budde et al. 2010). The regulation of the mTOR pathway seems to be crucial during cystic development. Epithelial cells lining the cysts are observed to be larger in size. In addition the cells expressing mutated PC1 protein have increased levels of cell proliferation and abnormally activated mTOR signaling due to a disrupted interaction between PC1 and TSC2, which controls mTOR activity (Shillingford, Murcia et al. 2006). The study by Boehlke and colleagues showed that in kidney epithelial cells, the localization of Lkb1 on the axoneme upon fluid flow activates AMPK, inhibits mTORC1 and controls cell volume and this process is reduced in non-ciliated mutant cells (Figure 6) (Boehlke, Kotsis et al. 2010). In non- ciliated cystic renal epithelial cells or cell lines with mutations in ciliogenesis-related proteins, the regulation of cell volume is disrupted. The crosstalk between cilia-dependent cell size regulation and autophagy will be addressed by our study.

Interestingly, Belibi et al. showed that the presence of autophagosomes in hypoxic areas caused by the cysts in different rodent models of PKD is due to a blockage of autophagic flux (Belibi, Zafar et al. 2011). In addition, mouse kidney cells that express mutant PKD1 failed to induce autophagy in response to glucose starvation (Rowe, Chiaravalli et al. 2013). These studies suggest that autophagy is impaired in PKD patients either due to a reduction of autophagosome biogenesis or due to impaired autophagic flux (Ravichandran and Edelstein 2014). Data from a recent study imply that autophagy plays a role in PKD by

altering ciliary protein turnover by degrading PC2 (Cebotaru, Cebotaru et al. 2014). In support of this hypothesis, the expression level of PC1 is critical for regulation of normal tubule formation in cultured cells and mouse models (Chapin and Caplan 2010). PC1 overexpression is observed in cystic epithelia of ADPKD patients (Ong, Harris et al. 1999), and its overexpression disrupts tubule formation in renal epithelial cells in mice (Patel, Hajarnis et al. 2012). In contrast, the down-regulation of PC1 induces cystogenesis in mouse models (Lantinga-van Leeuwen, Dauwerse et al. 2004). A major conclusion of these studies is that the dosage of PKD1 expression is important for cystogenesis and autophagy might play an important role for the turnover of the protein. Cebotaru et al. showed that PC1 regulates the degradation of PC2 by autophagy through a mechanism dependent on histone deacetylase- 6 (HDAC6). HDAC6 has an important role in selective autophagy as described earlier. Interestingly a pathogenic mutant of PC1 failed to induce autophagy-dependent PC2 degradation. Together these findings suggest that autophagy must be considered as a factor in the development of renal ciliopathies.

AIMS OF THE STUDY

II. AIMS OF THE STUDY

The dialog between the cell and its extracellular milieu is vital for the cell fate and the associated tissue. Primary cilia (PC), an organelle that evolved as a sensory antenna in most of the vertebrate cells may be the key to our understanding of the dialog of extracellular signals and the cell. PC integrates and transduces outside signals and triggers subsequent response pathways for the cell to adapt to incoming signals thus functions as an interface between extra- and intra-cellular milieu. Autophagy is an evolutionarily conserved adaptation mechanism that is vital for the cell homeostasis in response to various stress conditions in different organs and tissues. These stress conditions may include starvation, a well-known autophagic stimulus, or mechanical stress. A basic observation in relation to these two cellular mechanisms was based on the temporal coincidence of serum deprivation stimuli. Serum deprivation induces cilia growth starting in the first hours, up to 2 days in vitro. Autophagy is also induced upon serum starvation in the first hours and maintained during the full duration of starvation. The aim of this study was to establish a novel crosstalk between primary cilia (PC) and autophagy, in order to define the role of autophagy during ciliogenesis as well as analyze the role of PC in the regulation of autophagic activity in response to extracellular inputs.

The first part of our study asked several questions in order to determine the machinery between these two adaptive responses:

• Does autophagy activation upon serum deprivation depend on the presence of primary cilium? • Does primary cilium dependent signaling regulate autophagy? • Does autophagy have a role in cilia growth?

To answer these questions we used various ciliated cell culture models which were pharmacologically or genetically impaired for different ciliary proteins expressions or ciliary signalizations. IFT88 or IFT20 impaired cells are used to disrupt ciliogenesis. Hedgehog pathway or PDGF pathway were desregulated in vitro as well. To establish the role of autophagy in ciliogenesis different autophagy deficient mouse embryonic fibroblasts were used as models.

After establishing the functional interaction between PC and autophagy with the identification of the Hedgehog signaling mechanism and autophagic machinery behind the reciprocal role of the two cellular components in relation to nutrient deprivation, we continued our research to define the physiological function of the dialog between primary cilium and autophagy. In kidney tubules, epithelial cells are under constitutive fluid flow to filter urine. PC was shown to function as a sensory antenna in response to fluid flow to control various signaling pathways such including kidney epithelial cell size regulation which is critical for proper functioning of the kidney. . Thus the role of PC in kidney cells prompted us to further analyze the possible role of autophagy as a response to fluid flow in ciliated kidney epithelial cells. The second part of our study aimed to answer the following questions:

• Is autophagy activated in kidney epithelial cells under constant fluid flow? • Does this mechanical stress induced autophagy depend on the presence of primary cilia? • Does autophagy has a role in cell size regulation in kidney epithelial cells?

In order to answer these questions we used an in vitro assay where we applied constant laminar flow on different cell lines. We used MDCK cells that are derived from the distal tubule of the canine kidney and kidney epithelial cells (KEC) derived from wild type and IFT88 hypomorphic mice kidney and knock-downed various autophagy genes by lentiviral transfection.

MATERIALS AND METHODS

III. Material and Methods

Cell culture:

Cell Lines Provider

KEC WT Dr. G. Pazour (University of Massachusetts) KEC IFT88-/- (Tg737-/-)

MDCK

MDCK GFP-LC3 Dr. Isabelle Beau (INSERM UMRS 984)

MDCK RFP-GFP-LC3

MEF Dr. N. Mizushima (Tokyo University) MEF Atg5-/-

MEF Ptc1-/- Dr. M.P. Scott (Stanford University)

All cells were cultured at 37°C and 5%CO2 in Dulbecca’s Modified Eagle’s medium (GIBCO) supplemented with 10% fetal bovine serum (FBS).

Plasmids: myc tagged Gli1 and myc tagged Gli2 were provided from Addgen. The transfection was performed using Lipofectamine 2000 (Invitrogen). siRNA: Mouse siSMO (5’GUGGAGAAGAUCAAUCUA3’ and 5’AUAGAUUGAUCUUCUCCAC3’) were purchased from Eurogentec. The transfection was performed using Lipofectamine 2000 (Invitrogen).

Transgenic cell line: For inducible knockdown of Atg5 and Beclin1, KEC cells were transduced with lentivirus purchased from Sigma Mission Library (Clone ID

TRCN0000099431 and TRCN0000087290 respectively). In order to induce the target gene knock-down, the cells were cultured with Isopropyl β-D-1-thgalactopyranoside (IPTG) (5mM) for 72hours. The transfer vector plasmids containing the IFT20 and IFT88 shRNA sequence were purchased from Sigma Mission Library (Clone ID NM_018854.4- 417s21c1 and NM_009376.1-762s1c1).

Pharmacological agonists, inhibitors, and autophagy reagents:

Reagent Concentration Company

3-methyladenine 10mM Sigma

Bafilomycin A1 100 nM LC Laboratories

Chloroquine 50 µm Sigma

Cyclopamine 40 nm Sigma

Isopropyl β-D-1-thgalactopyranoside 5mM Sigma

Leupeptin 100 µm Fisher BioReagents

Purmorphamine 5 µm Calbiochem

Rapamycin 100 µm Calbiochem

Flow-chamber and imaging: MDCK or KEC cells were seeded into closed perfusion chambers (Microslide 0.6luer I channel dimensions 50 °— 5 °— 0.4 mm with iBiTreat from Ibidi), and cultured for at least 3 days to be polarized and induce maximum ciliogenesis. The medium was changed 2 times a day. The chamber was then connected to a computer- controlled setup containing an air-pressure pump and a two-way switching valve (Ibidi pump system 10902), pumping 20 ml of phenol-red-free cell- culture medium unidirectionally between two reservoirs through the flow channel at a rate corresponding to a shear stress of approximately 0.75–1 dyn per cm2.

Autophagic Measurements: All autophagic measurements and treatments were performed as described in (Dupont, Orhon et al. 2014). See Annexe for detailed methodology of autophagy.

Quantitative RT-PCR:

Total RNA was extracted from KEC and cDNA was synthetized using a Cells-to-Ct Kit (Applied Biosystems), according to the manufacturer’s instructions. Real time PCR was performed using SYBR Green Master Mix (Applied Biosystems), and products were detected on a Applied Biosystems ViiA 7™ Real-Time PCR System. The relative extent of P62, Gli1, Ptch expression was calculated using the 2(∆ ∆C(t)) method. Conditions for real time PCR were: initial denaturation for 10 min at 95°C, followed by amplification cycles with 15 s at 95°C and 1 min at 60°C.

Primer Sequence/Company

Mouse P62-F : GCCAGAGGAACAGATGGAGT

Mouse P62-R : TCCGATTCTGGCATCTGTAG

Mouse β-Actine-F GGCCAACCGTGAAAAGATGA

Mouse β-Actine-R ACCAGAGGCATACAGGGACAG

Mouse Gli1 MIX (F+R) Qiagen (QT00173537) Stock 10X

Mouse Gli2 MIX (F+R) Qiagen (QT01062236) Stock 10X

Mouse Ptch1 MIX (F+R) Qiagen (QT00149135) Stock 10X

Mouse Smo MIX (F+R) Qiagen (QT00111769) Stock 10X

Mouse Hhip MIX (F+R) Qiagen (QT02392292) Stock 10X

Mouse Bcl2 MIX (F+R) Qiagen (QT00147518) Stock 10X

Antibodies, immunoblotting, immunofluorescence:

Anti- Company cat no: ATG14 MBL #PD026 clathrin heavy chain BD Science #610500 acetylated-tubulin sigma Clone 6-11B-1 #T7451 actin Millipore Clone C4 MAB1501 AMBRA1 Novus #FLJ20294 Atg16L1 sigma T5326 Atg16L1 MCL #PM040 ATG5/12 Novus #NB110-53818 Atg7 Cell signaling #2631S Atg7 Serotec #AHP1651 beclin Novus #NB500-249 beta-tubulin sigma #T8328 cell mask Invitrogen #C10046 detyrosinated alpha-tubulin Abcam #Ab-48389 GABARAP Santa Cruz #sc-28938 Ift20 Proteintech #13615-1-AP Ift20 abcam #ab73110 Ift88 Proteintech #13967-1-AP Ift88 Novus #NB100-2471 LAMP-2 Iowa Hybridoma Bank LC3B Sigma #L7543 LC3B Cell signaling #2775 p62 Abnova #H00008878-MOI p70 S6K Cell signaling #9202 phospho- p70 S6K Cell signaling Th389 #L7543 SMO abcam #ab38686 ULK-1 sigma clone 2H8 #WH0008408M1 VPS15 Novus #pAb NBP1-30463 VPS34 Invitrogen #38-2100 γ-tubulin Santa cruz #sc-10732

Cells grown on microslide were fixed in 100% cold methanol, blocked, permeabilized and incubated with primary antibodies followed with fluorophore-conjugated secondary antibodies. Mounting medium contained DAPI (4′,6-diamidino-2-phenylindole) to highlight the nucleus were used.

Antibody was revealed with Super Signal West Dura chemilluminescent substrate (Pierce).

Imaging: Immunofluorescence conventional microscopy was carried out with a Zeiss ApoTome 2 piloted with Zen software 2012. Confocal microscopy was carried out using a confocal Leica TCS SP5 (LAS AF version 2.6.0) (laser wavelengths 488 nm, 561 nm and 633 nm). Quantification of the yellow and red puncta was performed by ImageJ software 1.47n (NIH) were used to quantify the. Data are from 3 independent experiments in which, each time, more than 300 basal body were analysed. Pearson’s colocalization coefficients between ATG16L1 and γ-tubulin were also derived using ImageJ software 1.47n (NIH). Pearson’s coefficient was from three independent experiments with five fields per experiment for a total of 15 fields contributing to the cumulative result.

3D reconstruction images were modeled as mixed renderings using Imaris Software for cell volume quantification following membraine staining using Cell Mask. For quantification of the number of ciliated cells and cell volume Z-stack sections were acquired and orthogonal views generated using AxioVision Rel. 4.8. Number of fluorescent puncta and recruitment of proteins to basal body (γ-tubulin positive structures) were quantified using ImageJ (NIH) after thresholding of the images. Quantification of co-localization was performed using Image J (NIH) in individual frames after thresholding, and co-localization was calculated with the JACoP plugin of the same program in merged images.

Statistical analyses: Statistical analysis was carried out in Excel (Microsoft Cooperation). Because of variability between cell lines of different genetic background, only genetically identical cell lines were compared statistically. P values were calculated by unpaired t-test from the mean data of each flow channel (n), representing ten visual fields each. All values are given as mean of n ± s.e.m. Results represent data from a minimum oh three independent experiments unless otherwise stated.

RESULTS

VI. RESULTS

1. Objective 1: Study of the functional interaction between autophagy and primary cilium

The objective of this collaborative study was to identify a previously unknown interplay between autophagy and the primary cilia and show that Hedgehog (Hh) signaling pathway is the regulator of crosstalk between the two cellular adaptive mechanisms upon serum deprivation.

• To determine the role of primary cilia on autophagy, different ciliated cell lines were used to observe autophagic activity upon inhibiting ciliogenesis by various methods in presence or absence of serum. Mouse embryonic fibroblasts (MEFs) were used as models to induce ciliogenesis by serum removal and to inhibit IFT machinery by knocking-down IFT20, an essential component of the IFT-B complex. Kidney epithelial cell lines generated from wild-type mice or mice with hypomorphic mutation of the IFT88 allele (IFT88-/-), another key component of the IFT-B complex, were also used. Our study led to the conclusion that autophagy induction upon serum removal is controlled by IFT-dependent ciliogenesis.

• To identify the signaling pathway that regulates autophagy activation by the PC we concentrated on the Hh pathway, one of the best characterized ciliary signaling pathways, which respond to starvation in different settings. We observed the activation of autophagy pathway upon induction of Hh signaling using pharmacological agonist or through constitituve activation in MEF cells lacking Ptc1 receptor (Ptc1-/-) which possess higher percentage of ciliogenesis. In order to define the ciliary-Hh dependency of autophagic activation we used different IFT protein knock-downs and pharmacological inhibition by antagonist cyclopamine. Rescue experiments by overexpression of Hh transcription factor Gli1 was used to establish autophagic induction in IFT88-/- cells.

• In order to identify the molecular mechanism of cilia-dependent autophagy induction in response to serum deprivation, various autophagy proteins were observed by

immunofluorescence at the axoneme or basal body (BB) of the PC in presence or absence of serum. While a number of core autophagic proteins were localized at the BB in presence of serum, induction of ciliogenesis and subsequent autophagy by serum removal recruited additional Atg proteins to the axoneme. The specifity of the starvation stimuli for the recruitment of autophagic machinery to the PC was observed in IFT88-/- cells.

• The observation of early autophagic marker Atg16L1 recruitment to the BB upon serum removal and Hh activation led to the search of the possible interaction of the protein with IFT20 in order to characterize the mechanism behind the recruitment of Atg16L1 and its dependence on IFT mechanism.

• In order to determine the role of autophagy in ciliogenesis, different autophagy deficient MEFs (Atg5-/-, Atg7-/- and Atg14-/-) were used in presence or absence of serum. Following the observation that the loss of autophagy alone was sufficient for ciliogenesis, we wanted to establish that the loss of autophagy did not impair ciliary function. We illustrated the intact Hh signalization at the cilia in autophagy deficient cells. To explain the inhibitory role of autophagy on ciliogenesis, we observed the degradation of ciliary components by autophagy, primarily IFT20 which is degraded by basal autophagy but not by induced autophagy. The differences of the length of the cilia between wild-type and autophagy deficient cells were used to identify the role of autophagy in regulation the cilia length since Atg5-/- cells presented longer cilia.

1.1. Results

ARTICLE doi:10.1038/nature12639

Functional interaction between autophagy and ciliogenesis

Olatz Pampliega1, Idil Orhon2,3, Bindi Patel1, Sunandini Sridhar1, Antonio Dı´az-Carretero1, Isabelle Beau3, Patrice Codogno2,3, Birgit H. Satir4, Peter Satir4 & Ana Maria Cuervo1,4,5

Nutrient deprivation is a stimulus shared by both autophagy and the formation of primary cilia. The recently discovered role of primary cilia in nutrient sensing and signalling motivated us to explore the possible functional interactions between this signalling hub and autophagy. Here we show that part of the molecular machinery involved in ciliogenesis also participates in the early steps of the autophagic process. Signalling from the cilia, such as that from the Hedgehog pathway, induces autophagy by acting directly on essential autophagy-related proteins strategically located in the base of the cilium by ciliary trafficking proteins. Whereas abrogation of ciliogenesis partially inhibits autophagy, blockage of autophagy enhances primary cilia growth and cilia-associated signalling during normal nutritional conditions. We propose that basal autophagy regulates ciliary growth through the degradation of proteins required for intraflagellar transport. Compromised ability to activate the autophagic response may underlie some common ciliopathies.

Autophagy is a cellular catabolic process that contributes to quality enhanced and cilia grow longer. We conclude that ciliary signalling control and maintenance of the cellular energetic balance through the pathways, such as Hh, may specify a cilia-mediated autophagy, closely turnover of proteins and organelles in lysosomes1. Induction of autop- related to autophagosome assembly at the plasma membrane and that, hagy recruits proteins and lipids from different intracellular membranes2 in turn, basal autophagy negatively modulates ciliary growth through to initiate the formation of autophagosomes, double-membrane vesicles turnover of essential ciliogenesis proteins (Extended Data Fig. 1). that sequester cytoplasmic material and deliver it through vesicular fusion to lysosomes for degradation3. Disruption of IFT compromises autophagy The primary cilium is a non-motile signalling organelle that grows To determine the role of primary cilia on autophagy, we used two in a specific region of the plasma membrane and senses, among other cellular models with compromised ciliogenesis: mouse embryonic 2 things, changes in the enrichment of nutrients in the environment4. fibroblasts (MEFs) stably knocked down for IFT20 (IFT20 ) and kid- Cargo trafficking along the ciliary axoneme (intraflagellar transport; ney epithelial cells (KECs) from mice with a hypomorphic mutant 2 2 IFT) is maintained through motor proteins (kinesins and dyneins) IFT88 (Ift88 / ), two IFT-B complex components required for cilio- and two large multiprotein complexes (IFT particles A and B)5. Some genesis6 (Extended Data Fig. 2a). After 24 h of serum removal, cells 2 subunits of these complexes can be found in other cellular compart- with detectable primary cilia were reduced by 60% in IFT20 MEFs 2 2 ments such as Golgi, from where they facilitate mobilization of spe- and almost absent in the Ift88 / KECs (Fig. 1a). cific cargo to the basal body and cilium for ciliogenesis and ciliary Rates of lysosomal protein degradation after serum deprivation, a signalling6. The primary cilium coordinates a variety of signalling stimulus that activates autophagy, were reduced in the two ciliogenesis- pathways including the Hedgehog (Hh) pathway, which requires impaired models (Extended Data Fig. 2b). To test whether autophagy IFT-mediated recruitment to the basal body and axoneme of smooth- was defective we analysed flux through the autophagic pathway using ened (SMO) and the transcription factors GLI1 and GLI27,8. lysosomal proteolysis inhibitors and measured changes in levels of In many types of cultured cells, deprivation of serum from the LC3-II, an autophagosome component degraded with the cargo in media induces almost linear growth of primary cilia for up to 2 days9. lysosomes11. The characteristic upregulation of the autophagic flux 2 Induction of autophagy also occurs in the first hours that follow after serum removal was reduced in IFT20 MEFs whereas basal auto- serum removal, and can be sustained for the full duration of the phagic flux (rich media) was only discretely reduced (Fig. 1b). Although starvation period10. Despite the temporal coincidence of the forma- a high percentage of KECs are ciliated under basal conditions and tion of autophagosomes and primary cilia, any possible relation serum removal has small effect on ciliogenesis (Extended Data Fig. 2c), between the autophagic and ciliogenesis machineries remains we still observed a reduction in autophagic flux upon serum removal in 2 2 unknown. Likewise, the possible involvement of autophagy in ciliary Ift88 / KECs when compared to wild type (Fig. 1b), indicating that dynamics has not been analysed. primary cilia are not necessary to maintain the high basal autophagic In this study we investigate the functional interaction between flux of these cells (Extended Data Fig. 2d), but are essential to sustain primary cilia and autophagic induction. We have found that max- proper autophagic flux during nutrient deprivation. imum activation of autophagy in response to nutrient deprivation To directly analyse autophagosome formation and their clearance requires the strategic location of components of the autophagic by lysosomes, we transfected cells with a pH-sensitive reporter machinery at the ciliary base in a IFT and Hh signalling-dependent (mCherry–GFP–LC3; fusion of mCherry, green fluorescent protein manner. In contrast, when autophagy is compromised, ciliogenesis is and LC3) that highlights autophagosomes as yellow puncta and

1Department of Development and Molecular Biology, Albert Einstein College of Medicine, Bronx, New York 10461, USA. 2INSERM U845; Paris-Descartes University, 75014 Paris, France. 3INSERM U984; University Paris-Sud 11; 92296 Chaˆtenay-Malabry, France. 4Department of Anatomy and Structural Biology, Albert Einstein College of Medicine, Bronx, New York 10461, USA. 5Institute for Aging Studies, Albert Einstein College of Medicine, Bronx, New York 10461, USA.

Figure 1 | Blockage of IFT reduces autophagic a Ctrl IFT20– b Serum+ Serum– Ctrl IFT20– 10 activity. a, Percentage of ciliated cells in control Ac-tubulin – – 2 CtrlIFT20 Ctrl IFT20 ** (Ctrl) and IFT20 MEFs (*P 5 0.030, n 5 3) and 2 2 PI ––+ +++ – – wild-type (WT) and Ift88 / KECs (**P 5 0.003, 5 5 LC3-I 25 cells each, n 3) after 24 h of serum removal. LC3-Il Arrows, cilia. b, LC3 immunoblot (left) for LC3-II lux 44.3 ± 7.3% 17.5 ± 3.7%* autophagic flux quantification (right) in Ctrl and 0 2 –/– Actin IFT20 MEFs (top, **P 5 0.0003, n 5 7), and in WT Ift88 Serum+ Serum– 2 2 WT and Ift88 / KECs (bottom, *P 5 0.035, Ac-tubulin WT Ift88–/– WT Ift88–/– n 5 9). PI, protease inhibitors; S, serum. 10 * 2/2 S + +–– ++–– c, Autophagic flux in WT and Ift88 KECs by 8 PI –+ –+ – + – + mCherry–GFP–LC3 puncta quantification. 6 Puncta: yellow, autophagosomes; red only, 76.4 ± 6.2% 8.6 ± 8.5% LC3-I 4 autophagolysosomes; total, both together LC3-II 5 5 LC3-II lux 2 (**P 0.008, *P 0.025). d, LC3 immunoblot c 5 Actin 0 (left) for quantification of autophagosome Yellow ** Serum+ Serum– * 5 5 e 4 Red formation (right). ( P 0.048, n 4). , Electron Total d microscopy images (left) and morphometric 3 * Ctrl IFT20– Ctrl IFT20– quantification of autophagic vacuoles (AV, right) 3 2 2 n.s. S ++–– ++–– in Ctrl and IFT20 MEFs (top, **P 5 0.001, 10 n.s. PI t(h) 2 6 2 6 2 6 2 6 2 fields in 2 experiments) and KECs WT and 1 2 2 * Ift88 / (bottom, *P 5 0.037, 5 fields). Arrows, LC3-I Area puncta/cell area 0 1 WT Ift88–/– WT Ift88–/– LC3-II autophagic vacuoles (AV, black) and endosomal compartments (yellow). f, Percentage of ciliated Serum+ Serum– 0 Tubulin MEFs after induction of cilia resorption with

Times LC3-II PI 6 h/2 h Serum+ Serum– 2 50 ng ml 1 platelet-derived growth factor (PDGF) f e – treatment. (**P 5 0.001, ***P 5 0.0009, 121, 230 Ctrl IFT20 Ctrl 60 IFT20– and 162 cells, n 5 5). g, LC3 immunoblot (top) for autophagic flux quantification (bottom) upon 8 40 ** *** PDGF treatment (*P 5 0.005, n 5 3). Scale bars in 6 a,10mm. n.s., statistically non-significant. ** 20 6 6 4 ** Mean s.d. in b, d and mean s.e.m. in other

Ciliated cells (%) panels. 0 2 S+ S– S– PDGF 0 0.2 μm 0.2 μm Serum+ Serum– g

Area occupied by AVs (%) PDGF –/– PI – + – + WT Ift88 WT Ift88–/– * LC3-I 2.5 LC3-Il 2.0 Actin 1.5 8 6 1.0 4 * 0.5 2 LC3-II lux

2 μm 0.2 μm2m 2 μm 0.2 μm AV/cell (times serum+) 0 0 Serum+ Serum– None PDGF

autophagolysosomes (post-lysosomal fusion) as red puncta12. Basal IFT88 knockdown cells did not result in further reduction of autophagy levels of autophagic vacuoles were comparable in wild-type and (Extended Data Fig. 3a–d), supporting that blockage of anterograde 2 2 2 2 Ift88 / KECs, but upon removal of serum Ift88 / KECs showed IFT (aIFT) compromises activation of inducible autophagy. Consistent a significantly lower content of autophagolysosomes (Fig. 1c and with this notion, cilia resorption induced by treatment with platelet- Extended Data Fig. 2e). This reduction in autophagolysosomes was derived growth factor14 (Fig. 1f) markedly decreased starvation-induced not paralleled by an increase in autophagosomes, and the overall autophagic flux in MEFs, KECs and in RGC-5 cells, a retinal ganglion 2 2 content of LC3-positive vesicles in serum-deprived Ift88 / KECs cell line with robust induction of ciliogenesis in response to serum was lower than in wild-type cells (Extended Data Fig. 2f), indicating removal (Fig. 1g and Extended Data Fig. 3e). that reduced autophagic flux was mainly owing to lower auto- Overall these findings support a reduced ability for autophagosome phagosome synthesis rather than blockage in their degradation. synthesis in cells with compromised ciliogenesis that limits their capa- Autophagosome formation (assessed as the increase in LC3-II levels city to upregulate autophagy in response to nutrient deficiency. at two time points after inhibition of lysosomal proteolysis) was also 2 reduced in IFT20 MEFs (Fig. 1d). Ultrastructural analysis confirmed Ciliary Hh signalling induces autophagy 2 that despite normal autophagosome morphology, IFT20 and Altered aIFT blocks ciliogenesis and prevents the recruitment of essential 2 2 Ift88 / cells failed to expand the autophagic compartment in res- signalling molecules to the ciliary region6. We proposed that malfunc- ponse to serum deprivation (Fig. 1e). tioning of ciliary signalling may be behind the defective starvation- Contrary to the restricted location of IFT88 to the ciliary base13, induced autophagy in cells with compromised IFT. Although primary IFT20isalsopresentinothercellularcompartments6.However,the cilia downregulate signalling through the mammalian target of rapa- effect of IFT20 on autophagy seems to be primarily related to its mycin (mTOR) pathway15, a well-known negative regulator of autop- 2 2 2 function in ciliogenesis because additional knockdown for IFT20 in hagy, treatment of IFT20 and Ift88 / cells with the mTOR inhibitor

2 | NATURE | VOL 000 | 00 MONTH 2013 ©2013 Macmillan Publishers Limited. All rights reserved ARTICLE RESEARCH a b Figure 2 | Hedgehog signalling requires the None Purmo Ctrl IFT20– IFT88– 20 primary cilia to regulate autophagy. a, GFP–LC3 GFP–LC3 * 4.0 * 15 3.5 Yellow puncta pattern (left) and quantification (right) in Red 10 3.0 MEFs treated with purmorphamine (Purmo; top, Total 2 2 2.5 5 5 / 5 Serum + 5 *P 0.028, n 4), in Ptc MEFs (middle, *P 2.0 n.s. 5 0 n.s. 0.015, n 4) and after myc–GLI1 overexpression Control Purmo 1.5 (bottom, *P 5 0.012, n 5 3). b, Autophagic flux WT Ptc–/– 1.0 (folds % cell area) GFP–LC3 12 * Fluorescent puncta 0.5 by mCherry–GFP–LC3 puncta quantification 9 0 upon purmorphamine treatment in MEFs in the None Purmo None Purmo None Purmo 6 presence of serum. Yellow, autophagosomes; red

Serum + 3 only, autophagolysosomes; total, both c Ctrl SMO– None 0 Purmo compartments. (*P 5 0.0323, 40 fields). None Purmo None Purmo WT Ptc–/– 10 Ctrl myc–GLI1 c, LC3 immunoblot (left) for autophagic flux 12 PI – + – + – + – + 8 *** GFP–LC3 * n.s. quantification (right) in Ctrl (**P 5 0.001) and

GFP–LC3 puncta per cell section 9 LC3-I 6 MEFs knocked down for SMO (SMO2)(**P 5 6 4 * LC3-II ** 0.0001) upon purmorphamine treatment (*P 5 3 LC3-II lux Serum + 2 0.009, n 5 4). d, Autophagic flux by mCherry– 0 0 Actin – GFP–LC3 puncta quantification in WT and Ctrl myc–GLI1 Control SMO 2 2 Ift88 / KECs upon purmorphamine treatment d e 5 5 WT Ift88–/– Ift88–/– None (*P 0.046, **P 0.008, 40 fields). e,LC3 GLI1 5 None myc–GLI1 6 immunoblot (left) for autophagic flux ** Yellow ** 2/2 n.s. 5 quantification (right) in Ift88 KECs 4 Red S+ S– S+ S– 4 overexpressing myc–GLI1 (**P 5 0.00006, n 5 3). 3 * Total PI–+–+–+–+ * 3 f, Autophagic flux by mCherry–GFP–LC3 puncta 2 LC3-I 2 quantification in MEFs treated with cyclopamine LC3-II lux LC3-II 1 ** 1 upon serum removal (*P 5 0.011, **P 5 0.013, 25 * 6 0 0 fields). Scale bars, 10 mm. Mean s.d. in c, f and

None Purmo Serum– None Purmo Serum– Actin Serum+ Serum– 6 Fluorescent puncta (% cell area) mean s.e.m. in other panels.

f Serum+ Serum– Serum– Cyclo 2.5 * Yellow 2.0 Red 1.5 Total * 1.0 ** 0.5 n.s. Fluorescent puncta (fold % area per cell) 0 Serum+ Serum– Serum– None Cyclo rapamycin failed to restore normal autophagic activity in these cells which made us propose an involvement of aIFT in the delivery of the (data not shown); thereby reduced autophagy in ciliogenesis-defective autophagic machinery to this membrane. cells was not due to enhanced mTOR signalling. Co-immunostaining for different autophagy-related proteins We next focused on the Hedgehog (Hh) signalling pathway because (ATGs, green) and acetylated tubulin (red) to highlight the ciliary of its dependence on the primary cilium and intact IFT7,8.Treatmentof axoneme in serum-deprived KECs revealed that of the twelve ATGs serum-supplemented MEFs or KECs with purmorphamine, an SMO analysed, five ATGs2,3 (ATG16L, AMBRA1, LC3, GABARAP and agonist that activates expression of Hh downstream factors (Extended VPS15) localized as discrete puncta along the ciliary axoneme, clearly Data Fig. 4a, c), returned content of LC3-positive compartments and distinguishable from the surrounding cytosol signal with deconvolu- autophagic flux to values observed upon serum removal (Fig. 2a–d and tion immunofluorescence and three-dimensional (3D) wire mod- Extended Data Fig. 4d–f). Purmorphamine failed to induce autophagy 2 2 2 2 elling (Fig. 3a, Extended Data Fig. 5a, b and Supplementary Videos in IFT20 or IFT88 MEFs and in Ift88 / KECs, supporting depend- 2 1–5). Similar colocalization was observed using fluorescent-tagged ence on IFT (Fig. 2b, d and Extended Data Fig. 4f) and in SMO cells versions of the ciliary protein inversin18 or of ATGs (Extended Data (Fig. 2c), which as expected showed reduced purmorphamine-mediated Fig. 5c, d). Other more abundant ATGs, such as ATG14, were not upregulation of GLI1 and GLI2 (Extended Data Fig. 4c). Two other detectable in the axoneme, and knockdown against cilia-associate conditions that activate Hh signalling, Patched-1 receptor knockout ATGs abolished their staining (Extended Data Fig. 5b, d) supporting 2/2 (Ptc ,alsoknownasPtch1; constitutive activation of Hh signalling specificity of the staining. These ATGs were also detectable in primary and ciliogenesis8) and overexpression of GLI1 also upregulated autop- cilia isolated by the ‘peeling off’ procedure (Fig. 3b). hagy (Fig. 2a and Extended Data Fig. 4d, g–j), and in fact, overexpres- A larger number of ATGs associated with the basal body (highlighted sion of GLI1 is sufficient to partially rescue the autophagic defect in 2/2 with anti-gamma tubulin). In addition to the five cilia-associated ATGs, Ift88 cells (Fig. 2e). Conversely, two interventions that reduce Hh we also found ATG14, VPS34, ATG7 and ATG5, but not Beclin 1 or signalling, knockdown of SMO and treatment with Hh antagonist ULK-1, at the base of the axoneme (Fig. 3c; LAMP-2 is used a negative cyclopamine, reduced starvation-induced autophagy (Fig. 2c, f and control). Similar colocalization experiments in serum-supplemented Extended Data Fig. 4k). cells, to repress autophagy, and in cells with compromised ciliogenesis 2 2 Altogether, these data reveal a positive regulatory effect of Hh (Ift88 / cells) revealed three types of association of ATGs with the signalling on autophagy, and support that the inability to activate basal body: serum- and aIFT-dependent, such as VPS34 and ATG16L; autophagy in cells with defective IFT originates, at least in part, from serum-independent but aIFT-dependent, as ATG7 and ATG14; and the loss of Hh signalling. serum- and aIFT-independent, as ATG5 and LC3 (Fig. 3d–f and Extended Data Fig. 6). Autophagic machinery localizes at the cilia Disruption of IFT also altered the overall intracellular distribution of Besides ciliogenesis, IFT also participates in recruitment of non-ciliary ATGs that localize at the basal body in an aIFT-dependent manner. proteins to the plasma membrane16, a site of autophagosome formation17, Thus, ATG7 was normally observed in KECs as discrete cytosolic puncta

a Figure 3 | Autophagy-related proteins associate Initiation complex Elongation complex Autophagosome with ciliary structures in a serum-dependent manner. a, Co-immunostaining (top) and 3D reconstruction (middle and bottom) for the indicated autophagy-related proteins (ATGs, green) and acetylated tubulin (red, ciliary axoneme marker) in KECs maintained in the absence of

Ac-tubulin serum for 24 h. Yellow arrows, colocalization. Ac-tubulin Ac-tubulin Ac-tubulin

Ac-tubulin b, Schematic representation of the ATGs associated

LC3 with the axoneme (top left). Co-immunostaining VPS15 ATG16L AMBRA1

GABARAP for ATGs and acetylated tubulin in cilia isolated by peeling off. Individual channels and merged images are shown. c, Co-immunostaining (bottom) and 3D reconstruction (top) for ATGs (green) and b d e f gamma-tubulin (red, basal body marker) in the Isolated cilia WT Ift88–/– same cells and conditions. Arrows, colocalization AMBRA1 Ac-Tub Merge VPS34 ATG7 ATG5 (yellow), no colocalization (white). d–f, Percentage 40 80 80 LC3 † n.s. of cells with colocalizing ATGs in the basal body ATG16L 30 60 60 2/2 AMBRA1 n.s. (BB) in WT and Ift88 KECs in the presence or 20 40 40 VPS15 ** n.s. absence of serum. d, Serum- and IFT-dependent GABARAP 10 20 ** *** 20 association of ATGs with the BB (VPS34, {P 5 0 0 0 5 5 5 S+ S– S+ S– S+ S– 0.04, **P 0.002, n 4; ATG16L, {{P 0.0003, 5 5 ATG16L ATG14 **P 0.0009, 15 cells each per experiment, n 7). LC3 Ac-tubMerge GABARAP Ac-tub Merge LC3 50 †† 80 80 n.s. e, IFT-dependent but serum-independent 40 n.s. 5 60 n.s. 60 association of ATGs with the BB (**P 0.001, 30 26 40 40 ***P 5 9.33768 3 10 , n 5 4; *P 5 0.018, 15 cells 20 Basal body colocalization (%) Basal body colocalization (%) Basal body colocalization (%) 5 ** 20 20 each per experiment, n 5). f, BB association of 10 * 5 0 0 0 ATGs independent on serum or IFT (ATG5, n 5; S+ S– S+ S– S+ S– LC3, n 5 8; 15 cells each per experiment). Scale c 6 ULK-1 VPS34 VPS15 Beclin 1 AMBRA1 ATG14 bars, 10 mm. Mean s.e.m. Induction Initiation complex

ATG7 ATG5 ATG16L LC3 GABARAP LAMP-2 Lysosome Autophagosome Elongation complex

that markedly increased upon starvation, but it adopted a diffuse reticu- IFT-dependent trafficking of ATG16L 2/2 lar pattern unchanged by starvation in Ift88 KECs (Extended Data To gain a better understanding of the mechanisms behind the ciliary Fig. 7a). Disruption of IFT also reduced the basal association of ATG14, regulation of starvation-induced autophagy, we focused on ATGs that key for autophagy initiation3, with the Golgi and its mobilization into associated preferentially with ciliary structures upon serum removal. discrete cytosolic puncta upon serum removal (Extended Data Fig. 7a). We focused on ATG16L because it was almost undetectable in the Interestingly, IFT88 deficiency also blunted the starvation-induced basal body in serum-supplemented cells (Fig. 3d), starvation induced changes in the intracellular location of ATGs that localize to the basal its IFT-dependent association with the basal body and its presence in body independently of IFT88, such as LC3 and ATG5 (Fig. 3f). this location increased in two models with enhanced ciliary Hh sig- The association with the plasma membrane of LC3 and GABARAP nalling (Extended Data Fig. 8a, b). We proposed that active recruit- (both involved in autophagosome membrane formation/elongation3), ment of ATG16L to the ciliary base during starvation could be the clearly noticeable in wild-type cells upon serum removal, was no trigger for ciliary-induced autophagy. longer observed in the IFT88-deficient cells (Extended Data Fig. 7a). We first investigated the contribution of IFT20 to ciliary delivery of Likewise, the often ‘hook-like’ organization of ATG5 (involved in ATG16L because IFT20 also participates in trafficking of ciliary mem- autophagosome membrane elongation) at the basal body was also brane proteins from the Golgi to the base of the cilium6. To analyse 2 2 abrogated in Ift88 / KECs (Extended Data Fig. 7b, c). The abundance separately both functions of IFT20, we compared its association with 2 2 of ATG5 hook-like structures under basal conditions in cells with ATG16L in cells with intact or disrupted aIFT (Ift88 / KECs). We 2 2 constitutive Hh signalling (Ptc / ; Extended Data Fig. 7d) indicates found that IFT20 and ATG16L colocalized in small cytosolic vesicles the need of a functional cilium for ATG5 clustering in this location. that become more abundant upon serum removal and that this coloca- These findings support a role for aIFT in the cellular relocation of lization was only partially reduced when aIFT was disrupted (Extended the autophagic machinery to sites of autophagosome formation. Data Fig. 8c). Both proteins co-immunoprecipitated independently of

4 | NATURE | VOL 000 | 00 MONTH 2013 ©2013 Macmillan Publishers Limited. All rights reserved ARTICLE RESEARCH a b Figure 4 | IFT20 regulates trafficking of IP : IFT20 IP : ATG16L WT Ift88–/– WT Ift88–/– ATG16L to the cilium. a, b, Immunoblot for the 2/2 Serum + Serum – Serum + Serum – Serum + Serum – Serum + Serum – indicated proteins of WT and Ift88 KECs subjected to co-immunoprecipitation of IFT20 Inp IP FT Inp IP FT Inp IP FT Inp IP FT Inp IP FT Inp IP FT Inp IP FT Inp IP FT (a) and ATG16L (b). Inp, 1/10 input; IP, immunoprecipitate; FT, 1/10 flow-through. c, Immunoblot for ATG16L and clathrin in MEF ATG16L ATG16L homogenate (Hom), and the pellets from 1 h centrifugation at 100,000g (organelles (Org)), IFT88 300,000g (300k) and 500,000g (vesicles, 500k). ULK-1 d, Immunoblot for IFT20 and ATG16L in the same 2 2 IFT20 fractions isolated from Ctrl, IFT20 and IFT88 MEFs in the presence (top) and absence (bottom) IFT20 of serum. e, Immunogold electron microscopy for c d Ctrl IFT20– IFT88– ATG16L (15-nm particles) and IFT20 (10-nm Serum + Serum – Vesicles Vesicles Vesicles particles) in vesicles from MEFs isolated by the Vesicles Vesicles same method as in c. Right, quantification of IFT20 HHOO300k500k 300k500k H O300k500k and ATG16L presence in the vesicles (11 fields). Hom Org 300k 500k Hom Org 300k 500k 6 IFT20 Mean s.e.m. Serum +

ATG16L ATG16L

IFT20 Clathrin ATG16L Serum – e ATG16L (15 nm) + IFT20 (10 nm)

IFT20 15.3 ± 2.1

23.4 ± 1.8 61.2 ± 2.5 ATG16L

Both

the presence of IFT88 (Fig. 4a), and IFT20, but not IFT88, could be Autophagy activation reduces cilia growth pulled-down with ATG16L (Fig. 4b). Starvation did not change the Although inducible autophagy relies on IFT and cilia function, we association between ATG16L and IFT20, but changed the location of found that autophagy is not required for cilia formation. In fact, 2 2 the IFT20- and ATG16L-positive vesicles from the proximity of the Atg5 / MEFs formed cilia both longer and faster than wild-type Golgi towards a more cytosolic location (Extended Data Fig. 8c), indi- MEFs upon serum removal, and a higher percentage of them grew cating that the same signal that enhances IFT20-mediated trafficking cilia even in basal conditions (Fig. 5a–d). Knockdown of two other of proteins from the Golgi to the cilia may also be responsible for ciliary ATGs and even chemical inhibition of autophagy were sufficient to delivery of ATG16L. Other ATGs (VPS15, ATG7) have minimal colo- induce ciliogenesis in different cell types under basal conditions, calization with IFT20, or colocalize only at the Golgi (ATG14) whereas upregulation of autophagic activity with rapamycin did not (Extended Data Fig. 8d), and none of them binds directly to IFT20 affect either basal or inducible ciliary growth (Extended Data Fig. 9). 2 2 (Extended Data Fig. 8e–g). Scanning electron microscopy revealed that in Atg5 / MEFs star- The amount of ATG16L in isolated clathrin-enriched vesicles19 was vation induced formation of long and narrow cilia that come out of 2 2 very low under basal conditions, but it increased markedly upon smaller ciliary pockets, and confirmed that Atg5 / MEFs grow prim- serum removal (Fig. 4c). Although the vesicular content of IFT20 ary cilia in serum-supplemented media, although shorter and with did not change with starvation (Fig. 4d), knockdown of IFT20 fewer surface-adhered vesicles (presumably exosomes) than the cilia reduced starvation-induced loading of ATG16L in the vesicles grown during starvation (Fig. 5c and Extended Data Fig. 10). (Fig. 4d). Interestingly, ATG16L content was also reduced in vesicles Activation of ciliary Hh signalling by starvation or treatment with isolated from serum-deprived cells knocked down for IFT88 (Fig. 4d), purmorphamine showed efficient recruitment of SMO to the cilia 2 2 explaining the reduced colocalization between IFT20 and ATG16L in in Atg5 / cells and increased expression of the downstream tran- 2 2 Ift88 / cells. Electron microscopy and double immunogold labelling scription factors GLI1 and GLI2 in these cells even under basal for IFT20 and ATG16L confirmed the vesicle nature of the fractions conditions, supporting full functionality of the cilia induced by autop- and the coincidence of both proteins in more than 60% of the labelled hagic blockage (Fig. 5e, f and Extended Data Fig. 11a). vesicles (Fig. 4e and Extended Data Fig. 8h). To start investigating the basis for this inhibitory effect of autop- We propose that, during starvation, ATG16L reaches the base of hagy on ciliogenesis, we compared levels and distribution of IFT 2 2 the cilia through the Golgi-to-cilia shuttling function of IFT20, and proteins in wild-type and Atg5 / MEFs and found that basal levels 2 2 then it accesses the ciliary region by IFT88-dependent mechanisms20. of IFT20, but not IFT88, were markedly higher in Atg5 / MEFs Overall, our results support a novel role for the centrioles as organ- (Fig. 5g). Because chemical inhibition of lysosomal proteolysis in ization centres for the autophagic machinery and for IFT proteins as control cells reproduces a similar increase in IFT20 levels (Fig. 5h), common components shared by both ciliogenesis and starvation- and IFT20 could be detected in isolated autophagic vacuoles (Fig. 5i), induced autophagosome formation. we propose that basal levels of IFT20, and consequently ciliary

a c with lysosomal proteolysis inhibitors in wild-type MEFs confirmed 30 * WT absence of lysosomal degradation of IFT20 during the first 12 h of 20 serum removal, when inducible autophagy is maximally activated (Extended Data Fig. 11c, d). In light of these findings, we propose that 10 basal autophagy contributes to the regulation of ciliogenesis and the length of an already formed cilium, at least in part through the degra-

Ciliated cells (%) 0 dation of IFT20. The switch from basal to induced autophagy seems to WT Atg5–/– spare cytosolic IFT20 from degradation and allow its active engage- b WT ment in the vesicular trafficking required for ciliogenesis. Atg5–/– 80 –/– * Serum – Atg5 Discussion 60 In this work, we identify a previously unknown reciprocal relation- 40 ship between primary cilium and autophagy whereby components of the ciliary machinery are shared with autophagy for starvation- 20 induced autophagosome biogenesis. In contrast, basal autophagy Ciliated cells (%) 0 inhibits ciliogenesis by limiting trafficking to the cilium of compo- 0 10 20 30 40 50 nents required for ciliary growth. The temporal coincidence of activa- Starvation time (h) –/– –/– tion of ciliogenesis and autophagy at the early times of nutritional d ** Atg5 Atg5 deprivation may thus serve as a self-regulatory brake for each of these 4 * ††

m) processes (Extended Data Fig. 1). μ 3 The localization of autophagy initiating ATGs at the ciliary base 2 and the active recruitment of ATG16L to this location upon starvation

1 Serum + suggests that sensing of nutrient deficiency and activation of signal-

Cilia length ( ling from the cilium could initiate a cilia-mediated autophagic pro- 0 gram (see additional Discussion 1 in Supplementary Information). –/– –/– j WT Atg5 Atg5 The presence along the cilium axoneme of the pre-autophagosomal Serum– Serum+ ATG16L and integral autophagosome membrane ATGs—for example, e f g h LC3 and GABARAP—and their IFT-dependent enrichment at the WT 3 WT † Serum+ Serum+ 1,000 –/– –/– plasma membrane suggest that cilia-mediated autophagy may induce Atg5 Atg5 WT Atg5–/– – NL 2 autophagosome formation from this location, making the ciliary 100 *** pocket, characterized by high vesicular activity21, an attractive place mRNA ** IFT20 1 IFT20 for autophagosome formation (see additional Discussion 2 in Sup- mRNA (log) 10 * Gli2 n.d. I I plementary Information). LC3 LC3 Gli1 II II 1 0 – – Purmo – The negative regulatory role of basal autophagy on ciliogenesis None None identified here may take place through different mechanisms. We Purmo Purmo S+ S– Actin Actin have found that basal autophagy controls trafficking of ciliary proteins i j k by limiting the amount of IFT20 accessible for shuttling between the Cyt Hom APG WT 24 h 2.5 –/– Golgi and the ciliary base (see additional Discussion 3 in Supplemen- Atg5–/– * WT Atg5 2.0 S+ S– S+ S– tary Information). As induction of autophagy depends on the same

IFT20 1.5 IFT20 protein that autophagy degrades, this could represent a novel

1.0 IFT20 mechanism for self-containment of the autophagic process. I 0.5 The primary cilium is constitutively present in most tissues, but the

LC3 II Actin

IFT20 (folds WT S+) 0 fact that ciliogenesis blockage in cells with permanent cilia compromises S+ 2 h 24 h their starvation-induced autophagy supports that it is not the mere Figure 5 | Ciliogenesis is enhanced in autophagy-defective cells. a, b, Ciliated presence of a cilium but the activation of ciliary signalling by starvation 2 2 1 Atg5 / MEFs in serum (*P 5 0.038, 25 cells each per experiment, n 5 3) that contributes to autophagic induction. We have also identified the 2 (a) or serum (*P 5 0.006, nonlinear fit regression, 25 cells each per regulatory effect of the ciliary Hedgehog signalling pathway, for which experiment, n 5 3) (b). c, Scanning electron microscopy (SEM) of primary an emerging role in autophagy regulation has been proposed22–25 (see cilia. Arrows, cilia-associated vesicles; arrowheads, ciliary pocket. d, Cilia additional Discussion 4 in Supplementary Information). The possible 5 5 5 length quantification from SEM (*P 0.027 n 28; **P 0.001, participation of the cilium in the integration of other autophagy- {{P 5 0.0001, n 5 23). e, f, GLI1 (*P 5 0.04, n 5 3) (e) and GLI2 5 5 5 5 inducing signals and the effect of autophagy in IFT20 functions beyond (***P 0.0005, **P 0.0046, {P 0.0132, n 3) (f) messenger RNA 16 2/2 2/2 ciliogenesis require further investigation. expression in Atg5 MEFs. g, h, IFT20 immunoblot in Atg5 (g) and WT MEFs with lysosomal inhibitors (NL) (h). i, IFT20 immunoblot in cytosol (Cyt), homogenate (Hom) and autophagosomes (APG). j, IFT20 protein levels METHODS SUMMARY 2 2 2 2 2 2 2 2 / * 5 5 k / MEFs defective in autophagy (Atg5 / ) were from N. Mizushima, Ift88 / (also in Atg5 MEFs ( P 0.038, n 6). , IFT20 immunoblot in Atg5 2 2 MEFs. n.s., statistically non-significant. Mean 6 s.d. in d and j and known as Tg737 / ) mouse kidney epithelial cells were from G. Pazour, and the mean 6 s.e.m. in other panels. RGC-5 cell line from P. Boya. Cell culture, removal of serum and treatment with chemical modulators of autophagy were performed as described before26. Intracellular growth, are regulated at least in part through IFT20 degradation by protein degradation was measured as described before27. Autophagic flux was 11 basal autophagy. Although most IFT20 associates with Golgi, the measured as changes in levels of LC3-II upon inhibition of lysosomal proteolysis . 2 2 basal increase in cellular IFT20 levels in Atg5 / MEFs seems to occur Lentivirus-mediated short hairpin RNA silencing was performed as described before26. Cilia were isolated by the peel-off technique28 and autophagosomes and in the small cytosolic vesicles only visible in control cells during autophagolysosomes by floatation in metrizamide gradients as previously described29. starvation-induced ciliogenesis (Extended Data Fig. 11b). Interestingly, 2/2 Cells grown on coverslips were processed for immunofluorescence following the differences in the IFT20 content between wild-type and Atg5 standard procedures26 and imaged in an Apotome.2 system using an Axiovert MEFs were no longer observed early after serum removal, but become 200 fluorescence microscope (Carl Zeiss). Transmission electron microscopy apparent again after 24 h in serum-deprived media (Fig. 5j, k). Studies was performed as before26 and scanning electron microscopy following standard

6 | NATURE | VOL 000 | 00 MONTH 2013 ©2013 Macmillan Publishers Limited. All rights reserved ARTICLE RESEARCH procedures30. Gel densitometry, morphometric analysis and quantification of 19. Sridhar, S. et al. The lipid kinase PI4KIIIb preserves lysosomal identity. EMBO J. 32, colocalization was done using Image J (NIH). Protein quantification, co-immu- 324–339 (2012). noprecipitation, electrophoresis and immunoblot and semiquantitative real-time 20. Richey, E. A. & Qin, H. Dissecting the sequential assembly and localization of 26 intraflagellar transport particle complex B in Chlamydomonas. PLoS ONE 7, PCR were performed as previously described . Immunoblot membranes were e43118 (2012). developed using the LAS-3000 Imaging System (Fujifilm). Results are shown as 21. Satir, P. The new biology of cilia: review and annotation of a symposium. Dev. Dyn. mean 6 s.e.m. or mean 6 s.d., and Student’s t-test for unpaired data was used for 241, 426–430 (2012). statistical analysis and one-way analysis of variance (ANOVA) was used for mul- 22. Li, H. et al. Sonic hedgehog promotes autophagy of vascular smooth muscle cells. tiple comparisons. A value of P , 0.05 was considered statistically significant. Am. J. Physiol. Heart Circ. Physiol. 303, H1319–H1331 (2012). 23. Jimenez-Sanchez, M. et al. The Hedgehog signalling pathway regulates autophagy. Online Content Any additional Methods, Extended Data display items and Source Nature Comm. 3, 1200 (2012). Data are available in the online version of the paper; references unique to these 24. Wang, Y., Han, C., Lu, L., Magliato, S. & Wu, T. Hedgehog signaling pathway sections appear only in the online paper. regulates autophagy in human hepatocellular carcinoma cells. Hepatology 58, 995–1010 (2013). Received 20 August 2012; accepted 10 September 2013. 25. Petralia, R. S. et al. Sonic hedgehog promotes autophagy in hippocampal neurons. Biol. Open 2, 499–504 (2013). Published online 2 October 2013. 26. Singh, R. et al. Autophagy regulates lipid metabolism. Nature 458, 1131–1135 (2009). 1. Mizushima, N., Levine, B., Cuervo, A. M. & Klionsky, D. J. Autophagy fights disease 27. Auteri, J. S., Okada, A., Bochaki, V. & Dice, J. F. Regulation of intracellular protein through cellular self-digestion. Nature 451, 1069–1075 (2008). degradation in IMR-90 human diploid fibroblasts. J. Cell. Physiol. 115, 167–174 2. Hamasaki, M., Shibutani, S. T. & Yoshimori, T. Up-to-date membrane biogenesis in (1983). the autophagosome formation. Curr. Opin. Cell Biol. 25, 455–460 (2013). 28. Huang, B. Q. et al. Isolation and characterization of cholangiocyte primary cilia. Am. 3. Yang, Z. & Klionsky, D. J. Mammalian autophagy: core molecular machinery and J. Physiol. Gastrointest. Liver Physiol. 291, G500–G509 (2006). signaling regulation. Curr. Opin. Cell Biol. 22, 124–131 (2010). 4. Satir, P., Pedersen, L. B. & Christensen, S. T. The primary cilium at a glance. J. Cell 29. Marzella, L., Ahlberg, J. & Glaumann, H. Isolation of autophagic vacuoles from rat Sci. 123, 499–503 (2010). liver: morphological and biochemical characterization. J. Cell Biol. 93, 144–154 5. Taschner, M., Bhogaraju, S. & Lorentzen, E. Architecture and function of IFT (1982). complex proteins in ciliogenesis. Differentiation 83, S12–S22 (2012). 30. Wada, K. et al. Application of photoacoustic microscopy to analysis of biological 6. Follit, J. A., Xu, F., Keady, B. T. & Pazour, G. J. Characterization of mouse IFT complex components in tissue sections. Chem. Pharm. Bull. (Tokyo) 34, 1688–1693 (1986). B. Cell Motil. Cytoskeleton 66, 457–468 (2009). Supplementary Information is available in the online version of the paper. 7. Liem, K. F. et al. The IFT-A complex regulates Shh signaling through cilia structure and membrane protein trafficking. J. Cell Biol. 197, 789–800 (2012). Acknowledgements We thank N. Mizushima and G. Pazour for providing MEFs and 8. Rohatgi, R., Milenkovic, L. & Scott, M. P. Patched1 regulates Hedgehog signaling at I. R. Veland and S. T. Christensen for providing the plasmid for expression of GFP– the primary cilium. Science 317, 372–376 (2007). inversin. We also thank the personnel at the Analytical Imaging Facility for their 9. Kiprilov, E. N. et al. Human embryonic stem cells in culture possess primary cilia technical assistance with TEM, J. Kraut and N. Rodriguez for their early assistance in with hedgehog signaling machinery. J. Cell Biol. 180, 897–904 (2008). cilia staining optimization and S. Kaushik for critically reviewing this manuscript. This 10. Onodera, J. & Ohsumi, Y. Autophagy is required for maintenance of amino acid work was supported by grants from the National Institute of Health National Institute on levels and protein synthesis under nitrogen starvation. J. Biol. Chem. 280, Aging AG031782 and AG038072, National Institute of Diabetes and Digestive and 31582–31586 (2005). Kidney Diseases DK098408 (to A.M.C.), Institut National de la Sante´ et de la Recherche 11. Tanida, I., Minematsu-Ikeguchi, N., Ueno, T. & Kominami, E. Lysosomal turnover, Medicale (to P.C. and I.B.), Institut National du Cancer and L’Agence Nationale de la but not a cellular level, of endogenous LC3 is a marker for autophagy. Autophagy 1, Recherche (to P.C.), and the generous support of Robert and Renee Belfer (to A.M.C.). 84–91 (2005). O.P. was supported by a Basque Government Postdoctoral Fellowship. 12. Klionsky, D. J. et al. Guidelines for the use and interpretation of assays for monitoring autophagy. Autophagy 8, 445–544 (2012). Author Contributions O.P. designed and performed most of the experiments, analysed 13. Pazour, G. J., San Agustin, J. T., Follit, J. A., Rosenbaum, J. L. & Witman, G. B. the data and contributed to writing the paper; I.O. and I.B. performed and analysed the Polycystin-2 localizes to kidney cilia and the ciliary level is elevated in orpk mice experiments related to Hh signalling; B.P. performed the electron microscopy studies with polycystic kidney disease. Curr. Biol. 12, R378–R380 (2002). and morphometric analysis; S.S. performed the cytosolic vesicle experiments; A.D.-C. 14. Schneider, L. et al. PDGFRaa signaling is regulated through the primary cilium in assisted with cell culture; P.C. conceived the part of the study related to Hh signalling, fibroblasts. Curr. Biol. 15, 1861–1866 (2005). provided interpretation to the data and contributed to the writing and revision of the 15. Boehlke, C. et al. Primary cilia regulate mTORC1 activity and cell size through Lkb1. paper; B.H.S. and P.S. set the bases for the rationale of the study, provided feedback in Nature Cell Biol. 12, 1115–1122 (2010). the interpretation of the data and revised the written manuscript; A.M.C. coordinated 16. Finetti, F. et al. Intraflagellar transport is required for polarized recycling of the the study, designed experiments, analysed data and contributed to the writing and TCR/CD3 complex to the immune synapse. Nature Cell Biol. 11, 1332–1339 revision of the paper. (2009). 17. Ravikumar, B., Moreau, K., Jahreiss, L., Puri, C. & Rubinsztein, D. C. Plasma Author Information Reprints and permissions information is available at membrane contributes to the formation of pre-autophagosomal structures. www.nature.com/reprints. The authors declare no competing financial interests. Nature Cell Biol. 12, 747–757 (2010). Readers are welcome to comment on the online version of the paper. Correspondence 18. Veland, I. R. et al. Inversin/nephrocystin-2 is required for fibroblast polarity and and requests for materials should be addressed to A.M.C. directional cell migration. PLoS ONE 8, e60193 (2013). ([email protected]) or P.S. ([email protected]).

METHODS 100k pellet was further subjected to two successive centrifugations; 300,000g and 2 2 Cells and reagents. MEFs were prepared as previously described26. Atg5 / MEFs 500,000g, and the pellets from each of these centrifugations were collected as the 1 1 2 2 2 2 were a gift from N. Mizushima, WT (Tg737 / ), Ift88 / (Tg737 / ) mouse 300k and 500k crude vesicular fractions, respectively. 2 2 kidney epithelial cells were from G. Pazour, Ptc / MEFs were from P. Scott Co-immunoprecipitation assays. Cytosol was resuspended in a buffer contain- u and RGC-5 the cell line from P. Boya. Inversin–GFP 3T3s were from I. R. ing 0.5% NP-40, and incubated over-night with the desired antibody at 4 C. The u fraction bound to antibodies was precipitated using either protein A- or protein Veland and S. T. Christensen. All cells were cultured at 37 C and 5% CO2 in Dulbecco’s modified Eagle’s medium (Sigma) supplemented with 10% fetal bovine G-coated beads, eluted and subjected to SDS–PAGE as described in the general serum (FBS) and 1% penicillin/streptomycin/fungizone. Serum removal was per- methods. formed by thoroughly washing the cells with Hanks’ balanced salt solution Immunocytochemistry and morphometric analysis. Cells grown on coverslips (Invitrogen) and placing them in serum-free medium. Where indicated, cells were were fixed either in 4% paraformaldehyde or 100% cold methanol, blocked and treated with the macroautophagy inhibitor 3-methyladenine (10 mM) (Sigma) or permeabilized, and incubated with primary antibodies (1:200 except for acety- with NH Cl (20 mM) and leupeptin (100 mM) (Fisher BioReagents) to inhibit lated tubulin which was 1:1,000) and fluorophore-conjugated secondary antibod- 4 9 lysosomal proteolysis or with rapamycin (100 mM) (Calbiochem) to activate ies (1:200) sequentially. Mounting medium contained DAPI (4 ,6-diamidino-2- autophagy. For cilia resorption experiments cytochalasin D (0.5 mM) (Tocris) phenylindole) to highlight the nucleus. Images were acquired with an ApoTome.2 2 and PDGF (50ng ml 1) (R&D systems) were used. Purmophamine was from system using an Axiovert 200 fluorescence microscope (Carl Zeiss) equipped with a 3 Calbiochem and cyclopamine from Sigma. The plasmid for EGFP–LC3 was from 63 1.4 NA oil objective lens and red (excitation 570/30 nm, emission 615/30 nm), Addgene, and inversin–GFP was from I. R. Veland and S. T. Christensen. The cyan (excitation 365/50 nm and emission 530/45 nm) and green (excitation 475/ antibodies against acetylated tubulin, beta-tubulin, gamma-tubulin (mouse) and 40 nm and emission 535/45 nm) filter sets (Chroma), and prepared using Adobe ULK-1 were from Sigma; against IFT20 and IFT88 from Proteintech (rabbit) and Photoshop 6.0 software (Adobe Systems). 3D reconstruction images were modelled Abcam (mouse); against AMBRA1, ATG5/12, beclin, IFT88 and VPS15 from as mixed renderings using the Inside4D module for AxioVision Rel. 4.8. after apply- Novus; against ATG7 (immunoblot) and LC3 from Cell Signaling; against ing the Nyquist sampling criteria. For quantification of the number of ciliated cells ATG7 (immunofluorescence) from Serotec; against GABARAP and gamma-tubulin Z-stack sections were acquired and orthogonal views generated using AxioVision (rabbit) from Santa Cruz; against LC3 (immunogold), ATG14, and ATG16(L) from Rel. 4.8. Cilia length was measured using AxioVision Rel. 4.8 and number of fluor- MBL; against actin and detyrosinated alpha-tubulin (Glu-tubulin) from AbCam; escent puncta using ImageJ (NIH) after thresholding of the images. Quantification of against VPS34 from Invitrogen; against SMO from AbCam, against LAMP-2 from colocalization was performed using Image J (NIH) in individual frames after thresh- the Iowa Hybridoma Bank and against clathrin heavy chain (mouse) from BD olding, and colocalization was calculated with the JACoP plugin of the same program Biosciences. The rest of the chemicals were from Sigma. in merged images. Autophagic measurements. Intracellular protein degradation was measured by Transmission and scanning electron microscopy and morphometric analysis. metabolic labelling and pulse-chase experiments as described before27.Briefly,con- Cells grown to confluence were pelleted and fixed in 2.5% glutaraldehyde in 2 fluent cells labelled with [3H]leucine (2 mCi ml 1) (NENPerkinElmer Life Sciences) 100 mM sodium cacodylate (SC) pH 7.43 at room temperature for 45 min. The for 48 h at 37 uC were extensively washed and maintained in medium supplemented pellet was then rinsed in SC, post-fixed in 1% osmium tetroxide in SC followed by or not with serum with an excess of unlabelled leucine (2.8 mM). Aliquots of the 1% uranyl acetate, dehydrated through a graded series of ethanol, and embedded medium taken at different time points were precipitated in trichloroacetic acid and in LX112 resin (LADD Research Industries). Ultrathin sections were cut on a proteolysis was measured as the amount of acid-precipitable radioactivity (proteins) Reichert Ultracut E, stained with uranyl acetate followed by lead citrate. transformed in acid-soluble (amino acids and small peptides) at each time. This Immunogold labelling was performed in ultrathin sections of isolated cytosolic assay quantifies the degradation of the pool of long-lived proteins that become vesicles fixed in 4% paraformaldehyde/0.1% glutaraldehyde in sodium cacody- radiolabelled during the 48-h pulse. Lysosomal degradation was determined as late, dehydrated and embedded in Lowicryl. Grids were washed in 50 mM glycine the percentage of degradation sensitive to inhibition by ammonium chloride and in PBS blocked, either single labelled with ATG16L or IFT20 or double labelled leupeptin. Autophagic flux was measured by immunoblot as changes in levels of with ATG16L and IFT20 antibody for 2 h, washed extensively and incubated with LC3-II upon inhibition of lysosomal proteolysis (net flux) and autophagosome the gold-conjugated secondary antibodies (1:100) for 2 h. Control grids were formation as the increase in LC3-II levels at two consecutive times during lysosomal incubated with the secondary antibody alone or with an irrelevant immuno- proteolysis inhibition11. Autophagosome content was evaluated as the number of globulin G. After extensive washing, samples were fixed a second time for fluorescent puncta in cells transfected with GFP–LC3 or after immunostaining with 5 min in 2% glutaraldehyde, washed and negatively stained with 1% uranyl acet- antibodies against endogenous LC3. Autophagic flux was also tested by transient ate for 15 min. All grids were viewed on a JEOL 1200EX transmission electron transfection of the mCherry–GFP–LC3 plasmid, which was a gift from T. Johansen. microscope at 80 kV. Scanning electron microscopy was performed by the Constructs were transfected using either Lipofectamine 2000 or Optifect reagent Analytical Imaging facility at the Albert Einstein College of Medicine, following (Invitrogen) according to manufacturer’s instructions. Quantification of the yellow standard protocols30. Images were acquired on a Zeiss Supra 40 Field Emission and red puncta was performed by Green and Red Puncta Colocalization Macro for Scanning Electron Microscope at 5 kV. Morphometric analysis of transmitted ImageJ(D.J.SwiwarskimodifiedbyR.K.Dagda). and scanning electron micrographs was done using Image J (NIH). Lentivirus-mediated shRNA silencing. The transfer vector plasmids containing mRNA quantification. Semi-quantitative PCR was performed after extracting the Ift20 and Ift88 shRNA sequence were purchased from Sigma Mission Library total RNA using the RNeasy Protect Mini Kit (Qiagen) following the manufac- (Clone ID NM_018854.4-417s21c1 and NM_009376.1-762s1c1). Lentiviral part- turer’s indications. The first strand cDNA was synthesized from 0.5 mg of the total icles were generated by co-transfection with the third-generation packaging con- RNA with the SuperScript II RNase H Reverse Transcriptase (Invitrogen) and structs pMDLg/pRRE and pRSV-REV, and as envelope the G glycoprotein of the oligo(dT)12–18 primers. The expression levels were normalized to levels of VSV (pMD2.G) into HEK293T cells as described before26. Cultured cells were b-actin in the same samples after amplification using the SYBR Green PCR kit 2 transduced by addition of packed virus at a titre of 2.63 3 106 units ml 1. (PE Biosystems) and the following primers from the QuantiTect Primer Assay Isolation of subcellular factions. Cilia were isolated by peel-off technique as (Quiagen): Gli1 (QT00173537) Gli2 (QT01062236); Ptc1 (QT00149135) and previously described28. Briefly, a poly-L-lysine-coated coverslip was placed on mouse b-actin 59-GGCTGTATTCCCCTCCATCG-39. Differences between sam- top of kidney epithelial cells cultured in absence of serum. After application of ples were calculated using the DDCt method. light pressure, the poly-L-lysine-coated coverslip containing the isolated cilia was Statistical analysis. Results are shown as mean 6 s.e.m. or mean 6 s.d., and lifted up and processed for immunofluorescence. Autophagosomes and autop- represent data from a minimum of three independent experiments unless other- hagolysosomes were isolated by differential centrifugation and floatation in wise stated. Student’s t-test for unpaired data was used for statistical analysis and metrizamide gradients as previously described29. The cytosolic fraction was one-way analysis of variance (ANOVA) was used for multiple comparisons. For obtained by centrifugation for 1 h at 100,000g of the supernatant obtained after cilia growing curves nonlinear regression was performed applying the least separating the autophagosome-autophagolysosome-enriched fraction. squares fitting method with Sigma Plot software. A value of P , 0.05 was con- Isolation of cytosolic vesicles. Vesicles were isolated from 10-cm confluent sidered statistically significant. plates of MEF cells as described before19. Briefly, cells collected in PBS were General methods. Cells were solubilized in RIPA buffer (150 mM NaCl, 50 mM disrupted in homogenizing buffer (0.25 M sucrose, 20 mM MOPS, pH 7.3) by Tris pH 5 8, 1% NP-40, 0.5% NaDoc, 0.1% SDS) and protein concentration was nitrogen cavitation (nitrogen bomb; Parr Instrument Company) applying a pres- determined by the Lowry method. After SDS–PAGE and immunoblot, the pro- sure of 3,500 p.s.i. for 7 min. Cell homogenate was centrifuged at 2,500g for teins recognized by the specific antibodies were visualized by chemiluminescent 15 min to pellet the cell debris (heavy particles, or fractions of cells) and super- HRP substrate from Pierce in a Fujifilm Las-300 Imager. Densitometric quan- natant was centrifuged at 100,000g for 60 min in a TLA 110 rotor (Beckman) to tification of the bands was done using the square method with ImageJ (NIH), and pellet all major organelles and was designated as 100k. The supernatant of this all values were corrected by actin or tubulin as labelled in the figures.

Extended Data Figure 1 | Schematic model of the interplay between under these conditions molecules of SMO that are mobilized to the ciliary autophagy and ciliogenesis. Starvation conditions (serum2) increase membrane after release of the inhibitory effect of Ptc. Activated SMO induces intraflagellar transport (IFT) and ciliogenesis and Hedgehog (Hh) signalling. the downstream transcription factors GLI1/GLI2 which further favours We show in this work that both events increase autophagosome formation recruitment and assembly of ATGs in the base of the cilium. Normal nutritional interdependently. ATG16L is trafficked in IFT20-enriched vesicles to the base conditions (serum1) lead to decreased Hh and IFT and suppression of of the cilia where most ATG localize. Arrival of ATG16L to this ATG assembly starvation-induced autophagy. Basal autophagy contributes to maintain and/or its transport along the axoneme seems to initiate the elongation of ciliogenesis to a minimum through the degradation of IFT20, which in turns ATG5 structures and presence of lipid-binding ATGs (such as LC3 and reduces trafficking of ATG16L into the ATG assembly at the base of the cilium, GABARAP) in the cilia and at the plasma membrane. The cilium also contains preventing induction of ciliary autophagy.

Extended Data Figure 2 | Ciliogenesis and LC3 in different cell types. ganglion cells (RGC-5; **P 5 0.004, ##P 5 0.0002, n 5 5) maintained in the a, Immunoblot for IFT20 and IFT88 and quantification in MEFs control (Ctrl) presence or absence of serum. d, Immunoblot for LC3 in MEFs and KECs or knocked down (2) for IFT20 (**P 5 0.007, n 5 7), or kidney epithelial cells maintained in the presence of serum and treated or not with protease inhibitors 2 2 2 2 (KECs) wild-type (WT) or knockout for IFT88 (Ift88 / ; **P 5 0.0003, n 5 3). (PI). e, Single-channel images and merge of WT and Ift88 / KECs transfected ADU, arbitrary densitometric units (n 5 3). b, Rates of degradation of long- with the mCherry–GFP–LC3 reporter. f, Immunofluorescence for LC3 in WT 2 2 lived proteins in the same cells expressed as a percentage of total proteolysis and Ift88 / KECs upon serum removal. Quantification of LC3-positive (top) or as a percentage of lysosomal degradation (sensitive to inhibition of puncta (*P 5 0.04, n 5 3). Scale bars, 10 mm. Mean 6 s.e.m. unless otherwise lysosomal proteases) (*P 5 0.021, n 5 3). c, Ciliated cells in MEFs stated. (**P 5 0.003, n 5 5), KECs (*P 5 0.011, *P 5 0.021, n 5 5) and retinal

Extended Data Figure 3 | Autophagy in IFT88 and IFT20 double- presence or absence of serum. d, Quantification of LC3-II levels (S1, knockdown cells. a, Immunoblot for the indicated proteins of MEFs Ctrl or **P 5 0.0008, **P 5 0.004, n 5 4; S2, *P 5 0.034, n 5 5) and LC3-II flux (S1, 2 2 2 knockdown for IFT88 (IFT88 ) or for IFT88 and IFT20 (IFT20 /IFT88 ). n 5 4; S2, *P 5 0.016, n 5 5) by densitometry of blots as the ones shown in b, Percentage of ciliated cells after 24-h serum starvation and representative c. Mean 6 s.d. e, LC3 flux in a retinal ganglion cell line (RGC-5) and in KEC images of cilia (acetylated tubulin, green) and basal body (c-tubulin, red) in the treated or not with platelet-derived growth factor (PDGF). Scale bars, 10 mm. 2 2 same cells(IFT88 , **P 5 0.00001; IFT20-88 , **P 5 0.0006, n 5 5). Arrows Mean 6 s.e.m. unless otherwise stated. indicate cilia. c, LC3 flux immunoblot in the same cell lines maintained in

2 2 Extended Data Figure 4 | Ciliary hedgehog signalling modulates autophagy. control cells treated or not with purmorphamine, and in Ptc / cells. Right, a, Scheme of chemical and genetic approaches to modulate hedgehog (Hh) quantification (n 5 3). e, mCherry–GFP–LC3 reporter in cells treated with 2 signalling. Both the agonist purmorphamine (Purmo) and genetic ablation of purmorphamine. f, LC3 flux in IFT88 MEFs treated with purmorphamine. 2 2 the Patched-1 receptor (Ptc / ) result in the recruitment of SMO to the cilia Right, quantification of LC3-II flux (Mean 6 s.d.; n 5 3). g, Ptc mRNA levels in 2 2 2 and initiation of expression of GLI factors. Knockdown of SMO or treatment MEF WT and Ptc1 / (n 5 3). h, Immunofluorescence for acetylated tubulin 2 2 2 2 with the SMO antagonist cyclopamine (Cyclo) suppress activation of in MEF WT and Ptc / . i, Percentage ciliated cells in WT and Ptc / MEFs downstream effectors. b, mRNA expression of Hh downstream effector genes (*P 5 0.014, ##P 5 0.0009, n 5 3). j, Relative mRNA expression by RT–PCR of relative after PDGF-induced ciliary resorption relative to untreated Serum2 Hh downstream effector genes in control and myc–GLI1 cells (Gli1, cells (Gli1, **P 5 0.0006, **P 5 0.007; Gli2, **P 5 0.0001; Ptc, **P 5 0.002; **P 5 0.0017; Bcl2 **P 5 0.004; Ptc **P 5 0.0006; Hhip *P 5 0.03; Gli2 n 5 3). c, mRNA levels for downstream target genes of Hh signalling measured *P 5 0.028; n 5 3). k, LC3 flux immunoblot in MEFs treated or not with Cyclo by RT–PCR after the indicated treatments (Smo; Ctrl *P 5 0.028, and PI. Quantification of LC3-II flux (n 5 3). Differences with Ctrl (*) or with purmorphamine *P 5 0.027; Gli1, ctrl **P 5 0.006, purmorphamine Serum 1 (#) are significant for P , 0.05. n.s., statistically non-significant. *P 5 0.037; Bcl-2, Ctrl **P 5 0.007; n 5 3). d, Immunofluorescence for LC3 in Mean 6 s.e.m. unless otherwise stated. Scale bars, 10 mm.

Extended Data Figure 5 | Presence of ATGs at the primary cilia. a, Co- cells transiently transfected with either GFP–LC3 or GFP–inversin and in immunostaining for the indicated autophagy-related proteins (ATGs; green) GFP–inversin 3T3 stable cell lines. d, Immunostaining for ATG16L in NRK Ctr and acetylated tubulin (red) of mouse kidney epithelial cells maintained in the or knockdown for ATG16L transiently transfected with GFP–inversin to absence of serum for 24 h. b, 3D reconstruction of the co-staining for ATG16L highlight the primary cilia. Individual channels, merge and 3D reconstitution of 2 and acetylated tubulin in the cilia. ATG14 is shown an as example of absence of the co-staining are shown. e, Immunoblot for ATG16L in ATG16L NRK cells. colocalization. 0.2-mm Z-stack are shown from the surface to the bottom part of Scale bars, 10 mm. the cilium. c, Staining with ATG16L or acetylated tubulin, and 3D images in

Extended Data Figure 6 | Autophagy-related proteins associate with the (yellow) or no colocalization (white) in the centriole. c, Co-immunostaining of 2 2 basal body in a serum-dependent manner. a, b, Co-immunostaining for ATG7 (green) with gamma tubulin (red) in WT and Ift88 / KECs maintained ATGs and gamma tubulin of kidney epithelial cells (KECs) maintained in in the presence or absence of serum for 24 h. Scale bars, 10 mm. serum-free (a) or serum-supplemented media (b). Arrows, colocalization

Extended Data Figure 7 | Changes in intracellular distribution of ATGs in KECs. Arrows indicate clusters of ATG5. c, Immunostaining for ATG5 in WT 2 2 kidney epithelial cells with defective IFT. a, Immunostaining for ATG7, and Ift88 / KECs maintained in the absence of serum for 24 h. 2 2 ATG14, LC3 and GABARAP in WT and Ift88 / kidney epithelial cells (KECs) d, Immunostaining for ATG5 (green) and acetylated tubulin (red) in WT or 2 2 maintained in the presence or absence of serum for 24 h. b, ATG5 co- Ptc / MEFs. Yellow arrow, ATG5 over ciliary structures. Scale bars, 10 mm. immunostaining (green) with gamma tubulin or acetylated tubulin (red) in

Extended Data Figure 8 | Interaction of IFT20 with ATGs. a,Co- IFT20 (red) with VPS15, ATG7 and ATG14 (green) in WT KECs during immunostaining for ATG16L and c-tubulin in kidney epithelial cells (KECs) starvation. Insets show boxed areas at higher magnification. Yellow arrows, 2 2 WT or Ift88 / maintained in the presence or absence of serum for 24 h. colocalization. Percentage of colocalization is indicated. e–g, Immunoblot for Arrows, colocalization (yellow) or no colocalization (white) at the basal body the indicated proteins after coimmunoprecipitation for IFT20 (e, f) or ATG14 (BB). b, Percentage of WT MEFs untreated (2), treated with purmorphamine (g) in the same cells. h, Immunogold electron microscopy for ATG16L and 2 2 (*P 5 0.01, n 5 4) or Ptc / (*P 5 0.005, n 5 4) showing colocalization of IFT20 in isolated cytosolic vesicles. Full-field images of double immunogold ATG16L and BB in the absence of serum. c, Co-immunostaining for IFT20 and staining for IFT20 (10 nm) and ATG16L (15 nm) captured by transmission ATG16L in the same cells as in a. Insets show split channels of boxed areas at electron microscopy. ATG16L alone (yellow), IFT20 alone (blue), both proteins higher magnification. Arrows, colocalization. Right: quantification of the (red). Right image shows absence of gold particles in a region of vesicle-free colocalization (##P 5 0.004, **P 5 0.004, n 5 4). d, Co-immunostaining for film. Mean 6 s.e.m. unless otherwise stated. Scale bars, 10 mm.

2 Extended Data Figure 9 | Effect of blockage of autophagy on ciliogenesis. percentage of ciliated cells in ATG14 cells. Cells from Ift88 null mice are used a, Co-immunostaining for acetylated tubulin and gamma tubulin in MEFs from as negative control in e,(n 5 4). f, Percentage of ciliated MEFs upon treatment 2 2 WT or ATG5 null mice (Atg5 / ) maintained in the absence of serum for the with 3-methyladenine (3MA) or rapamycin (Rapa) in the presence or absence indicated periods of time. Arrows, cilia. b, Immunostaining for acetylated and of serum (n 5 3). g, Percentage of ciliated retinal ganglionar cells (RGC-5) at gamma tubulin in MEFs control or knocked down (2) for ATG7, and the indicated times of treatment with 3MA or rapamycin in the presence of maintained in the presence or absence of serum. Arrows, cilia. c, Quantification serum (n 5 3). Scale bars, 10 mm. Arrows; cilia. All values are mean 6 s.e.m. 2 of the percentage of ciliated cells in ATG7 (n 5 5). d, Immunostaining for unless otherwise stated. Differences with Ctrl (*) or with Serum1 (#) are 2 acetylated tubulin in KEC control or ATG14 . e, Quantification of the significant for P , 0.05. n.s., statistically non-significant.

Extended Data Figure 10 | Ultrastructure of the primary cilia in autophagy- Arrows, cilia-associated vesicles. Arrowhead, ciliary pocket. deficient cells. Scanning electron microscopy images of embryonic fibroblasts c–e, Morphometric analysis of the cilia; cilia diameter (S2, **P 5 0.0002, 2 2 from WT and ATG5-defective mice (Atg5 / ) grown in the absence or n 5 28; S1, **P 5 0.026, n 5 24) (c), area of the ciliary pocket (S2, presence of serum. a, Surface of the different cells to highlight lower levels of **P 5 0.009, n 5 28; S1, **P 5 0.0001; n 5 24) (d), and exosome diameter 2 2 2 2 villi in the Atg5 / cells. Red arrow, primary cilia. b, Details of primary cilia. and number of exosomes per cilia (e) in WT and Atg5 / MEFs. Mean 6 s.d.

Extended Data Figure 11 | Enhanced ciliogenesis in autophagy-deficient proteins in WT MEFs treated or not with ammonium chloride and leupeptin cells. a, Co-immunostaining for SMO and acetylated tubulin in WT and (N/L) and collected at different times after serum removal. d, Time-course of 2 2 2 2 Atg5 / MEFs treated or not with purmorphamine. Arrows, colocalization changes in IFT20 protein levels in WT and Atg5 / cells during serum removal (yellow) or no colocalization (white). b, Immunofluorescence for IFT20 in MEF relative to levels in serum supplemented WT MEFs. Time-course of changes in 2 2 WT and Atg5 / . Arrows, IFT20 cytosolic vesicles. Scale bars, 10 mm. All values LC3-II flux is plotted as discontinuous line. are mean 6 s.e.m. unless otherwise stated. c, Immunoblot for the indicated

1.1. Discussion 1: Functional interaction between autophagy and primary cilium

In this work, we have identified a novel reciprocal relationship between the primary cilium and autophagy. We were able to show that induction of autophagosome formation depended on serum, functional cilia, and the ciliary Hedgehog pathway. We have observed that components of the autophagic machinery are located at the axoneme and basal body of the cilium (see Figure 3 of article and Table 3 of the manuscript for localization of different autophagic machinery components to the cilia). The localization of autophagy initiating Atg proteins at the base of the cilium suggests that sensing of nutrient deficiency by the cilium, or activation of signaling from this organelle could initiate a class of cilia- mediated autophagic induction mechanism. This hypothesis is consistent with the reduced autophagy in cells with compromised ciliogenesis, and with the recruitment of one of the early markers of autophagy, Atg16L1 to the BB upon starvation. A functional PC is required for the recruitment of Atg16L1 to the BB upon serum removal. IFT20 protein is also delivered to the ciliary axoneme from the Golgi apparatus in Atg16L1-containing vesicles. The delivery of IFT20 to the cilia is crucial for cilia assembly and localization of various ciliary proteins including polycystin 2 (PC2). The Hh signaling emanating from the cilia triggers the Golgi-to-cilia shuttling of these IFT20 and Atg16L1 positive vesicles. This serum- dependent relocation of specific Atg proteins to pre-existing basal bodies leads to maximal activation of autophagy.

The recently described contribution of the plasma membrane to autophagosome formation (Moreau, Ravikumar et al. 2011; Puri, Renna et al. 2013) and the presence of a subset of Atg proteins along the ciliary membrane make a direct regulatory effect of PC on plasma membrane autophagosome biogenesis an attractive possibility. The ciliary presence of the pre-autophagosomal Atg16L1 and Atg proteins, such as LC3 and GABARAP that remain in the limiting membrane of the autophagosome and the enrichment of this later group at the plasma membrane in an IFT-dependent manner suggest that cilia-mediated autophagy may be closely related to the formation of autophagosomes from the plasma membrane. The hypothesis of the base of the cilium which is continuous to the plasma membrane as a novel site for autophagosome biogenesis is supported by the previously known role of the ciliary pocket as this site was identified to have high activity in vesicular trafficking (Molla-Herman,

Ghossoub et al. 2010). However it should be noted that Beclin-1, a core component of the autophagic machinery is not observed to be recruited to the cilia which may question the canonical pathway of cilia-dependent autophagy induction.

In contrast of the cilia-dependent autophagy induction in response to serum removal, we have observed that basal autophagy has an inhibitory effect on ciliogenesis as chemical or genetic abrogation of autophagy under normal nutritional conditions is sufficient to induce ciliogenesis. This inhibitory effect of constitutive autophagy occurs likely by limiting the trafficking of components required for ciliary growth to this organelle, such as IFT machinery. IFT machinery is required for the regulation of cilia length in order to deliver additional axonemal subunits at the distal tip (Scholey 2008). We observed that basal autophagy degrades IFT20 thereby limits the shuttling of the IFT20 positive vesicles between the Golgi and PC and reduce ciliary growth. The switch between basal and induced autophagy and how IFT20 escapes autophagic degradation upon starvation needs further investigation. This inverse effect of ciliogenesis and autophagy on each other suggests that the temporal coincidence of their activation at the early times of serum deprivation may serve as a self- regulatory brake. Thus, the absence of nutrients would favor formation of the cilium and this will upregulate the autophagic process, which in turn will limit ciliary growth and prevent in this way massive upregulation of the autophagic process itself (Extended Data Fig. 1 of article).

It is also shown that negative regulators of ciliogenesis such as the centriolar satellite protein OFD1 can be degraded by serum starvation induced-autophagy (Tang, Lin et al. 2013). OFD1 functions as a suppressor of ciliogenesis at the centriolar satellites in basal conditions. Consequently, the degradation of the centriolar satellite pool of OFD1 by autophagy induces ciliogenesis. The dual role of autophagy as positive and negative regulator of ciliogenesis seems to be determined by the type of the autophagic process engaged and the conditions. Interestingly, as starvation persists, IFT20 becomes again the preferred cargo for autophagy, and degradation of the ciliary vesicular components imposes a negative control on ciliary growth. The switch in autophagic cargo for basal and inducible autophagy, determined in basal conditions, IFT20, a positive regulator of ciliogenesis, is degraded by autophagy while OFD1 is located in the centriolar satellites and inhibits ciliogenesis via BBS4 sequestering. Upon serum starvation of cultured cells, ciliogenesis is induced following the degradation of OFD1 by autophagy, and, in this state, IFT20 protein regulates ciliary trafficking. This switch between basal and induced cilia-dependent autophagy suggests the

91 presence of an important link between autophagy and ciliogenesis for cell homeostasis. An interesting aspect that requires clarification is the contradictory reports of cilia length in autophagy-deficient MEFs between the two studies under basal conditions. The varying degrees of confluence of the cells on the day of the experiment– known to be an important factor during ciliogenesis – might explain the presence of the contradictory results.

Although cultured cells induce ciliogenesis in response to nutrient deprivation, the primary cilium is constitutively present in cells of most tissues. The fact that blockage of ciliogenesis in cells with abundant cilia in basal conditions does not affect their basal autophagy but compromises the ability of the cells to activate starvation-induced autophagy, supports the idea that it is not the presence of a cilium but the activation of ciliary signaling by starvation that contributes to autophagic induction.

In this study, we have identified the regulatory effect of the Hh signaling pathway, one of the best-characterized cilium dependent signaling pathways, on autophagy. Because proteins exiting the cilium find themselves at the basal body where most of the regulatory Atg proteins reside, we propose that the arrival of these signaling proteins may act as a trigger for autophagosome formation. In support of this idea, constitutive upregulation of Hh signaling in permanently ciliated Ptc1-/- cells enhances association of Atg proteins to the basal body, elongation of the Atg5 hook-like structures and abundance of LC3 along the primary cilium. In addition to these local changes, constitutive activation of Hh is enough to reproduce changes in the overall cellular distribution of Atg proteins, such as Atg7, further enhancing the possibility of a cilium mediated signal amplification to activate the autophagic machinery. We propose that the four Atg proteins located in the primary cilium - Vps15, AMBRA, LC3 and GABARAP – also reach this compartment in an IFT-dependent manner. LC3 and GABARAP may gain access to the cilium by lateral diffusion after reaching the plasma membrane since as we have shown in this work, nutrient starvation induces the recruitment of these two proteins towards the plasma in an IFT-dependent manner. Transport of some or all of these molecules into the cilium may depend on their lipid conjugation, a property shared by the four of them.

In an intriguing manner, it was previously shown that in ciliated glial cells, cilia- dependent Hh signaling favored cell survival under serum starvation independent of cell proliferation and transcriptional regulation (Yoshimura, Kawate et al. 2011). Our study explains that in serum deprivation conditions, ciliary Hh-dependent induced autophagy

92 enables the survival of cells which can include fibroblasts and epithelial cells as well ciliated adult central nervous system neurons. The question of how Sonic hedgehog (Shh) molecules are secreted in these stressed neurons and if autophagy could play a role in this state remains to be answered.

However, the relationship between autophagy and Hh signaling appears to be complex. The study of Jimenez-Sanchez et al. showed that Hh signaling represses autophagy in fibroblasts, HeLa cells, and Drosophila melanogaster (Jimenez-Sanchez, Menzies et al. 2012). These opposite effect of Hh signaling on autophagy could be in part determined by the relative contribution of downstream Gli proteins in each cell type as some of them have opposite effects in other cellular processes. For example, our study shows that Gli1 overexpression alone is capable to rescue the autophagic phenotype, whereas the inhibitory effect on autophagy has been more closely linked to changes in Gli2. Interestingly, HeLa cells are poorly ciliated cells. Until recently it was known that that in Drosophila melanogaster PC is not required for Hh signaling (Goetz and Anderson 2010). In fact, a recent study shows that in Drosophila olfactory sensory neurons, the Hh pathway is mediated via the PC and suggests that there can be two Hh pathways in the organism, cilia-dependent and cilia-independent (Kuzhandaivel, Schultz, et al. 2014). It would be important to consider the presence or absence of cilia of the model studied in this research in order to conclude the Hh-autophagy crosstalk. In support of the hypothesis that autophagy is stimulated by a cilia-dependent Hh pathway, Shh treatment induced autophagy in ciliated vascular smooth muscle cells (Li, Li et al. 2012) and in ciliated hippocampal neurons (Petralia, Schwartz et al. 2013). In this respect, ciliary Hh signaling may exert its stimulatory effect on autophagy by directly influencing Atg proteins located at the base of this organelle, and thus be different from the signaling originating from other cellular locations. The role of Hh signaling in autophagy regulation is likely to be dependent on multiple parameters such as cell growth, the presence of PC and localization of signaling components.

2. Objective 2: Study of the role of autophagy in integration of mechanical stress via primary cilium

Autophagy contributes to the maintenance of homeostasis of different tissues and organs through the turnover of proteins and organelles (Mizushima, Levine et al. 2008). Autophagy is known to have physiological roles in different organs, tissues or stages of development at basal levels or upon integration of different extracellular inputs. These inputs include starvation, DNA damage, infection, hypoxia or mechanical stress. We showed the interaction between the primary cilium (PC) and autophagy in basal conditions and in response to serum starvation. There are three types of mechanical stress in different tissues and organs; compression, stretching or shear stress induced by fluid flow (King, Veltman et al. 2011). Fluid flow can be sensed by primary cilia in the circulatory system and epithelial cells in the kidney tubules. Previous studies show that PC is important in calcium signaling, maintenance of planar polarity and cell size regulation in kidney epithelial cells under fluid flow.

The objective of this study was to investigate the role of the interplay between autophagy and primary cilia in a physiological state, specifically in response to mechanical stress. We used an in vitro model to induce constant flow on ciliated kidney epithelial cells and studied the role of autophagy by various methods.

• To determine if fluid flow in kidney epithelial cells induces an autophagic response we induced a 1 dyn/cm2 fluid shear stress using an IBIDI pump to create a constant fluid flow on MDCK cells. In order to observe the autophagic response we used MDCK cells expressing GFP-LC3 or RFP-GFP-LC3 and observed autophagy initiation and autophagic flux. • In order to determine if autophagy activation upon fluid flow was PC-dependent, we used kidney epithelial cells (KEC) derived from wild-type and IFT88 hypomorphic mice which have impaired ciliogenesis. We observed endogenous LC3 and its lipidation to quantify autophagosome initiation and autophagic flux in response to flow. To confirm that flow induced reduction of cell size and volume depended on PC we used our IFT88-/- cells.

• To define the role of autophagy during cell size regulation in response to fluid flow, we used lentiviral transfection to establish inducible shBeclin1 and shAtg5 KEC and applied flow to observe cell size regulation. • To confirm that autophagic machinery is recruited to the basal body (BB) of the cilium upon stimuli we observed the co-localization of early autophagic marker (Atg16L1) with the γ-tubulin (BB) and established that the recruitment of Atg16L1 upon flow was dependent on functional PC using WT and IFT88-/- cells. • To define the signalization pathway controlling the autophagy-dependent cell size regulation we first studied the Hedgehog pathway and mTOR pathway upon fluid flow.

2.1. Results

Flow stimulates the autophagic flux in MDCK cells

To determine the effect of fluid flow in kidney epithelial cells, first we used the well- known ciliated MDCK cell line. These cells, derived from the collecting duct of the canine kidney are commonly used to study kidney epithelial cell physiology. We have applied a physiologically relevant constant laminar fluid flow corresponding to 1 dyn/cm2 on the cells in a flow chamber to mimic the physiological conditions in kidney tubules in vitro. We observed the autophagosome formation by using stably transfected GFP-LC3 cells. Quantification of GFP-LC3 dots upon fluid flow for 4 hours, 1 day, 2 days and 4 days revealed a significant increase of autophagosome formation compared to resting cells cultures in static conditions. This increase was rapid after 4 hours and stabilized up to 4 days of flow (Figure 17 A-B). An increase of autophagosome number could correspond to an induction of autophagy by also to a blockage of autophagosome maturation into autophagolysosomes. To see if fluid flow disturbed autophagic flux, we used MDCK cells stably transfected with RFP- GFP-LC3, a pH-sensitive reporter that allows the visualization of autophagosomes before fusion with lysosomes as yellow puncta and autophagolysosomes as red puncta (since GFP is sensitive to the low lysosomal pH). The quantification of resting cells and flow-induced cells showed that a significant increase in yellow puncta is accompanied by a lower level but significant increase of red puncta. The overall content of autophagic vacuoles in cell subjected to flow was significantly higher compared to resting cells (Figure 17 C-D). These results suggested that fluid flow induced activation of autophagy and increase autophagosome biogenesis.

As previously shown MDCK cell volume is decreased under fluid shear stress (Boehlke, Kotsis et al. 2010; Heo, Sachs et al. 2012). In order to confirm cell size reduction in our time-course analysis of autophagy, we quantified the volume of cells subjected to flow, by 3D reconstruction after membrane-labelling. We observe a linear decrease in cell volume which becomes significantly smaller after 2 days of fluid flow (Figure 17 E-F). These observations suggest that autophagy is activated immediately following a mechanical stress stimuli on kidney epithelial cells which is followed by a volume reduction.

Inhibition of ciliogenesis impairs flow-induced autophagy

To determine the role of primary cilia (PC) on flow-induced autophagy, we used a cellular model with impaired ciliogenesis, the kidney epithelial cell line (KEC) derived from mice with a hypomorphic mutation of the IFT88 allele (IFT88-/-), an essential component of the IFT-B complex of PC (Figure 18 A). We applied fluid flow on wild-type and cilia- defected IFT88-/- cells starting from 4 hours up to 4 days and we quantified endogenous LC3- II puncta. We confirmed the induction of autophagy in wild-type cells by a significant increase of LC3 puncta after 1 day which stabilized during 4 days of flow (Figure 18 C-D). In contrast, IF88-/- cells did not show a significant increase of LC3 puncta at any time-point and when compared to ciliated cells subjected to flow, there were significant decrease in autophagosome levels (Figure 18 C-D). To confirm that this reduction of autophagosome content in IFT88-/- is accompanied by a defective autophagic activity, we analyzed the lipidation of LC3-II upon treatment with chloroquine, a lysomotropic agent blocking the maturation of autophagosomes into autophagolysosomes, which allows the measure autophagic flux in the cells. We found that cilia-defective cells had an impairment of their autophagic activity compared to ciliated cells subjected to 1 day of flow (Figure 18 E-F). To observe autophagy activation we also looked at the expression of autophagy substrate p62. Surprisingly we observed an increase of p62 expression in response to fluid flow in WT and IFT88-/- cells (Supp Figure 1A-B). However, the increase of protein expression was due to an increase of p62 transcription in response to mechanical stress. As a positive control we used the increase of p62 mRNA levels after prolonged starvation stimuli(Sahani, Itakura et al. 2014) (Supp Figure C). In this context, we can conclude that p62 is not a good substrate to monitor autophagy activation.

We quantified the cell volume in WT and IFT88-/- KECs and confirmed that cilia- defective cells did not have a reduction of cell volume compared to WT after 4days of flow (Figure 18 G-H). We observed that cells lacking the mechano-sensory property of PC, could no longer stimulate autophagy and regulate their cell volume. In addition they exhibited a more fibroblastic phenotype (data not shown) upon flow in vitro.

Inhibition of autophagy impairs primary cilium-dependent cell size regulation

In order to test if autophagy played a role in cell volume regulation in kidney epithelial cells under flow we generated a KEC cell line with inducible shRNA Atg5. When the cells are treated with IPTG, we observed a knock-down of the expression of the core autophagic protein Atg5 and therefore a dramatic defect in LC3-II lipidation consequently in autophagosome formation (Figure 19A). When the cells are subjected to flow in presence of IPTG, we observed less LC3 puncta than cells with functional autophagy as expected (Figure 19 B). Quantification of cell volume revealed that autophagy deficient shAtg5 cells remained larger under flow compared to WT cells (Figure 19C-D). These observations were confirmed by preliminary data obtained with KEC shBeclin1 cells subjected to flow (data not shown).

Overall, these data suggest that cilium-dependent activation of autophagy upon mechanical stress induced by fluid flow regulates cell size reduction.

Autophagic machinery (Atg16L1) recruited to the basal body upon fluid flow

As previously shown, PC-dependent autophagy activation by starvation is regulated by the recruitment of autophagic machinery to the axoneme and basal body (BB) of the cilia (Pampliega, Orhon et al. 2013). The most intriguing transport of autophagic protein seems to be the early-autophagic marker Atg16L1 who is actively recruited to the BB upon starvation. We hypothesized that Atg16L1 would be localized to the base of the cilium during a mechanical stress activation as well. Co-localization of Atg16L1 and γ-tubulin to stain BB, revealed that there was a significant increase of Atg16L1 at the BB upon flow induction and the presence of the protein became more prominent at the base when subjected to flow (Figure 20 A-C). This data supports the hypothesis of autophagosome biogenesis induction upon starvation, or mechanical stress, in a cilia dependent manner, near the plasma membrane where PC and its basal body resides.

Signaling regulation of primary cilium-dependent autophagy activation during flow

We wanted to determine the mechanism regulating autophagy activation under mechanical stress via PC. We first focused on the Hedgehog (Hh) pathway which we know to have a positive regulatory control on autophagy during serum starvation. We followed the transcriptional activation of Gli1, a transcription factor downstream of the pathway, and Ptc1, receptor of the pathway, in cells subjected to fluid flow. No significant increase of Gli1 and Ptc1 mRNA level was observed (Figure 21 A). We also studied the recruitment of Smo protein to the axoneme of PC, which is recruited to the cilia upon Hh activation by Smo agonist purmorphamine (Purmo) treatment. Under fluid flow, we did not observe any transcriptional activation of the Hh pathway and no localization of Smo to the axoneme (Figure 21 B). These data suggest that Hh pathway is not activated by mechanical stress therefore do not regulate the activation of autophagy in these conditions.

Secondly we studied mTORC1 pathway, one of the main regulatory pathways of autophagy that can control cell growth as well. Previously, mTORC1 was shown to be down- regulated after 6 days of flow (Boehlke, Kotsis et al. 2010). Since autophagy was induced rapidly at 4hours, and more significantly after 1 day of fluid flow (Supp Figure 1D), and before cell volume reduction, we studied mTORC1 pathway regulation in short-term. Following 1 day of flow, an increase in the expression of downstream proteins p70S6 and in phosphorylation of p70S6 Kinase (S6K) was observed (Figure 21 C). This data suggests that autophagy activation in response to fluid shear stress is mTORC1 independent. Other regulatory pathways should be investigated to define the mechanism behind the cilium- dependent autophagy activation in response to fluid flow.

Overall, we show that fluid flow induces autophagy activation and increases autophagosome biogenesis in a cilium dependent manner. Cell volume regulation in cells subjected to flow depends on the presence of autophagy activation and this regulation is independent of ciliary Hedgehog pathway or mTORC1 inhibition.

GFP-LC3

no Flow Flow

RFP-GFP-LC3

Figure 17: Flow stimulates the autophagic flux in MDCK cells

(A,B) Time course microscopy analysis of MDCK GFP-LC3 under flow conditions. Cells were subjected to flow or not from 4 hours to 4 days (d0, 4hrs, d1, d2, d4) then fixed and analyzed by apotome fluorescence microscopy. Scale bar, 10 µm. (B) Number of GFP+ puncta per cell were quantified by Image J software (see Materials and Methods for more details), *, p<0.05. (C,D) Microscopy analysis of MDCK RFP-GFP-LC3 under flow conditions. Cells were subjected to flow or not for 1 day then fixed. Scale bar, 10 µm. (D) Number of GFP+ RFP+ puncta (yellow) and GFP- RFP+ puncta (red) per cell were quantified by Image J software (see Materials and Methods for more details), *, p<0.05 (E,F) Time course microscopy analysis of cell size. 3D reconstruction of MDCK cells subjected to flow after membrane staining. Dark blue; small cells, white; large cells. (F) Mean cell volumes of MDCK cells under flow compared with non-flow, *, p<0.05.

100

Figure 18: Inhibition of ciliogenesis impairs flow-induced autophagy

(A,B) Microscopy analysis of wild type (WT) KEC and IFT88-/- KEC. Cells were fixed, labeled with acetyl-tubulin antibody and analyzed by apotome fluorescence microscopy. Arrowheads, primary cilium. (B) Number of ciliated cells were quantified by Image J software (see Materials and Methods for more details), *, p<0.05. (C,D) Time course microscopy analysis of wild type (WT) KEC and IFT88-/- KEC under flow conditions. Cells were subjected to flow or not from 4 hours to 4 days (do, 4hrs, d1, d2, d4) then fixed, labeled with anti-LC3 antibody and analyzed by apotome fluorescence microscopy. Images show cells subjected to no flow or 1 day of flow. Scale bar, 10 µm. (D) Number of endogenous LC3 puncta per cell were quantified by Image J software (see Materials and Methods for more details), *, p<0.05. (E,F) Wild type (WT) KEC and IFT88-/- KEC cells were subjected to flow or not (d0) for 1 day (d1) with or without chloroquine (CQ). LC3-II and actin were determined by immunoblotting (E) and densitometry (F). Immunobloting data: means means ± s.e., *, p<0.05 (t-test). (G-H) Microscopy analysis of cell size. 3D reconstruction of Wild type (WT) KEC and IFT88-/- KEC cells subjected to flow or not (d0) during 4 days (d4) before membrane staining. Dark blue; small cells, white; large cells. (H) Mean cell volumes of cells of Wild type (WT) KEC and Ift88-/- KEC under flow compared with non-flow, *, p<0.05.

101

Figure 19: Inhibition of autophagy impairs primary cilium cell size regulation

(A) Inducible shRNA Atg5 KEC were treated or not with IPTG and subjected to flow or not (d0) for 4 days (d4) Atg5 and LC3 expression were analyzed by immunoblotting. (B) Inducible shRNA Atg5 KEC were treated or not with IPTG and subjected to flow or not (d0) for 4 days (d4) After fixation, the cells were labeled with anti-LC3 antibody and analyzed by apotome fluorescence microscopy. Scale bar, 10 µm (C) Inducible shAtg5 KEC cells were subjected to 4 days of flow in absence (-) or presence (+) of IPTG to knock-down Atg5 expression. Mean cell volumes of cells are analywed after fixation and membrane staining. (D) 3D reconstruction of Inducible shAtg5 KEC cells treated or not with IPTG and subjected to flow during 4 days (d4) after membrane staining. Dark blue; small cells, white; large cells.

102

Figure 20: Autophagic machinery (ATG16L1) recruited to the basal body upon fluid flow

(A-E) Microscopy analysis of wild type (WT) KEC cells. Cells were subjected to flow or not (d0) for 1 day (d1) then fixed, labeled with γ-tubulin and Atg16L1 antibodies and analyzed by confocal fluorescence microscopy. Arrowheads, co-localization between Atg16L1 and γ-tubulin. (B) 3D reconstruction of Wild type (WT) KEC cells subjected to flow or not.(C) Two-fluorescence channel line tracings corresponding to dashed lines in the merge of panel A. (D) Quantification of the recruitment of Atg16L1 to basal body. Atg16L1 positive basal body were quantified by Image J software (see Materials and Methods for more details). (E) Pearson’s coefficients for Atg16L1 and γ- tubulin

103

S6K P

S6K tot

S6 P

S6 tot

Actin Actin

Figure 21: Signaling regulation of primary cilium-dependent autophagy activation during flow

(A) Time course QPCR analysis of wild type (WT) KEC cells under flow conditions. Cells were subjected to flow or not (d0) from 4 hours to 4 days (do, 4hrs, d1, d2, d4). As a positive control for the Hh pathway, cells were treated with purmophamine (purmo) for 24 hours. Gli1 and Ptc1 mRNA levels were quantified by real-time RT-PCR, normalized with respect to the β-actin gene, and presented as fold increases. All histograms were obtained from at least three independent experiments. *, p<0.05. (B) Cells were subjected to flow or not (d0) from 4 hours or treated with purmo for 24 hours, thex fixed and labelled for Smo and acetyl-tubulin for localization of Smo to the axoneme of cilia. (C) Wild type (WT) KEC cells were subjected to flow or not (d0) for 1 day (d1). S6K, phosphorylated S6K (S6K P), S6 and phosphorylated S6 (S6 P) were revealed by immunoblotting.

104

Supplementary Figure 1.

(A-B) Wild type (WT) KEC and IFT88-/- KEC cells were subjected to flow or not (d0) for 1 day (d1) with or without chloroquine (CQ), and P62/actin ratios determined by immunoblotting (A) and densitometry (B). Immunobloting data: means means ± s.e., *, p<0.05 (t-test). (C) QPCR analysis of wild type (WT) KEC cells under flow conditions. Cells were subjected to flow or not (d0) for 1 day (d1). As a positive control for the p62 upregulation, cells were starved for 4 hours. p62 mRNA levels were quantified by real-time RT-PCR, normalized with respect to the β-actin gene, and presented as fold increases compared to d0. All histograms were obtained from at least three independent experiments. *, p<0.05. (D, E) Time course immunoblot analysis of wild type (WT) KEC cells under flow conditions. Cells were subjected to flow or not (d0) not for 4 hours (4 hrs) or 1 days (d1) with or without chloroquine (CQ), and LC3-II/actin ratios determined by immunoblotting (D) and densitometry (E). Immunobloting data: means means ± s.e., *, p<0.05 (t-test).

105

2.2. Discussion 2: The role of autophagy in integration of fluid flow via primary cilium in the homeostasis of kidney epithelial cells

In this work, we have shown that mechanical stress, specifically fluid flow, induces autophagy in a primary cilium (PC) dependent manner. In kidney epithelial cells, primary cilium as a mechanosensory organelle transduces laminal fluid flow and recruits Atg16L1 to the basal body triggering the induction of autophagy. This flow dependent autophagy activation in turn regulates the decrease of cell volume and maintains the physiological function of kidney cells. In cells with impaired ciliogenesis, flow could not stimulate the induction of autophagosome biogenesis and in turn couldn’t regulate cell size. When autophagy was disrupted by Beclin1 and Atg5 depletion, flow could not induce a decrease of cell volume showing that the cell size regulation by fluid flow depends on autophagy activation in kidney epithelial cells.

Autophagy activation by fluid flow

It’s been previously shown that fluid flow induced cell volume decrease in vitro, and contributed to the maintenance of polarity of kidney epithelial cells (Boehlke, Kotsis et al. 2010; Heo, Sachs et al. 2012). Autophagy is known to have be an important mechanism for cellular homeostasis of podocytes however the physiological role of autophagy in kidney tubules which are under constant flow during urine filtration is not yet known. Fluid flow creates a shear stress in kidney epithelial cells which is an important stress for the kidney tubules serving for the filtration of urine and thus maintenance of various cellular responses, including cell size regulation. When we apply a physiologically relevant fluid flow of 1dyn/cm2 on MDCK and kidney epithelial cells (KEC) in vitro, autophagy activation is quite rapid, starting from 4 hours and maintained for the duration of the flow up to 4 days (Figure 17 and Figure 18). This suggest that constitutive autophagy in kidney epithelial cells that are under constant flow is an important adaptive response for the maintenance of cell homeostasis. In these cells, primary cilia function as a mechanosensor for various signaling pathway regulation such as calcium signaling, cytoskeletal re-arrangement, polarity, maintenance of differentiation, mTOR signaling and cell size regulation. In our study we found that autophagy activation upon fluid flow is dependant on the presence of primary cilia.

106

This observation supports our hypothesis that PC function as an antenna to integrate extracellular signals, in our system, the mechanical stress to induce autophagy as an adaptive response. There are previous reports that mechanical stress may induce autophagy. Compression is a mechanical stress that bone and cartilage face primarily. Compression of a pressure of 0.25 kPA (equal to 2500 dyn/cm2) was applied on Dictyostelium and breast cancer epithelial cells. LC3-II accumulation was observed by immunofluorescence and autophagosome biogenesis and autophagic flux were quantified by western blot. This autophagy induction was mTOR independent (King, Veltman et al. 2011). The endothelial cells and cardiomyocytes face stretching and shear stress caused by blood flow which is increased upon hypertrophy. Lin et al. researched the effect of stretch on cardiomyocytes and observed the induction of autophagy and activation of p38MAPK (Lin, Tang et al. 2014)

Our study showed that fluid shear stress is also an autophagic stimulus for kidney epithelial cells. As there are different cell types that are under fluid flow in the body (see Table 1), the role of autophagy under different mechanical stress in different ciliated tissues should be studied as a mechanical stress response.

Cell volume regulation by autophagy under fluid flow

Observation of severe impairment of flow induced autophagy was severely impaired in IFT88-/- cells which have impaired ciliogenesis; led us to conclude that PC triggers the initiation of autophagy possibly by the recruitment of early autophagic markers, such as Atg16L1, to the base of the cilia for autophagosome biogenesis (Figure 18 and Figure 20). It is known that the tissues under fluid flow are ciliated and the PC regulates various signaling pathways. Autophagy response in those tissues is probably mediated by PC and the impairment of the PC results in autophagy impairment that interferes with the maintenance of cellular homeostasis. In kidney epithelial cells, PC transduces different signalizations in response to flow including calcium signaling, cytoskeletal re-arrangement, planar cell polarity, differentiation, mTOR signaling and autophagy.

By studying the kinetic of autophagy activation upon flow, we observed a rapid induction of autophagosome formation which decreases after 2 days of flow accompanied but remains significantly higher under flow compared to static culture (Figure 17). This suggests a transient “hyperactivation” of autophagosome biogenesis followed by stabilization of the

107 autophagy activation under constant fluid flow. The kinetics of cell volume reduction under flow in two different cell lines show that the volume reduction rapidly but becomes significant after 2 days, suggesting that the reduction begins following the activation of autophagy. As we observe in autophagy deficient cells that the cells are unable to reduce their volume and that their size remains similar to resting state, it supports our hypothesis that autophagy activation is upstream of volume regulation. The study by Heo et al. showed that shear stress induced rapid volume reduction in MDCK cells and this reduction is regulated by the efflux of water through aquaporin channels induced by changes in intracellular hydrostatic pressure (Heo, Sachs et al. 2012). This flow induced shrinkage was followed by irreversible cellular deformation and independent of Ca2+ influx or K+-Cl- transport. The data shown by Boehlke and colleagues show that mTOR pathway is inhibited in the cells after 4 days of fluid shear stress. This previous data supports our results that rapid autophagy activation by fluid flow can be independent of mTOR. Autophagy can possibly regulate cell mass by active degradation of intracellular content and may regulate water transport. Aquaporin-2 (AQP-2) is located at the PC of the apical membrane of the collecting duct cells and in kidney epithelial cells derived from IFT88-/- mice, it is shown that AQP-2 is abundantly present on the membranes thus increasing water transport which may inhibit the cell volume regulation (Saigusa et al. 2012). A previous study showing that aquaporin-5 was degraded by autophagy in submandibular gland of rats increases the possibility that aquaporins on the kidney epithelial cells are good candidates as autophagy substrates (Azlina, Javkhlan et al. 2010). The hypothesis would be that autophagy activated under fluid flow by PC, degrades cilia- associated proteins including AQP-2, inhibiting its translocation to the membrane and the subsequent water efflux, thus regulating the cell volume.

Another mechanosensory channel is the non-selective cation channel, the vanniloid receptor TRPV4, localized at the PC (Patel 2014). This channel has a role in osmolality and regulatory volume decrease as well as in controlling flow mediated Ca2+ signaling via its interaction of polycystin-2. However the TRPV4 depletion does not lead to cystogenesis in vivo (Wu, Gao et al. 2007). During the induction of flow, other signalizations may get involved and regulated by the PC, including the recruitment of LKB1 to the axoneme and P- AMPK at the BB inhibiting mTOR pathway which may contribute to the constitutive activation of autophagy as well as cell size regulation. The downstream effector of the mTOR pathway includes S6K, a Ser/Thr kinase that can phosphorylate 4EBP-1 (eukaryotic initiation factor 4E binding protein) which represses protein translation contributing to cell growth.

108

In order to confirm the hypothesis that autophagy regulates cell volume by aquaporin degradation, the inhibition of aquaporins by Hg2+ should be induced under fluid flow. The localization and expression of aquaporins in autophagy deficient cells are also important to determine the role of autophagy in aquaporin degradation or translocation. In order to determine if autophagy is upstream of mTOR1 inhibition during flow, AMPK and mTOR depleted cells should be studied for autophagic response during fluid flow.

Another important cilia-dependent signalization controlled tightly in kidney epithelial cell is planar cell polarity (PCP). Apart from the apical/ basal polarity that defines cellular orientation, kidney epithelial cells are polarized within the plane of the tissue, orthogonal to the apical/basal axis which is PCP. PCP, specifically a component of PCP which is the oriented cell division is critical for the development of nephron and the maintenance of proper tubule physiology. Oriented cell division, characterized by division parallel to the axis of tubule elongation and the plane of the kidney epithelium, is vital following kidney injury (Fedeles and Gallagher 2012). In PKD models, oriented cell division is observed to be impaired, leading to dilated and cystic tubules, especially in adult late onset of cystogenesis in autosomal dominant polycystic kidney disease (ADPKD) patients (Carroll and Das 2011). PKD phenotype results from the loss of cilia and since PCP is controlled by the non-canonical Wnt pathway regulated through the PC, the disrupted PCP would be an important phenotype of the disease (Fedeles and Gallagher 2012). Loss of inversin, the main regulatory ciliary protein of PCP pathway, leads to randomized oriented cell division and perturbs the directional tubule resulting in cyst formation (Germino 2005; Saburi, Hester et al. 2008). The role of mechanical stress in PCP is not studied in vertebrates, however in Xenopus larval skin it’s been shown that multi-ciliated epithelial cells respond to fluid flow by orientation of BB and direction of ciliary beating (Mitchell, Jacobs et al. 2007). It would be important to study the PCP regulation in response to fluid flow in mammalian kidney cells to define the specific role of this pathway physiologically, especially during the development of PKD. In the liver, hepatocyte polarization is important for the tissue function and hepatocytes can lose their polarized state upon stress. Autophagy has been shown to have an important role during repolarization by increasing ATP levels for mitochondrial function in the maintenance of hepatocyte phenotype (Fu, Mitra et al. 2013). These observations suggest that flow-induced autophagy may have a role in planar cell polarity and epithelial phenotype of kidney tubule cells. Autophagy impairment under fluid flow should be studied to investigate whether this process has an effect of the maintenance of PCP.

109

The reciprocal role of cilia and PCP during the development of PKD still remains to be discovered. However it seems that perturbation of the tubular cell physiology by impaired cilia lead to the deregulation of various pathways, such as water transport, PCP, calcium signaling or mTOR. A tight regulation of these cellular pathways is highly critical for kidney tubules and the main extracellular input of this specialized tissue is the urine filtration thus making the study of autophagic response that could coordinate the regulation of various pathways via PC is crucial to understand the physiology of kidney cells and development of cilia related kidney pathologies.

110

GENERAL DISCUSSION

AND PERSPECTIVES

111

V. GENERAL DISCUSSION AND PERSPECTIVES

How do ATG proteins recruited to the primary cilium contribute to the biogenesis of phagophores and autophagosomes?

Our two studies show that autophagy is induced by extracellular signals in a primary cilia- dependent manner. This activation of autophagy was accompanied by the localization of various autophagic proteins at the axoneme and basal body (BB) (see Table 3). Especially the presence of early autophagic protein Atg16L1 and the enrichment of LC3 at the plasma membrane in an IFT dependent manner suggest that autophagosome biogenesis is controlled by the plasma membrane. Extending from the plasma membrane and having high vesicular activity, the ciliary pocket accentuates its interaction with the plasma membrane as a source for autophagosome formation at the base of the cilium. Autophagy induction is regulated by active recruitment of Atg16L1 to the BB via IFT20 vesicles. The nature and mechanism of this vesicular trafficking should be further investigated in order to further understand the reciprocal relationship between PC and autophagosome biogenesis. The recent identification of an Atg16L1 binding motif, [YW]-X3-[ED]-X4-[YWF]-X2-L, is a good candidate to investigate the interaction between IFT20 and Atg16L1. Boada-Romero et al. showed that this motif interacts with the WD-repeat domain of Atg16L1, and may target membranous compartments for autophagic degradation (Boada-Romero, Letek et al. 2013). The ongoing study in the lab investigates the interaction between IFT20 and Atg16L1 to reveal how these proteins are transported to the BB together.

112

Table 3: Structure and signaling from the cilium and relationship with the autophagic pathway

Serum deprivation induces autophagy activation via primary cilia (PC) by recruitment of autophagic machinery to the PC. IFT20, which interacts with Atg16L1, regulates the trafficking of ciliary components from the Golgi to the cilium. The table shows the localization of the ATG machinery induction in the basal body and/or in the axoneme in complete medium or upon serum starvation. ‘ND’ means ‘not determined’ (from results Pampliega et al. 2013)

The absence of the core autophagic protein Beclin1 with γ-tubulin (BB) colocalization may be explained by two possibilities. One possibility is that PC activates a non- canonical autophagy which can bypass Beclin1 complex formation during autophagosome formation. The second possibility is that since we can observe Vps34 at the BB, Beclin1 can be found on the ciliary pocket during autophagosome formation. Due to lack of a marker to visualize ciliary pocket, a 3D reconstruction of the Beclin1 localization on the base of the cilium, along with live microscopy to follow Atg localizations will paint a clearer picture for autophagosome initiation site in relation to ciliary pocket. Since we observe the induction of cilia–dependent autophagy quite rapidly, the research of autophagosome biogenesis site remains intriguing and important to reveal the mechanism behind autophagy induction by PC.

113

Is cilium dependent autophagy a selective process?

We’ve previously observed that basal autophagy is responsible for the degradation of IFT20 in turn regulating cilia length. It’s also shown that autophagy degrades OFD1 during starvation and PC2. It’s recently been demonstrated that autophagy can also degrade motile cilia-associated proteins in the lung, such as IFT88, Centrin1, ARL13 and Pericentrin which are involved in the development of chronic obstructive pulmonary disease (COPD) (Lam, Cloonan et al. 2013). Epithelial cells in the airway are ciliated with multiple motile cilia nucleated by a basal body. The presence of cilia in the respiratory track is critical for normal lung function, and alterations in these motile cilia contribute to the development of pathologies such as chronic obstructive pulmonary disease (COPD), emphysema, and chronic bronchitis (Barnes 2000). In COPD patients who were chronic cigarette smokers, a significant cilia shortening is observed that leads to impaired mucociliary clearance (Leopold, O'Mahony et al. 2009; Hessel, Heldrich et al. 2014). Cilium shortening leads to excess mucus production in the lung epithelial cells and interferes with the protection of airways from infections. The lung tissues of COPD patients have been previously shown to accumulate autophagosomes (Chen, Kim et al. 2008). Lam et al. (Lam, Cloonan et al. 2013) demonstrated epithelial cilia shortening in chronic cigarette smoker that correlated with increased autophagic activity and an increase in the expression of various proteins involved in cilia regulation. In support of the proposed role of autophagy in the negative regulation of ciliary growth, autophagy-impaired cells, Becn1+/-, Maplc3b-/-, and Hdac6-/Y cells and mice were protected from cilia shortening. Ubiquitinated protein aggregates are cleared by a HDAC-6-dependent autophagy. In addition, various ciliary proteins, such as IFT88, ARL13, centrin1, and pericentrin, accumulate in autophagosomes. Lam et al. proposes a new role for selective autophagy in degradation of ciliary proteins and in the regulation of motile cilia length in response to chronic cigarette smoking stress in lung epithelial cells. This hypothesis is supported by previous observations of hyper-responsiveness in lung-specific Atg7 knock-out mice with abnormal ciliogenesis and long cilia formation in mTOR activated cells (DiBella, Park et al. 2009; Inoue, Kubo et al. 2011). Constitutive autophagy seems to regulate the turnover of proteins involved in cilia function, thus controlling sensitivity of the cell in response to different stressors, such as cigarette smoking in MC or starvation in PC. Therefore, inhibition of autophagy in lung epithelial cells may have therapeutic relevance in COPD, chronic bronchitis, or emphysema patients. The study by Lams et al. (Lam, Cloonan et al. 2013) demonstrated that cross-talk between autophagy and cilia is not limited to the PC but also extends to the MC. The

114 degradation of ciliary components, such as OFD1, PC2, IFT20, and IFT88, suggest that selective forms of autophagy, defined by the term “ciliophagy”, may serve to sequester certain ciliary structures (Cloonan, Lam et al. 2014) (see Table 4 for examples of ciliary proteins degraded by autophagy). The preliminary observation that these proteins possess LC3 interacting domains (LIR) suggests that selective autophagy is responsible for the turnover of cilia and cilia length (data not shown).

115

Cellular Autophagy Associated Localizatio Ciliary Role Ref substrate Ciliopathy n

Golgi, IFTB anterograde IFT complex and ciliary (Pampliega, IFT201 basal body, assembly via vesicular PKD3 Orhon et al. and transport of proteins to the 2013) axoneme basal body

Oral-facial- distal Cilia length and centriolar (Tang, Lin et al. OFD12 digital centriole recruitment of IFT88 2013) syndrome

axoneme (Cebotaru, and PC2 Ca2+-specific cation channel PKD3 Cebotaru et al. endoplasmi 2014) c reticulum

Ciliary assembly and IFTB (Lam, Cloonan et IFT88 axoneme COPD4 anterograde IFT complex al. 2013)

Ca2+ signaling in the (Lam, Cloonan et centrin1 basal body COPD4 connecting cilium al. 2013)

(Lam, Cloonan et ARL13 axoneme IFTA assembly stability COPD4 al. 2013)

(Lam, Cloonan et Pericentrin basal body Ciliary assembly COPD4 al. 2013)

Table 4: Cilia-related proteins degraded by autophagy (from review Orhon et al. in press)

1IFT20 is an autophagy substrate under basal condition

2OFD1 is an autophagy substrate under autophagy stimulation

3 COPD: chronic obstructive pulmonary disease

4 PKD: polycystic kidney disease

116

Do signals that trigger cilium-dependent autophagy depend on the stimuli (chemical, mechanical) or do all stimuli converge to a single signaling pathway upstream of the autophagy machinery?

We identified that serum starvation and fluid shear stress are extracellular stimuli that activate autophagy in a PC dependent manner. There are various other extracellular signals that activate autophagy in vitro and in vivo such as amino acid deprivation, glucose deprivation, hypoxia, infection. PC as a sensory antenna also functions like a signaling coordination center for various cellular pathways through chemical or mechanical stimuli interacting with different channels or receptors on the cilia. It would be compelling to define if other autophagic stimuli depended on the presence of PC in different ciliated tissues. We also observe that starvation induced autophagy depended on ciliary Hedgehog pathway (Hh) and is mTOR independent whereas in kidney epithelial cells fluid flow induced autophagy is independent of the Hh pathway activation. This observation leads to the hypothesis that cilia- dependent autophagy activation is regulated by different signalizations depending on the stimuli.

What is the physiological relevance of PC-mediated autophagy in kidney physiopathology?

As we have discussed, kidney epithelial cells under constant flow mediate their cell size via cilia-dependent autophagy. This regulation contributes to the proper function of kidney tubules. Various pathologies show impaired autophagy as described in this manuscript. Cell size regulation is particularly important for the development of polycystic kidney diseases as the cells lining the cysts are larger, with impaired cilia, impaired orientation during mitosis and fibroblastic phenotype resulting in the perturbation of tubule dilation upon cystogenesis (Figure 6) (Grantham, Geiser et al. 1987; Pazour, Dickert et al. 2000; Brown and Murcia 2003). In PKD mice models, it was previously observed that autophagy flux was impaired (Belibi, Zafar et al. 2011; Ravichandran and Edelstein 2014). We have identified that cells with impaired cilia, derived from IFT88-/- mice which develop cystic kidneys, have impaired autophagy activation in response to flow. We can argue that the cells with impaired cilia will not respond to fluid flow through PC mediated autophagy activation which will

117 interfere with cell size regulation and homeostasis, and may contribute to the disease development.

During ischemic injuries autophagy activation is also important for cell survival. In progressive renal fibrosis models, upon unilateral ureteral obstruction (UUO), the remaining healthy kidney shows an induction of autophagy which remains to be activated following transient tubule dilation. This observation is supportive of our hypothesis that fluid shear stress induces autophagy. We aim to confirm our in vitro results in vivo by using two approaches. First experiment aims to reduce fluid flow following ligation of the mouse kidney and observe autophagy response and epithelial cell size. The second experiment aims to increase fluid flow in one kidney by doing a nephrectomy of the one kidney and observe autophagic response in the other kidney in a time dependent manner.

During kidney transplants, cold storage and re-warming of the donor kidney are risk factors for delayed graft function and the inhibition of autophagy induce apoptosis and necrosis. However, warm ischemia-reperfusion induced autophagy that serves to protect tubular cell survival (Jain, Keys et al. 2014). The injury to the donor organ and during ischemia/reperfusion is caused by the cease of blood flow. The clinical trials on hypothermic pulsatile perfusion for organ preservation during transplantations show benefits of the continuous perfusion for the live tissue. The study concentrates on the inflammatory response inhibition upon fluid flow of 1 dyn/cm2 (Timsit, Adams et al. 2013). The mechanism behind the beneficial effect of this organ preservation can be autophagy. As continuous perfusion will induce autophagy, the continuous turnover of proteins, organelles and cell size regulation will avoid injury. Our study could lead to the path of identification of the mechanism behind this preservation method and contribute to the clinical trials to prevent kidney injury.

118

BIBLIOGRAPHY

119

A Aberger, F. and I. A. A. Ruiz (2014). "Context-dependent signal integration by the GLI code: the oncogenic load, pathways, modifiers and implications for cancer therapy." Semin Cell Dev Biol 33: 93-104. Adolph, T. E., M. F. Tomczak, et al. (2013). "Paneth cells as a site of origin for intestinal inflammation." Nature 503(7475): 272-6. Ando, J. and K. Yamamoto (2013). "Flow detection and calcium signalling in vascular endothelial cells." Cardiovasc Res 99(2): 260-8. Angers, S. and R. T. Moon (2009). "Proximal events in Wnt signal transduction." Nat Rev Mol Cell Biol 10(7): 468-77. Artal-Martinez de Narvajas, A., T. S. Gomez, et al. (2013). "Epigenetic regulation of autophagy by the methyltransferase G9a." Mol Cell Biol 33(20): 3983-93. Avidor-Reiss, T. and J. Gopalakrishnan "Cell Cycle Regulation of the Centrosome and Cilium." Drug Discov Today Dis Mech 10(3-4): e119-e124. Avidor-Reiss, T. and J. Gopalakrishnan (2013). "Cell Cycle Regulation of the Centrosome and Cilium." Drug Discov Today Dis Mech 10(3-4): e119-e124. Axe, E. L., S. A. Walker, et al. (2008). "Autophagosome formation from membrane compartments enriched in phosphatidylinositol 3-phosphate and dynamically connected to the endoplasmic reticulum." J Cell Biol 182(4): 685-701. Azad, M. B., Y. Chen, et al. (2009). "Regulation of autophagy by reactive oxygen species (ROS): implications for cancer progression and treatment." Antioxid Redox Signal 11(4): 777-90. Azlina, A., P. Javkhlan, et al. (2010). "Roles of lysosomal proteolytic systems in AQP5 degradation in the submandibular gland of rats following chorda tympani parasympathetic denervation." Am J Physiol Gastrointest Liver Physiol 299(5): G1106-17.

Baba, M., M. Osumi, et al. (1997). "Two distinct pathways for targeting proteins from the cytoplasm to the vacuole/lysosome." J Cell Biol 139(7): 1687-95. Bandyopadhyay, U., S. Sridhar, et al. (2010). "Identification of regulators of chaperone- mediated autophagy." Mol Cell 39(4): 535-47. Barnes, P. J. (2000). "Chronic obstructive pulmonary disease." N Engl J Med 343(4): 269-80. Bastos, A. P., K. Piontek, et al. (2009). "Pkd1 haploinsufficiency increases renal damage and induces microcyst formation following ischemia/reperfusion." J Am Soc Nephrol 20(11): 2389-402. Belibi, F., I. Zafar, et al. (2011). "Hypoxia-inducible factor-1alpha (HIF-1alpha) and autophagy in polycystic kidney disease (PKD)." Am J Physiol Renal Physiol 300(5): F1235-43. Benmerah, A. (2012). "The ciliary pocket." Curr Opin Cell Biol 25(1): 78-84.

120

Birgisdottir, A. B., T. Lamark, et al. (2013). "The LIR motif - crucial for selective autophagy." J Cell Sci 126(Pt 15): 3237-47. Boada-Romero, E., M. Letek, et al. (2013). "TMEM59 defines a novel ATG16L1-binding motif that promotes local activation of LC3." EMBO J 32(4): 566-82. Boehlke, C., F. Kotsis, et al. (2010). "Primary cilia regulate mTORC1 activity and cell size through Lkb1." Nat Cell Biol 12(11): 1115-22. Boya, P., F. Reggiori, et al. (2013). "Emerging regulation and functions of autophagy." Nat Cell Biol 15(7): 713-20. Briscoe, J. and P. P. Therond (2013). "The mechanisms of Hedgehog signalling and its roles in development and disease." Nat Rev Mol Cell Biol 14(7): 416-29. Brown, N. E. and N. S. Murcia (2003). "Delayed cystogenesis and increased ciliogenesis associated with the re-expression of polaris in Tg737 mutant mice." Kidney Int 63(4): 1220-9.

Cadwell, K., J. Y. Liu, et al. (2008). "A key role for autophagy and the autophagy gene Atg16l1 in mouse and human intestinal Paneth cells." Nature 456(7219): 259-63. Carroll, T. J. and A. Das (2011). "Planar cell polarity in kidney development and disease." Organogenesis 7(3): 180-90. Cebotaru, V., L. Cebotaru, et al. (2014). "Polycystin-1 negatively regulates Polycystin-2 expression via the aggresome/autophagosome pathway." J Biol Chem 289(10): 6404- 14. Chan, E. Y., S. Kir, et al. (2007). "siRNA screening of the kinome identifies ULK1 as a multidomain modulator of autophagy." J Biol Chem 282(35): 25464-74. Chapin, H. C. and M. J. Caplan (2010). "The cell biology of polycystic kidney disease." J Cell Biol 191(4): 701-10. Chen, H., M. Fan, et al. (2012). "The histone H3 lysine 56 acetylation pathway is regulated by target of rapamycin (TOR) signaling and functions directly in ribosomal RNA biogenesis." Nucleic Acids Res 40(14): 6534-46. Chen, Y., E. McMillan-Ward, et al. (2008). "Oxidative stress induces autophagic cell death independent of apoptosis in transformed and cancer cells." Cell Death Differ 15(1): 171-82. Chen, Z. H., H. P. Kim, et al. (2008). "Egr-1 regulates autophagy in cigarette smoke-induced chronic obstructive pulmonary disease." PLoS One 3(10): e3316. Cherra, S. J., 3rd, S. M. Kulich, et al. (2010). "Regulation of the autophagy protein LC3 by phosphorylation." J Cell Biol 190(4): 533-9. Choi, A. M., S. W. Ryter, et al. (2013). "Autophagy in human health and disease." N Engl J Med 368(7): 651-62. Clement, C. A., K. D. Ajbro, et al. (2013). "TGF-beta signaling is associated with endocytosis at the pocket region of the primary cilium." Cell Rep 3(6): 1806-14. Clement, D. L., S. Mally, et al. (2013). "PDGFRalpha signaling in the primary cilium regulates NHE1-dependent fibroblast migration via coordinated differential activity of

121

MEK1/2-ERK1/2-p90RSK and AKT signaling pathways." J Cell Sci 126(Pt 4): 953- 65. Cloonan, S. M., H. C. Lam, et al. (2014). ""Ciliophagy": The consumption of cilia components by autophagy." Autophagy 10(3): 532-4. Corbit, K. C., A. E. Shyer, et al. (2008). "Kif3a constrains beta-catenin-dependent Wnt signalling through dual ciliary and non-ciliary mechanisms." Nat Cell Biol 10(1): 70- 6. Cowley, B. D., Jr. (2008). "Calcium, cyclic AMP, and MAP kinases: dysregulation in polycystic kidney disease." Kidney Int 73(3): 251-3. Criollo, A., L. Senovilla, et al. (2009). "The IKK complex contributes to the induction of autophagy." EMBO J 29(3): 619-31. Cuervo, A. M. and E. Wong (2014). "Chaperone-mediated autophagy: roles in disease and aging." Cell Res 24(1): 92-104.

Davies, P. F., T. Mundel, et al. (1995). "A mechanism for heterogeneous endothelial responses to flow in vivo and in vitro." J Biomech 28(12): 1553-60. Delling, M., P. G. DeCaen, et al. (2013). "Primary cilia are specialized calcium signalling organelles." Nature 504(7479): 311-4. Deribe, Y. L., T. Pawson, et al. (2010). "Post-translational modifications in signal integration." Nat Struct Mol Biol 17(6): 666-72. Deter, R. L., P. Baudhuin, et al. (1967). "Participation of lysosomes in cellular autophagy induced in rat liver by glucagon." J Cell Biol 35(2): C11-6. Di Bartolomeo, S., M. Corazzari, et al. (2010). "The dynamic interaction of AMBRA1 with the dynein motor complex regulates mammalian autophagy." J Cell Biol 191(1): 155- 68. DiBella, L. M., A. Park, et al. (2009). "Zebrafish Tsc1 reveals functional interactions between the cilium and the TOR pathway." Hum Mol Genet 18(4): 595-606. Dice, J. F. (1990). "Peptide sequences that target cytosolic proteins for lysosomal proteolysis." Trends Biochem Sci 15(8): 305-9. Dohi, E., S. Tanaka, et al. (2012). "Hypoxic stress activates chaperone-mediated autophagy and modulates neuronal cell survival." Neurochem Int 60(4): 431-42. Dooley, H. C., M. Razi, et al. (2014). "WIPI2 links LC3 conjugation with PI3P, autophagosome formation, and pathogen clearance by recruiting Atg12-5-16L1." Mol Cell 55(2): 238-52. Dressler, G. R. (2006). "The cellular basis of kidney development." Annu Rev Cell Dev Biol 22: 509-29. Du, Z., Y. Duan, et al. (2004). "Mechanosensory function of microvilli of the kidney proximal tubule." Proc Natl Acad Sci U S A 101(35): 13068-73. Dunn, W. A., Jr. (1990). "Studies on the mechanisms of autophagy: maturation of the autophagic vacuole." J Cell Biol 110(6): 1935-45.

122

Dupont, N. and F. Lafont (2009). "How autophagy regulates the host cell signaling associated with the postpartum bacteria cocoon experienced as a danger signal." Autophagy 5(8): 1222-3. Dupont, N., I. Orhon, et al. (2014). "Autophagy and autophagic flux in tumor cells." Methods Enzymol 543: 73-88. Ebato, C., T. Uchida, et al. (2008). "Autophagy is important in islet homeostasis and compensatory increase of beta cell mass in response to high-fat diet." Cell Metab 8(4): 325-32. Egan, D. F., D. B. Shackelford, et al. (2011). "Phosphorylation of ULK1 (hATG1) by AMP- activated protein kinase connects energy sensing to mitophagy." Science 331(6016): 456-61. Eggenschwiler, J. T. and K. V. Anderson (2007). "Cilia and developmental signaling." Annu Rev Cell Dev Biol 23: 345-73. Egorova, A. D., P. P. Khedoe, et al. (2011). "Lack of primary cilia primes shear-induced endothelial-to-mesenchymal transition." Circ Res 108(9): 1093-101. Egorova, A. D., K. van der Heiden, et al. (2012). "Primary cilia as biomechanical sensors in regulating endothelial function." Differentiation 83(2): S56-61. English, L., M. Chemali, et al. (2009). "Autophagy enhances the presentation of endogenous viral antigens on MHC class I molecules during HSV-1 infection." Nat Immunol 10(5): 480-7. Eskelinen, E. L. (2005). "Maturation of autophagic vacuoles in Mammalian cells." Autophagy 1(1): 1-10. Essig, M., F. Terzi, et al. (2001). "Mechanical strains induced by tubular flow affect the phenotype of proximal tubular cells." Am J Physiol Renal Physiol 281(4): F751-62. Ezaki, J., N. Matsumoto, et al. (2011). "Liver autophagy contributes to the maintenance of blood glucose and amino acid levels." Autophagy 7(7): 727-36.

Fedeles, S. and A. R. Gallagher (2012). "Cell polarity and cystic kidney disease." Pediatr Nephrol 28(8): 1161-72. Ferkol, T. W. and M. W. Leigh (2012). "Ciliopathies: the central role of cilia in a spectrum of pediatric disorders." J Pediatr 160(3): 366-71. Follit, J. A., R. A. Tuft, et al. (2006). "The intraflagellar transport protein IFT20 is associated with the Golgi complex and is required for cilia assembly." Mol Biol Cell 17(9): 3781- 92. Fritton, S. P. and S. Weinbaum (2009). "Fluid and Solute Transport in Bone: Flow-Induced Mechanotransduction." Annu Rev Fluid Mech 41: 347-374. Fu, D., K. Mitra, et al. (2013). "Coordinated elevation of mitochondrial oxidative phosphorylation and autophagy help drive hepatocyte polarization." Proc Natl Acad Sci U S A 110(18): 7288-93.

123

Fujioka, Y., N. N. Noda, et al. (2010). "Dimeric coiled-coil structure of Saccharomyces cerevisiae Atg16 and its functional significance in autophagy." J Biol Chem 285(2): 1508-15. Fujita, N., M. Hayashi-Nishino, et al. (2008). "An Atg4B mutant hampers the lipidation of LC3 paralogues and causes defects in autophagosome closure." Mol Biol Cell 19(11): 4651-9. Fujitani, Y., T. Ueno, et al. (2010). "Autophagy in health and disease. 4. The role of pancreatic beta-cell autophagy in health and diabetes." Am J Physiol Cell Physiol 299(1): C1-6. Fullgrabe, J., D. J. Klionsky, et al. (2014). "The return of the nucleus: transcriptional and epigenetic control of autophagy." Nat Rev Mol Cell Biol 15(1): 65-74.

Ganley, I. G., P. M. Wong, et al. (2011). "Distinct autophagosomal-lysosomal fusion mechanism revealed by thapsigargin-induced autophagy arrest." Mol Cell 42(6): 731- 43. Ge, L., D. Melville, et al. (2013). "The ER-Golgi intermediate compartment is a key membrane source for the LC3 lipidation step of autophagosome biogenesis." Elife 2: e00947. Geeraert, C., A. Ratier, et al. (2010). "Starvation-induced hyperacetylation of tubulin is required for the stimulation of autophagy by nutrient deprivation." J Biol Chem 285(31): 24184-94. Geisler, S., K. M. Holmstrom, et al. (2010). "PINK1/Parkin-mediated mitophagy is dependent on VDAC1 and p62/SQSTM1." Nat Cell Biol 12(2): 119-31. Geng, J., M. Baba, et al. (2008). "Quantitative analysis of autophagy-related protein stoichiometry by fluorescence microscopy." J Cell Biol 182(1): 129-40. Geng, J. and D. J. Klionsky (2010). "The Golgi as a potential membrane source for autophagy." Autophagy 6(7): 950-1. Germino, G. G. (2005). "Linking cilia to Wnts." Nat Genet 37(5): 455-7. Ghossoub, R., A. Molla-Herman, et al. (2011). "The ciliary pocket: a once-forgotten membrane domain at the base of cilia." Biol Cell 103(3): 131-44. Godel, M., B. Hartleben, et al. (2011). "Role of mTOR in podocyte function and diabetic nephropathy in humans and mice." J Clin Invest 121(6): 2197-209. Goetz, S. C. and K. V. Anderson (2010). "The primary cilium: a signalling centre during vertebrate development." Nat Rev Genet 11(5): 331-44. Goetz, J. G., E. Steed, et al. (2014). "Endothelial cilia mediate low flow sensing during zebrafish vascular development." Cell Rep 6(5): 799-808. Gong, C., C. Bauvy, et al. (2013). "Beclin 1 and autophagy are required for the tumorigenicity of breast cancer stem-like/progenitor cells." Oncogene 32(18): 2261-72, 2272e 1-11. Gonzalez, J., M. Essig, et al. (2013). "[Renal urinary shear stress: a novel actor in nephropathies]." Med Sci (Paris) 29(3): 279-85.

124

Gordon, P. B., I. Holen, et al. (1993). "Dependence of hepatocytic autophagy on intracellularly sequestered calcium." J Biol Chem 268(35): 26107-12. Grabias, B. M. and K. Konstantopoulos (2013). "The physical basis of renal fibrosis: effects of altered hydrodynamic forces on kidney homeostasis." Am J Physiol Renal Physiol 306(5): F473-85. Gradilone, S. A., A. I. Masyuk, et al. (2007). "Cholangiocyte cilia express TRPV4 and detect changes in luminal tonicity inducing bicarbonate secretion." Proc Natl Acad Sci U S A 104(48): 19138-43. Grantham, J. J., J. L. Geiser, et al. (1987). "Cyst formation and growth in autosomal dominant polycystic kidney disease." Kidney Int 31(5): 1145-52. Guo, J. Y., H. Y. Chen, et al. (2011). "Activated Ras requires autophagy to maintain oxidative metabolism and tumorigenesis." Genes Dev 25(5): 460-70. Guo, Y., C. Chang, et al. (2012). "AP1 is essential for generation of autophagosomes from the trans-Golgi network." J Cell Sci 125(Pt 7): 1706-15.

Hailey, D. W., A. S. Rambold, et al. (2010). "Mitochondria supply membranes for autophagosome biogenesis during starvation." Cell 141(4): 656-67. Hamai, A. and P. Codogno (2012). "New targets for acetylation in autophagy." Sci Signal 5(231): pe29. Hamasaki, M., N. Furuta, et al. (2013). "Autophagosomes form at ER-mitochondria contact sites." Nature 495(7441): 389-93. Hanada, T., N. N. Noda, et al. (2007). "The Atg12-Atg5 conjugate has a novel E3-like activity for protein lipidation in autophagy." J Biol Chem 282(52): 37298-302. Hara, T., K. Nakamura, et al. (2006). "Suppression of basal autophagy in neural cells causes neurodegenerative disease in mice." Nature 441(7095): 885-9. Hara, T., A. Takamura, et al. (2008). "FIP200, a ULK-interacting protein, is required for autophagosome formation in mammalian cells." J Cell Biol 181(3): 497-510. Harding, T. M., K. A. Morano, et al. (1995). "Isolation and characterization of yeast mutants in the cytoplasm to vacuole protein targeting pathway." J Cell Biol 131(3): 591-602. Hartleben, B., M. Godel, et al. (2010). "Autophagy influences glomerular disease susceptibility and maintains podocyte homeostasis in aging mice." J Clin Invest 120(4): 1084-96. Hayashi-Nishino, M., N. Fujita, et al. (2009). "A subdomain of the endoplasmic reticulum forms a cradle for autophagosome formation." Nat Cell Biol 11(12): 1433-7. He, M., R. Subramanian, et al. (2014). "The kinesin-4 protein Kif7 regulates mammalian Hedgehog signalling by organizing the cilium tip compartment." Nat Cell Biol 16(7): 663-72. Heo, J., F. Sachs, et al. (2012). "Shear-induced volume decrease in MDCK cells." Cell Physiol Biochem 30(2): 395-406. Hessel, J., J. Heldrich, et al. (2014). "Intraflagellar transport gene expression associated with short cilia in smoking and COPD." PLoS One 9(1): e85453.

125

Hierck, B. P., K. Van der Heiden, et al. (2008). "Primary cilia sensitize endothelial cells for fluid shear stress." Dev Dyn 237(3): 725-35. Hoey, D. A., S. Tormey, et al. (2012). "Primary cilia-mediated mechanotransduction in human mesenchymal stem cells." Stem Cells 30(11): 2561-70. Huang, K., D. R. Diener, et al. (2007). "Function and dynamics of PKD2 in Chlamydomonas reinhardtii flagella." J Cell Biol 179(3): 501-14. Huber, T. B., C. L. Edelstein, et al. (2012). "Emerging role of autophagy in kidney function, diseases and aging." Autophagy 8(7): 1009-31. Hui, C. C. and S. Angers (2011). "Gli proteins in development and disease." Annu Rev Cell Dev Biol 27: 513-37.

Ichimura, Y., T. Kirisako, et al. (2000). "A ubiquitin-like system mediates protein lipidation." Nature 408(6811): 488-92. Ingham, P. W., Y. Nakano, et al. (2011). "Mechanisms and functions of Hedgehog signalling across the metazoa." Nat Rev Genet 12(6): 393-406. Inoki, K., H. Mori, et al. (2011). "mTORC1 activation in podocytes is a critical step in the development of diabetic nephropathy in mice." J Clin Invest 121(6): 2181-96. Inoue, D., H. Kubo, et al. (2011). "Inducible disruption of autophagy in the lung causes airway hyper-responsiveness." Biochem Biophys Res Commun 405(1): 13-8. Inoue, K., H. Kuwana, et al. (2010). "Cisplatin-induced macroautophagy occurs prior to apoptosis in proximal tubules in vivo." Clin Exp Nephrol 14(2): 112-22. Ishibashi, K., N. Fujita, et al. (2011). "Atg16L2, a novel isoform of mammalian Atg16L that is not essential for canonical autophagy despite forming an Atg12-5-16L2 complex." Autophagy 7(12): 1500-13. Itakura, E., C. Kishi, et al. (2008). "Beclin 1 forms two distinct phosphatidylinositol 3-kinase complexes with mammalian Atg14 and UVRAG." Mol Biol Cell 19(12): 5360-72. Itakura, E. and N. Mizushima (2010). "Characterization of autophagosome formation site by a hierarchical analysis of mammalian Atg proteins." Autophagy 6(6): 764-76. Itakura, E., C. Kishi-Itakura, et al. (2012). "The hairpin-type tail-anchored SNARE syntaxin 17 targets to autophagosomes for fusion with endosomes/lysosomes." Cell 151(6): 1256-69. Iwata, A., B. E. Riley, et al. (2005). "HDAC6 and microtubules are required for autophagic degradation of aggregated huntingtin." J Biol Chem 280(48): 40282-92.

Jager, S., C. Bucci, et al. (2004). "Role for Rab7 in maturation of late autophagic vacuoles." J Cell Sci 117(Pt 20): 4837-48. Jain, S., D. Keys, et al. (2014). "Inhibition of autophagy increases apoptosis during re- warming after cold storage in renal tubular epithelial cells." Transpl Int.

126

Jeong, H., F. Then, et al. (2009). "Acetylation targets mutant huntingtin to autophagosomes for degradation." Cell 137(1): 60-72. Jiang, H., D. Cheng, et al. (2010). "Protein kinase C inhibits autophagy and phosphorylates LC3." Biochem Biophys Res Commun 395(4): 471-6. Jiang, J. and C. C. Hui (2008). "Hedgehog signaling in development and cancer." Dev Cell 15(6): 801-12. Jiang, M., K. Liu, et al. (2010). "Autophagy is a renoprotective mechanism during in vitro hypoxia and in vivo ischemia-reperfusion injury." Am J Pathol 176(3): 1181-92. Jimenez-Sanchez, M., F. M. Menzies, et al. (2012). "The Hedgehog signalling pathway regulates autophagy." Nat Commun 3: 1200. Joubert, P. E., G. Meiffren, et al. (2009). "Autophagy induction by the pathogen receptor CD46." Cell Host Microbe 6(4): 354-66. Jung, C. H., C. B. Jun, et al. (2009). "ULK-Atg13-FIP200 complexes mediate mTOR signaling to the autophagy machinery." Mol Biol Cell 20(7): 1992-2003. Jung, H. S., K. W. Chung, et al. (2008). "Loss of autophagy diminishes pancreatic beta cell mass and function with resultant hyperglycemia." Cell Metab 8(4): 318-24.

Kabeya, Y., N. Mizushima, et al. (2000). "LC3, a mammalian homologue of yeast Apg8p, is localized in autophagosome membranes after processing." EMBO J 19(21): 5720-8. Kallakuri, S., J. A. Yu, et al. (2014). "Endothelial Cilia Are Essential for Developmental Vascular Integrity in Zebrafish." J Am Soc Nephrol. Keady, B. T., Y. Z. Le, et al. (2011). "IFT20 is required for opsin trafficking and photoreceptor outer segment development." Mol Biol Cell 22(7): 921-30. Kee, H. L., J. F. Dishinger, et al. (2012). "A size-exclusion permeability barrier and nucleoporins characterize a ciliary pore complex that regulates transport into cilia." Nat Cell Biol 14(4): 431-7. Kenzelmann Broz, D., S. Spano Mello, et al. (2013). "Global genomic profiling reveals an extensive p53-regulated autophagy program contributing to key p53 responses." Genes Dev 27(9): 1016-31. Kestler, H. A. and M. Kuhl (2008). "From individual Wnt pathways towards a Wnt signalling network." Philos Trans R Soc Lond B Biol Sci 363(1495): 1333-47. Khaldoun, S. A., M. A. Emond-Boisjoly, et al. (2013). "Autophagosomes contribute to intracellular lipid distribution in enterocytes." Mol Biol Cell 25(1): 118-32. Kim, J., Y. C. Kim, et al. (2013). "Differential regulation of distinct Vps34 complexes by AMPK in nutrient stress and autophagy." Cell 152(1-2): 290-303. Kim, J., M. Kundu, et al. (2011). "AMPK and mTOR regulate autophagy through direct phosphorylation of Ulk1." Nat Cell Biol 13(2): 132-41. Kim, W. Y., S. A. Nam, et al. (2011). "The role of autophagy in unilateral ureteral obstruction rat model." Nephrology (Carlton) 17(2): 148-59. Kimura, T., Y. Takabatake, et al. (2011). "Autophagy protects the proximal tubule from degeneration and acute ischemic injury." J Am Soc Nephrol 22(5): 902-13.

127

King, J. S., D. M. Veltman, et al. (2011). "The induction of autophagy by mechanical stress." Autophagy 7(12): 1490-9. Klionsky, D. J. (2007). "Autophagy: from phenomenology to molecular understanding in less than a decade." Nat Rev Mol Cell Biol 8(11): 931-7. Komatsu, M., S. Waguri, et al. (2006). "Loss of autophagy in the central nervous system causes neurodegeneration in mice." Nature 441(7095): 880-4. Komatsu, M., S. Waguri, et al. (2007). "Homeostatic levels of p62 control cytoplasmic inclusion body formation in autophagy-deficient mice." Cell 131(6): 1149-63. Komatsu, M., S. Waguri, et al. (2005). "Impairment of starvation-induced and constitutive autophagy in Atg7-deficient mice." J Cell Biol 169(3): 425-34. Kotoulas, O. B., S. A. Kalamidas, et al. (2004). "Glycogen autophagy." Microsc Res Tech 64(1): 10-20. Kotoulas, O. B., S. A. Kalamidas, et al. (2006). "Glycogen autophagy in glucose homeostasis." Pathol Res Pract 202(9): 631-8. Kotsis, F., R. Nitschke, et al. (2008). "Flow modulates centriole movements in tubular epithelial cells." Pflugers Arch 456(6): 1025-35. Kozminski, K. G., K. A. Johnson, et al. (1993). "A motility in the eukaryotic flagellum unrelated to flagellar beating." Proc Natl Acad Sci U S A 90(12): 5519-23. Kuma, A., M. Hatano, et al. (2004). "The role of autophagy during the early neonatal starvation period." Nature 432(7020): 1032-6. Kuzhandaivel, A., S. W. Schultz, et al. (2014). "Cilia-mediated hedgehog signaling in Drosophila." Cell Rep 7(3): 672-80.

Lam, H. C., S. M. Cloonan, et al. (2013). "Histone deacetylase 6-mediated selective autophagy regulates COPD-associated cilia dysfunction." J Clin Invest 123(12): 5212- 30. Lantinga-van Leeuwen, I. S., J. G. Dauwerse, et al. (2004). "Lowering of Pkd1 expression is sufficient to cause polycystic kidney disease." Hum Mol Genet 13(24): 3069-77. Lecker, S. H., A. L. Goldberg, et al. (2006). "Protein degradation by the ubiquitin-proteasome pathway in normal and disease states." J Am Soc Nephrol 17(7): 1807-19. Lee, H. K., J. M. Lund, et al. (2007). "Autophagy-dependent viral recognition by plasmacytoid dendritic cells." Science 315(5817): 1398-401. Lee, I. H., L. Cao, et al. (2008). "A role for the NAD-dependent deacetylase Sirt1 in the regulation of autophagy." Proc Natl Acad Sci U S A 105(9): 3374-9. Lee, I. H. and T. Finkel (2009). "Regulation of autophagy by the p300 acetyltransferase." J Biol Chem 284(10): 6322-8. Lee, J. A., A. Beigneux, et al. (2007). "ESCRT-III dysfunction causes autophagosome accumulation and neurodegeneration." Curr Biol 17(18): 1561-7. Lee, J. Y., H. Koga, et al. (2010). "HDAC6 controls autophagosome maturation essential for ubiquitin-selective quality-control autophagy." EMBO J 29(5): 969-80.

128

Lee, M. J., D. Feliers, et al. (2007). "A role for AMP-activated protein kinase in diabetes- induced renal hypertrophy." Am J Physiol Renal Physiol 292(2): F617-27. Leopold, P. L., M. J. O'Mahony, et al. (2009). "Smoking is associated with shortened airway cilia." PLoS One 4(12): e8157. Levine, B. (2005). "Eating oneself and uninvited guests: autophagy-related pathways in cellular defense." Cell 120(2): 159-62. Li, H., J. Li, et al. (2012). "Sonic hedgehog promotes autophagy of vascular smooth muscle cells." Am J Physiol Heart Circ Physiol 303(11): H1319-31. Li, W. W., J. Li, et al. (2012). "Microautophagy: lesser-known self-eating." Cell Mol Life Sci 69(7): 1125-36. Liang, C., J. S. Lee, et al. (2008). "Beclin1-binding UVRAG targets the class C Vps complex to coordinate autophagosome maturation and endocytic trafficking." Nat Cell Biol 10(7): 776-87. Lin, L., C. Tang, et al. (2014). "Mechanical stress triggers cardiomyocyte autophagy through angiotensin II type 1 receptor-mediated p38MAP kinase independently of angiotensin II." PLoS One 9(2): e89629. Lin, S. Y., T. Y. Li, et al. (2012). "GSK3-TIP60-ULK1 signaling pathway links growth factor deprivation to autophagy." Science 336(6080): 477-81. Liu, S., B. Hartleben, et al. (2012). "Autophagy plays a critical role in kidney tubule maintenance, aging and ischemia-reperfusion injury." Autophagy 8(5): 826-37. Liu, W., S. Xu, et al. (2003). "Effect of flow and stretch on the [Ca2+]i response of principal and intercalated cells in cortical collecting duct." Am J Physiol Renal Physiol 285(5): F998-F1012. Longatti, A., C. A. Lamb, et al. (2012). "TBC1D14 regulates autophagosome formation via Rab11- and ULK1-positive recycling endosomes." J Cell Biol 197(5): 659-75. Longatti, A. and S. A. Tooze (2010). "Recycling endosomes contribute to autophagosome formation." Autophagy 8(11): 1682-3. LoVerso, F., S. Carnio, et al. (2014). "Autophagy is not required to sustain exercise and PRKAA1/AMPK activity but is important to prevent mitochondrial damage during physical activity." Autophagy 10(11).

Malone, A. M., C. T. Anderson, et al. (2007). "Primary cilia mediate mechanosensing in bone cells by a calcium-independent mechanism." Proc Natl Acad Sci U S A 104(33): 13325-30. Manil-Segalen, M., E. Culetto, et al. (2014). "Interactions between endosomal maturation and autophagy: analysis of ESCRT machinery during Caenorhabditis elegans development." Methods Enzymol 534: 93-118. Masiero, E., L. Agatea, et al. (2009). "Autophagy is required to maintain muscle mass." Cell Metab 10(6): 507-15. Massey, A. C., S. Kaushik, et al. (2006). "Consequences of the selective blockage of chaperone-mediated autophagy." Proc Natl Acad Sci U S A 103(15): 5805-10.

129

Massey, A. C., C. Zhang, et al. (2006). "Chaperone-mediated autophagy in aging and disease." Curr Top Dev Biol 73: 205-35. Masyuk, A. I., T. V. Masyuk, et al. (2006). "Cholangiocyte cilia detect changes in luminal fluid flow and transmit them into intracellular Ca2+ and cAMP signaling." Gastroenterology 131(3): 911-20. Mathew, R. and E. White (2011). "Autophagy, stress, and cancer metabolism: what doesn't kill you makes you stronger." Cold Spring Harb Symp Quant Biol 76: 389-96. Matsunaga, K., E. Morita, et al. (2010). "Autophagy requires endoplasmic reticulum targeting of the PI3-kinase complex via Atg14L." J Cell Biol 190(4): 511-21. Matsunaga, K., T. Saitoh, et al. (2009). "Two Beclin 1-binding proteins, Atg14L and Rubicon, reciprocally regulate autophagy at different stages." Nat Cell Biol 11(4): 385-96. May-Simera, H. L. and M. W. Kelley (2012). "Cilia, Wnt signaling, and the cytoskeleton." Cilia 1(1): 7. McAlpine, F., L. E. Williamson, et al. (2013). "Regulation of nutrient-sensitive autophagy by uncoordinated 51-like kinases 1 and 2." Autophagy 9(3): 361-73. McEwan, D. G. and I. Dikic (2011). "The Three Musketeers of Autophagy: phosphorylation, ubiquitylation and acetylation." Trends Cell Biol 21(4): 195-201. McGlashan, S. R., C. J. Haycraft, et al. (2007). "Articular cartilage and growth plate defects are associated with chondrocyte cytoskeletal abnormalities in Tg737orpk mice lacking the primary cilia protein polaris." Matrix Biol 26(4): 234-46. McGrath, J., S. Somlo, et al. (2003). "Two populations of node monocilia initiate left-right asymmetry in the mouse." Cell 114(1): 61-73. Mehrpour, M., A. Esclatine, et al. (2010). "Autophagy in health and disease. 1. Regulation and significance of autophagy: an overview." Am J Physiol Cell Physiol 298(4): C776-85. Meijer, A. J. and P. Codogno (2011). "Autophagy: regulation by energy sensing." Curr Biol 21(6): R227-9. Meijer, A. J., S. Lorin, et al. (2014). "Regulation of autophagy by amino acids and MTOR- dependent signal transduction." Amino Acids. Metcalf, D. and A. M. Isaacs (2010). "The role of ESCRT proteins in fusion events involving lysosomes, endosomes and autophagosomes." Biochem Soc Trans 38(6): 1469-73. Michiorri, S., V. Gelmetti, et al. (2010). "The Parkinson-associated protein PINK1 interacts with Beclin1 and promotes autophagy." Cell Death Differ 17(6): 962-74. Miller, B. C., Z. Zhao, et al. (2008). "The autophagy gene ATG5 plays an essential role in B lymphocyte development." Autophagy 4(3): 309-14. Mitchell, B., R. Jacobs, et al. (2007). "A positive feedback mechanism governs the polarity and motion of motile cilia." Nature 447(7140): 97-101. Mizushima, N. and M. Komatsu (2011). "Autophagy: renovation of cells and tissues." Cell 147(4): 728-41. Mizushima, N. and B. Levine (2010). "Autophagy in mammalian development and differentiation." Nat Cell Biol 12(9): 823-30. Mizushima, N., B. Levine, et al. (2008). "Autophagy fights disease through cellular self- digestion." Nature 451(7182): 1069-75.

130

Mizushima, N., T. Noda, et al. (1998). "A protein conjugation system essential for autophagy." Nature 395(6700): 395-8. Mizushima, N., Y. Ohsumi, et al. (2002). "Autophagosome formation in mammalian cells." Cell Struct Funct 27(6): 421-9. Mizushima, N., A. Yamamoto, et al. (2001). "Dissection of autophagosome formation using Apg5-deficient mouse embryonic stem cells." J Cell Biol 152(4): 657-68. Mizushima, N., T. Yoshimori, et al. (2011). "The role of Atg proteins in autophagosome formation." Annu Rev Cell Dev Biol 27: 107-32. Molla-Herman, A., R. Ghossoub, et al. (2010). "The ciliary pocket: an endocytic membrane domain at the base of primary and motile cilia." J Cell Sci 123(Pt 10): 1785-95. Moreau, K., B. Ravikumar, et al. (2011). "Autophagosome precursor maturation requires homotypic fusion." Cell 146(2): 303-17. Moresi, V., M. Carrer, et al. (2012). "Histone deacetylases 1 and 2 regulate autophagy flux and skeletal muscle homeostasis in mice." Proc Natl Acad Sci U S A 109(5): 1649-54. Mortimore, G. E. and C. M. Schworer (1977). "Induction of autophagy by amino-acid deprivation in perfused rat liver." Nature 270(5633): 174-6.

Nakai, A., O. Yamaguchi, et al. (2007). "The role of autophagy in cardiomyocytes in the basal state and in response to hemodynamic stress." Nat Med 13(5): 619-24. Narendra, D., L. A. Kane, et al. (2010). "p62/SQSTM1 is required for Parkin-induced mitochondrial clustering but not mitophagy; VDAC1 is dispensable for both." Autophagy 6(8): 1090-106. Nauli, S. M., F. J. Alenghat, et al. (2003). "Polycystins 1 and 2 mediate mechanosensation in the primary cilium of kidney cells." Nat Genet 33(2): 129-37. Nauli, S. M., Y. Kawanabe, et al. (2008). "Endothelial cilia are fluid shear sensors that regulate calcium signaling and nitric oxide production through polycystin-1." Circulation 117(9): 1161-71. Nishida, K., S. Kyoi, et al. (2009). "The role of autophagy in the heart." Cell Death Differ 16(1): 31-8. Okamoto, K. (2014). "Organellophagy: eliminating cellular building blocks via selective autophagy." J Cell Biol 205(4): 435-45.

Ong, A. C., P. C. Harris, et al. (1999). "Polycystin-1 expression in PKD1, early-onset PKD1, and TSC2/PKD1 cystic tissue." Kidney Int 56(4): 1324-33.

131

Orsi, A., M. Razi, et al. (2012). "Dynamic and transient interactions of Atg9 with autophagosomes, but not membrane integration, are required for autophagy." Mol Biol Cell 23(10): 1860-73.

Paludan, C., D. Schmid, et al. (2005). "Endogenous MHC class II processing of a viral nuclear antigen after autophagy." Science 307(5709): 593-6. Pampliega, O., I. Orhon, et al. (2013). "Functional interaction between autophagy and ciliogenesis." Nature 502(7470): 194-200. Pankiv, S., T. H. Clausen, et al. (2007). "p62/SQSTM1 binds directly to Atg8/LC3 to facilitate degradation of ubiquitinated protein aggregates by autophagy." J Biol Chem 282(33): 24131-45. Patel, A. (2014). "The Primary cilium calcium channels and their role in flow sensing." Pflugers Arch. Patel, A. and E. Honore (2010). "Polycystins and renovascular mechanosensory transduction." Nat Rev Nephrol 6(9): 530-8. Patel, V., S. Hajarnis, et al. (2012). "MicroRNAs regulate renal tubule maturation through modulation of Pkd1." J Am Soc Nephrol 23(12): 1941-8. Pattingre, S., A. Tassa, et al. (2005). "Bcl-2 antiapoptotic proteins inhibit Beclin 1-dependent autophagy." Cell 122(6): 927-39. Pazour, G. J., B. L. Dickert, et al. (2000). "Chlamydomonas IFT88 and its mouse homologue, polycystic kidney disease gene tg737, are required for assembly of cilia and flagella." J Cell Biol 151(3): 709-18. Pedersen, L. B. and J. L. Rosenbaum (2008). "Intraflagellar transport (IFT) role in ciliary assembly, resorption and signalling." Curr Top Dev Biol 85: 23-61. Pedersen, L. B., I. R. Veland, et al. (2008). "Assembly of primary cilia." Dev Dyn 237(8): 1993-2006. Petralia, R. S., C. M. Schwartz, et al. (2013). "Sonic hedgehog promotes autophagy in hippocampal neurons." Biol Open 2(5): 499-504. Polager, S., M. Ofir, et al. (2008). "E2F1 regulates autophagy and the transcription of autophagy genes." Oncogene 27(35): 4860-4. Praetorius, H. A. and K. R. Spring (2001). "Bending the MDCK cell primary cilium increases intracellular calcium." J Membr Biol 184(1): 71-9. Praetorius, H. A. and K. R. Spring (2003). "Removal of the MDCK cell primary cilium abolishes flow sensing." J Membr Biol 191(1): 69-76. Prasad, S., J. P. McDaid, et al. (2009). "Pkd2 dosage influences cellular repair responses following ischemia-reperfusion injury." Am J Pathol 175(4): 1493-503. Pua, H. H., I. Dzhagalov, et al. (2007). "A critical role for the autophagy gene Atg5 in T cell survival and proliferation." J Exp Med 204(1): 25-31. Punnonen, E. L., S. Autio, et al. (1992). "Autophagy, cathepsin L transport, and acidification in cultured rat fibroblasts." J Histochem Cytochem 40(10): 1579-87. Puri, C., M. Renna, et al. (2013). "Diverse autophagosome membrane sources coalesce in recycling endosomes." Cell 154(6): 1285-99.

132

Raben, N., V. Hill, et al. (2008). "Suppression of autophagy in skeletal muscle uncovers the accumulation of ubiquitinated proteins and their potential role in muscle damage in Pompe disease." Hum Mol Genet 17(24): 3897-908. Radoshevich, L. and J. Debnath (2010). "ATG12-ATG3 and mitochondria." Autophagy 7(1): 109-11. Radoshevich, L., L. Murrow, et al. (2010). "ATG12 conjugation to ATG3 regulates mitochondrial homeostasis and cell death." Cell 142(4): 590-600. Randall-Demllo, S., M. Chieppa, et al. (2013). "Intestinal epithelium and autophagy: partners in gut homeostasis." Front Immunol 4: 301. Ravichandran, K. and C. L. Edelstein (2014). "Polycystic kidney disease: a case of suppressed autophagy?" Semin Nephrol 34(1): 27-33. Ravikumar, B., K. Moreau, et al. (2010). "Plasma membrane contributes to the formation of pre-autophagosomal structures." Nat Cell Biol 12(8): 747-57. Ravikumar, B., S. Sarkar, et al. (2010). "Regulation of mammalian autophagy in physiology and pathophysiology." Physiol Rev 90(4): 1383-435. Reiter, J. F., O. E. Blacque, et al. (2012). "The base of the cilium: roles for transition fibres and the transition zone in ciliary formation, maintenance and compartmentalization." EMBO Rep 13(7): 608-18. Resnick, N., H. Yahav, et al. (2003). "Fluid shear stress and the vascular endothelium: for better and for worse." Prog Biophys Mol Biol 81(3): 177-99. Richetta, C. and M. Faure (2013). "Autophagy in antiviral innate immunity." Cell Microbiol 15(3): 368-76. Robbins, D. J., D. L. Fei, et al. (2012). "The Hedgehog signal transduction network." Sci Signal 5(246): re6. Robinson, P. A. and H. C. Ardley (2004). "Ubiquitin-protein ligases." J Cell Sci 117(Pt 22): 5191-4. Rodat-Despoix, L., J. Hao, et al. (2013). "Shear stress-induced Ca(2)(+) mobilization in MDCK cells is ATP dependent, no matter the primary cilium." Cell Calcium 53(5-6): 327-37. Rogov, V., V. Dotsch, et al. (2014). "Interactions between autophagy receptors and ubiquitin- like proteins form the molecular basis for selective autophagy." Mol Cell 53(2): 167- 78. Rohde, J. R. and M. E. Cardenas (2003). "The tor pathway regulates gene expression by linking nutrient sensing to histone acetylation." Mol Cell Biol 23(2): 629-35. Rosenbaum, J. L. and G. B. Witman (2002). "Intraflagellar transport." Nat Rev Mol Cell Biol 3(11): 813-25. Rowe, I., M. Chiaravalli, et al. (2013). "Defective glucose metabolism in polycystic kidney disease identifies a new therapeutic strategy." Nat Med 19(4): 488-93. Rubinsztein, D. C., P. Codogno, et al. (2012). "Autophagy modulation as a potential therapeutic target for diverse diseases." Nat Rev Drug Discov 11(9): 709-30.

133

Rubinsztein, D. C., T. Shpilka, et al. (2012). "Mechanisms of autophagosome biogenesis." Curr Biol 22(1): R29-34. Rusten, T. E. and A. Simonsen (2008). "ESCRT functions in autophagy and associated disease." Cell Cycle 7(9): 1166-72.

Saburi, S., I. Hester, et al. (2008). "Loss of Fat4 disrupts PCP signaling and oriented cell division and leads to cystic kidney disease." Nat Genet 40(8): 1010-5. Saftig, P., W. Beertsen, et al. (2008). "LAMP-2: a control step for phagosome and autophagosome maturation." Autophagy 4(4): 510-2. Sahani, M. H., E. Itakura, et al. (2014). "Expression of the autophagy substrate SQSTM1/p62 is restored during prolonged starvation depending on transcriptional upregulation and autophagy-derived amino acids." Autophagy 10(3): 431-41. Sakoh-Nakatogawa, M., K. Matoba, et al. (2013). "Atg12-Atg5 conjugate enhances E2 activity of Atg3 by rearranging its catalytic site." Nat Struct Mol Biol 20(4): 433-9. Satir, P. and S. T. Christensen (2007). "Overview of structure and function of mammalian cilia." Annu Rev Physiol 69: 377-400. Satir, P., L. B. Pedersen, et al. (2010). "The primary cilium at a glance." J Cell Sci 123(Pt 4): 499-503. Sawamoto, K., H. Wichterle, et al. (2006). "New neurons follow the flow of cerebrospinal fluid in the adult brain." Science 311(5761): 629-32. Scherz-Shouval, R., E. Shvets, et al. (2007). "Reactive oxygen species are essential for autophagy and specifically regulate the activity of Atg4." EMBO J 26(7): 1749-60. Schneider, J. L. and A. M. Cuervo (2013). "Liver autophagy: much more than just taking out the trash." Nat Rev Gastroenterol Hepatol 11(3): 187-200. Schneider, J. L. and A. M. Cuervo (2014). "Autophagy and human disease: emerging themes." Curr Opin Genet Dev 26C: 16-23. Schneider, J. L., Y. Suh, et al. (2014). "Deficient chaperone-mediated autophagy in liver leads to metabolic dysregulation." Cell Metab 20(3): 417-32. Schneider, L., C. A. Clement, et al. (2005). "PDGFRalphaalpha signaling is regulated through the primary cilium in fibroblasts." Curr Biol 15(20): 1861-6. Schneider, L., C. M. Stock, et al. (2009). "The Na+/H+ exchanger NHE1 is required for directional migration stimulated via PDGFR-alpha in the primary cilium." J Cell Biol 185(1): 163-76. Scholey, J. M. (2008). "Intraflagellar transport motors in cilia: moving along the cell's antenna." J Cell Biol 180(1): 23-9. Schwartz, E. A., M. L. Leonard, et al. (1997). "Analysis and modeling of the primary cilium bending response to fluid shear." Am J Physiol 272(1 Pt 2): F132-8. Sebti, S., C. Prebois, et al. (2014). "BAT3 modulates p300-dependent acetylation of p53 and autophagy-related protein 7 (ATG7) during autophagy." Proc Natl Acad Sci U S A 111(11): 4115-20.

134

Sengupta, A., J. D. Molkentin, et al. (2009). "FoxO transcription factors promote autophagy in cardiomyocytes." J Biol Chem 284(41): 28319-31. Settembre, C. and A. Ballabio (2011). "TFEB regulates autophagy: an integrated coordination of cellular degradation and recycling processes." Autophagy 7(11): 1379-81. Settembre, C., R. De Cegli, et al. (2013). "TFEB controls cellular lipid metabolism through a starvation-induced autoregulatory loop." Nat Cell Biol 15(6): 647-58. Shaw, J., N. Yurkova, et al. (2008). "Antagonism of E2F-1 regulated Bnip3 transcription by NF-kappaB is essential for basal cell survival." Proc Natl Acad Sci U S A 105(52): 20734-9. Shiba, D., Y. Yamaoka, et al. (2009). "Localization of Inv in a distinctive intraciliary compartment requires the C-terminal ninein-homolog-containing region." J Cell Sci 122(Pt 1): 44-54. Shibutani, S. T. and T. Yoshimori (2014). "A current perspective of autophagosome biogenesis." Cell Res 24(1): 58-68. Shillingford, J. M., N. S. Murcia, et al. (2006). "The mTOR pathway is regulated by polycystin-1, and its inhibition reverses renal cystogenesis in polycystic kidney disease." Proc Natl Acad Sci U S A 103(14): 5466-71. Simpson, F., M. C. Kerr, et al. (2009). "Trafficking, development and hedgehog." Mech Dev 126(5-6): 279-88. Singh, R. and A. M. Cuervo (2011). "Autophagy in the cellular energetic balance." Cell Metab 13(5): 495-504. Singh, R., S. Kaushik, et al. (2009). "Autophagy regulates lipid metabolism." Nature 458(7242): 1131-5. Singla, V., M. Romaguera-Ros, et al. (2010). "Ofd1, a human disease gene, regulates the length and distal structure of centrioles." Dev Cell 18(3): 410-24. Sorokin, S. (1962). "Centrioles and the formation of rudimentary cilia by fibroblasts and smooth muscle cells." J Cell Biol 15: 363-77. Sorokin, S. P. (1968). "Centriole formation and ciliogenesis." Aspen Emphysema Conf 11: 213-6. Stephan, J. S., Y. Y. Yeh, et al. (2009). "The Tor and PKA signaling pathways independently target the Atg1/Atg13 protein kinase complex to control autophagy." Proc Natl Acad Sci U S A 106(40): 17049-54. Stolz, A., A. Ernst, et al. (2014). "Cargo recognition and trafficking in selective autophagy." Nat Cell Biol 16(6): 495-501. Sung, C. H. and M. R. Leroux (2013). "The roles of evolutionarily conserved functional modules in cilia-related trafficking." Nat Cell Biol 15(12): 1387-97. Suzuki, C., Y. Isaka, et al. (2008). "Participation of autophagy in renal ischemia/reperfusion injury." Biochem Biophys Res Commun 368(1): 100-6. Suzuki, K., T. Kirisako, et al. (2001). "The pre-autophagosomal structure organized by concerted functions of APG genes is essential for autophagosome formation." EMBO J 20(21): 5971-81.

135

Takeda, S. and K. Narita (2011). "Structure and function of vertebrate cilia, towards a new taxonomy." Differentiation 83(2): S4-11. Takiar, V. and M. J. Caplan (2011). "Polycystic kidney disease: pathogenesis and potential therapies." Biochim Biophys Acta 1812(10): 1337-43. Tanaka, Y., G. Guhde, et al. (2000). "Accumulation of autophagic vacuoles and cardiomyopathy in LAMP-2-deficient mice." Nature 406(6798): 902-6. Tang, Z., M. G. Lin, et al. (2013). "Autophagy promotes primary ciliogenesis by removing OFD1 from centriolar satellites." Nature 502(7470): 254-7. Tasdemir, E., M. C. Maiuri, et al. (2008). "Regulation of autophagy by cytoplasmic p53." Nat Cell Biol 10(6): 676-87. Thumm, M., R. Egner, et al. (1994). "Isolation of autophagocytosis mutants of Saccharomyces cerevisiae." FEBS Lett 349(2): 275-80. Timsit, M. O., W. J. Adams, et al. (2013). "Flow is critical for maintaining a protective phenotype in renal proximal tubular cells." Am J Transplant 13(6): 1617-8. Tooze, S. A., A. Abada, et al. (2014). "Endocytosis and autophagy: exploitation or cooperation?" Cold Spring Harb Perspect Biol 6(5): a018358. Tsukada, M. and Y. Ohsumi (1993). "Isolation and characterization of autophagy-defective mutants of Saccharomyces cerevisiae." FEBS Lett 333(1-2): 169-74. Tsukamoto, S., A. Kuma, et al. (2008). "Autophagy is essential for preimplantation development of mouse embryos." Science 321(5885): 117-20.

Valdor, R., E. Mocholi, et al. (2014). "Chaperone-mediated autophagy regulates T cell responses through targeted degradation of negative regulators of T cell activation." Nat Immunol. Van der Vaart, A. and F. Reggiori (2010). "The Golgi complex as a source for yeast autophagosomal membranes." Autophagy 6(6): 800-1. Varjosalo, M. and J. Taipale (2008). "Hedgehog: functions and mechanisms." Genes Dev 22(18): 2454-72. Vergne, I., S. Singh, et al. (2006). "Autophagy in immune defense against Mycobacterium tuberculosis." Autophagy 2(3): 175-8.

Walz, G., K. Budde, et al. (2010). "Everolimus in patients with autosomal dominant polycystic kidney disease." N Engl J Med 363(9): 830-40. Wang, R. C., Y. Wei, et al. (2012). "Akt-mediated regulation of autophagy and tumorigenesis through Beclin 1 phosphorylation." Science 338(6109): 956-9. Wang, Z. and M. E. Choi (2013). "Autophagy in kidney health and disease." Antioxid Redox Signal 20(3): 519-37.

136

Wang, Z. and M. E. Choi (2014). "Autophagy in kidney health and disease." Antioxid Redox Signal 20(3): 519-37. Wann, A. K., N. Zuo, et al. (2012). "Primary cilia mediate mechanotransduction through control of ATP-induced Ca2+ signaling in compressed chondrocytes." FASEB J 26(4): 1663-71. Watson, R. O., P. S. Manzanillo, et al. (2012). "Extracellular M. tuberculosis DNA targets bacteria for autophagy by activating the host DNA-sensing pathway." Cell 150(4): 803-15. Wei, Y., S. Pattingre, et al. (2008). "JNK1-mediated phosphorylation of Bcl-2 regulates starvation-induced autophagy." Mol Cell 30(6): 678-88. Weidberg, H., E. Shvets, et al. (2010). "LC3 and GATE-16/GABARAP subfamilies are both essential yet act differently in autophagosome biogenesis." EMBO J 29(11): 1792-802. Weinbaum, S., Y. Duan, et al. (2010). "Mechanotransduction in the renal tubule." Am J Physiol Renal Physiol 299(6): F1220-36. Weinbaum, S., Y. Duan, et al. (2011). "An Integrative Review of Mechanotransduction in Endothelial, Epithelial (Renal) and Dendritic Cells (Osteocytes)." Cell Mol Bioeng 4(4): 510-537. Williams, A., S. Sarkar, et al. (2008). "Novel targets for Huntington's disease in an mTOR- independent autophagy pathway." Nat Chem Biol 4(5): 295-305. Winyard, P. and D. Jenkins (2011). "Putative roles of cilia in polycystic kidney disease." Biochim Biophys Acta 1812(10): 1256-62. Wirth, M., J. Joachim, et al. (2013). "Autophagosome formation--the role of ULK1 and Beclin1-PI3KC3 complexes in setting the stage." Semin Cancer Biol 23(5): 301-9. Wu, L., X. Gao, et al. (2007). "Dual role of the TRPV4 channel as a sensor of flow and osmolality in renal epithelial cells." Am J Physiol Renal Physiol 293(5): F1699-713. Wu, L., Y. Zhang, et al. (2012). "The effect of resveratrol on FoxO1 expression in kidneys of diabetic nephropathy rats." Mol Biol Rep 39(9): 9085-93.

Xie, Z., U. Nair, et al. (2008). "Atg8 controls phagophore expansion during autophagosome formation." Mol Biol Cell 19(8): 3290-8. Xu, Y., S. Ruan, et al. (2013). "Autophagy and apoptosis in tubular cells following unilateral ureteral obstruction are associated with mitochondrial oxidative stress." Int J Mol Med 31(3): 628-36.

Yang, Z. and D. J. Klionsky (2010). "Eaten alive: a history of macroautophagy." Nat Cell Biol 12(9): 814-22. Yla-Anttila, P., H. Vihinen, et al. (2009). "3D tomography reveals connections between the phagophore and endoplasmic reticulum." Autophagy 5(8): 1180-5.

137

Yoshimura, K., T. Kawate, et al. (2011). "Signaling through the primary cilium affects glial cell survival under a stressed environment." Glia 59(2): 333-44. Young, A. R., E. Y. Chan, et al. (2006). "Starvation and ULK1-dependent cycling of mammalian Atg9 between the TGN and endosomes." J Cell Sci 119(Pt 18): 3888-900. Yue, W., A. Hamai, et al. (2013). "Inhibition of the autophagic flux by salinomycin in breast cancer stem-like/progenitor cells interferes with their maintenance." Autophagy 9(5): 714-29.

Zalckvar, E., H. Berissi, et al. (2009). "DAP-kinase-mediated phosphorylation on the BH3 domain of beclin 1 promotes dissociation of beclin 1 from Bcl-XL and induction of autophagy." EMBO Rep 10(3): 285-92. Zhang, H. and G. L. Semenza (2008). "The expanding universe of hypoxia." J Mol Med (Berl) 86(7): 739-46. Zhong, Y., Q. J. Wang, et al. (2009). "Distinct regulation of autophagic activity by Atg14L and Rubicon associated with Beclin 1-phosphatidylinositol-3-kinase complex." Nat Cell Biol 11(4): 468-76. Zhou, J., W. Liao, et al. (2012). "FOXO3 induces FOXO1-dependent autophagy by activating the AKT1 signaling pathway." Autophagy 8(12): 1712-23. Zhuo, J. L. and X. C. Li (2013). "Proximal nephron." Compr Physiol 3(3): 1079-123.

138

ANNEXE

139

List of Publications

Orhon I., Zaidan M., Beau I., Bauvy C., Friedlander G., Terzi F., Codogno P., Dupont N. The Role of autophagy in flow sensing via the primary cilium in kidney epithelial cells (in preparation)

Orhon I., Dupont N., Pampliega O., Cuervo AM., Codogno P. Autophagy and regulation of cilia function and assembly Cell Death and Differentiation. 2014 Oct 31(review)(in press)

Dupont N., Orhon I., Bauvy C., Codogno P.. Autophagy and autophagic flux in tumor cells. Methods in Enzymology. 2014;543:73-88

Pampliega O, Orhon I, Patel B, Sridhar S, Díaz-Carretero A, Beau I, Codogno P, Satir BH, Satir P, Cuervo AM. Functional interaction between autophagy and ciliogenesis. Nature. 2013 Oct 10;502(7470):194-200.

Malik SA, Orhon I, Morselli E, Criollo A, Shen S, Mariño G, BenYounes A, Bénit P, Rustin P, Maiuri MC, Kroemer G. BH3 mimetics activate multiple pro-autophagic pathways. Oncogene. 2011 Sep 15;30(37):3918-29

Tasdemir E, Maiuri MC, Orhon I, Kepp O, Morselli E, Criollo A, Kroemer G. p53 represses autophagy in a cell-cycle dependent fashion. Cell Cycle. 2008 Oct;7(19):3006-11.

140

CHAPTER FOUR

Autophagy and Autophagic Flux in Tumor Cells

Nicolas Dupont, Idil Orhon, Chantal Bauvy, Patrice Codogno1 INSERM U1151-CNRS UMR 8253, Institut Necker Enfants-Malades, University Paris Descartes, Paris cedex 14, France 1Corresponding author: e-mail address: [email protected]

Contents 1. Introduction 74 2. Autophagy Initiation and Autophagic Flux 75 2.1 LC3 Western blot 75 2.2 Assaying the autophagic flux in cultured cell lines using the tandem probe 79 3. Autophagic Flux and Degradation 82 3.1 SQSTM1/p62 Western blot 83 3.2 Proteolysis 84 4. Conclusion 86 Acknowledgments 87 References 87

Abstract Macroautophagy (hereafter referred to as autophagy), a central mechanism mediating the lysosomal degradation of cytoplasmic components, can be stimulated by a wide panel of adverse stimuli, including a panoply of anticancer agents. The central autophagic organelle is the autophagosome, a double membrane-bound vacuole that sequesters the cytoplasmic material destined to disposal. The ultimate destiny of the autophagosome is to fuse with a lysosome, resulting in the degradation of the autophagic cargo. In this setting, it is important to discriminate whether a particular stimulus actually promotes autophagy or it simply blocks the fusion of autophagosomes with lysosomes. To this aim, the methods that assess autophagy should assess not only the number of autophagosomes but also the so-called autophagic flux, that is, the clearance of the autophagy cargo from the lysosomal compartment. Here, we present a compendium of methods to assess the autophagic flux in cultured malignant cells. This approach should allow for the study of the intimate link between autophagy and oncometabolism in several experimental paradigms.

1. INTRODUCTION Macroautophagy (referred to below as autophagy) is a form of autophagy that degrades cellular macromolecules and cytoplasmic structures in eukaryotic cells (Boya, Reggiori, & Codogno, 2013; Yang & Klionsky, 2010). Autophagy begins with the formation of a phagophore or isolation membrane that subsequently elongates to produce double membrane-bound vacuoles known as autophagosomes, which sequester the cytoplasmic material in bulk or in a selective manner via autophagy adapters such as SQSTM1/ p62 (Lamb, Yoshimori, & Tooze, 2013). The discovery of autophagy- related (Atg) proteins (Mizushima, Yoshimori, & Ohsumi, 2011), which are involved in autophagosome formation, was a major breakthrough in the understanding of autophagy and its importance in physiology and pathology (Mizushima & Komatsu, 2011; Ravikumar et al., 2010; Rubinsztein, Codogno, & Levine, 2012). In mammalian cells, once the autophagosome has been formed, it acquires acidic and degradative capacities by merging with endocytic compartments to form a compartment known as the amphisome (Stromhaug & Seglen, 1993). The final stage of autophagy is the fusion of autophagic vacuoles (amphisomes or autophagosomes) with lysosomes to form autolysosomes, where the autophagy cargo is totally degraded. The role of autophagy in cancer is complex and context dependent (Liu & Ryan, 2012; Lorin, Hamaı¨, Mehrpour, & Codogno, 2013; White, 2012). Defects in macroautophagy promote DNA damage and genomic instability. In contrast, macroautophagy also protects cancer cells from metabolic stress by providing substances that allow them to maintain their metabolism. Moreover, several drugs used in cancer therapy stimulate autophagy, which in many instances constitutes a form of resistance to therapy (Lorin et al., 2013; Rubinsztein et al., 2012). However, in some cases, cancer treatments can trigger cell death involving autophagy (Lorin et al., 2013; Rubinsztein et al., 2012). Whatever the outcome of autophagy, the mere visualization of autophagosomes is not sufficient to conclude that autophagy has been stimulated (Klionsky et al., 2012; Mizushima, Yoshimori, & Levine, 2010). Here, two sets of information are of fundamental importance before any conclusion can be reached about the stimulation of autophagy: (1) Determination of the increase in the number of autophagosomes formed (induction of autophagy), and (2) the assay of the autophagic flux, that is, the clearance of the autophagy cargo by the lysosomal compartment. An increase in the number of autophagosomes can be the consequence of an Autophagy and Autophagic Flux 75 increase in the induction of autophagy and in autophagic flux, but it can also be the consequence of a blockade of the autophagic flux without any stimulation of the induction step (Klionsky et al., 2012; Mizushima et al., 2010). In this chapter, we describe some of the methods most commonly used for determining the induction of autophagy and autophagic flux. Many of the methods described are based on the detection of protein LC3 (a member of the Atg8 family) (Mizushima et al., 2011). During autophagy, LC3-I is converted into LC3-II by the formation of a covalent link between the C-terminus of LC3-I and the polar head of phosphatidylethanol amine. LC3-II and other members of the Atg8 family are involved in the elongation of the preautophagosomal membrane and in the sealing of the membrane (Mizushima & Komatsu, 2011; Mizushima et al., 2011; Weidberg et al., 2010). The fraction of LC3-II associated with the inner face of the autophagosome is transported into the lysosomal compartment where it is degraded. We will therefore describe LC3-based methods used to analyze the induction of autophagy and autophagic flux. We will also describe methods based on the degradation of the autophagy cargo, such as SQSTM1/p62, which is transported to the lysosome via the interaction between its LC3-Interaction Region (LIR) with LC3-II (Birgisdottir, Lamark, & Johansen, 2013) and the degradation of radiolabeled, long-lived proteins used to analyze the autophagic flux.

2. AUTOPHAGY INITIATION AND AUTOPHAGIC FLUX 2.1. LC3 Western blot 2.1.1 Materials and reagents 1. Dulbecco’s Modified Eagle’s Medium (DMEM, 4.5 g/l glucose, glutamax) (Invitrogen, Life Technologies, ref. 31966) supplemented with 10% fetal bovine serum (Invitrogen, Life Technologies). 2. Hanks’ balanced salt solution (HBSS) without sodium bicarbonate (Invitrogen, Life Technologies, ref. 14025) or Earle’s balanced salt solution (EBSS) (Invitrogen, Life Technologies, ref. 24010-043). 3. Protease Inhibitors: Bafilomycin A1 (Sigma B1793, final concentration of 100 mM)/chloroquine (Sigma C6628, final concentration of 50 mM). 4. Phosphate-buffered saline (PBS): 137 mM NaCl, 2.7 mM KCl, 4.3 mM Na2HPO4, 1.4 mM KH2PO4, pH 7.3. 5. Laemmli (2Â) sample buffer (3.75% SDS, 25% glycerol, 25% b-mercaptoethanol, 0.125% bromophenol blue, and 31.5% of 1.5 M buffer Tris–HCl (containing 0.4% SDS), pH 8.8). 76 Nicolas Dupont et al.

6. RIPA buffer (150 mM NaCl, 1% NP-40, 0.5% sodium deoxycholate, 0.1% SDS, 50 mM Tris, pH 8). 7. PBS–Tween 20 (0.1%) (Sigma-Aldrich, ref. P7949). 8. Blocking solution (5% milk PBS–Tween 20). 9. Antibody solution (2% milk PBS–Tween 20). 10. LC3B antibody (Sigma-Aldrich, ref. L7543).

2.1.2 Protocol The day before the experiment, cells are seeded in six-well plates to be used near confluence: 1. In most cases, leave the cells for 24 h to adhere before applying any stimuli. Several starvation conditions can be used to stimulate autophagy, and these include Earle’s Balanced Salt Solution (EBSS), serum deprivation (medium without FBS), amino acid starvation, or glucose starvation (see Notes A and B). 2. In order to observe the autophagic flux, Bafilomycin A1 (100 or 200 nM) or hydroxychloroquine (50 mM) is added directly to the medium for 2–4 h (see Note C) before the experiment is stopped. 3. Discard the medium and wash the cells with ice-cold PBS, while the plates are kept on ice. Aspirate all the PBS and then store the plates at À80 C for Western blotting (see Note D). 4. On the day of the Western blot experiment, thaw the stored cells on ice. Add around 100 ml (the volume depends on the cell type) of Laemmli 2Â directly to the wells at room temperature. It is important to note that cell lysis must be carried out on the same day as the Western blotting in order to avoid degradation of the LC3 (see Note E). 5. Scrape the cells using a cell scraper and collect the cell suspension in a microcentrifuge tube. Boil the samples at 100 C for 5 min. 6. Load the samples onto a 12.5% SDS-PAGE. It is important to note that a gel at less than 12.5% would not be sufficient to separate LC3-I (16 kDa) and LC3-II (14 kDa). 7. Carry out a wet transfer using a transfer buffer consisting of 24 mM Tris and 190 mM glycine, for 1 h at 120 V. Flowchart for LC3 Western blot sample preparation Cells are seeded in the usual medium in six-well plates and used near confluence. Autophagy and Autophagic Flux 77

Aspirate media

ON ICE 1ϫ wash with ice-cold PBS Aspirate the PBS

Store the plates @ –80 °C without delay

@ RT add Laemmli 2ϫ sample buffer and scrape into a microcentrifuge tube

Boil @ 100 °C for 5 min

Spin-down the samples

Load onto 12.5% SDS-PAGE

The relative band intensity of the LC3-II levels is quantified using ImageJ or some other densitometry software. In order to quantify the changes in autophagy initiation, the levels of LC3-II are calculated over the expression of actin (or some other housekeeping protein). The increase or decrease in the expression of LC3-II, the autophagosome-bound form of LC3, serves to monitor the formation of mature autophagosomes triggered by different stimuli. The total accumulation of LC3-II is observed by blocking the fusion of autophagosome with lysosomes by adding lysosomal inhibitors (e.g., Baf A1) (Tanida, Minematsu-Ikeguchi, Ueno, & Kominami, 2005). In response to autophagic stimuli, such as starvation, autophagosome synthesis would be expected to increase, which is indeed observed by the increase in lipidated LC3-II levels (LC3-II normalized versus actin) under these conditions (Fig. 4.1) (see Note F). Autophagic flux is quantified from the ratio of LC3-II levels in the presence or absence of lysosomal inhibitors. Another method of monitoring autophagic activity is to use two different time points of inhibition. By using two time- points, the turnover of the LC3-II protein can be observed over time. This methodisusefultodeterminetherateofautophagosomebiosynthesisovertime (Rubinsztein et al., 2009). 78 Nicolas Dupont et al.

A B 1.8 1.6 1.4 1.2 LC3-I 1 0.8 LC3-II

LC3-II / actin LC3-II 0.6 0.4

Actin 0.2 0 EBSS – – + + EBSS – – ++ Baf A1 ––+ + Baf A1 ––+ + Figure 4.1 LC3-II turnover during autophagy. LC3 expression levels in response to 2 h of EBSS treatment, with or without Bafilomycin A1 (100 nM). HeLa cells were cultured in DMEM for 24 h and then starved for 2 h in EBSS medium with or without the lysosomal inhibitor Bafilomycin A1 (Baf A1) in the medium. (A) Cells were lysed and proteins resolved by SDS-PAGE. LC3 was detected by immunoblotting. (B) LC3-II expression is normalized versus actin expression. Increased levels of LC3-II lipidation are observed in response to starvation, in the presence of Baf A1, which shows the induction of autophagosome synthesis in response to EBSS treatment.

2.1.3 Notes and potential pitfalls A. HBSS is used when the cells are incubated in a humidified chamber at 37 C in the absence of CO2 ; otherwise EBSS is used. The reason is that EBSS, unlike HBSS, contains bicarbonate, and the pH of the medium is governed by the Henderson–Hasselbalch equation: pH¼pKa +log À À [HCO3 ]/[CO2]¼6.1+log[HCO3 ]/[CO2]. B. A necessary precaution to ensure the correct stimulus: during the experiment, it is important to wash the cells with preheated medium containing the desired stimulus (EBSS, HBSS, or other) three times, incubating for at least 2 min during each wash in order to induce autophagic activation. C. Bafilomycin A1, a specific inhibitor of vacuolar type H+-ATPase (V-ATPase), inhibits the acidification of lysosomes and therefore blocks the last step of autophagy, that is, the fusion of the autophagosome with a lysosome, leading to an accumulation of early autophagosome vacuoles (Yamamoto et al., 1998). The time window for the incubation of cells with lysosomal inhibitors is critical and depends on the cell type involved. It is advisable to use a saturating concentration of Bafilomycin A1 (Klionsky, Elazar, Seglen, & Rubinsztein, 2008). The inhibitor treatment time is important for monitoring the blockade of the Autophagy and Autophagic Flux 79

autophagic flux because at early time-points (2–4 h) the drug seems to have a greater blocking effect. However, for treatments lasting more than 6 h, impairment of the fusion of autophagosomes with late endosomes and lysosomes is observed (Klionsky et al., 2008). It is advisable to use two different time-points for the inhibitor treatment in order to observe the change in autophagic flux. Hydroxychloroquine can be used for longer incubations if needed. D. Some treatments such as cell death inducers can lead to unequal loading with Laemmli buffer. In this step, other lysis buffers, such as RIPA buffer, can be used on the samples in order to quantify protein concentration prior to Western blot experiment. Collect the cells in cold PBS on ice in microcentrifuge tubes, centrifuge, and discard the supernatant. Add 50 ml of RIPA buffer to lyse the cells, centrifuge, and collect the supernatant. On the day of the experiment, add 50 mg of protein with sample buffer and boil the samples at 100 C for 5 min. The protein concentration can be measured using a standard protein assay kit (Pierce BCA Protein Assay Kit #23335). E. LC3-II is more sensitive to repeated freeze–thaw cycles of the samples, so it is advisable to use fresh samples in order to avoid any degradation of the protein. It is known that a third band of LC3 can be observed in some experiments after the repeated use of lysates (Klionsky et al., 2012). It is therefore advisable to use fresh cell lysates to analyze LC3 turnover. F. LC3 turnover is cell type and stress dependent. It should be noted that the turnover of LC3 protein from LC3-I to LC3-II can also be used as a marker for autophagosome formation by determining the LC3-II/LC3- I ratio; however, it is preferable to avoid this determination. Depending on the tissue or cell lines used, the abundance of LC3-I may change. The choice of LC3 antibody is also important, and the detection of the two forms of the LC3 protein may change depending on the affinity of the antibody. This means that it is preferable to quantify the turnover of LC3 from the ratio of the intensity of LC3-II over a housekeeping gene rather from than LC3-II/LC3-I (Klionsky et al., 2012).

2.2. Assaying the autophagic flux in cultured cell lines using the tandem probe An mRFP-GFP tandem fluorescent-tagged LC3 has been specially designed to monitor the autophagic flux. This method is based on differential quenching of the fluorescence emitted by the red and green probes in the 80 Nicolas Dupont et al.

Figure 4.2 Monitoring the autophagic flux by the tandem mRFP-GFP LC3 probe. HeLa cells stably expressing mRFP-GFP LC3 were or were not subjected to starvation (EBSS) for 2 hours, fixed, and then examined by confocal microscopy. Bar indicates 10 mm.

lysosomal acidic compartment (for details, see Kimura, Noda, & Yoshimori, 2007). Briefly, autophagosomes appear yellow (GFP+RFP+) and autolysosomes red (GFPÀRFP+), because of the quenching of the GFP- LC3 signal at an acidic pH (Fig. 4.2). An alternative method to monitor autophagic flux by confocal microscopy is to quantify lysosomal proteins in autophagolysosomes (see Note A).

2.2.1 Materials and reagents 1. DMEM (4.5 g/l glucose, glutamax) (Invitrogen, Life Technologies, ref. 31966) supplemented with 10% fetal bovine serum (Invitrogen, Life Technologies). 2. PBS: 137 mM NaCl, 2.7 mM KCl, 4.3 mM Na2HPO4, 1.4 mM KH2PO4, pH 7.3. 3. HBSS (Invitrogen, Life Technologies, ref. 14025) or EBSS (Invitrogen, Life Technologies, ref. 24010-043). 4. 3-Methyladenine (3-MA, Sigma-Aldrich, ref. M9281). 5. Paraformaldehyde32% (PAF, Electron Microscopy Science, ref. 15714-S). 6. Dako fluorescent mounting medium (Dako, ref. S3023).

2.2.2 Protocol The protocol described below was originally validated for use in HeLa cells (Kimura, Fujita, Noda, & Yoshimori, 2009) and can be optimized as appropriate for the cell system used. Stable or transiently transfected mRFP-GFP LC3 cells can be used (the plasmid expressing mRFP-GFP LC3 is available Autophagy and Autophagic Flux 81 on addgene) (see Notes B–). This protocol is designed for stably transfected mRFP-GFP LC3 HeLa cells. The day before the experiment, the cells are seeded in six-well plates with a cover slip (previously sterilized) and used near confluence (usually the next day): 1. On the day of the experiment, the cells are washed three times with HBSS (or EBSS) (see Note E) and then incubated for at least 2 h (for some cell lines, this period can be extended to 4 or 6 h) with fresh HBSS (or EBSS). The negative control consists of cells incubated with full medium for the same periods of time. Throughout the period of autophagy stimulation, 3-MA can be added at a final concentration of 10 mM to inhibit de novo formation of autophagic vacuoles (Seglen & Gordon, 1982; also see Note F). 2. Next, remove the medium by aspiration, wash the cells three times with PBS, fixed with 4% PAF (previously diluted from 32% PAF in PBS) for 10 min at room temperature, and then wash again with PBS. 3. To quench the fluorescence signal from free aldehyde groups, incubate the cells with NH4Cl 50 mM in PBS for 10 min at room temperature, and then wash again with PBS. 4. After mounting cover slips with mounting medium on microscope slides, pictures are obtained using a confocal microscope (see Note G and Fig. 4.2); count the yellow (GFP+RFP+) and red (GFPÀRFP+) dots using appropriate computer software, such as ImageJ.

2.2.3 Notes and potential pitfalls A. The quantification of the colocalization between a lysosomal protein such as LAMP2 and LC3 could be used via a standard immunofluorescence protocol (fixation using 4% PFA (10 min) and permeabilization using 0.1% Triton X-100 (5 min)) as an alternative method to monitor autophagic flux. The primary antibodies used for this alternative method are from MBL (PM036) and from Santa Cruz (clone H4B4) for LC3 and LAMP2, respectively. B. It has been reported that weak fluorescence of EGFP is still present in acidic lysosomes (pH between 4 and 5), leading to some yellow signals (GFP+RFP+) from autolysosomes, which can result in misinterpreta- tion of the autophagic flux results. It is therefore preferable to choose a monomeric green fluorescent protein that is more acid sensitive than EGFP, such as the mTagRFP-mWasabi-LC3 reporter (Zhou et al., 2012), in which mWasabi is more acid sensitive than EGFP. 82 Nicolas Dupont et al.

C. Another necessary precaution in interpreting the tandem fluorescent marker is that colocalization of GFP and mRFP can also be seen in the context of impaired proteolytic degradation within autolysosomes or altered lysosomal pH. The accumulation of yellow dots does not therefore always mean that the fusion of the lysosome with the autophagosome has been inhibited. D. A limitation on the use of the tandem fluorescent marker is that green dots can also be observed (GFP+RFPÀ). If they are, this is due to unfolding of the RFP proteins. These green puncta should not be taken into consideration. E. HBSS is used when the cells are incubated in a humidified chamber at 37 C in the absence of CO2; otherwise, EBSS is used. The reason is that EBSS, unlike HBSS, contains bicarbonate, and the pH of the medium is governed by the Henderson–Hasselbalch equation: pH¼pKa +log À À [HCO3 ]/[CO2]¼6.1+log[HCO3 ]/[CO2]. F. 3-MA blocks autophagy by inhibiting class-III phosphatidylinositol 3-kinase (Petiot, Ogier-Denis, Blommaart, Meijer, & Codogno, 2000). However, it should be kept in mind that 3-MA is a phosphatidylinositol 3-kinase inhibitor (Blommaart, Krause, Schellens, Vreeling-Sindelarova, & Meijer, 1997), which interferes with other intracellular trafficking pathways that are dependent on phosphatidylinositol 3-kinase. 3-MA also affects some other intracellular events (Tolkovsky, Xue, Fletcher, & Borutaite, 2002). It is a good idea to prepare a more concentrated stock solution, for example, 50 mM in water, culture medium, or balanced salt solution. It may be necessary to heat the solution under a warm water faucet to dissolve such a high concentration of 3-MA. Once prepared, the solution can be stored frozen, but the 3-MA may precipitate on thawing. G. The quantification of “yellow” and “red-only” dots can be automated using an automatic microscope (such as Cellomics from Thermoscientific) that can be used to assess a huge population of cells (1000 or more) and provide more accurate data SQSTM1/p62 than can be achieved by manual assessment of a few selected cells.

3. AUTOPHAGIC FLUX AND DEGRADATION SQSTM1/p62 is another protein marker that can be used to monitor autophagic flux. The interaction of SQSTM1/p62 with LC3 via the LIR domain means that the degradation of SQSTM1/p62 is specific to autophagy. Increased levels of SQSTM1/p62 are associated with the Autophagy and Autophagic Flux 83 inhibition of autophagy, whereas decreased SQSTM1/p62 levels are associated with the activation of autophagy. The expression and localization of endogenous SQSTM1/p62 or ectopically expressed SQSTM1/p62 can also be observed using immunofluorescence with commercially available antibodies (for details see Bjorkoy et al., 2009 and see Notes A–C).

3.1. SQSTM1/p62 Western blot 3.1.1 Materials and reagents Materials and reagents used to perform SQSTM1/p62 Western blot are identical to the materials and reagents used to perform LC3 Western blot (cf. Section 2.1.1). The SQSTM1/p62 antibody used is the BD science ref. 610830.

3.1.2 Protocol The protocol for SQSTM1/p62 Western blot is identical to the protocol for LC3 Western blot (cf. Section 2.1.2) with the exception of the way to prepare the SDS-PAGE. We recommend to use 8% or 10% SDS-PAGE for high molecular weigh proteins such as SQSTM1/p62. In response to autophagic stimuli, such as starvation, SQSTM1/p62 is degraded as a result of the activation of autophagy, which is observed at decreased levels of the protein (Fig. 4.3).

A B 1 EBSS – ++– Baf A1 – – + + 0.8

0.6 p62 0.4 p62 / actin

0.2

Actin 0 EBSS – ++– Baf A1 – – + + Figure 4.3 SQSTM1/p62 degradation during autophagy. SQSTM/p62 expression levels in response to 2 h of EBSS treatment, with or without Bafilomycin A1 (100 nM). HeLa cells were cultured in DMEM for 24 h and then starved for 2 h in EBSS medium with or without the lysosomal inhibitor Bafilomycin A1 (Baf A1) in the medium. (A) Cells were lysed and proteins resolved by SDS-PAGE. SQSTM1/p62 was detected by immunoblotting. (B) SQSTM1/p62 levels are normalized versus actin expression. Decreasing levels of the SQSTM1/p62 protein are observed in response to starvation in the presence of Baf A1. This decrease indicates the total degradation of SQSTM1/p62 by autophagolysosomes when autophagy is activated. 84 Nicolas Dupont et al.

3.1.3 Notes and potential pitfalls Most of the notes and pitfalls for LC3 Western blot can be applied for SQSTM1/p62 Western blot also (cf. Section 2.1.3); however, there are additional specific notes to look out for this assay: A. Transcriptional activation of SQSTM1/p62 may result in increased accumulation of SQSTM1/p62, which can be misleading for interpreting autophagy. Transcriptional regulation of SQSTM1/p62 can be misleading for interpreting autophagic activity depending on protein levels. SQSTM1/p62 can be upregulated under different conditions, thus increasing the Western blot levels independently of any autophagic degradation. In some cases, autophagic stimuli can induce the expression of the SQSTM1/p62 gene and protein, as in the case of muscle atrophy induced by cancer. One way to avoid misinterpreta- tions is to use a stable cell line expression EGFP-tagged SQSTM1/ p62 under the control of an inducible promoter to monitor SQSTM1/p62 degradation. Other solutions are suggested in detail in the guidelines for monitoring autophagy (Klionsky et al., 2012), including a radioactive pulse-chase measurement (Bjorkoy et al., 2009). B. SQSTM1/p62 degradation is a good readout for monitoring autophagic activity; however, because of the context- and stimulus-dependent behavior of the protein, it is recommended that an SQSTM1/p62 assay be used with different autophagy methods, as well as different SQSTM1/p62 measurements, such as pulse chase or immunofluorescence (for details see Bjorkoy et al., 2009). C. SQSTM1/p62 aggregates are insoluble in NP-40 or Triton X-100 solutions, and therefore cell lysates prepared with lysis buffers containing these detergents may show lower SQSTM1/p62 levels in Western blots. In order to monitor the correct level of SQSTM1/p62 expression, it is recommended to avoid using lysis buffers containing these detergents, but rather to use one that contain SDS, as described in the protocol (Lim et al., 2011).

3.2. Proteolysis 3.2.1 Materials and reagents 1. DMEM (4.5 g/l glucose, glutamax) (Invitrogen, Life Technologies, ref. 31966) supplemented with 10% fetal bovine serum (Invitrogen, Life Technologies). Autophagy and Autophagic Flux 85

2. PBS: 137 mM NaCl, 2.7 mM KCl, 4.3 mM Na2HPO4, 1.4 mM KH2PO4, pH 7.3. 3. HBSS (Invitrogen, Life Technologies, ref. 14025) or EBSS (Invitrogen, Life Technologies, ref. 24010-043). 14 4. L-[U- C] valine (266 mCi/mmol, 9.84 GBq/mmol, Amersham Biosciences, ref. CFB75). 5. “Cold” valine (Merck, ref. 1.08495.0100). 6. Trichloroacetic acid (TCA) (Sigma-Aldrich, ref. T4885). 7. 0.2 M NaOH. 8. 3-MA (Sigma-Aldrich, ref. M9281).

3.2.2 Protocol This protocol described below was originally validated for use with various different cancer cells (see Bauvy, Meijer, & Codogno, 2009). Cells are seeded in the usual medium in six-well plates and used near confluence: 1. Intracellular proteins are labeled for 18 h at 37 C with 0.2 mCi/ml of 14 L-[U- C] valine (spec. activity: 266 mCi/mmol) in complete medium (Pulse). 2. Any unincorporated radioactivity is eliminated by rinsing the cells three times with PBS. 3. The cells are then incubated with fresh complete medium containing 10 mM cold valine (see Note A) for 1 h to degrade short-lived proteins (in some cell lines, this period can be extended to 24 h). 4. After this period, the medium is removed and replaced with the appropriate fresh complete medium supplemented with 10 mM cold valine, and incubated for a further 4 h (longer incubations in the chase medium are also possible, and in some cases it may be desirable to use multiple points). Throughout the chase period, 3-MA (see Note B) can be added at a final concentration of 10 mM to inhibit the de novo formation of autophagic vacuoles (Seglen & Gordon, 1982). To stimulate autophagy HBSS (or EBSS) (see Note C) containing 10 mM cold valine and 0.1% of bovine serum albumin can be added. 5. The medium is then precipitated overnight after adding TCA at a final concentration of 10%. 6. After centrifuging the culture medium for 10 min at 470Âg at 4 C, the acid-soluble radioactivity is measured by liquid scintillation counting. 7. The cells are washed twice with cold TCA (w/v), plus 10 mM cold valine to make sure that no radioactivity has remained adsorbed to the 86 Nicolas Dupont et al.

denatured proteins. The cell pellet is then dissolved at 37 C in 0.2 M NaOH for 2 h. 8. The radioactivity is then measured by liquid scintillation counting. The rate of degradation of longer-lived proteins is calculated from the ratio of the acid-soluble radioactivity in the medium to that in the acid- precipitable cell fraction.

3.2.3 Notes and potential pitfalls A. Valine (amino acids) is used because this amino acid does not interfere with autophagy in most cell types. Since amino acids such as leucine are physiological inhibitors of autophagy (Meijer & Dubbelhuis, 2004), the choice of the amino acid used during the pulse chase is important. B. 3-MA blocks autophagy by inhibiting class-III phosphatidylinositol 3-kinase (Petiot et al., 2000). However, it should be kept in mind that 3-MA is a phosphatidylinositol 3-kinase inhibitor (Blommaart et al., 1997), which also interferes with other intracellular trafficking pathways dependent on the phosphatidylinositol 3-kinases. 3-MA also affects some other intracellular events (Tolkovsky et al., 2002). 3-MA is routinely used at a final concentration of 10 mM. It is a good idea to prepare a stronger stock solution, for example,100 mM in water. It may be necessary to heat the solution under a warm water faucet to dissolve such a high concentration of 3-MA.The prepared solution can be stored frozen, but the 3-MA may precipitate on thawing. C. HBSS is used when the cells are incubated in a humidified chamber at 37 C in the absence of CO2; otherwise, EBSS is used. The reason is that EBSS, unlike HBSS, contains bicarbonate, and the pH of the medium is governed by the Henderson–Hasselbalch equation: pH¼pKa +log À À [HCO3 ]/[CO2]¼6.1+log[HCO3 ]/[CO2].

4. CONCLUSION The assays presented here have been widely used. However, the gold standard assay remains the detection of autophagic structures (autophagosomes, amphisomes, and autolysosomes) by electron microscopy (Eskelinen, Reggiori, Baba, Kovacs, & Seglen, 2011). Other methods for measuring autophagic flux have also been developed, such as fluorescence-activated cell sorter-based methods (Shvets, Fass, & Elazar, Autophagy and Autophagic Flux 87

2008). Researchers should perform several autophagic assays before reaching any final conclusions about whether the initiation of autophagy or/and autophagy flux is affected under their experimental conditions.

ACKNOWLEDGMENTS N. D. is recipient of a Fondation pour la Recherche Me´dicale (FRM) fellowship. Studies in P. C.’s laboratory are supported by institutional funding from INSERM, University Paris Descartes, and grants from ANR and INCa.

REFERENCES Bauvy, C., Meijer, A. J., & Codogno, P. (2009). Assaying of autophagic protein degradation. Methods in Enzymology, 452, 47–61. Birgisdottir, A. B., Lamark, T., & Johansen, T. (2013). The LIR motif—Crucial for selective autophagy. Journal of Cell Science, 126(Pt 15), 3237–3247. Bjorkoy, G., Lamark, T., Pankiv, S., Overvatn, A., Brech, A., & Johansen, T. (2009). Mon- itoring autophagic degradation of p62/SQSTM1. Methods in Enzymology, 452, 181–197. Blommaart, E. F., Krause, U., Schellens, J. P., Vreeling-Sindelarova, H., & Meijer, A. J. (1997). The phosphatidylinositol 3-kinase inhibitors wortmannin and LY294002 inhibit autophagy in isolated rat hepatocytes. European Journal of Biochemistry, 243(1–2), 240–246. Boya, P., Reggiori, F., & Codogno, P. (2013). Emerging regulation and functions of autophagy. Nature Cell Biology, 15(7), 713–720. Eskelinen, E. L., Reggiori, F., Baba, M., Kovacs, A. L., & Seglen, P. O. (2011). Seeing is believing: The impact of electron microscopy on autophagy research. Autophagy, 7(9), 935–956. Kimura, S., Fujita, N., Noda, T., & Yoshimori, T. (2009). Monitoring autophagy in mammalian cultured cells through the dynamics of LC3. Methods in Enzymology, 452, 1–12. Kimura, S., Noda, T., & Yoshimori, T. (2007). Dissection of the autophagosome maturation process by a novel reporter protein, tandem fluorescent-tagged LC3. Autophagy, 3(5), 452–460. Klionsky, D. J., Abdalla, F. C., Abeliovich, H., Abraham, R. T., Acevedo-Arozena, A., Adeli, K., et al. (2012). Guidelines for the use and interpretation of assays for monitoring autophagy. Autophagy, 8(4), 445–544. Klionsky, D. J., Elazar, Z., Seglen, P. O., & Rubinsztein, D. C. (2008). Does bafilomycin A1 block the fusion of autophagosomes with lysosomes? Autophagy, 4(7), 849–950. Lamb, C. A., Yoshimori, T., & Tooze, S. A. (2013). The autophagosome: Origins unknown, biogenesis complex. Nature Reviews. Molecular Cell Biology, 14(12), 759–774. Lim, J., Kim, H. W., Youdim, M. B., Rhyu, I. J., Choe, K. M., & Oh, Y. J. (2011). Binding preference of p62 towards LC3-ll during dopaminergic neurotoxin-induced impairment of autophagic flux. Autophagy, 7(1), 51–60. Liu, E. Y., & Ryan, K. M. (2012). Autophagy and cancer—Issues we need to digest. Journal of Cell Science, 125(Pt 10), 2349–2358. Lorin, S., Hamaı¨, A., Mehrpour, M., & Codogno, P. (2013). Autophagy regulation and its role in cancer. Seminars in Cancer Biology, 23(5), 361–379. Meijer, A. J., & Dubbelhuis, P. F. (2004). Amino acid signalling and the integration of metabolism. Biochemical and Biophysical Research Communications, 313(2), 397–403. Mizushima, N., & Komatsu, M. (2011). Autophagy: Renovation of cells and tissues. Cell, 147(4), 728–741. 88 Nicolas Dupont et al.

Mizushima, N., Yoshimori, T., & Levine, B. (2010). Methods in mammalian autophagy research. Cell, 140(3), 313–326. Mizushima, N., Yoshimori, T., & Ohsumi, Y. (2011). The role of Atg proteins in autophagosome formation. Annual Review of Cell and Developmental Biology, 27, 107–132. Petiot, A., Ogier-Denis, E., Blommaart, E. F., Meijer, A. J., & Codogno, P. (2000). Distinct classes of phosphatidylinositol 30-kinases are involved in signaling pathways that control macroautophagy in HT-29 cells. Journal of Biological Chemistry, 275(2), 992–998. Ravikumar, B., Sarkar, S., Davies, J. E., Futter, M., Garcia-Arencibia, M., Green- Thompson, Z. W., et al. (2010). Regulation of mammalian autophagy in physiology and pathophysiology. Physiological Reviews, 90(4), 1383–1435. Rubinsztein, D. C., Codogno, P., & Levine, B. (2012). Autophagy modulation as a potential therapeutic target for diverse diseases. Nature Reviews. Drug Discovery, 11(9), 709–730. Rubinsztein, D. C., Cuervo, A. M., Ravikumar, B., Sarkar, S., Korolchuk, V., Kaushik, S., et al. (2009). In search of an “autophagomometer”. Autophagy, 5(5), 585–589. Seglen, P. O., & Gordon, P. B. (1982). 3-Methyladenine: Specific inhibitor of autophagic/ lysosomal protein degradation in isolated rat hepatocytes. Proceedings of the National Academy of Sciences of the United States of America, 79(6), 1889–1892. Shvets, E., Fass, E., & Elazar, Z. (2008). Utilizing flow cytometry to monitor autophagy in living mammalian cells. Autophagy, 4(5), 621–628. Stromhaug, P. E., & Seglen, P. O. (1993). Evidence for acidity of prelysosomal autophagic/ endocytic vacuoles (amphisomes). Biochemical Journal, 291(Pt 1), 115–121. Tanida, I., Minematsu-Ikeguchi, N., Ueno, T., & Kominami, E. (2005). Lysosomal turnover, but not a cellular level, of endogenous LC3 is a marker for autophagy. Autophagy, 1(2), 84–91. Tolkovsky, A. M., Xue, L., Fletcher, G. C., & Borutaite, V. (2002). Mitochondrial disap- pearance from cells: A clue to the role of autophagy in programmed cell death and disease? Biochimie, 84(2–3), 233–240. Weidberg, H., Shvets, E., Shpilka, T., Shimron, F., Shinder, V., & Elazar, Z. (2010). LC3 and GATE-16/GABARAP subfamilies are both essential yet act differently in autophagosome biogenesis. EMBO Journal, 29(11), 1792–1802. White, E. (2012). Deconvoluting the context-dependent role for autophagy in cancer. Nature Reviews Cancer, 12(6), 401–410. Yamamoto, A., Tagawa, Y., Yoshimori, T., Moriyama, Y., Masaki, R., & Tashiro, Y. (1998). Bafilomycin A1 prevents maturation of autophagic vacuoles by inhibiting fusion between autophagosomes and lysosomes in rat hepatoma cell line, H-4-II-E cells. Cell Structure and Function, 23(1), 33–42. Yang, Z., & Klionsky, D. J. (2010). Eaten alive: A history of macroautophagy. Nature Cell Biology, 12(9), 814–822. Zhou, C., Zhong, W., Zhou, J., Sheng, F., Fang, Z., Wei, Y., et al. (2012). Monitoring autophagic flux by an improved tandem fluorescent-tagged LC3 (mTagRFP-mWasabi-LC3) reveals that high-dose rapamycin impairs autophagic flux in cancer cells. Autophagy, 8(8), 1215–1226.

I Orhon1,2,3, N Dupont1,2,3, O Pampliega4,5, AM Cuervo*,4,5 and P Codogno1,2,3

Motile and primary cilia (PC) are microtubule-based structures located at the cell surface of many cell types. Cilia govern cellular functions ranging from motility to integration of mechanical and chemical signaling from the environment. Recent studies highlight the interplay between cilia and autophagy, a conserved cellular process responsible for intracellular degradation. Signaling from the PC recruits the autophagic machinery to trigger autophagosome formation. Conversely, autophagy regulates ciliogenesis by controlling the levels of ciliary proteins. The cross talk between autophagy and ciliated structures is a novel aspect of cell biology with major implications in development, physiology and human pathologies related to defects in cilium function. Cell Death and Differentiation advance online publication, 31 October 2014; doi:10.1038/cdd.2014.171

Facts What are the determinants that switch the degradation of ciliary proteins between basal and induced autophagy? Autophagy is a degradative and recycling mechanism that Is there any specific function for cilium-dependent auto- allows cells to adapt to stress situations including nutrient phagy in ciliated cells? deprivation. Do signals that trigger cilium-dependent autophagy depend Primary cilia (PC) and motile cilia (MC) are sensory structures on the stimuli (chemical, mechanical) or do all stimuli at the cell surface. PC sense nutrient, growth factor and converge to a single signaling pathway upstream of the calcium changes in the extracellular environment and also autophagy machinery? respond to mechanical stress. MC are beating structures that Can modulation of autophagy contribute to the restoration contribute to innate defense in the lung, for example. of cilium functions? The PC is a site for the recruitment of autophagy-related (ATG) proteins that mediate autophagosome formation in Receptors, transporters and specialized plasma membrane response to cilium-dependent signaling. In turn, autophagy structures such as cilia constitute the interfaces between the regulates the biogenesis of cilia by degrading certain extracellular and intracellular milieu and allow the cells to trigger proteins involved in cilia formation. specific responses to adapt to changes imposed by the Modulation of autophagy represents a new therapeutic environment. Among these adaptive responses, macroauto- opportunity in diseases related to defective cilia function. phagy (hereafter referred to as ‘autophagy’) is a catabolic process conserved in eukaryotic cells by which the cell recycles its own constituents.1 Autophagy is involved in the adaptation to starvation, cell differentiation and development, the degradation Open Questions of aberrant structures, organelle turnover, and in tumor suppression, innate and adaptive immunity, lifespan extension, How are ATG proteins transported to the PC? and cell death. The process of autophagy has been the subject How do ATG proteins recruited to the PC contribute to the of tremendous scientific interest in recent years because of the biogenesis of phagophores and autophagosomes? importance of autophagy in physiology and development and its Ciliary proteins are degraded by autophagy. How selective dysregulation in various human pathologies.2,3 is this degradative pathway? Do some of these proteins Cilia are microtubule-based structures present at the contain motifs that result in selective engulfment by surface of many cells.4,5 Depending on the microtubule autophagosomes? composition and function, cilia fall into two classes: motile

1INSERM U1151-CNRS UMR 8253, Paris, France; 2Institut Necker Enfants-Malades (INEM), Paris, France; 3Université Paris Descartes, Sorbonne Paris Cité, Paris, France; 4Department of Development and Molecular Biology, Albert Einstein College of Medicine, Bronx, NY,USA and 5Institute for Aging Studies, Albert Einstein College of Medicine, Bronx, NY, USA *Corresponding author: AM Cuervo, Albert Einstein College of Medicine, 1300 Morris Park Avenue, Bronx, NY 10461, USA. Tel: +1 718 430 2689. Fax: +1 718 430 8975; Email: [email protected] or P Codogno, INSERM, U1151-CNRS UMR 8253, 14 rue Maria Helena Vieira Da Silva, CS61431, Paris Cedex 14, 75993, France. Tel: +33 1 72606476; Fax: +33 1 72606399; E-mail: [email protected] Abbreviations: ADPKD, autosomal dominant polycystic kidney disease; AMPK, AMP-activated protein kinase; APC, adenomatosis polyposis complex; ATG, autophagy- related; COPD, chronic obstructive pulmonary disease; Gli, glioma-associated oncogene family; HDAC6, histone deacetylase-6; Hh, hedgehog; IFT, intraflagellar transport; MC, motile cilium; OFD1, oral-facial-digital syndrome type 1; PC1, polycystin 1; PC2, polycystin 2; PC, primary cilium; PCP, planar cell polarity; PDGF, platelet-derived growth factor; PDGFRα, platelet-derived growth factor receptor alpha; Ptc1, patched 1; PKD, polycystic kidney disease; Shh, sonic hedgehog; Smo, smoothened; Sufu, suppressor of fused protein; Vangl2, Van Gogh-like 2 protein Received 26.6.14; revised 5.9.14; accepted 10.9.14; Edited by D Rubinsztein Autophagy in ciliary pathophysiology I Orhon et al 2

cilia (MC) and primary cilia (PC).6 MC are beating organelles intraflagellar transport (IFT) proteins, along with proteins of that are present at the surface of epithelial cells in the airway, in different signaling pathways, are selected and concentrated the reproductive tract and in the brain. The PC is generally before entering the axoneme.16 considered as a non-motile organelle that grows from the Continuous trafficking of cargo molecules along the ciliary centriole and protrudes from the plasma membrane of most axoneme is controlled by the evolutionarily conserved IFT cell types.4,5 It functions as a transducer of inputs from the mechanism.17 Trafficking along the axoneme is maintained extracellular environment into different cellular signaling through the action of two motor protein families: kinesins, pathways, acting as a sensory antenna.5 responsible for anterograde IFT, and dyneins, responsible for Recent studies have shown that signals emanating from retrograde IFT. In addition, two large complexes of 420 ciliated structures control the formation of autophagosomes.7 proteins called IFT A and IFT B are also essential for IFT, A role for autophagy in ciliogenesis and control of cilia length exemplified by mutations in IFT B, which lead to the absence of has also been demonstrated.7–9 The aim of this review is to cilia or to shortened cilia, whereas defects in IFT A proteins present the recent findings that uncover the interplay between lead to a bulged cilia.18 Extraciliary functions of some of these – cilia and autophagy and to discuss their potential impact on subunits have been also described,19 21 as for example, cell homeostasis and oncilia-related disease. continuous shuttling of the protein IFT20 between the Golgi and the basal body, which facilitates the mobilization of specific cargo. Cilium: Structure and Signaling

Structure. Conventionally, cilia are classified as motile and Ciliary signaling. The major function of PC is to sense non-motile according to their microtubule pattern.4,10 MC is extracellular stimuli (Figure 1). Thus, PC responds to different formed by a microtubule-based axoneme of nine outer sensory modalities such as mechanical stimuli (e.g., a shear microtubule doublets and two single central microtubules stress that results in bending of the cilia) or chemosensation (Figure 1). Two dynein motors and a radial spoke are (e.g., specific ligand, growth factor, hormone or morphogen attached to the doublets along the axoneme, which generate recognition). A variety of signaling pathways are coordinated fluid and beat-like movements. MC are found on the surfaces through this organelle during development, tissue home- of the epithelial cells lining the airways and the reproductive ostasis, cell migration, cellular differentiation, cell cycle and tracts, as well as on epithelial cells of the ependyma and in apoptosis. Examples are the Hedgehog (Hh) pathway, the brain. Motility of the cilium in these epithelial cells is platelet-derived growth factor (PDGF) pathway, the Wnt 2+ responsible for mucociliary clearance and ependymal flow. pathway and the Ca signaling cascade (see Figure 1 for 5,22 PC are solitary and non-motile, and consist of an axoneme an overview of ciliary structure and signaling). of nine outer doublet microtubules extending from a basal body that is derived from the older (mother) centriole of the Hh signaling. The Hh pathway is involved in regulation of 4 centrosome. Two main ciliogenesis models have been cell proliferation, cell fate determination, the epithelial-to- 11 proposed. In quiescent polarized cells, the centrioles are mesenchymal transition and in homeostasis maintenance in sited close to the plasma membrane where the mother adult cells.23,24Mammalian Hh signaling pathway requires the centriole serve as a docking center for the growth of the PC as its major signaling components are localized to the axoneme. PC project from the apical surfaces of the plasma cilium, and IFT proteins are essential for trafficking of Hh membrane of a wide range of polarized cells including stem, molecules.18 epithelial, endothelial, connective-tissue muscle cells and There are three secreted Hh ligand family member proteins: neurons. In non-polarized cells, for instance, an intracellular Sonic Hh (Shh), Indian Hh and desert Hh.25,26 Hh signaling ciliary vesicle grows around the mother centriole together with depends on the balance between activator and repressor the axoneme (Figure 1). This vesicle will later fuse with the forms of the glioma-associated oncogene family (Gli) of zinc- plasma membrane, resulting in the protrusion of the PC into finger transcription factors. The basic signaling pathway the extracellular space and the generation of the ciliary involves a 12-transmembrane domain receptor protein pocket.11 Continuous with the plasma membrane, but func- Patched 1 (Ptc1), the 7-transmembrane domain protein tionally different, the ciliary pocket is a discrete structure that Smoothened (Smo), the suppressor of fused protein (Sufu) contributes to the function of the PC. It is characterized by the and the Gli transcription factors. Hh signaling is suppressed in presence of clathrin-coated vesicles resulting from the the absence of Hh ligand through the physiological inhibitory vesicular trafficking that originates and ends in this effect of Ptc1 over Smo. When the pathway is activated by Hh – region.12 14 Although it is not isolated from the cytoplasm, ligand binding to Ptc1, Smo is released from Ptc1 and the PC is a highly compartmentalized organelle (Figure 1), re-localizes to the PC axoneme where it inhibits Sufu activity. surrounded by a specialized region of the plasma membrane Sufu controls the processing and translocation of Glis from the that holds a different concentration of channels and receptors. cilium to the nucleus. The ciliary axoneme indeed is separated from the rest of the The Gli protein family includes Gli1, Gli2 and Gli3, which intracellular compartments by the ‘ciliary necklace’, which is balances between activator (GliA) and repressor forms (GliR). located at the base of the cilium and connects the transition Activator forms of Gli (GliA) directly induce transcription of Hh zone fibers of the basal body to the ‘ciliary pore complex’.15 target genes, including Ptc1 and Gli gene family This complex is a highly organized structure molecularly members.27–29 The activity of Gli1 depends on Gli2, and Gli1 similar to the nuclear pore, where different precursors and and Gli2 act primarily as transcriptional activators. Gli3 exists

Cell Death and Differentiation Autophagy in ciliary pathophysiology I Orhon et al 3

Figure 1 Structure and signaling from the cilium and relationship with the autophagic pathway. Cilia are microtubule-based structures that project from the plasma membrane (PM) to the extracellular environment and act as sensory antennas. Cilia are formed of a basal body, a transition zone (transition fibers and ciliary necklace) and an axoneme. Inset A schematically shows the PC structure of nine microtubule doublets and the MC structure of nine microtubules doublets and two single central doublets. Inset B presents two ciliogenesis models for the PC: (1) Representation of the ciliogenesis in polarized cells, where the basal body docks directly to the PM. (2) Representation of the intracellular pathway during which a ciliary vesicle grows around the mother centriole and the axoneme. This vesicle later fuses with the PM and forms the ciliary pocket (in blue). The PC function as sensors signaling pathway ligands, which dock withreceptors (such as Ptc1 and PDGFRα) located on the axoneme, orions (e.g., Ca2+) to mediate transcription of target genes (see the text for details). PC also respond to mechanical stimuli (e.g., fluid flow/shear stress) via, for example, the recruitment of Lkb1 and AMPK at the axoneme and basal body, respectively, in order to regulate the cell volume through mTOR. Ciliogenesis and the autophagic pathway are interconnected. Several ciliary proteins are degraded by autophagy (see Table 1). IFT20, which interacts with ATG16L, regulates the trafficking of ciliary components from the Golgi to the cilium. The table shows the localization of the ATG machinery induction in the basal body and/or in the axoneme in complete medium or upon serum starvation. ‘ND’ means ‘not determined’ (from results Pampliega et al.7) in two forms: a non-processed, full-length form that can PDGF pathway. PDGF receptor alpha (PDGFRα), a widely function as transcriptional activator, and a truncated amino- expressed protein, localizes specifically to the PC in – terminal fragment that acts as transcriptional repressor.27 29 quiescent fibroblasts in a ligand-dependent manner, and Also, it has been recently described that the activity of Kif7 at induces activation of Akt, Mek1/2, Erk1/2 and Rsk pathways the cilium creates a specialized compartment at the top of the at the axoneme and in the basal body.4,5 PC-dependent cilia where the activity of the Gli proteins is regulated.30 The PDGFRα pathway induces MAPK activation, as well as bifunctional Gli proteins have roles in embryogenesis and regulates the ubiquitous plasma membrane Na+/H+ adult homeostasis as their target genes regulate cell exchanger NHE1. Both MAPK and NHE1 ultimately proliferation (cyclin and Myc genes), cell death (Bcl-2), regulate cell proliferation, survival and migration during differentiation (genes encoding Forkhead family transcription embryogenesis.32,33 Defects in PDGFRα pathway can lead factors) and stem cell renewal (Bmi-1). Target genes regulated to different human pathologies including gastrointestinal by the Hh pathway depend on tissue and cell types.23,31 stromal tumors, lung tumors and ovarian carcinoma.34

Cell Death and Differentiation Autophagy in ciliary pathophysiology I Orhon et al 4

Wnt pathway. Canonical and non-canonical Wnt pathways Autophagy and Cilium are crucial during development.35,36 The non-canonical Wnt pathway is controlled by a membrane protein called Van The autophagic pathway Gogh-like 2 (Vangl2) and regulates cytoskeletal changes, cell Role of autophagy: Autophagy is a constitutive degradative adhesion, migration and polarity.37 Vangl2 functions in process that eliminates non-functional or redundant orga- 1,45–47 asymmetric positioning of MC and in centrosome positioning, nelles and protein aggregates in the cytoplasm. Levels which is important for cytoskeletal rearrangement. Mutations of autophagy are upregulated in response to sublethal stress, of different ciliary proteins, such as Kif3a, Ift88, Ofd1 (the enabling an organism to adapt to stress conditions, to protein mutated in oral-facial-digital syndrome type 1), generate intracellular nutrients and energy, and to eliminate cellular damage. In most cases, stimulation of autophagy polycystin 1 (PC1) and polycystin 2 (PC2), result in 46 dysregulation of planar cell polarity (PCP) and are associated protects cells from insults and contributes to cell survival. During nutrient deprivation, autophagy is essential to with hyper- or de-activation of the Wnt pathway depending of generate metabolites and to maintain ATP levels in various the tissue, timing, and localization of these proteins. tissues such as the liver.48,49 At birth, soon after the Components of the PCP pathway, such as Vangl2 and interruption of the maternal nutrient supply via the placenta, inversin (also called nephrocystin 2), localize at the PC, autophagy allows neonates to adapt this abrupt nutritional specifically at the basal body, during development. Inversin modification.50 serves as a molecular switch between canonical and non- Although in general autophagy is an in bulk process, it can canonical Wnt pathways by regulating Disheveled protein also be highly selective toward cellular structures and degradation at the basal body by the adenomatosis polyposis 37 microorganisms invading the cytoplasm. The paradigm for complex (APC). selective autophagy is the recognition of the cargo by The canonical Wnt pathway is regulated by β-catenin and autophagy receptors (p62/SQSTM1, NBR1, NDP52, opti- controls proliferation, apoptosis and cell fate determination via neurin and BNIP3L), which connect the cargo to the activation of target genes. The connection between the autophagy machinery (reviewed in references Birgisdottir canonical Wnt pathway and PC seems to be linear. Activation et al.51; Okamoto52; Rogov et al.53; and Stolz et al.54). of Wnt target genes depends on ligand binding to the Frizzled Autophagy has an essential role in multitude of physiological receptor at the plasma membrane and subsequent β-catenin processes including cell differentiation and development, translocation into the nucleus. The stability of β-catenin depends cellular quality control, tumor suppression, innate and adaptive on its degradation by the APC complex at the basal body. immunity, lifespan extension, and cell death.47,55 Its role in Recent studies have established the association of Wnt path- disease and its potential targeting with therapeutic agents have ways with ciliary components such as kinesins. These findings been extensively described in recent reviews.2,3,46,47,56 shed light on the contribution of the interaction of Wnt pathway Regulation of autophagy: The core molecular machinery with the PC in the regulation of cytoskeletal architecture.37 engaged in autophagosome formation, as well as the signaling pathways that stimulate autophagy have been reviewed in detail elsewhere.57–61 Fifteen autophagy-related Calcium signaling. The Ca2+-specific cation channel PC2, (ATG) proteins constitute the core machinery of autophago- 59 which forms an ion channel complex with the G-protein some formation. In mammalian cells, these ATG proteins coupled receptor PC1 localized in the axoneme, regulates are recruited to form a phagophore, or isolation membrane, calcium signaling.38,39 PC1 and PC2 are necessary for the which subsequently elongates to form the autophagosome calcium response and nitric oxide release in endothelial cells (Figure 2). Recent studies have shown that phagophore 40 elongation takes place at the endoplasmic reticulum in a in response to blood flow. Extracellular flow-induced structure called the omegasome characterized by the calcium influx leads to Ca2+ release from the endoplasmic presence of the phosphatidylinositol 3-phosphate binding reticulum via ryanodine and inositol triphosphate receptors, protein DFCP1.58,61,62 In addition, other membrane sources resulting in AMP release.41 Kidney epithelial cells respond to also contribute to the biogenesis of the autophagosome, urinary flow in the lumen of the renal tubules by bending of including those of the endosomes and the Golgi apparatus, the PC, which acts as a mechanosensory antenna, increas- 2+ the contact site between the endoplasmic reticulum and ing membrane permeability to Ca associated with the mitochondria, endoplasmic reticulum transition elements, and generated shear stress on the tubule lining cells, and, the plasma membrane (for recent reviews, Lamb et al.58; therefore, regulating the intracellular calcium response. It Shibutani and Yoshimori61; and Puri et al.63). Each functional has also been shown that different renal cells have additional module from the initiation of the autophagosome formation to 2+ 42,43 cilia-independent mechanosensitive Ca responses. the elongation/closure step involves different ATG proteins. 44 2+ Recently, Delling et al. showed by using a ratiometric Ca The ULK1 complex (mammalian ortholog of yeast Atg1) and sensor for individual cilia and patch-clamp methods that the the phosphatidylinositol 3 kinase (PIK3) complex I (which PC is a specialized organelle regarding intracellular ions. The contains PIK3C3/VPS34 and its adapter VPS15, ATG14L study demonstrated that ciliary calcium channels, specifically and Beclin 1, yeast Atg6) initiate autophagosome formation. the heteromeric PKD1L1–PKD2L1 channel, maintain a high A variety of signaling pathways that sense nutrient availability, Ca2+ concentration in the cilium with respect to cytoplasm in ATP levels, growth factors and reactive oxygen species resting state, thus regulating a steady diffusion of this control autophagy by regulating the activity of the ULK1 signaling molecule into the cytoplasm. complex and/or the activity of the PIK3 complex 1.58

Cell Death and Differentiation Autophagy in ciliary pathophysiology I Orhon et al 5

PC and Hh activation are required for the recruitment of LYSOSOME ATG16L to the basal body upon serum removal, which is delivered in IFT20-containing vesicles. This serum- dependent relocation of specific ATG proteins to

Phagophore Autophagosome Autolysosome pre-existing basal bodies leads to maximal activation of autophagy. Moreover, the contribution of the plasma membrane to autophagosome formation76,77 together with the presence of a subset of ATG proteins along the AUTOPHAGOSOME AUTOPHAGOSOME ciliary membrane, suggest a direct regulatory effect of the FORMATION STEP MATURATION / FUSION STEP PC on the biogenesis of autophagosomes from the plasma INITIATION / NUCLEATION ELONGATION / CLOSURE membrane.

ULK1 complex WIPIs ATG12-ATG5—ATG16L Rab GTPases (Rab7, Rab11) However, the relationship between autophagy and Hh

PIK3 complex I ATG9 LC3 SNAREs (syntaxin 17, SNAP-29, VAMP8) signaling appears to be complex. The study of Jiménez- Sánchez et al.78 showed that Hh signaling represses ESCRT (SKD1, CHMP2B, CHMP4B) autophagy in fibroblasts, HeLa cells and Drosophila melano- HOPS complex (VPS16, VPS33A, VPS39) gaster. Interestingly, HeLa cells are poorly ciliated cells, and LAMP (LAMP-2B, LAMP-2C) PC is not required for Hh signaling in Drosophila melano- 18 Figure 2 The autophagic pathway. There are two important stages in autophagy gaster. In support of the hypothesis that autophagy is pathway: (i) autophagosome formation, which includes the initiation/nucleation stimulated by a cilia-dependent Hh pathway, Shh treatment (formation of the phagophore) and elongation/closure, and (ii) autophagosome fusion induced autophagy in ciliated vascular smooth muscle cells79 with the lysosome (The autophagosomes can merge with endocytic compartments and in ciliated hippocampal neurons.80 The role of Hh before the fusion with the lysosome). ATG proteins involved in the formation of signaling in autophagy is probably dependent on multiple autophagosome and a non-exhaustive list of the protein families involved in the maturation/fusion step are shown in italics parameters including cell growth and the presence of PC. In addition to the modulatory effect of PC on autophagy Phosphatidylinositol 3-phosphate, which is produced by the activation, ciliogenesis is also regulated by autophagy. enzymatic activity of PIK3 class III (PIK3C3/VPS34), recruits Autophagy has been shown to degrade components essential WIPI1/2 (homologs of yeast Atg18) at the phagophore. One of for ciliogenesis, such as IFT20, and also negative regulators of the functions of WIPI2 is to control the transport of the multi- ciliogenesis such as the centriolar satellite protein OFD1 9 membrane spanning ATG9 between the phagophore and a (see Figure 1). This allows the autophagic process to control peripheral endosome/Golgi localization.64 The trafficking of the length of the PC in response to the environmental changes. The degradation of IFT20 occurs predominantly via basal ATG9 to the phagophore is an early event that occurs soon 7 58,65 autophagy, and that of OFD1 is observed early during serum after autophagy induction. Another function of WIPI2 is to 9 recruit ATG16L to contribute to the elongation of the starvation-induced autophagy. OFD1 is encoded by the gene phagophore.66 The last functional module consists of the involved in the X-linked ciliopathy, the OFD1. OFD1 functions two ubiquitin-like conjugation systems: ATG12–ATG5 (this as a suppressor of ciliogenesis at the centriolar satellites. conjugate interacts with ATG16L) and the LC3–phosphatidy- Consequently, the degradation of the centriolar satellite pool of lethanolamine (yeast Atg8), which are involved in the OFD1 by autophagy induces ciliogenesis (Figure 3). elongation and closure of the autophagosomal membrane. Interestingly, as starvation persists, IFT20 becomes again a Once formed, most autophagosomes mature into autolyso- preferred cargo for autophagy, and degradation of the ciliary somes after merging with endocytic/lysosomal compartments vesicular components imposes a negative control on ciliary growth (Figure 3). This cargo switch between basal and with the help of Rab GTPases (Rab7, Rab11), SNAREs, – – inducible autophagy determines that in basal conditions, ESCRT proteins,67 70 the HOPS complex71 73 and some of IFT20, a positive regulator of ciliogenesis, is degraded by the lysosomal membrane glycoproteins LAMP-2 variants autophagy, whereas OFD1 is located at the centriolar (Figure 2).74 The very last stage of autophagy is the satellites inhibiting ciliogenesis via sequestration of BBS4 degradation and recycling of components from the lysosomal (Table 1). Upon serum starvation, ciliogenesis is induced, and lumen to the cytosol, as well as the mammalian target of OFD1 is degraded by autophagy, whereas IFT20 protein rapamycin (mTOR)-regulated restoration of lysosomes.75 In contributes to ciliary trafficking (Figure 3). This switch between addition to the maturation of autophagosomes, it is worth basal and inducible ciliophagy highlights the importance of the noting that SNAREs and elements of the endocytic machinery interplay between autophagy and ciliogenesis for cell also contribute to the formation of autophagosomes.61,71,76 homeostasis. An interesting aspect that requires further clarification is the Autophagy and PC. Until recently, the functional interaction differences of ciliary length in autophagy-deficient MEFs between autophagy and PC had not been characterized. between the two studies under basal conditions.7,9 Differ- Pampliega et al.7 demonstrated that induction of autophago- ences in the state of confluence of the cells – known to some formation during starvation depends on functional cilia influence ciliogenesis – could be behind the discrepant and the Hh pathway, and observed that components of the findings. More broadly, additional studies are needed to autophagic machinery are located at the axoneme and basal determine the influence of the physiological context on the body of the cilium (see Figure 1 for localization of different interplay between PC-related signaling and the autophagic autophagic machinery components to the cilia). A functional pathway.

Cell Death and Differentiation Autophagy in ciliary pathophysiology I Orhon et al 6

Autophagy and ciliopathies. Cilia defects are associated of abnormal renal epithelial cell growth, together with a with a wide range of pathologies called ‘ciliopathies’ and can disturbed fluid transport resulting in end-stage renal disease affect many organs and systems.18 Polycystic kidney disease in 50% of patients. (PKD) is a genetic disorder that causes formation of fluid- The role of PC is complex in PKD. It has been shown that filled cysts in the renal tubules.81Autosomal dominant PKD mutations in various ciliary proteins, including IFT88 (also (ADPKD), the most common form of PKD, is due to mutations known as Tg737, Polaris, or Orpk), IFT20, IFT140, and a in two genes: PKD1 (around 85% of cases) and PKD2 subunit of kinesin-II Kif3-α, interfere with cilia structure and can (around 15% of cases) that encode PC1 and PC2 ciliary result in cyst formation. Mutations of PC1 and PC2, however, proteins, respectively.82 PC1 and PC2 are critical for cellular result in functional defects in cilium-dependent signaling, like 2+ repair and controlled growth, as well as for division of tubule defective mechanosensation that decreases intracellular Ca 85 cells after kidney injury, thus explaining the injury sensitivity of and reduces intracellular cAMP clearance. The role of IFT88 ADPKD kidneys.83,84 ADPKD leads to kidney enlargement (Tg737) has been extensively studied in PKD. Loss of the with increasing numbers and size of cysts over time because ciliary protein IFT88 induces embryonic lethal PKD, and reintroduction of the protein ameliorates the phenotype.86,87 Mice with a hypomorphic mutation in the gene encoding IFT88 (Tg737) show a similar phenotype to autosomal recessive PKD, with cystic kidneys, hepatic biliary disease, large cysts in the collecting duct and inability to concentrate urine.85,86 In ADPKD models, loss of cilia reduce cystogenesis; however, this reduction is independent of mTOR or cAMP signaling. The role of autophagy in PKD is under discussion because of the diverse effects of the pharmacological agents used in PKD models, as well as in the trials with the mTOR complex 1 inhibitor rapamycin and other autophagy inducers.88,89 It was previously reported that mTOR signaling is activated in PKD, and rapamycin was identified as an effective therapeutic agent against cystogenesis in rat and mice PKD models. However, rapamycin (everolimus) was not a successful treatment in clinical trials with PKD patients.88–90 The regulation of the mTOR pathway is crucial during cystic development. Epithelial cells lining the cysts expressing mutated PC1 protein have Figure 3 Autophagy as a novel modulator of ciliogenesis. (1) Basal autophagy increased levels of cell proliferation and abnormally activated prevents ciliary growth under basal conditions through degradation of proteins involved in intraflagellar trafficking such as IFT20, a protein that shuttles from Golgi to mTOR signaling because of a disrupted interaction between cilia to contribute to ciliogenesis. (2) Early upon nutrient removal, induction of PC1 and tuberin, a product of TSC2 gene, which controls autophagy favors degradation of the endogenous inhibitor of ciliogenesis OFD1. This mTOR activity.91 Boehlke et al.92 showed that in kidney switch in autophagic cargo allows an increase in the delivery of IFT20 to the base of epithelial cells, the ciliary axoneme-dependent Lkb1/AMP- the cilia and promotes ciliogenesis. Ciliary growth leads to activation of signaling activated protein kinase (AMPK) pathway controls cell volume pathways that promote recruitment of specific ATG proteins to complexes already present at the ciliary base to promote maximal activation of inducible autophagy. (3) If upstream of the inhibition of the mTOR pathway. In non-ciliated starvation is sustained, there is a switch toward IFT20 degradation to prevent cystic renal epithelial cells, or cell lines with mutations in unlimited growth of the cilia and excessive activation of the autophagic process ciliogenesis-related proteins, this regulation of cell volume is

Table 1 Examples of ciliary proteins degraded by autophagy

Autophagy substrate Cellular localization Ciliary role Associated ciliopathy Ref

IFT20a Golgi, basal body IFT B anterograde IFT complex PKD Pampliega et al.7 and axoneme and ciliary assembly via vesicular transport of proteins to the basal body OFD1b Distal centriole Cilia length and centriolar recruitment of IFT88 Oral-facial-digital Tang et al.9 syndrome PC2 Axoneme and Ca2+-specific cation channel PKD Hessel et al.100 endoplasmic reticulum IFT88 Axoneme Ciliary assembly and IFT B anterograde IFT COPD Lam et al.8 complex Centrin1 Basal body Ca2+ signaling in the connecting cilium COPD Lam et al.8 ARL13 Axoneme IFT A assembly stability COPD Lam et al.8 Pericentrin Basal body Ciliary assembly COPD Lam et al.8

Abbreviations: COPD, chronic obstructive pulmonary disease; PKD, polycystic kidney disease aIFT20 is an autophagy substrate under basal condition bOFD1 is an autophagy substrate under autophagy stimulation

Cell Death and Differentiation Autophagy in ciliary pathophysiology I Orhon et al 7 disrupted. Whether autophagy has a role in the cilium- cells may have therapeutic relevance in COPD, chronic dependent regulation of kidney epithelial cell volume remains bronchitis or emphysema patients. to be investigated. In addition, Belibi et al.93 showed that the presence of autophagosomes in cystic kidneys in response to hypoxia is due Conclusions and Future Directions to a blockage of autophagic flux. In addition, mouse kidney cells Cilia-mediated autophagy can be induced by different stimuli that express a mutated PKD1 failed to induce autophagy in because of the fact that the cilium can sense various response to glucose starvation.94 These studies suggest that extracellular changes, such as nutrient deprivation, calcium autophagy is altered in PKD patients either because of a influx, presence of morphogens or flow-induced shear stress. reduction in the biogenesis of autophagosomes, or because of The autophagic machinery is recruited to the ciliary structures limitations on the autophagic flux. A recent study reports that in vitro in response to serum starvation, suggesting a novel autophagy has a role in PKD by altering ciliary protein cellular localization for autophagosome biogenesis regulated turnover.95 In support of this hypothesis, the expression level by active recruitment of ATG16L1 via IFT20 vesicles of PC1 is critical for regulation of normal tubule formation in (Figure 1). The nature and mechanism of this vesicular cultured cells and mouse models.81 PC1 overexpression is trafficking should be further investigated in order to complete observed in cystic epithelia of ADPKD patients,96 and its our understanding of the reciprocal relationship between PC overexpression disrupts tubule formation in renal epithelial cells and autophagosome biogenesis. A growing number of ciliary in mice.97 In contrast, the downregulation of PC1 induces structures (the axoneme, basal body and the ciliary pocket) cystogenesis in mouse models.98 A major conclusion of these have been identified as platforms for multiple signaling studies is that the dosage of PC1 is important for cystogenesis. pathways (see Figure 1). Thus, it would be interesting to Recently Cebotaru et al.95 showed that PC1 controls the examine the interplay between autophagy and the transduc- degradation of PC2 by autophagy through a mechanism tion of different pathways in a cilia-dependent manner. Several dependent on histone deacetylase-6 (HDAC6). A pathogenic studies cited in this review examined the turnover of ciliary mutant of PC1 failed to induce autophagy-dependent PC2 proteins as autophagy substrates in different cellular contexts. degradation.95 Together these findings suggest that auto- The degradation of ciliary components can suggest that a phagy must be considered as a factor in the development of selective form of autophagy, defined by the term ‘ciliophagy’, renal ciliopathies. acts on certainciliary structures (see Table 1 for examples of ciliary proteins degraded by autophagy).105 The study by Lams et al.8 demonstrated that cross-talk between autophagy Autophagy and MC. Epithelial cells in the airway are ciliated and cilia is not limited to the PC but also extends to the MC. with multiple MC nucleated by a basal body. The presence of Cilia/flagella have a role in many processes including cilia in the respiratory track is critical for normal lung function, development and sensory reception in invertebrate organisms and alterations in these MC contribute to the in development (Caenorhabditis elegans, Drosophila melanogaster).21,106 of pathologies such as chronic obstructive pulmonary 99 These models have contributed to our understanding of the disease (COPD), emphysema and chronic bronchitis. In – function of autophagy,107 110 and it would be useful to COPD, patients who were chronic cigarette smokers, a investigate the interplay between autophagy and ciliated significant cilia shortening is observed that leads to impaired structures in these organisms to better understand its mucociliary clearance.100,101 Cilium shortening leads to physiological and developmental impact. excess mucus production in the lung epithelial cells and The coordinated action of autophagy and ciliogenesis interferes with the protection of airways from infections. The during nutritional stress is likely just the tip of the iceberg of lung tissues of COPD patients have been previously shown the process now known as ciliophagy that has been implicated to accumulate autophagosomes.102 Lam et al.8 demonstrated in normal functioning of various ciliated organs such as brain, that cilia shortening in chronic cigarette smoker correlates kidney, skin, bone, colon and blood vessels and in the with a HDAC6-dependent autophagy degradation of ciliary molecular basis of pathologies related to defects in ciliary proteins. In support of the proposed role of autophagy in the functions. negative regulation of ciliary growth, autophagy-impaired + − − − cells, Becn1 / , Maplc3b / and Hdac6-/Y cells, were protected from cilia shortening. In addition, ciliary proteins, Acknowledgements. Work in our laboratories is supported by institutional such as IFT88, ARL13, centrin1 and pericentrin, are funding from INSERM, CNRS, University Paris Descartes and grants from ANR and sequestered in autophagosomes (Table 1). Lam et al.8 INCa (to PC) and grants from the National Institutes of Health, NIA, NIDDK and NINDS propose a new role for selective autophagy in degradation (to AMC). ND is recipient of a Fondation pour la Recherche Medicale (FRM) fellowship. of ciliary proteins and in the regulation of MC length in response to chronic cigarette smoking stress in lung epithelial 1. Yang Z, Klionsky DJ. Eaten alive: a history of macroautophagy. Nat Cell Biol 2010; 12: cells. This hypothesis is supported by previous observations 814–822. of hyper-responsiveness in lung-specific Atg7 knock-out mice 2. Choi AM, Ryter SW, Levine B. Autophagy in human health and disease. N Engl J Med 2013; 368: 1845–1846. with abnormal ciliogenesis and long cilia formation in mTOR- 3. Rubinsztein DC, Codogno P, Levine B. Autophagy modulation as a potential therapeutic activated cells.103,104 Basal autophagy regulates the turnover target for diverse diseases. Nat Rev Drug Discov 2012; 11: 709–730. of proteins involved in cilia function, thus controlling 4. Satir P, Christensen ST. Overview of structure and function of mammalian cilia. Annu Rev Physiol 2007; 69: 377–400. sensitivity of the cell to different stressors, such as cigarette 5. Satir P, Pedersen LB, Christensen ST. The primary cilium at a glance. J Cell Sci 2010; 123: smoking. Therefore, inhibition of autophagy in lung epithelial 499–503.

Cell Death and Differentiation Autophagy in ciliary pathophysiology I Orhon et al 8

6. Christensen ST, Pedersen LB, Schneider L, Satir P. Sensory cilia and integration of signal 39. Patel A, Honore E. Polycystins and renovascular mechanosensory transduction. Nat Rev transduction in human health and disease. Traffic 2007; 8:97–109. Nephrol 2010; 6: 530–538. 7. Pampliega O, Orhon I, Patel B, Sridhar S, Diaz-Carretero A, Beau I et al. Functional 40. Nauli SM, Kawanabe Y, Kaminski JJ, Pearce WJ, Ingber DE, Zhou J. Endothelial cilia are interaction between autophagy and ciliogenesis. Nature 2013; 502: 194–200. fluid shear sensors that regulate calcium signaling and nitric oxide production through 8. Lam HC, Cloonan SM, Bhashyam AR, Haspel JA, Singh A, Sathirapongsasuti JF et al. polycystin-1. Circulation 2008; 117: 1161–1171. Histone deacetylase 6-mediated selective autophagy regulates COPD-associated cilia 41. Winyard P, Jenkins D. Putative roles of cilia in polycystic kidney disease. Biochim Biophys dysfunction. J Clin Invest 2013; 123: 5212–5230. Acta 2011; 1812: 1256–1262. 9. Tang Z, Lin MG, Stowe TR, Chen S, Zhu M, Stearns T et al. Autophagy promotes primary 42. Liu W, Xu S, Woda C, Kim P, Weinbaum S, Satlin LM. Effect of flow and stretch on the [Ca2 ciliogenesis by removing OFD1 from centriolar satellites. Nature 2013; 502: 254–257. +]i response of principal and intercalated cells in cortical collecting duct. Am J Physiol Renal 10. Takeda S, Narita K. Structure and function of vertebrate cilia, towards a new taxonomy. Physiol 2003; 285: F998–F1012. Differentiation 2012; 83:S4–11. 43. Rodat-Despoix L, Hao J, Dandonneau M, Delmas P. Shear stress-induced Ca(2)(+) 11. Ghossoub R, Molla-Herman A, Bastin P, Benmerah A. The ciliary pocket: a once-forgotten mobilization in MDCK cells is ATP dependent, no matter the primary cilium. Cell Calcium membrane domain at the base of cilia. Biol Cell 2011; 103: 131–144. 2013; 53: 327–337. 12. Benmerah A. The ciliary pocket. Curr Opin Cell Biol 2012; 25:78–84. 44. Delling M, DeCaen PG, Doerner JF, Febvay S, Clapham DE. Primary cilia are specialized 13. Clement CA, Ajbro KD, Koefoed K, Vestergaard ML, Veland IR, Henriques de Jesus MP et calcium signalling organelles. Nature 2013; 504: 311–314. al. TGF-beta signaling is associated with endocytosis at the pocket region of the 45. Boya P, Reggiori F, Codogno P. Emerging regulation and functions of autophagy. Nat Cell primary cilium. Cell Rep 2013; 3: 1806–1814. Biol 2013; 15: 713–720. 14. Molla-Herman A, Ghossoub R, Blisnick T, Meunier A, Serres C, Silbermann F et al. 46. Levine B, Kroemer G. Autophagy in the pathogenesis of disease. Cell 2008; 132:27–42. The ciliary pocket: an endocytic membrane domain at the base of primary and motile cilia. 47. Mizushima N, Komatsu M. Autophagy: renovation of cells and tissues. Cell 2011; 147: – J Cell Sci 2010; 123: 1785–1795. 728 741. 15. Reiter JF, Blacque OE, Leroux MR. The base of the cilium: roles for transition fibres and the 48. Mizushima N, Yamamoto A, Matsui M, Yoshimori T, Ohsumi Y. In vivo analysis of transition zone in ciliary formation, maintenance and compartmentalization. EMBO Rep autophagy in response to nutrient starvation using transgenic mice expressing a – 2012; 13: 608–618. fluorescent autophagosome marker. Mol Biol Cell 2004; 15: 1101 1111. 16. Kee HL, Dishinger JF, Blasius TL, Liu CJ, Margolis B, Verhey KJ. A size-exclusion 49. Meijer AJ, Lorin S, Blommaart EF, Codogno P. Regulation of autophagy by amino permeability barrier and nucleoporins characterize a ciliary pore complex that regulates acids and mTOR-dependent signal transduction. Amino Acids 2014; e-pub ahead of print 1 transport into cilia. Nat Cell Biol 2012; 14: 431–437. June 2014. 17. Rosenbaum JL, Witman GB. Intraflagellar transport. Nat Rev Mol Cell Biol 2002; 3: 50. Kuma A, Hatano M, Matsui M, Yamamoto A, Nakaya H, Yoshimori T et al. The role of – 813–825. autophagy during the early neonatal starvation period. Nature 2004; 432: 1032 1036. — 18. Goetz SC, Anderson KV. The primary cilium: a signalling centre during vertebrate 51. Birgisdottir AB, Lamark T, Johansen T. The LIR motif crucial for selective autophagy. – development. Nat Rev Genet 2010; 11: 331–344. J Cell Sci 2013; 126: 3237 3247. 19. Follit JA, Tuft RA, Fogarty KE, Pazour GJ. The intraflagellar transport protein IFT20 is 52. Okamoto K. Organellophagy: eliminating cellular building blocks via selective autophagy. J Cell Biol 2014; 205: 435–445. associated with the Golgi complex and is required for cilia assembly. Mol Biol Cell 2006; 17: 53. Rogov V, Dotsch V, Johansen T, Kirkin V. Interactions between autophagy receptors and ubiquitin- 3781–3792. like proteins form the molecular basis for selective autophagy. Mol Cell 2014; 53:167–178. 20. Keady BT, Le YZ, Pazour GJ. IFT20 is required for opsin trafficking and photoreceptor outer 54. Stolz A, Ernst A, Dikic I. Cargo recognition and trafficking in selective autophagy. Nat Cell segment development. Mol Biol Cell 2011; 22: 921–930. Biol 2014; 16: 495–501. 21. Sung CH, Leroux MR. The roles of evolutionarily conserved functional modules in 55. Deretic V, Saitoh T, Akira S. Autophagy in infection, inflammation and immunity. Nat Rev cilia-related trafficking. Nat Cell Biol 2013; 15: 1387–1397. Immunol 2013; 13: 722–737. 22. Eggenschwiler JT, Anderson KV. Cilia and developmental signaling. Annu Rev Cell Dev 56. Schneider JL, Cuervo AM. Autophagy and human disease: emerging themes. Curr Opin Biol 2007; 23: 345–373. Genet Dev 2014; 26C:16–23. 23. Briscoe J, Therond PP. The mechanisms of Hedgehog signalling and its roles in 57. Abada A, Elazar Z. Getting ready for building: signaling and autophagosome biogenesis. development and disease. Nat Rev Mol Cell Biol 2013; 14: 416–429. EMBO Rep 2014; 15: 839–852. 24. Simpson F, Kerr MC, Wicking C. Trafficking, development and hedgehog. Mech Dev 2009; 58. Lamb CA, Yoshimori T, Tooze SA. The autophagosome: origins unknown, biogenesis 126: 279–288. complex. Nat Rev Mol Cell Biol 2013; 14: 759–774. 25. Ingham PW, Nakano Y, Seger C. Mechanisms and functions of Hedgehog signalling across 59. Mizushima N, Yoshimori T, Ohsumi Y. The role of Atg proteins in autophagosome – the metazoa. Nat Rev Genet 2011; 12: 393 406. formation. Annu Rev Cell Dev Biol 2011; 27: 107–132. 26. Varjosalo M, Taipale J. Hedgehog: functions and mechanisms. Genes Dev 2008; 22: 60. Rubinsztein DC, Shpilka T, Elazar Z. Mechanisms of autophagosome biogenesis. Curr Biol – 2454 2472. 2012; 22:R29–R34. 27. Aberger F, Ruiz IAA. Context-dependent signal integration by the GLI code: the oncogenic 61. Shibutani ST, Yoshimori T. A current perspective of autophagosome biogenesis. Cell Res load, pathways, modifiers and implications for cancer therapy. Semin Cell Dev Biol 2014; 2014; 24:58–68. – 33C:93 104. 62. Axe EL, Walker SA, Manifava M, Chandra P, Roderick HL, Habermann A et al. 28. Hui CC, Angers S. Gli proteins in development and disease. Annu Rev Cell Dev Biol 2011; Autophagosome formation from membrane compartments enriched in phosphatidylinositol – 27: 513 537. 3-phosphate and dynamically connected to the endoplasmic reticulum. J Cell Biol 2008; 29. Robbins DJ, Fei DL, Riobo NA. The Hedgehog signal transduction network. Sci Signal 182: 685–701. 2012; 5: re6. 63. Puri C, Renna M, Bento CF, Moreau K, Rubinsztein DC. ATG16L1 meets ATG9 in recycling 30. He M, Subramanian R, Bangs F, Omelchenko T, Liem KF Jr., Kapoor TM et al. The endosomes: additional roles for the plasma membrane and endocytosis in autophagosome kinesin-4 protein Kif7 regulates mammalian Hedgehog signalling by organizing the cilium biogenesis. Autophagy 2014; 10: 182–184. – tip compartment. Nat Cell Biol 2014; 16: 663 672. 64. Orsi A, Razi M, Dooley HC, Robinson D, Weston AE, Collinson LM et al. Dynamic and 31. Jiang J, Hui CC. Hedgehog signaling in development and cancer. Dev Cell 2008; 15: transient interactions of Atg9 with autophagosomes, but not membrane integration, are – 801 812. required for autophagy. Mol Biol Cell 2012; 23: 1860–1873. 32. Clement DL, Mally S, Stock C, Lethan M, Satir P, Schwab A et al. PDGFRalpha signaling in 65. Zavodszky E, Vicinanza M, Rubinsztein DC. Biology and trafficking of ATG9 and ATG16L1, the primary cilium regulates NHE1-dependent fibroblast migration via coordinated two proteins that regulate autophagosome formation. FEBS Lett 2013; 587: 1988–1996. differential activity of MEK1/2-ERK1/2-p90RSK and AKT signaling pathways. J Cell Sci 66. Dooley HC, Razi M, Polson HE, Girardin SE, Wilson MI, Tooze SA. WIPI2 links LC3 2013; 126: 953–965. conjugation with PI3P, autophagosome formation, and pathogen clearance by recruiting 33. Schneider L, Stock CM, Dieterich P, Jensen BH, Pedersen LB, Satir P et al. The Na+/H+ Atg12-5-16L1. Mol Cell 2014; 55: 238–252. exchanger NHE1 is required for directional migration stimulated via PDGFR-alpha in the 67. Bento CF, Puri C, Moreau K, Rubinsztein DC. The role of membrane-trafficking small primary cilium. J Cell Biol 2009; 185: 163–176. GTPases in the regulation of autophagy. J Cell Sci 2013; 126: 1059–1069. 34. Schneider L, Clement CA, Teilmann SC, Pazour GJ, Hoffmann EK, Satir P et al. 68. Itakura E, Kishi-Itakura C, Mizushima N. The hairpin-type tail-anchored SNARE syntaxin 17 PDGFRalphaalpha signaling is regulated through the primary cilium in fibroblasts. Curr Biol targets to autophagosomes for fusion with endosomes/lysosomes. Cell 2012; 151: 2005; 15: 1861–1866. 1256–1269. 35. Angers S, Moon RT. Proximal events in Wnt signal transduction. Nat Rev Mol Cell Biol 69. Rusten TE, Simonsen A. ESCRT functions in autophagy and associated disease. Cell 2009; 10: 468–477. Cycle 2008; 7: 1166–1172. 36. Kestler HA, Kuhl M. From individual Wnt pathways towards a Wnt signalling network. Philos 70. Tooze SA, Abada A, Elazar Z. Endocytosis and autophagy: exploitation or cooperation? Trans R Soc Lond B Biol Sci 2008; 363: 1333–1347. Cold Spring Harb Perspect Biol 2014; 6: a018358. 37. May-Simera HL, Kelley MW. Cilia, Wnt signaling, and the cytoskeleton. Cilia 2012; 1:7. 71. Jiang P, Nishimura T, Sakamaki Y, Itakura E, Hatta T, Natsume T et al. The HOPS complex 38. Patel A. The Primary cilium calcium channels and their role in flow sensing. Pflugers Arch mediates autophagosome-lysosome fusion through interaction with syntaxin 17. Mol Biol 2014; e-pub ahead of print 26 April 2014. Cell 2014; 25: 1327–1337.

Cell Death and Differentiation Autophagy in ciliary pathophysiology I Orhon et al 9

72. Manil-Segalen M, Lefebvre C, Jenzer C, Trichet M, Boulogne C, Satiat-Jeunemaitre B et al. 92. Boehlke C, Kotsis F, Patel V, Braeg S, Voelker H, Bredt S et al. Primary cilia regulate The C. elegans LC3 acts downstream of GABARAP to degrade autophagosomes by mTORC1 activity and cell size through Lkb1. Nat Cell Biol 2010; 12: 1115–1122. interacting with the HOPS subunit VPS39. Dev Cell 2014; 28:43–55. 93. Belibi F, Zafar I, Ravichandran K, Segvic AB, Jani A, Ljubanovic DG et al. Hypoxia-inducible 73. Takats S, Pircs K, Nagy P, Varga A, Karpati M, Hegedus K et al. Interaction of the HOPS factor-1alpha (HIF-1alpha) and autophagy in polycystic kidney disease (PKD). Am J complex with Syntaxin 17 mediates autophagosome clearance in Drosophila. Mol Biol Cell Physiol Renal Physiol 2011; 300: F1235–F1243. 2014; 25: 1338–1354. 94. Rowe I, Chiaravalli M, Mannella V, Ulisse V, Quilici G, Pema M et al. Defective glucose 74. Saftig P, Beertsen W, Eskelinen EL. LAMP-2: a control step for phagosome and metabolism in polycystic kidney disease identifies a new therapeutic strategy. Nat Med autophagosome maturation. Autophagy 2008; 4: 510–512. 2013; 19: 488–493. 75. Yu L, McPhee CK, Zheng L, Mardones GA, Rong Y, Peng J et al. Termination of 95. Cebotaru V, Cebotaru L, Kim H, Chiaravalli M, Boletta A, Qian F et al. Polycystin-1 autophagy and reformation of lysosomes regulated by mTOR. Nature 2010; 465: negatively regulates polycystin-2 expression via the aggresome/autophagosome pathway. – 942 946. J Biol Chem 2014; 289: 6404–6414. 76. Puri C, Renna M, Bento CF, Moreau K, Rubinsztein DC. Diverse autophagosome 96. Ong AC, Harris PC, Davies DR, Pritchard L, Rossetti S, Biddolph S et al. Polycystin-1 – membrane sources coalesce in recycling endosomes. Cell 2013; 154: 1285 1299. expression in PKD1, early-onset PKD1, and TSC2/PKD1 cystic tissue. Kidney Int 1999; 56: 77. Moreau K, Ravikumar B, Renna M, Puri C, Rubinsztein DC. Autophagosome precursor 1324–1333. – maturation requires homotypic fusion. Cell 2011; 146: 303 317. 97. Patel V, Hajarnis S, Williams D, Hunter R, Huynh D, Igarashi P. MicroRNAs regulate renal 78. Jiménez-Sánchez M, Menzies FM, Chang YY, Simecek N, Neufeld TP, Rubinsztein DC. tubule maturation through modulation of Pkd1. J Am Soc Nephrol 2012; 23: 1941–1948. The Hedgehog signalling pathway regulates autophagy. Nat Commun 2012; 3: 1200. 98. Lantinga-van Leeuwen IS, Dauwerse JG, Baelde HJ, Leonhard WN, van de Wal A, Ward 79. Li H, Li J, Li Y, Singh P, Cao L, Xu LJ et al. Sonic hedgehog promotes autophagy of CJ et al. Lowering of Pkd1 expression is sufficient to cause polycystic kidney disease. Hum vascular smooth muscle cells. Am J Physiol Heart Circ Physiol 2012; 303: H1319–H1331. Mol Genet 2004; 13: 3069–3077. 80. Petralia RS, Schwartz CM, Wang YX, Kawamoto EM, Mattson MP, Yao PJ. Sonic 99. Barnes PJ. Chronic obstructive pulmonary disease. N Engl J Med 2000; 343: 269–280. hedgehog promotes autophagy in hippocampal neurons. Biol Open 2013; 2: 499–504. 100. Hessel J, Heldrich J, Fuller J, Staudt MR, Radisch S, Hollmann C et al. Intraflagellar 81. Chapin HC, Caplan MJ. The cell biology of polycystic kidney disease. J Cell Biol 2010; 191: transport gene expression associated with short cilia in smoking and COPD. PLoS One 701–710. 2014; 9: e85453. 82. Takiar V, Caplan MJ. Polycystic kidney disease: pathogenesis and potential therapies. 101. Leopold PL, O'Mahony MJ, Lian XJ, Tilley AE, Harvey BG, Crystal RG. Smoking is Biochim Biophys Acta 2011; 1812: 1337–1343. associated with shortened airway cilia. PLoS One 2009; 4: e8157. 83. Bastos AP, Piontek K, Silva AM, Martini D, Menezes LF, Fonseca JM et al. Pkd1 102. Chen ZH, Kim HP, Sciurba FC, Lee SJ, Feghali-Bostwick C, Stolz DB et al. Egr-1 regulates haploinsufficiency increases renal damage and induces microcyst formation following autophagy in cigarette smoke-induced chronic obstructive pulmonary disease. PLoS One ischemia/reperfusion. J Am Soc Nephrol 2009; 20: 2389–2402. 84. Prasad S, McDaid JP, Tam FW, Haylor JL, Ong AC. Pkd2 dosage influences cellular repair 2008; 3: e3316. responses following ischemia-reperfusion injury. Am J Pathol 2009; 175: 1493–1503. 103. DiBella LM, Park A, Sun Z. Zebrafish Tsc1 reveals functional interactions between the – 85. Cowley BD Jr. Calcium, cyclic AMP, and MAP kinases: dysregulation in polycystic kidney cilium and the TOR pathway. Hum Mol Genet 2009; 18: 595 606. disease. Kidney Int 2008; 73: 251–253. 104. Inoue D, Kubo H, Taguchi K, Suzuki T, Komatsu M, Motohashi H et al. Inducible disruption 86. Pazour GJ, Dickert BL, Vucica Y, Seeley ES, Rosenbaum JL, Witman GB et al. of autophagy in the lung causes airway hyper-responsiveness. Biochem Biophys Res – Chlamydomonas IFT88 and its mouse homologue, polycystic kidney disease gene tg737, Commun 2011; 405:13 18. ‘ ’ are required for assembly of cilia and flagella. J Cell Biol 2000; 151: 709–718. 105. Cloonan SM, Lam HC, Ryter SW, Choi AM. Ciliophagy : the consumption of cilia – 87. Brown NE, Murcia NS. Delayed cystogenesis and increased ciliogenesis associated with components by autophagy. Autophagy 2014; 10: 532 534. the re-expression of polaris in Tg737 mutant mice. Kidney Int 2003; 63: 1220–1229. 106. Scholey JM, Anderson KV. Intraflagellar transport and cilium-based signaling. Cell 2006; 88. Huber TB, Edelstein CL, Hartleben B, Inoki K, Jiang M, Koya D et al. Emerging role of 125: 439–442. autophagy in kidney function, diseases and aging. Autophagy 2012; 8: 1009–1031. 107. Levine B, Klionsky DJ. Development by self-digestion: molecular mechanisms and 89. Wang Z, Choi ME. Autophagy in kidney health and disease. Antioxid Redox Signal 2014; biological functions of autophagy. Dev Cell 2004; 6:463–477. 20: 519–537. 108. Mulakkal NC, Nagy P, Takats S, Tusco R, Juhasz G, Nezis IP. Autophagy in Drosophila: 90. Ravichandran K, Edelstein CL. Polycystic kidney disease: a case of suppressed from historical studies to current knowledge. Biomed Res Int 2014; 2014: 273473. autophagy? Semin Nephrol 2014; 34:27–33. 109. Tracy K, Baehrecke EH. The role of autophagy in Drosophila metamorphosis. Curr Top Dev 91. Shillingford JM, Murcia NS, Larson CH, Low SH, Hedgepeth R, Brown N et al. The mTOR Biol 2013; 103: 101–125. pathway is regulated by polycystin-1, and its inhibition reverses renal cystogenesis in 110. Yang P, Zhang H. You are what you eat: multifaceted functions of autophagy during polycystic kidney disease. Proc Natl Acad Sci USA 2006; 103: 5466–5471. C. elegans development. Cell Res 2014; 24:80–91.

Cell Death and Differentiation