Functional Analysis of Bicc1 and Novel Interacting Proteins in Renal Tubule Cell Metabolism and Polycystic Kidneys

Functional analysis of Bicc1 and novel interacting proteins in renal tubule cell metabolism and polycystic kidneys

THÈSE NO 8266 (2017)

PRÉSENTÉE LE 19 DÉCEMBRE 2017 À LA FACULTÉ DES SCIENCES DE LA VIE UNITÉ DU PROF. CONSTAM PROGRAMME DOCTORAL EN APPROCHES MOLÉCULAIRES DU VIVANT

ÉCOLE POLYTECHNIQUE FÉDÉRALE DE LAUSANNE

POUR L'OBTENTION DU GRADE DE DOCTEUR ÈS SCIENCES

PAR

Lucía Carolina LEAL ESTEBAN

acceptée sur proposition du jury:

Prof. K. Sakamoto, président du jury Prof. D. Constam, directeur de thèse Prof. A. Boletta, rapporteuse Prof. E. Feraille, rapporteur Prof. E. Meylan, rapporteur

Suisse 2017

“You cannot swim for new horizons until you have courage to lose sight of the shore.” W. Faulkner.

To my mother.

Acknowledgements

At this point, when the experience of doing a Ph.D. is materialized in this dissertation, I look back, and I smile because it has been the biggest learning experience of my life. I was very lucky to come to EPFL, and I am very proud of my achievements during all these years. However, is because of all the people next to me that helped, taught and supported me to arrive at this point, that I can write these words.

First, I would like to thank Prof. Constam for giving me the opportunity to be part of his lab and for making me responsible for this project. Thank you for the scientific discussions and for having such a nice group that made me feel pleased to spend hours and hours in the lab.

I am very grateful to the professors Alessandra Boletta, Eric Feraille, Etienne Meylan and Kei Sakamoto, for having accepted to be part of my thesis jury and for taking the time to review this work.

Special thanks to the UNIL metabolomics facility, to Dr. Julijana Ivanisevic and Tony Tiev, for their help in the analysis of the samples and for welcoming me during some weeks in their lab. To Romain Hamelin and Diego Chiappe at EPFL proteomics facility for the mass spectrometry analysis. Thanks to Dr. Raphaël Doenlen and Roy Combe from the Center of Phenogenomics at EPFL for the analysis, and to Hongbo Zhang for technical assistance with the seahorse XF analyzer.

To all the UPCDA former and present members, thanks for being such an excellent group to work with. I would like to thank Benjamin, for his supervision, help, and support and particularly for being my scientific role model during these years. Simon thanks for always offering your help. To Prudence, who was my person in and outside the lab, I am so grateful to had you there when I needed a friend. Thanks to Pierpaolo, who was my brother in this experience, for so many moments together, for encouraging me in the hardest times in and outside the lab. Olivier and Manuela, thanks for making me laugh every day. Manu thanks for taking care of me.

Additionally, many thanks to my friend Zuzana, she was my family here and one of my biggest supports. To my Colombian community, Angela, Marcela, Natali, Mariela, Jhonny, Nicolas, and Wilson: THANK YOU for being a great representation at EPFL and for your being next to me in so many unforgettable moments.

To the friends I met outside the academic world. Thanks for showing me the “other side of life” as I liked to call it. I enjoyed very much hanging out and traveling with them, and I am thankful because they showed me in many ways what is really important in life. To Luis, thanks for listening and helping me for four years. Big thanks to all, present and former, members of BioScience Network Lausanne, for being an amazing association and hardworking people. It was wonderful working with and learning from them in the improvement of the scientific community in Lausanne by organizing so many nice events.

My friends from Colombia, for the Skype sessions, phone calls, and the annual meetings.

Thank you for showing me your support and love in the distance. Especially, I would like to thank Eloisa, who always was next to me and helped me in the most difficult moment of this journey.

Great THANKS to my family in Colombia and Canada, for taking care of me no matter how far we have been and for transmitting me day by day the strength to continue.

Finally, I dedicate this work and these years to my mother. She saw me starting my Ph.D. but she passed away soon after and cannot see me accomplishing it. To her, all my love and gratitude for making me the woman I am today and for always encouraging me to pursue my dreams and my happiness. Lo hicimos!

iii

Summary

Genetic disorders known as ciliopathies develop polycystic kidneys, including autosomal dominant and autosomal recessive polycystic kidney diseases (ADPKD and ARPKD). Several signaling pathways including the cAMP/PKA pathway are implicated in driving cystic growth. However, the mechanism underlying cyst formation remains incompletely understood. Renal cysts also arise in patients and in mouse models with mutations in the RNA-binding protein Bicaudal-C (Bicc1). Bicc1 binds specific mRNAs such as adenylate cyclase VI (AC6) mRNA for translational repression. Conducting a structure-function analysis, we demonstrated that Bicc1 self-polymerizes through its sterile alpha motif (SAM domain) to form cytoplasmic clusters, which specifically localize and silence bound mRNAs. Altered glucose and lipid metabolism also sustain cystic growth in polycystic kidneys, but the cause of these perturbations is unclear. Here, we found that Bicc1 is essential for normal expression of the gluconeogenic enzymes FBP1 and PEPCK specifically in kidneys, and to maintain euglycemia. In a proteomic interactome screen using a knock-in cell line, we found that TAP- tagged Bicc1 binds the C-Terminal to Lis-Homology domain (CTLH) complex, the mammalian orthologue of the glucose-induced degradation (Gid) complex that triggers degradation of

FBP1 in S. cerevisiae. Bicc1 stimulates FBP1 expression in vitro in cultured murine inner medullary collecting duct (mIMCD3) cells, albeit independently of its interaction with the CTLH complex. Instead, Bicc1 and CTLH complex seem to mutually downregulate each other. Bicc1 interactome is also enriched with metabolic related proteins. Here we found that Bicc1 coimmunoprecipitates with rate-limiting metabolic enzymes and contributes to maintain their protein levels, and that deletion of Bicc1 significantly alters the renal metabolome while enhancing autophagy. Altogether, these results implicate Bicc1 as an important metabolic regulator in the etiology of polycystic kidneys.

Keywords: Bicaudal-C, polycystic kidney disease, gluconeogenesis, CTLH complex, renal metabolism

Resumé

Les ciliopathies, dont les polykystoses rénales autosomiques dominantes et autosomiques négatives, sont des maladies génétiques caractérisées par le développement de kystes. De nombreuses voies de signalisation, dont la voie AMPc/PKA, sont impliquées dans la croissance des kystes, cependant les mécanismes conduisant à leur formation restent à ce jour méconnus. Les kystes rénaux surviennent également chez des patients et des modèles de souris porteurs de mutations dans la protéine de fixation à l’ARN Bicaudal-C (Bicc1). Bicc1 interagit avec des ARNm spécifiques, tels que l’ARNm codant l’adenylate cyclase VI (AC6), afin de réprimer leur traduction. En conduisant une analyse de relation structure-fonction, nous avons démontré que Bicc1 auto-polymérise via son domaine sterile alpha motif (SAM) pour former des corpuscules cytoplasmiques qui concentrent et répriment spécifiquement les ARNm associés. L’altération du métabolisme du glucose et des lipides contribue également la croissance des kystes rénaux, mais la cause de ces perturbations reste inconnue. Dans cette étude, nous avons montré que Bicc1 est essentielle pour l’expression normale des enzymes de la néoglucogenèse FBP1 et PEPCK spécifiquement dans les reins, ainsi que pour le maintien de l’euglycémie. En utilisant une approche protéomique à partir d'une lignée cellulaire knock-in, nous avons découvert que la protéine étiquetée TAP-Bicc1 interagit avec le complexe C-Terminal to Lis-Homology domain (CTLH), l’orthologue mammifère du complexe Glucose-induced degradation (Gid) de la levure S. cerevisiae. Indépendamment de son interaction avec le complexe CTLH, Bicc1 stimule l’expression de FBP1 in vitro dans des cellules de tubes collecteurs médullaires internes de souris (mIMCD3). En revanche, Bicc1 et le complexe CTLH semblent se réguler mutuellement de façon négative. L’interactome de Bicc1 est également enrichi en protéines liées au métabolisme. Dans cette étude, nous avons montré que Bicc1 co-immunoprécipite avec des enzymes métaboliques clefs et contribue au maintien de leurs niveaux protéiques, tandis que la délétion de Bicc1 altère significativement le métabolome rénal et augmente l’autophagie. Bicc1 et les facteurs régulateurs identifiés dans son interactome apparaissent comme de nouveaux candidats impliqués dans les perturbations métaboliques retrouvées dans les polykystoses rénales. Ces résultats positionnent Bicc1 comme un important régulateur du métabolisme dans l’étiologie des polykystoses rénales.

Mots clefs: Bicaudal-C, polykystose rénale, néoglucogenèse, complexe CTLH, métabolisme rénal

Table of content

Acknowledgements ...... ii

Summary ...... iv

Resumé ...... v

Table of content ...... vii

List of Figures ...... ix

Abbreviations ...... xi

1. Introduction ...... 1

1.1 The Kidneys ...... 1

1.2 Primary cilia and their role in the kidney ...... 13

1.3 Ciliopathies ...... 20

1.4 Bicaudal-C ...... 35

2. Thesis aims ...... 43

3. Results part I ...... 44

4. Results part II: Biochemical and functional analysis of Bicc1 interactome ...... 61

4.1. Bicc1 interactome screening ...... 61

4.2 Bicc1 co-purified ciliopathies related proteins ...... 66

4.3 Bicc1 co-purified the CTLH complex ...... 70

4.4 Discussion ...... 87

5. Results part III: Role of Bicc1 in cell metabolism ...... 95

5.1 Analysis of Bicc1 metabolic interactors and Bicc1-/- mouse model metabolome ...... 95

5.2 Additional metabolic alterations observed in Bicc1-/- mice ...... 114

5.3 Discussion ...... 125

6. Concluding remarks ...... 134

7. Materials and methods ...... 136

vii 8. Annex...... 149

References ...... 187

Curriculum Vitae ...... 223

viii List of Figures

Figure 1. Renal corpuscle ...... 2 Figure 2. The nephron ...... 4 Figure 3. Gross anatomy of the kidney ...... 5 Figure 4. Major substrates and enzymes participating in gluconeogenesis...... 10 Figure 5. Ketogenesis pathway ...... 12 Figure 6. Cilium fate during the cell cycle...... 14 Figure 7. Scheme of Polycystin 1 and Polycystin 2...... 23 Figure 8. Scheme of cyst development in an ADPKD kidney...... 25 Figure 9. Scheme of the intracellular signaling in collecting duct cells in ADPKD...... 26 Figure 10. Scheme showing localization of ciliary proteins in sub-ciliary compartments...... 32 Figure 11. MmBicc1 architecture...... 36 Figure 12. Bicc1 SAM domain can homopolymerize...... 37 Figure 13. Bicc1 expression in mouse kidneys...... 40 Figure 14. Doxycycline-inducible Bicc1-HA-StrepII expression and function...... 62 Figure 15. Bicc1 interacts with ANKS3 through its SAM domain independently of RNA. .... 67 Figure 16. Bicc1 interacts with NEK7 but its silencing activity is unaffected by NEK7 knockdown...... 69 Figure 17. Bicc1 lost increases the levels of Nek7 in mice kidneys...... 70 Figure 18. Bicc1 interaction with the mammalian CTLH complex subunits...... 72 Figure 19. Elevated protein levels of CTLH subunits in Bicc1-/- kidneys...... 73 Figure 20. Reduced glycemia and downregulation of FBP1 and PEPCK expression in Bicc1-/- kidneys...... 74 Figure 21. Endogenous Bicc1 in mIMCD3 cells increases FBP1 protein levels ...... 75 Figure 22. CRISPR-editing of Bicc1 in mIMCD3 cells...... 76 Figure 23. Bicc1 inactivation by CRISPR/Cas9 decreases FBP1 protein but no mRNA levels...... 76 Figure 24. Bicc1 regulate FBP1 levels independently of its interaction with the CTLH complex...... 77 Figure 25. Bicc1 is a target of the CTLH complex...... 79 Figure 26. Bicc1 SAM domain mediates binding with the CTLH complex...... 80 Figure 27. In vitro characterization of the novel Bicc1 self-polymerization mutant mut-I...... 81 Figure 28. Unpolymerized Bicc1 preferentially binds the CTLH subunits ...... 82 Figure 29. The CTLH complex co-fractionates with LMW Bicc1...... 83 Figure 30. Binding of CTLH subunits to each other in yeast two-hybrid assays...... 85

ix Figure 31. Bicc1 binds the CTLH complex via ARMc8 and independently of SAM polymerization...... 86 Figure 32. Regulatory interactions between CTLH complex and low molecular weight (LMW) Bicc1...... 89 Figure 33. “GO biological process” terms enriched for Bicc1 interactors...... 96 Figure 34. Bicc1 interacts in vitro with FASN, ACACA and CAD proteins...... 97 Figure 35. Bicc1 maintains protein levels of FASN, ACACA and CAD in the mice kidneys.. 98 Figure 36. Renal metabolomic analysis unambiguously distinguishes Bicc1-/- mice...... 100 Figure 38. Pathway analysis of all metabolites that were validated as de-regulated in Bicc1-/- kidneys...... 104 Figure 39. Purine biosynthesis ...... 105 Figure 40. Pyrimidine de novo synthesis...... 106 Figure 41. Arginine metabolism ...... 108 Figure 42. Tricarboxylic acid (TCA) cycle ...... 109 Figure 43. Glutamate and cystine transporters cooperate to regulate glutathione synthesis110 Figure 44. Synthesis of 3-hydroxybutyrate from fatty acids and ketogenic amino acids...... 111 Figure 45. Carnitine shuttle ...... 112 Figure 46. Bicc1 prevents lipid droplet accumulation in kidneys, but not in cultured mIMCD3 cells...... 115 Figure 47. Bicc1 reduces autophagy while increasing mTORC1 levels...... 117 Figure 48. Bicc1 reduces autophagic flux in mIMCD3 cells...... 119 Figure 49. Overview of the pathways assessed by Seahorse technology...... 120 Figure 50. Glycolysis and mitochondrial respiration profiles of sgBicc1 mIMCD3 cells...... 122 Figure 51. Plasma insulin and GLP-1 levels, ketone bodies, and urinary glucose concentrations in Bicc1-/- at postnatal day P2...... 124 Figure 52. Metabolic alterations in Bicc1-/- newborn mice found in this study ...... 125 Figure 53. Acidosis enhances glutaminolysis to restore acid-base balance...... 132

x Abbreviations EDTA: Ethylenediaminetetraacetic acid ALDOA: Aldolase, Fructose-Bisphosphate

A ESI-HESI: (Heated) Electrospray ionization

ALDOC: Aldolase,Fructose-Bisphosphate F2,6BP: Fructose 2,6-bisphosphate C FACS: Fluorescence-activated cell sorting ATP: Adenosine 5'- triphosphate FBP: Fructose 1,6-bisphosphate Bpk: Balb/c Polycystic Kidney FBP1: Fructose-1,6-bisphosphatase 1 BPGM: Bisphosphoglycerate Mutase FMRP: Fragile Mental Retardation Protein BSA: Bovine serum albumin G6PC: Glucose-6-Phosphatase Catalytic

Cas9: CRISPR associated protein 9 Subunit cDNA: complementary deoxyribonucleic GAL4 AD: Gal4 transcription factor’ acid activation domain

CFTR: Cystic fibrosis transmembrane GAL4 BD: Gal4 transcription factor’ binding conductance regulator domain

CoA: Coenzyme A GALM: Galactose Mutarotase

CREB: cAMP‐response element –binding GFP: Green fluorescent protein protein GLUT: Glucose transporter. SLC2A family CRISPR: Clustered regularly interspaced GTPase: Guanosine triphosphate short palindromic repeats hydrolase DAPI: 4',6'-diamidino-2-phenylindole GST: Glutathione S-transferase DBA: Dolichous Biflorus Agglutinin H&E: Hematoxylin and eosin DMSO: Dimethyl sulfoxide HEK: Human Embryonic Kidney DTT: Dithiothreitol Hh: Hedgehog

xi HILIC: hydrophilic interaction liquid OCR: Oxygen consumption rate chromatography OXPHOS: Oxidative phosphorylation HK1: Hexokinase 1 PCP: Planar cell polarity HRMS: High-resolution mass spectrometry qRT-PCR: Quantitative reverse transcriptase

IF: Immunofluorescence PCR

IFT: Intraflagellar transport RNP: Ribonucleoprotein

IGF1: Insulin-like growth factor 1 rtTA: Reverse tetracycline-controlled transactivator Jcpk: juvenile congenital polycystic kidney disease TCA: Citric acid (Krebs) cycle

LDHA: Lactate Dehydrogenase A UPLC: Ultraperformance liquid chromatography LTL: Lotus Tetragonolobus Lectin UTR: Untranslated region MDCK: Madine‐Darby canine kidney cells VPV2R: Vasopressin V2 receptor MEF: Mouse embryonic fibroblasts

WT: Wild-type mIMCD: Murine inner medullary collecting duct YAP: Yes-associated protein

MS/MS: Tandem mass spectrometry ZEB: Zinc finger E-box binding homeobox

family mTORC1: Mammalian target of rapamycin complex 1

NAD: Nicotinamide adenine dinucleotide

NADPH: Nicotinamide adenine dinucleotide phosphate

NPHP: Nephrocystin

O/N: Overnight

xii

1. Introduction

1.1 The Kidneys

The kidneys are retroperitoneal paired organs that function filtrating blood to excrete metabolic waste products. They also control electrolyte concentrations and the acid/base balance of the blood. Additionally, the kidneys play a pivotal role in blood pressure regulation and have hormonal functions. Perturbations of renal function by a variety of pathologies are life threatening conditions.

1.1.1 The nephron

The functional and fully developed unit of the kidney, the nephron, has two main components: the renal corpuscle and the renal tubule.

The renal corpuscle contains the capillaries of the glomerulus and Bowman’s capsule, and it is responsible for filtering the blood. The Bowman’s capsule is a cup- shaped double layer of simple squamous epithelium surrounding the glomerulus. This double layer is composed of podocytes, special epithelial cells that wrap around the endothelium of the capillaries to filtrate the blood while it is passing through the glomerulus. The podocytes have long foot projections called pedicels that leave slits in between, allowing blood to be filtered through them. Defects in the proteins required for the proper function of the podocytes, such as nephrin, podocalyxin or

CD2AP, allow the leak of large macromolecules like albumin or gamma globulins from the blood. Congenital defects in structural proteins of podocytes cause kidney failure and end-stage renal disease, e.g. the Finnish-type nephrosis, a congenital disorder caused by mutations in nephrin. In physiological conditions, small molecules including water, ionic salts and glucose can pass through the glomerular filtration barrier to form an ultrafiltrate. The ultrafiltrate is accumulated in the Bowman’s space (or capsular space) before flowing into the proximal convoluted tube, the beginning of the renal tubule (Fig. 1). Reviewed in (Greka and Mundel, 2012).

Figure 1. Renal corpuscle Schematic of a renal corpuscle. Main structural and functional components are depicted in high detail. Adapted from (McMahon, 2016).

The renal tubule is a complex epithelial structure intended to convert the glomerular filtrate into the urine that will be excreted.

The first curvy segment of the renal tubule is the proximal convoluted tube (PCT) composed of the first segment (S1), which starts at the glomeruli. Right after, the S2 segment continues until the first part of the proximal straight tubule. The S3 segment, the last segment of the PCT, reaches the medulla of the kidney. A well-developed brush border characterizes the apical membrane of the proximal tubule. It achieves an expansion of the surface area to execute its function: the vast reabsorption of water, ions (potassium, sodium, chloride, phosphate, and bicarbonate), and nutrients

(glucose and amino acids) initially filtered in the renal corpuscle back to the circulation. The basolateral membrane of the proximal tubule cells contains abundant mitochondria and interdigitations with very tight junctions. The high number of mitochondria are required to fulfill the energy demand of the cell to actively transport sodium ions. Proximal tubule cells also have a prominent Golgi apparatus important for the synthesis, sorting, and targeting of membrane components (such as transporters) to the surface of the plasma membrane (Curthoys and Moe, 2014).

Urine then continues to the thin descending and ascending loop of Henle. In this segment of the nephron, a concentration gradient is formed. The urine is concentrated in urea since water is highly reabsorbed due to the high permeability of the descending loop. The loop of Henle finishes with the thick ascending loop (TAL)

2 which is characterized by increased mitochondrial content. It contributes making the medullary interstitium hyperosmotic since it expresses Na-K-Cl cotransporters (NKCC2) (Mount, 2014).

The next segment of the tubule is the distal convoluted tube (DCT), where calcium is reabsorbed in response to the parathyroid hormone.

Finally, the urine coming from the distal convoluted tube of several nephrons enters the collecting system. The initial and cortical collecting tubules are composed of intercalated and principal cells. The intercalated cells are unusual due to the lack of primary cilium, but they can secrete protons or bicarbonate. Principal cells reabsorb

Na+ and secrete K+ (Roy et al., 2015). The medullary collecting duct continues the transport of electrolytes and the regulation of water and urea excretion. It is separated into the inner and the outer ducts, where the inner reach more deeply the medulla of the kidney (Fig. 2). The collecting ducts are characterized by a wide variation of water reabsorption, due to the hormonal influence. In case antidiuretic hormone (ADH) or vasopressin is present, aquaporins’ expression allows the reabsorption of water, inhibiting diuresis. On the other hand, in the absence of ADH, diuresis occurs.

Figure 2. The nephron Schematic showing a nephron. Plasma coming from the arterioles is filtered at the glomerulus and the ultrafiltrate continues all the way through the collecting system. Meanwhile, ions are excreted and absorbed, water and nutrients are retrieved. Concentrated urine arrives at the renal pelvis. The regions of the kidney, the cortex and outer and inner medulla are shown. Adapted from (Mullins et al., 2016).

1.1.2 Kidney anatomy

Detailed gross anatomy of the kidney is described in (Boron and Boulpaep, 2017). The kidneys are a pair of retroperitoneal bean-shaped organs protected by the ribs and the muscles of the back from external damage. They are covered by the renal capsule, a layer of fibrous tissue, which provides an outer shell to maintain the shape of the soft tissue inside. Some adipose tissue, the perirenal fat, acts as a protective

4 padding surrounding the kidneys. On top of each kidney is one adrenal gland, also surrounded by the perirenal fat.

At the concave surface of each kidney is the renal hilus. It serves as the entry of the renal artery and the exit of the renal vein and the ureter. From the renal artery originates the afferent arterioles in the kidneys, which after further branching give rise to a subpopulation of juxtamedullary glomeruli at the junction of the cortex and the medulla. The efferent arterioles of these glomeruli form a capillary network called the vasa recta. Inside, the hilus opens into the space called the renal sinus. It includes the extensions of the renal pelvis, the major and the minor calyces. The sinus also contains blood vessels and nerves. The lymph vessels drain the interstitial fluid from the cortex, which contains renal hormones. The lymph vessels leave the kidneys following the renal arteries exiting through the hilus (Fig. 3).

Figure 3. Gross anatomy of the kidney Scheme showing the internal architecture of the kidney in sectional view. The blue square represents a zoomed region containing details of the renal vasculature. Adapted from Boron and Boulpaep, Medical Physiology 3rd edition, 2017 (Boron and Boulpaep, 2017).

Internally, there are two basic regions in the kidney, the cortex, and the medulla. The presence of glomeruli and the convoluted tubular structures confer the cortex a

5 granular appearance region. The medulla a arrangement of tubules without glomeruli, is subdivided in renal pyramids, containing microscopic perforations at the tips to allow the urine to flow into the minor calyces. The minor calyces merge into the major calyces, and then into the renal pelvis at the center of the kidney. Finally at the renal hilus, the urine drains into the ureter.

1.1.3 Kidney functions

A) Filtration and reabsorption

Blood pressure pushes most of the blood plasma to go through the capillaries and the glomerular space. Epithelial cells of the renal tubules reabsorb important metabolites, such as glucose, amino acids, and ions, back into circulation. Osmotic pressure pushes water from the hypotonic filtrate back into the hypertonic blood. After the filtrate has passed across most of the nephron, it reaches the collecting duct where nearly all nutrients, ions and water have returned into circulation. Waste products and a small amount of water are left to form urine. Daily the kidney filters approximately 180 liters of plasma, and thus around 180 gr of glucose is filtered during the same period. In a physiological situation, almost all the glucose is reabsorbed into the circulation, leaving the urine free of glucose. Glucose reabsorption occurs at the proximal convoluted tubules by the sodium-glucose cotransporters (SGLTs), approximately 90% of the reabsorption is carried out by

SGLT2 a high capacity low-affinity glucose transporter. The remaining reabsorption is accomplished by SGLT1 located in the S3 segment of the proximal tubule (Quamme and Freeman, 1987). This active transportation requires energy derived from the electrochemical potential gradient by the transport of intracellular sodium into the circulation, a process mediated by the Na+K+-ATPase pumps. Glucose release is achieved by GLUTs (GLUT1/2), facilitative passive glucose transporters, located at the basolateral membrane of the proximal tubules. Reviewed in (Mather and Pollock, 2011).

B) Excretion of waste

One key function of the kidney is the excretion of metabolic waste products. Protein metabolism in the liver produces ammonia. The liver converts most of it into uric acid and urea, reducing toxicity for the body. Likewise, muscles produce creatinine as a product waste from creatine metabolism. Thus, creatinine, urea, and uric acid need to be cleared from the bloodstream to preserve body homeostasis. Urea is vastly reabsorbed in the inner medullary collecting duct (IMCD) lumen into the interstitium, due to the expression of the urea transporter UT-A1 at the apical surface. Similarly, UT-A2 is expressed in the thin descending loop of Henle also contributing to urea reabsorption (Klein et al., 2011).

C) Acid/base homeostasis

Maintaining systemic acid-base homeostasis is imperative for proper physiology. To achieve this, lungs and kidneys play important roles. The kidneys, control the blood pH by regulating the excretion of hydrogen ions (H+) and bicarbonate ions (HCO3-) in the urine. Ions are filtered in the glomerulus, and tubule cells selectively reabsorb

HCO3- while leaving H+ to be excreted. Once in the bloodstream, the HCO3- can neutralize H+ by forming carbonic acid, which can pass through the capillaries of the lungs and dissociate into carbon dioxide (CO2) (and water) and finally be exhaled

(Hamm et al., 2015). The proximal tubule can reabsorb around 80% of the filtered

HCO3- , in addition the the kidneys can produce new HCO3+ to compensate that consumed by acids (normal or pathologic). This production is achieved by proton secretion at the apical membrane by the NHE3, a sodium-hydrogen exchanger (Brant et al., 1995). Furthermore, other nephron segments contribute to acid-base homeostasis. An apical Na+/H+ exchanger at the TAL reabsorbs nearly 15% of the filtered load of HCO3-. In the distal nephron, regulation of acid-base transport is imperative because here is the final segment where urine composition is controlled.

Intercalated cells in the collecting duct segment of the nephron can increase H+ secretion by insertion of additional H+ATPase in the plasma membrane (Roy et al., 2015).

D) Hormone production

The kidneys also play an important endocrine role. The juxtaglomerular cells (JG cells) are smooth muscle cells near the afferent arterioles of the glomeruli, which synthesize renin in response to blood pressure decrease. The macula densa are cells localized at the DCT that stimulate JG cells when they detect low sodium concentration in the glomerular filtrate (Hackenthal et al., 1990).

Additionally, the proximal tubule cells can convert 25-hydroxyvitamin D (25D) into 1,25-dihydroxyvitamin D (circulating calcitriol). Calcitriol is a potent steroid that can bind to vitamin D receptors to activate pathways leading to expression of genes involved in mineral homeostasis (Ca+ and phosphorus metabolism) as well as in skeletal function (Dusso, 2011).

Moreover, peritubular fibroblasts from the renal cortex are the main source of adult erythropoietin (EPO) in response to hypoxia. EPO stimulates bone marrow generation of red blood cells and the transportation of hemoglobin (Jelkmann, 2011).

E) Blood pressure homeostasis

Kidney role in blood pressure balance depends on the control of the extracellular fluid compartment (blood plasma, lymph, and interstitial fluid), which depends on the plasma sodium concentration. The kidney is involved in the initiation of the renin- angiotensin-aldosterone system since they can produce renin and ultimately, regulate the outcome of the system, the production of angiotensin II and aldosterone.

Renin cleaves the circulating angiotensinogen to produce angiotensin I which later is hydrolyzed into angiotensin II at the endothelium. Angiotensin II stimulates the production of aldosterone by the adrenal gland; this promotes sodium reabsorption coupled with potassium exchange in the in the connecting tubule and collecting duct through the activation of the amiloride-sensitive sodium channel that is coupled to potassium secretion via apical ROMK (K+ channels) and consequently increases the extracellular fluid compartment and thus blood pressure. If the levels of renin are low, angiotensin II and aldosterone also decrease diminishing the extracellular fluid

8 compartment and decreasing blood pressure (Sparks et al., 2014). ADH promotes the expression of water channels in the collecting ducts.

F) Renal gluconeogenesis

Glucose regulation is critical for the survival of all kind of organisms. Mammalian glucose regulation allows to avoid two extreme situations that can produce systemic damage. On one hand, hyperglycemia may promote chronic defects such as neuropathy, retinopathy, nephropathy, and premature atherosclerosis (The Diabetes Control and Complications Trial/Epidemiology of Diabetes Interventions, 2005). On the other hand, the body has to defend itself from hypoglycemia which may cause neurological dysfunction, coma, seizures, cardiac arrhythmias, or death (Morales and

Schneider, 2014).

The involvement of the kidneys in glucose homeostasis was reported since the 1930s (Bergman and Drury, 1938). Nowadays it is known that the kidneys can regulate glucose homeostasis by at least in three processes: The reabsorption of glucose from glomerular filtrate, the uptake of glucose from the bloodstream to feed kidney’s own energy demands, and the release of glucose into the circulation by gluconeogenesis.

During gluconeogenesis, non-carbohydrate carbon sources, such as lactate, glycerol and gluconeogenic amino acids (such as alanine), are converted into glucose in a series of reactions in the mitochondria and the cytoplasm. Gluconeogenic substrates include gluconeogenic amino acids and lactate that can be converted to pyruvate.

Pyruvate is carboxylated by Pyruvate carboxylase to form oxaloacetate which is reduced to malate to allow its transportation outside the mitochondria. Once in the cytoplasm, malate is oxidized back to oxaloacetate and decarboxylated and phosphorylated by Phosphoenolpyruvate carboxykinase to produce phosphoenolpyruvate. The further reactions occur in the reverse way of glycolysis: fructose 1,6 biphosphate is converted to fructose 6-phosphate by Fructose 1,6- biphosphatase, the rate-limiting step of gluconeogenesis. Fructose 6-phosphate is converted to glucose 6-phosphate and in the endoplasmic reticulum is hydrolyzed by

Glucose 6-phosphatase to produce glucose and inorganic phosphate (Hatting et al., 2017) (Fig.4).

Figure 4. Major substrates and enzymes participating in gluconeogenesis. Red arrows and text depicting major enzymes and pathways involved in gluconeogenesis. AAT, alanine aminotransferase; fruc-1,6-bisphosphatase, fructose-1,6-bisphosphatase; glu-6-phosphatase, glucose-6-phosphatase; glyceraldehyde-3-P, glyceraldehyde-3- phosphate; glycerol-3-P, glycerol-3-phosphate; LDH, lactate dehydrogenase; OAA, oxaloacetate; PC, pyruvate carboxylase; PDH, pyruvate dehydrogenase. Taken from (Chung et al., 2015).

The rate-limiting enzymes for gluconeogenesis and glucose release, Phosphoenolpyruvate carboxykinase (PEPCK), Fructose 1,6- biphosphatase (FBP1), and Glucose-6-phosphatase (G6P) are expressed in the renal cortex. During gluconeogenesis, both the liver and the kidneys use lactate as gluconeogenic

10 substrate. However, besides lactate, kidneys prefer glutamine versus alanine as an amino acid substrate (Stumvoll et al., 1998).

In mammals, during postabsorptive states (considered 12-16 hrs after the last meal), the kidney contributes up to 25% to glucose release by gluconeogenesis, the remaining 75% of glucose release is coming from the liver, the classical organ responsible for gluconeogenesis. However, during post-prandial states (4-6 hrs after three main meals of the day), the kidneys increase their gluconeogenic capacity, up to approximately 60% of the glucose released into circulation (Meyer et al., 2002) facilitating the replenishment of glycogen by hepatic glycogenesis. Since renal gluconeogenesis is adapted to different stimuli like fasting, diabetes or hypoglycemia, its regulation is influenced by hormones including insulin and catecholamines (reviewed in (Gerich, 2010).

G) Renal ketogenesis

In response to fasting states, ketogenesis also takes place. The production of ketone bodies, mainly acetoacetate and β-hydroxybutyrate (3HB) and in a lesser extent acetone from the oxidation of fatty acids or ketogenic amino acids supply energy especially to the brain, heart, kidneys, and muscles. Interestingly, the kidney is also recognized as a ketogenic organ, it expresses the rate-limiting enzyme 3-hydroxy-3- methyl glutaryl-CoA (HMG-CoA) synthase 2 (HMGCS2), which expression is boosted in diabetic conditions (Zhang et al., 2011). HMGCS2 catalyzes the condensation of Acetyl-CoA with Acetoacetyl-CoA to form 3-hydroxy-3methylglutaryl-CoA (HMG-

CoA). HMG-CoA is an intermediate substrate for the generation of acetoacetate, the precursor of acetone and 3HB (Fukao et al., 2014) (Fig. 5).

Figure 5. Ketogenesis pathway Scheme representing the synthesis pathway of 3HB from fatty acids and ketogenic amino acids. In the mitochondria, Acyl-CoA is converted to Ac-CoA via β-oxidation. Ac-CoA is further metabolized into ketone bodies (acetoacetate, 3HB, and acetone). 3HB is the most stable ketone body in the bloodstream. FFA: fatty acids, ILE: isoleucine, LEU: leucine, LYS: lysine, THR: threonine, TRP: tryptophan, TYR: tyrosine, 3HB: β-hydroxybutyrate, 3HBDH: 3HB dehydrogenase, Ac-CoA: acetyl-CoA, AcetoAc-CoA: acetoacetyl-CoA, HMG-CoA: 3- hydroxy-3-methyl glutaryl-CoA, HMGCL: HMG-CoA lyase, HMGCS: HMG-CoA synthase. Taken from (Miyazaki et al., 2015).

1.2 Primary cilia and their role in the kidney

Cilia are highly conserved, complex and dynamic organelles found in almost all eukaryotic cells. They are broadly divided into motile or non-motile (primary) cilia depending on their structure. Motile cilia are present on specialized epithelia such as the pharynx or the oviduct, while primary cilia are monocilia on the cell surface of nearly all cells in the human body. In the nasal epithelium and the rod and cone cells of the retina, primary cilia are essential for odorant perception and sight, respectively. Interestingly, cilia can play critical roles in cell signaling.

1.2.1 Structure of the primary cilium

Protruding from the apical surface of almost all differentiated cells, primary cilia are highly conserved antenna-like organelles, which assemble and disassemble when cells exit and enter each mitotic cycle, respectively (Pan and Snell, 2007). The cilium originates from a basal body which corresponds to a centriole relocated at the plasma membrane when the cell exits the mitotic cycle. From each of the nine microtubule triplets forming the basal body, two microtubules extend in the protruding core of the cilium called the axoneme. The basal body is separated from the axoneme by the transition zone and the Inversin compartment, which act as gatekeepers at the ciliary compartment border (Omran, 2010). The cilium contains nine doublet microtubules but lacks the central pair of microtubules (9+0 structure) in the axoneme. The absence of elements involved in ciliary motility, the inner and outer dynein arms that power microtubule sliding, make the primary cilium a non-motile structure (Fig. 6A, B). Reviewed in (Satir et al., 2010).

A B

Figure 6. Cilium fate during the cell cycle. (A) Scheme of events during cell cycle progression: (a-f) the pair of centrioles can form centrosomes by organizing pericentriolar material (PMC). Closer to mitosis the centrioles separate from each other until they nucleate and organize microtubules which then organize the mitotic spindle. In cells that exited cell cycle (g), the pair of centrioles migrate to the cell surface, and from the basal body originated by the mother centriole protrudes the cilium. The figure is taken from (Conduit et al., 2015). (B) The detailed scheme of the square in (A). The basal body, triplet microtubule structure give rise to the axonemal doublet structure at the transition zone of a primary cilium. The outer microtubule is connected to the membrane by the proximal transition γ-shaped fibers (red) where the IFT proteins shuffle cargoes to and from the ciliary compartment. NPHP proteins localize at the transition zone likely as gatekeepers controlling access and exit of proteins from the ciliary compartment. Figure is taken from (Omran, 2010).

Ciliogenesis is a very dynamic process. Several events happen in the centriole to basal body transition (Jensen and Leroux, 2017). Cytoplasmic vesicles, called distal appendage vesicles (DAVs) start to accumulate near the distal appendages of the mother centriole where they dock. The ciliary vesicle, a membranous cap, on the tip of the mother centriole is produced by vesicular fusion (Sorokin, 1962). The centriole´s microtubule extension starts under this cap, while vesicular trafficking enlarges the cap allowing microtubules axoneme to grow inside a specialized membrane. By fusion with the ciliary sheath, this incipient cilium docks to the plasma membrane and thus, elongation continues through the axoneme protrusion. Reviewed in (Sánchez and Dynlacht, 2016).

Cilia and flagella rely on intraflagellar transport (IFT) to grow. Membrane proteins are targeted to the ciliary transition zone, where proteins enter the cilium before being

14 mobilized by the IFT particles machinery. For anterograde traffic, association with kinesin-2 proteins moves cargoes from the basal body to the tip of the cilium. In turn, during retrograde traffic, association with dynein-2 motor protein returns proteins from the tip to the base of the cilium (reviewed in (Rosenbaum and Witman, 2002)). Concomitant with the IFT system, the BBSome is a protein complex that coordinates the localization of transmembrane proteins in and out of the cilium. The BBSome organize the IFT related proteins and kinesin motors into a functional complex at the ciliary base to start anterograde traffic, which also must be remodeled after reaching the ciliary tip to continue with the retrograde transport. Cilia disassembly occurs after mitogen stimulation when cells leave G0 phase. HEF1 (NEDD9), Calmodulin and Aurora A kinase play roles by phosphorylating the histone deacetylase HDAC6, which in turn deacetylates axonemal microtubules to destabilize them (Plotnikova et al., 2012; Pugacheva et al., 2007). Additionally, deacetylation of cortactin and α- tubulin is needed for HDAC6 to enhance cilia disassembly (Ran et al., 2015).

1.2.2 The Ciliary Proteins

Around 1000 proteins have been identified as ciliary proteins (Gherman et al., 2006). Perhaps the most studied ciliary proteins belong to the IFT system. Mutations in IFT genes cause a range of phenotypes, including visceral situs inversions or heterotaxia, retinal degeneration, body curvature and/or cystic kidneys. Multiple assembled IFT proteins (IFT20, IFT88, and IFT57) are also present in T lymphocytes, despite the fact that these cells do not have primary cilia (Finetti et al., 2009), pointing to potential cilium-independent additional functions. Another class of transport proteins such as KIF3A and KIF3B participate in ciliogenesis as components of kinesin-2 anterograde motor of the IFT mechanism. In addition, depletion of KIF3A cause constitutive phosphorylation of Dishevelled (Dvl) by casein kinase I (CKI), suggesting KIF3A has a role in restraining canonical Wnt signaling (Corbit et al., 2008).

Szymanska and Johnson summarized the localization of ciliary proteins described by different studies (Szymanska and Johnson, 2012). highlighted protein complexes include the “nephronophthisis NPHP” module composed of NPHP1/nephrocystin-1,

NPHP4 and RPGRIP1L (which may play a role in epithelial morphogenesis and architecture), the “JBTS module” containing the ciliogenic proteins IQCB1 and CEP290, and the MKS module containing MKS1, CC2D2A and TCTN2 that are involved in neural tubule development in Sonic Hedgehog signaling. Also highlighted, the Tectonic module (TCTN1, TCNTC2, and TCTN3) and MKS proteins described as having a role in ciliary membrane composition depending on the tissue (Garcia- Gonzalo et al., 2011). Furthermore, a functional “MKS/MKSR module” containing MKS1, TMEM67, CC2D2A, B9D1, and B9D2 contributes to early ciliogenesis and when defective, cause Meckel-Gruber syndrome (MKS) (Williams et al., 2011). The module containing TMEM237/JBTS14 was identified as a cause of Joubert syndrome-related disorders. TMEM237 needs RPGRIP1L protein for proper localization to the cilium, suggesting a role for RPGRIP1L as a scaffolding or bridging protein at primary cilia (Huang et al., 2011). An additional ciliary segment, distal to the transition zone was known as the “Inversin compartment” contains a complex of INVS/Inversin, NEK8 (Never In Mitosis A-Related Kinase 8) and NPHP3 (Shiba et al., 2010) (Fig. 10). There are some inconsistencies among the studies reflecting perhaps differences in vitro or in vivo models or differences in the methods used to determine the composition of the compartments and modules, or the the dynamics of cilia assembly and protein transport.

Several proteins and protein complexes at the ciliary membrane mediate pivotal roles in different cell signaling pathways. For instance, the components of the Hedgehog signaling pathway or the Polycystins in Ca2+-cAMP signaling. Ciliary signaling pathways involved in renal homeostasis are described below.

1.2.3 Primary cilia-mediated signaling pathways in the kidney

A key function of primary cilia is to concentrate several receptors, ion channels and kinase proteins that coordinate chemo-, photo- and mechanosensation, allowing the cells to respond to and integrate specific environmental cues. Thus, they participate in vital developmental and homeostatic processes, including, among others, cell

16 differentiation, migration, re-entry into the cell cycle, planar cell polarity, and apoptosis (Kathem et al., 2014).

Ca2+/cAMP

At the apical surface of nephron epithelia, primary cilia are thought to sense the urine flow in the lumen of the tubules as shear stress. The primary cilium of cells receiving such mechanical stimulus can bend in response to fluid flow (Schwartz et al., 1997) and generate an intracellular calcium response (Praetorius and Spring, 2001), suggesting that the calcium rise may be mediated by mechanosensitive channels at the primary cilium. Such a role has been attributed to the ciliary protein Polycystin-1

(PC-1) encoded by the Pkd1 gene, because despite normal cilia length the

Pkd1del34/del34 mutant mouse renal tubule epithelial cells failed to induce Ca2+ influx (Nauli et al., 2003). Moreover, blockage of polycystin-2 protein (PC-2), using specific neutralizing antibodies, similarly inhibited flow-induced influx of extracellular Ca2+

(Nauli et al., 2003).

In a complex composed by A-kinase anchoring protein 150 (AKAP150), protein kinase A (PKA), adenylyl cyclases 5 and 6 (AC5/6), and phosphodiesterase 4C

(PDE4C), the C-terminus of PC-2 plays a role by interacting with the scaffold protein

AKAP150. Calcium-sensitive AC5/6 synthesize cyclic adenosine monophosphate (cAMP) from ATP and PDE4C catabolizes it (Choi et al., 2011). cAMP-dependent

PKA phosphorylates different targets involved in cellular metabolism, proliferation, chloride and fluid secretion or transcription of chemokines and cytokines (Harris and

Torres, 2014; Tengholm and Gylfe, 2017). Since impaired calcium and cAMP signaling affects multiple molecular pathways leading to cystic phenotypes, these mechanisms and the polycystin proteins will be described further in section 1.3. mTORC1 and LKB1

Mammalian target of rapamycin complex 1 (mTORC1) is a kinase that promotes transcription, protein synthesis, cell growth, and metabolism. The mTORC1 complex is sensitive to nutrient signals such as the availability of amino acids, growth factors or AMP-activated kinase (AMPK), among others (Zoncu et al., 2011). mTORC1

17 phosphorylation targets include the p70 S6 kinase and 4E-BP1. Upon phosphorylation, activation or inhibition of these targets respectively, stimulate protein synthesis and promote cell growth.

Fluid flow activates the ciliary liver-kinase B1 (LKB1) which in turn phosphorylates AMPK, leading to AMPK accumulation at the ciliary base and inhibition of mTORC1 (Boehlke et al., 2010). This suppression is mediated by the phosphorylation and activation of a negative regulator of mTORC1, the tuberous sclerosis complex (TSC) 1 and 2, which in turn stimulates the GTPase activity of Rheb to prevent it from activating mTORC1 (Wiczer et al., 2010). The mTORC1 complex is the main negative regulator of autophagy. mTORC1 inhibits autophagy initiation by phosphorylating ULK1/2. When ULK1/2 is phosphorylated at Ser758 by mTORC1, it cannot be activated by AMPK (Fantus et al., 2016).

Hippo signaling

The Hippo signaling pathway comprises a cascade of several serine/threonine- kinases involved in organ size and cell proliferation regulation. In mammals, the kinase module MST1/2 (orthologs of Drosophila Hippo) and the activating adaptor protein SAV1 interact with nephrocystin proteins (NPHPs) at the transition zone of the cilia (Habbig et al., 2011). MST1/2 and SAV1 form a complex to phosphorylate and activate LATS1/2, which in turn phosphorylates the transcription coactivators

YAP and TAZ leading to their sequestration in the cytoplasm. By contrast unphosphorylated YAP/TAZ translocate to the nucleus to promote transcription of proliferation genes by TEAD transcription factors. NPHP4 binding to LATS1 inhibits LATS-1-mediated phosphorylation of YAP/TAZ, allowing YAP/TAZ nuclear translocation (Habbig et al., 2011). In absence of NPHP4, Hippo signaling is derepressed thus increasing proliferation.

Interaction of NPHPs with MST1 and MST2 promotes ciliogenesis by promoting the accumulation of ciliary cargoes at the transition zone. Furthermore MST1/2 binding to AURKA, allows its dissociation from the AURKA/HDAC6 cilia-disassembly

18 complex (Kim et al., 2014). Accordingly, depletion of MST1/2 or SAV1 induces ciliopathy phenotypes in zebrafish.

Wnt pathways

Wnt ligands are secreted glycoproteins that ultimately diverse cellular processes, including cell proliferation, differentiation, survival, migration and planar cell polarity (Amerongen and Nusse, 2009; Massinen et al., 2011; Wallingford and Mitchell, 2011). Despite the overlap between receptors and effectors, Wnt signal transduction is classified as in Wnt/β-catenin (canonical) or Wnt/PCP and Wnt/Ca2+ (non- canonical) pathways. The canonical Wnt/β-catenin pathway induces the stabilization of cytoplasmic β-catenin, allowing it to translocate to the nucleus to regulate gene expression. Wnt proteins bind to membrane receptors of the Frizzled (Fdz) family together with the LDL receptor-related proteins 5 and 6 (LRP5/LRP6), which in turn recruits the cytoplasmic protein Dishevelled (Dvl). Consequently, this binding will disrupt the destruction complex (composed by APC, GSK-3β, and AXIN) that targets β-catenin for degradation. Once β-catenin is accumulated in the cytoplasm, it translocates to the nucleus to induce T-cell factor/lymphoid enhancing factor

(TCF/LEF) transcription factor and target gene expression of Axin2, Lef1, or c-Myc, promoting cell proliferation (reviewed by (MacDonald et al., 2009).

Non-canonical Wnt pathways involve the binding of non-canonical Wnt ligands to Fzd receptors, but without the participation of the LRP5/6 as co-receptor and independently of β-catenin. The Wnt/PCP pathway activates GTPases, RhoA cascade and c-Jun N-terminal kinase, which target genes that control establishment of planar cell polarity, the alignment of cell within the epithelial plane. On the other hand, the Ca2+/Wnt pathway activates three effectors: Ca2+ flux, Protein kinase C (PKC) and calcium/calmodulin-dependent protein kinase II (CamKII). Wnt/PCP and Wnt/Ca2+ signaling regulate convergent extension movements (Kühl et al., 2001).

Mutations in the ciliary proteins Kif3a, Ift88, or Ofd1, increase the activation of canonical Wnt pathway (Corbit et al., 2008). The balance of both Wnt signaling

19 pathways plays a role in kidney cyst formation in renal ciliopathies (Benzing et al., 2007).

1.3 Ciliopathies

Pathologies arising from ciliary structural or functional defects are named ciliopathies. They involve a broad spectrum of diseases affecting the normal development of vital organs. There are several causative genetic mutations leading to multi-systemic disorders, which can reach a prevalence as high as 1:2000 people. Importantly, the kidneys are commonly affected by these syndromes.

1.3.1 General characteristics

Ciliopathies are diseases caused by defects in motile or immotile cilia. In this chapter, only the ciliopathies arising from defects in immotile cilia will be described. They share several common characteristics like a renal cystic disease, neural tubule defects, intellectual disability, blindness, skeletal and muscular abnormalities, situs inversus, infertility and respiratory defects. Ciliopathies exhibit several signs and symptoms comprised several syndromes. Reviewed in (Oud et al., 2016).

For this thesis, this chapter will focus in more detail on Polycystic kidney diseases.

Polycystic kidney diseases (PKD) are genetic diseases characterized by the progressive development of fluid-filled cysts from the nephrons and which ultimately will impair renal functions. Kidney dysfunction is the main cause of morbidity in the two main forms of the disease, autosomal dominant and autosomal recessive PKD. In general, the impact of these hereditary renal diseases is very high since they represent a big fraction of adults and virtually all children patients requiring renal replacement therapy. Recently reviewed by (Mehta and Jim, 2017)).

1.3.2 Autosomal dominant polycystic kidney disease (ADPKD)

ADPKD is a frequent and the most common monogenic disorder affecting the kidneys. It affects all ethnic groups and it is recognized as a major cause of chronic

20 kidney disease accounting for approximately 10% of all end-stage renal disease (ESRD) cases with an estimated prevalence of 1:400 to 1:1000 births. ADPKD is the fourth most common cause of renal replacement therapy worldwide. The epidemiological aspects of ADPKD are recently reviewed by (Ong et al., 2015).

Two genes are the cause of ADPKD. Loss-of-function mutations in PKD1 (16p13.3) are found in 80-85% of the patients, while PKD2 (4q22.1) mutations account for 15- 20%. To date, 2323 and 278 mutations in PKD1 and PKD2 respectively, are annotated in the Autosomal dominant polycystic kidney disease Mutation database (www.pkdb.mayo.edu). Patients with PKD2 mutations have approximately 20 years longer survival than those with PKD1 mutations. Regarding the onset of ESRD in patients with mutated PKD1, it occurs at the median age of 58 years while in PKD2 patients it may be happening by the average age of 79 years. Additionally, truncating PKD1 mutations accelerate ESRD onset to the age of 55.6 years instead of 67.9 years, when compared to patients with non-truncating mutations (Cornec-Le Gall et al., 2013).

Renal fluid-filled cysts are the hallmark of ADPKD. However, extra-renal manifestations frequently accompanied disease progression, including liver and pancreatic cysts, cardiovascular defects and cerebral aneurysms (Braun, 2014). Patients frequently present hypertension and a significant decline in estimated glomerular filtration rate (eGFR), approximately 10-15 years before ESRD diagnosis.

Before the decline in eGFR, total kidney volume measured by magnetic resonance is employed to assess disease severity and as a parameter to estimate disease progression (Chapman et al., 2012).

The polycystins

Polycystin-1 (PC-1, encoded by PKD1) is a large transmembrane glycoprotein of about 460 kDa, containing a very large extracellular domain (approx. 3000 a.a.), 11 transmembrane domains and a C-terminal domain of approx. 200 a.a (Boletta et al., 2001). PC-1 comprises many domains involved in protein-protein and protein- carbohydrate interactions, including 16 PKD repeats likely having a role as ligand‐

21 binding sites in cell‐surface proteins (Bycroft et al., 1999; Consortium and others, 1995). PC-1 contains a G-protein coupled receptor proteolytic site (GPS) that produces an N-terminal fragment (the extracellular part of the protein) and the C- terminal fragment (CTF) comprising the remainder of the protein (Qian et al., 2002; Wei et al., 2007). Point mutations in the GPS site associate with polycystic kidneys, demonstrating the functionality of this cleavage (Yu et al., 2007). Additionally, the coiled-coil domain at the C-terminus can be cleaved at two different sites. The first generates the entire intracellular C-tail (CTT) which has been observed to interact with β-catenin in the nucleus and inhibits its ability to activate the transcription factor TCF (Lal et al., 2008). The second (p112) interacts and stimulates STAT6 transcriptional activity (Low et al., 2006).

The C-terminal coiled-coil domain also mediates binding of PC-1 to PC-2 (Qian et al.,

1997). Additionally PC-1 can colocalize and coimmunoprecipitate with actin-binding proteins, E-cadherin, and α- β-, and γ-catenin, suggesting that it connects extracellular matrix and actin cytoskeleton (Huan and Adelsberg, 1999). The C- terminal tail of PC-1 can also bind Na+, K+-ATPase, suggesting that PC-1 can indirectly regulate renal tubular fluid and electrolyte transport by increasing Na+/K+- ATPase activity (Zatti et al., 2005).

In adult human tissues, PC-1 is expressed in kidneys, in the distal nephron and collecting duct, as well as in biliary tree and pancretic ducts and many other organs including some which are not affected during ADPKD progression (Peters et al.,

1999). PC-1 normally localizes at the focal adhesion of the plasma membrane, desmosomes, tight junctions and in the primary cilia (PD et al., 1999; Ward et al., 2011). In ADPKD cyst-lining kidney epithelium PC-1 is largely limited to the cytoplasm (Palsson et al., 1996).

Polycystin-2 (PC-2, encoded by PKD2) is a 110 kDa protein, with 6 transmembrane domains and cytoplasmic N-terminal and C-terminal domains. In the mature kidney, PC-2 is expressed along the whole nephron, with the highest expression in the

22 medullary thick ascending loop of Henle and distal tubules (Markowitz et al., 1999). PC-2 has been shown to localize at the apical and basolateral membrane, endoplasmic reticulum, Golgi and primary cilia (Li et al., 2006; Yoder et al., 2002).

PC-2 belongs to the family of transient receptor potential (TRP) family (González- Perrett et al., 2001) was identified as a non-selective (high-permeable) Ca2+ channel. PC-2 plentifully to compartments that regulate calcium releases from intracellular stores (Koulen et al., 2002). PC-2, also known as TRPP2 interacts with TRPC1 and TRPV4, two other members of TRP family (Semmo et al., 2014).

At the plasma membrane and in cilia, PC-2 also form a complex with PC-1. It was demonstrated that the correct localization and the ion channel activity of PC-2, as well as the correct PC-1 maturation, depend on this interaction (Gainullin et al., 2015;

Hanaoka et al., 2000) (Fig. 7).

Figure 7. Scheme of Polycystin 1 and Polycystin 2. Scheme represents the different cleavage products of full-length PC-1 that are co- expressed in the cell and are associated with different functions. PC-2 is also depicted. Taken from (Boletta, 2009).

Molecular Mechanisms of ADPKD

Cyst generation in human ADPKD is sustained according to the “two-hit hypothesis” (Qian et al., 1996; Watnick et al., 1998). ADPKD patients are heterozygous, with one

PKD allele showing a germline mutation (first hit). Loss of the remaining normal PKD allele by a somatic mutation (second hit) in a very small percentage of cells (Baert, 1978) triggers cyst development. This model is supported by the fact that homozygous mice with full inactivation of mutations of Pkd1 or Pkd2 alleles are embryonically lethal and that renal cysts have a clonal origin (Lu et al., 1997; Qian et al., 1996). Additionally, the presence of hypomorphic alleles in compound heterozygous ADPKD patients, have an impact on the critical threshold where PC-1 or PC-2 are functional and indicates that gene dosage influences cyst development and severity of PKD (Hopp et al., 2012; Rossetti et al., 2009). Renal injury such as renal ischemia-reperfusion injury may contribute to reach this threshold, acting as a third hit contributing to accelerate development of ADPKD (Takakura et al., 2009).

In keeping with this idea, cyst growth in human ADPKD and in mouse models cyst depends on the stage of development. During developmental stages (newborns), cyst growth is accelerated compared to adult kidneys (Grantham et al., 2010; Piontek et al., 2007). Early onset of PKD, is also facilitated by additional mutations in other genes causing cystic phenotypes, such as HNF1B (Hepatocyte Nuclear Factor 1-

Beta) which is involved in renal cyst and diabetic syndrome of MODY-1 patients, or mutations in PKDHD1 the major causative gene of ARPKD (Bergmann et al., 2011).

Cyst development in PKD is attributed to abnormal fluid secretion and cell proliferation mechanisms (Harris and Torres, 2014). Cyst-lining renal epithelial cells lose their absorptive function as they increase rates of proliferation and apoptosis (Woo, 1995). Expanding cysts will eventually compress and impair the functioning of residual normal nephrons resulting in severe organ damage and inflammation mediated by macrophages that have been shown to promote cyst formation (Karihaloo et al., 2011; Swenson-Fields et al., 2013) (Fig. 8). In addition, planar cell polarity (PCP) regulate oriented cell division and convergent extension which are important processes in renal tubular morphology (Goggolidou, 2014). Defects in PCP are hypothesized to accelerate cyst formation (Germino, 2005). Many proteins that can produce renal cystic disease have roles in Wnt signaling. These include PC-1, NPHP2 (INVS) and NPHP3 which can downregulate canonical Wnt signaling or the

BBS proteins that can also regulate this pathway. Moreover, hyperactive canonical Wnt signaling gives rise to renal cyst (Simons et al., 2005) and mutations in Wnt4 have been detected in patients with renal dysplasia (Vivante et al., 2013). Altogether, these studies highlight the importance of a fine-tuned Wnt pathways activation in the kidneys to prevent cystogenesis. Reviewed in (Goggolidou, 2014).

Figure 8. Scheme of cyst development in an ADPKD kidney. ADPKD kidneys contain fluid-filled cyst derived from few number of nephrons. In the affected renal tubules, different pathophysiological mechanisms contribute to cyst enlargement. Cysts originate from single cells carrying a germline mutation in one of the alleles for the Pkd1 or Pkd2 genes. Every cell in the tubule carries a mutated allele, and a somatic mutation or haploinsufficiency of the second allele is believed to trigger the disease. Cysts continue growing even after detaching from the renal tubule that originated them. Figure taken from (Blanco and Wallace, 2013).

Downstream of ciliary defects, several signaling pathways promotes ADPKD pathogenesis. However, due to redundancies, feedback loops, and reinforcements signals they should be considered as network components instead of individual axes.

Figure 9. Scheme of the intracellular signaling in collecting duct cells in ADPKD. Taken from (Gastel and Torres, 2017).

Ciliary PC-1 has been proposed as a flow sensor, which in complex with PC-2, is activated by mechanical stimulation of cilia thus inducing calcium release from intracellular stores in the endoplasmic reticulum (Nauli et al., 2003; Yoder et al., 2002)

(Fig. 9). However, the role of Ca+2 responsive mechanosensor primary cilia are under debate since recent studies also showed a lack of Ca+2 influx mediated by fluid flow, arguing that mechanosensation may not be the decisive trigger of Ca+2 induced Ca+2 release (Delling et al., 2016). Instead, the PC-1/PC-2 complex has been shown activated by WNT ligands to mediate influx of extracellular calcium. WNT9B binding increased Ca+2 influx and together with WNT3A, induces whole-cell currents in WT but not in Pkd2−/− MEFs (Kim et al., 2016). In keeping with a key role for Ca+2 currents, the majority of studies that have assessed resting intracellular Ca+2, endoplasmic reticulum Ca+2 stores, or store-operated Ca+2 entry in primary cell cultures or microdissected samples, have shown reduced Ca+2 content in polycystic tissues (Torres and Harris, 2014).

A common characteristic of the polycystic disease is the accumulation of excess cAMP because reduced Ca+2 leads to hyperactivation of Adenylate Cyclase 5/6

(AC5/6) and inhibition of calcium/calmodulin-dependent phosphodiesterases (PDEs) (Wang et al., 2010). cAMP accumulation produces the activation of protein kinase A (PKA) that further contributes to the leak of calcium from intracellular stores by activating PC-2 in PKD1 or PKD2 mutated renal epithelial cells (Harris and Torres, 2014), thus maintaining high levels of cAMP/PKA signaling. PKA activation has a broad impact on cellular processes, including the disruption of tubulogenesis, activation of proliferative pathways, stimulation of chloride and fluid secretion or enhanced transcription of proinflammatory mediator dependent on STAT3-induction (Torres and Harris, 2014). Furthermore, cAMP positively mediates the expression of aquaporin-2 (Promeneur et al., 2000) and stimulates the chloride-ion driven fluid secretion in collecting duct cells (Wallace et al., 2001), thus contributing to the characteristic renal cyst expansion in polycystic diseases.

Renal epithelial tubulogenesis relies on canonical Wnt/β-catenin signaling. However, shortly after birth, canonical signaling ceases and is replaced by non-canonical Wnt/planar cell polarity (PCP) signaling (Carroll and Das, 2011). The cystic disease has been shown associated with enhanced canonical Wnt at the expense of non- canonical PCP signaling (Qian et al., 2005; Saburi et al., 2008). Inversin has been proposed to function as a molecular switch between the pathways (Simons et al.,

2005), likely by facilitating the recruitment of Dvl to Fz8 at the plasma membrane

(Lienkamp et al., 2010). By contrast, PKA indirectly stabilizes β-catenin promoting Wnt/β-catenin signaling (Li et al., 2000).

PKA signaling also regulates cell proliferation through MAPK signaling. In cultured normal renal tubular epithelial cells proliferation decreases in response to treatment with cAMP, since PKA inhibits the Ras/Raf-1/MEK/ERK pathway at the level of Raf- 1 (Yamaguchi et al., 2004). By contrast, PKD isolated cyst epithelial cells showed increased proliferation after cAMP treatment, by activating B-raf instead of inhibiting Raf-1. Moreover, M-1 renal cells expressing a PC-1 C-tail fragment hypothesized to disrupt calcium homeostasis exhibited cAMP-dependent activation of B-raf and ERK, and treatment with calcium repletion reversed their cAMP growth-stimulated

27 phenotype (Yamaguchi et al., 2004). In good agreement, inactivation of Pkd1 gene in kidney-specific conditional knockout mice has shown increased activation of ERK pathway (Shibazaki et al., 2008).

Known ERK targets that are likely disease-relevant include mTOR. Hyperactivation of mTOR signaling through ERK-dependent phosphorylation of tuberin (Distefano et al., 2009) has been shown to transcriptionally activate glycolysis (Rowe et al., 2013) and to inhibit AMPK, further increasing mTOR signaling (Düvel et al., 2010; Takiar et al., 2011). The mTORC1 pathway can be regulated as well by PC-1 since TSC1-2 complex activation requires phosphorylation by ERK kinase and PC-1 prevents the phosphorylation of tuberin at Ser664 by ERK (Distefano et al., 2009). More recently, it was shown that mTORC1 downregulates PC-1 protein levels, whereas mTOR inhibition by rapamycin treatment increases the trafficking of PC-1/2 to the cilium in TSC1 mutant cells (Pema et al., 2016). These results suggest that a cross-talk between TSC-PKD1 gene products is involved in the cystic phenotype of the tuberous sclerosis complex disease, where mTOR1 downregulates PC-1 via a negative feedback loop.

Altered cellular metabolism in ADPKD

Besides the signaling pathways that are deregulated in ADPKD kidneys, recent studies revealed that progression of PKD relies on profound metabolic reprogramming (Menezes et al., 2016; Rowe and Boletta, 2014; Rowe et al., 2013). Initially, transcriptomic and urine metabolomic analysis of a mouse model of early onset PKD after kidney specific Pkd1 knock-out, revealed that Pkd1-cystogenesis occurs in the context of gene networks like those that sustain normal kidney morphology and physiology, suggesting that the transcription factor HNF4α acts as a network node (Menezes et al., 2012). Interestingly, Pkd1 and HNF4α double mutants showed a more severe phenotype, indicating that upregulation of HNF4α target genes might constitute a protective mechanism.

Importantly, defective glucose metabolism was suggested as a hallmark of ADPKD.

More recently, analysis of murine Pkd1-/- MEFs and of mice lacking Pkd1 specifically

28 in kidney, revealed that loss of Pkd1 dramatically augments glucose consumption and energy production by aerobic glycolysis. These changes occurred concomitantly with hyperactivation of mTORC1, the master regulator of cellular metabolism transcriptional programs. Conversely, AMP-activated protein kinase (AMPK) was downregulated in ADPKD (Rowe et al., 2013). This is important because AMPK negatively regulates both mTOR and the cystic fibrosis transmembrane regulator (CFTR) chloride channel, which mediates fluid secretion into the cyst lumen at the apical membrane of the cystic cells (Davidow et al., 1996). AMPK directly inhibits CFTR by phosphorylation and indirectly inhibits mTOR by phosphorylation of TSC2 and Raptor (Gwinn et al., 2008; Hallows et al., 2000).

Another important metabolic sensor is the nicotinamide adenine dinucleotide–(NAD) dependent protein deacetylase sirtuin-1 (SIRT1). SIRT1 targets proteins involved in metabolic regulation and which are up-regulated in PKD mutant kidney cells by c-

Myc (Zhou et al., 2013). Although the exact mechanisms of sirtuin action are still under study, the sirtuins play critical roles diseases characterized by altered metabolism, suggesting that in PKD metabolic alterations are also intrinsic.

In healthy cells, nutrient deficiency, stress or reduced growth factors signaling can activate specific signaling pathways, including autophagy. Autophagy mediates the catabolism of intracellular organelles to restore energy homeostasis and ensure cell survival (Klionsky et al., 2016). However, in ADPKD autophagic flux appears to be suppressed in cyst lining cells, in part by mTORC1 hyperactivation (Belibi et al., 2011a). In cyst-lining cells autophagosomes increased due to a blockage in autophagosome-lysosome fusion. Murine Pkd1-/- renal cells under glucose starvation failed to induce autophagy and rather increased apoptosis (Rowe et al., 2013). Other ciliopathies also display decreased autophagy (Ravichandran and Edelstein, 2014).

Interestingly, in a different orthologous mouse model where Pkd1 inactivation was induced at the adult stage, Pkd1-/- cells revealed an abnormal fatty acid oxidation phenotype and the kidneys transcriptional profile was consistent with changes in lipid metabolism and whole metabolomic profile (Menezes et al., 2016). However, in this study, only a few metabolites and one lipid (diacylglycerol) were significantly altered.

1.3.3 Autosomal recessive polycystic kidney disease (ARPKD)

In contrast to ADPKD, the recessively inherited form of PKD primarily affects neonates and children and has an estimated incidence of 1:20000 births (Guay- Woodford and Desmond, 2003). ARPKD neonates have a high rate of death shortly after birth due to pulmonary hypoplasia and respiratory insufficiency. Fetuses are characterized by oligohydramnios, severe bilateral polycystic kidneys and biliary tract defects (Guay-Woodford and Desmond, 2003).

The responsible genes are polycystic kidney and hepatic disease 1, PKHD1 (6p12.3- p12.2), which encodes for fibrocystin (also called polyductin) and less frequently, DZIPL1 (Lu et al., 2017). Fibrocystin is expressed along the collecting ducts and biliary ducts and it is important for normal renal collecting ducts and intrahepatic bile duct differentiation (Guay-Woodford, 2006). In renal epithelial cells, it is localized to the basolateral plasma membrane, the primary cilium, and the centrosome

(Ibraghimov-Beskrovnaya and Bukanov, 2008). Fibrocystin (~460 kDa), has been reported in renal epithelial cells, to localize at the primary cilium, together with PC-2, and at basal bodies (Zhang et al., 2004).

Fibrocystin interaction with PC-2 is specific to the intracellular cytoplasmic domain, encoded mostly by the 67th exon, which also has other functional motifs like a ciliary targeting sequence and a nuclear localization signal (Hiesberger et al., 2006; Wu et al., 2006). However, a recent study used a novel mouse model with a triple HA-tag to fibrocystin C-terminus together with lox P sites flanking exon 67th. They showed in vivo the fibrocystin cleavage products but, contrary to previous in vitro studies, they failed to co-precipitate PC-2. Additionally, endogenous expression of fibrocystin was not detected at the primary cilia, and after Cre-mediated deletion, homozygous mice lacking only the fibrocystin C-terminus did not show any disease phenotype (Outeda et al., 2017).

Several ARPKD murine models have been reported with various Pkhd1 mutations (Moser et al., 2005; Ward et al., 2002). These models show biliary dysgenesis and fibrosis, as in human ARPKD but the renal cystogenesis shows a later onset and in a mild manner. Pancreatic cysts are also common.

In contrast to ADPKD, where cilia length is not affected, shortened and abnormal ciliary structures have been described in the cholangiocytes of the PCK rat, an ARPKD model where a spontaneous splicing loss of function mutation of Pkhd1 initiates hepatic cystogenesis (Muff et al., 2006).

1.3.4 Nephronophthisis

Nephronophthisis (NPHP) comprises a genetically and clinically heterogeneous group of autosomal recessive cystic kidney disorders that represente the most frequent genetic cause of ESRD in children and young adults (Salomon et al., 2009). Juvenile NPHP, the most common form of the disease, typically starts with polyuria, polydipsia, and anemia, already within the first 10 years of life. Among the NPHP genes identified so far, NPHP1 is the most commonly mutated gene (Bergmann, 2015).

Many NPHP mutations cause cystic kidney disease combined with extrarenal manifestations, such as liver fibrosis, retinitis pigmentosa, cerebellar hypoplasia, situs inversus or cardiac abnormalities (Benzing and Schermer, 2012).

NPHP genes encode different proteins (nephrocystins) that associate in macromolecular complexes (Williams et al., 2011). By tandem affinity purification and mass spectrometry analysis, seven NPHP proteins and the related Jouberin and MKS1 proteins were used for screening and identification of three NPHP modules with different intracellular localization (Sang et al., 2011). NPHP5-6 complexes occur and the MKS module were found at basal bodies. At the ciliary base (transition zone), the NPHP1-4-8 module is formed, whereas NPHP3 and NPHP9 (NEK8) localized to the inversin (NPHP2) compartment at so called inversin compartment (Shiba et al.,

2010) (Fig. 10).

Figure 10. Scheme showing localization of ciliary proteins in sub-ciliary compartments. Biochemical and genetic interaction data from four different studies and the model system used is shown on top of the figure. Dark lines correspond to interactions detected by MS/MS, dashed lines correspond to genetic interaction. Proteins in the white text correspond to common components identified by different studies. Figure is taken from (Szymanska and Johnson, 2012).

In a proteomic screen ectopically expressed NEK8 in HEK293T cells pulled down

ANSK6 (NPHP16) (Hoff et al., 2013). Interestingly, a missense mutation in ANKS6

(a.k.a SamCystin) is the genetic cause of the cystic kidney phenotype of the Han:SPRD cy/+ rat (Brown et al., 2005). Furthermore, ANKS6 was identified as a causal gene of nephronophthisis in six human families and the same study show that

ANKS6 has a role in renal development by preventing pronephric cyst formation (Hoff et al., 2013). ANKS6 has shown to be recruited by inversin into the inversin ciliary compartment to activate NEK8 thus stimulating its kinase activity. In turn, NEK8 mutations block ANKS6 localization to cilia (Czarnecki et al., 2015). In addition NEK8 blocks activation of the Hippo signaling effector YAP (Grampa et al., 2016). Therefore, ANKS6 emerges as a central component of the NEK8-Inversin-NPHP3 module that links cilia to growth signals mediated by the Hippo pathway.

In contrast to ANKS6 and NEK8, has been shown to activate YAZ instead of inhibiting (Habbig et al., 2011). NPHP4 directly binds Lats1 and inhibits Lats1-dependent phosphorylation of YAP and TAZ, thereby allowing these transcriptional regulators to

32 translocate to the nucleus and stimulate transcription of proliferation genes by TEAD transcription factors.

Independently, NPHP3 and NPHP4 were also described as regulators of Wnt signaling in the developing zebrafish pronephros (Burcklé et al., 2011).

1.3.5 PKD treatments

The different pathways identified as drivers of PKD suggest a range of possible therapeutic targets. However, so far only antagonists of vasopressin receptor 2 (VR2), such as Tolvaptan, has been approved for treatment of rapid disease progression in adult ADPKD patients. Since the antidiuretic hormone (ADH/AVP) can bind VR2 and promote water retention and that AVP is an agonist for cAMP production in collecting ducts, the use of Tolvaptan was supported to achieve reduction of cysts growth (Janssens et al., 2017).

The second main therapeutic agents tested were the rapalogues (Walz et al., 2010).

These inhibitors of mTOR pathway failed to succeed in clinical trials at least in part due to the many side effect observed in humans despite the promising results in murine models (Shillingford et al., 2006). A possible alternative is to use rapalogues in combination therapies to achieve more consistent effects. This strategy was proposed by using rapamycin and Cl-IB-MECA, an agonist of adenosine type 3 receptors (A3AR) known for its inhibitory effect on adenylyl cyclases. This combination has been shown to reduce cAMP levels and cell proliferation in cystic cells mediated by ERK and mTOR signaling (Aguiari et al., 2009). Combination with rapamycin, had synergistic effects on the inhibition of CREB, mTOR and ERK pathways, thus, indicating that combined therapies could synergized, at least in cultured ADPKD cell lines (Stephanis et al., 2017).

Remarkably, recent data support the hypothesis that dependence on aerobic glycolysis sensitized cystic cells to death after glucose deprivation (Rowe et al.,

2013). The improvement of cystic disease in a conditional Pkd1 inactivation mouse model treated with the nonmetabolizable glucose analog 2-Deoxy-D-Glucose

(Chiaravalli et al., 2016), suggested that aerobic glycolysis is a plausible treatment for PKD patients. This treatment also rescued the pAMPK and pAcetyl-CoA Carboxykinase levels that are intrinsically low in this mouse model (Rowe et al., 2013). In line with the critical role of metabolic pathways in PKD progression, it was shown that food restriction slows ADPKD progression, cyst area, fibrosis and inflammation by suppressing mTOR and by activating the Lkb1/AMPK pathway in

Pkd1 RC/RC (PKD1 R3277C) and Pkd2 WS25/- mouse models (Warner et al., 2016).

AMPK is of interest because of its role as master regulator of cellular energy balance among other therapeutic targets that emerged from PKD models. Especially metformin, a well-known AMPK activator, has been shown to reduce CFTR and mTOR levels, leading to slower cyst growth of cystic kidneys (Takiar et al., 2011). Interestingly, metformin or AICAR (5-aminoimidazole-4-carboxamide ribonucleoside) which also activates AMPK, inhibited ERK in Pkd1-/- cells (Rowe et al., 2013). Finally, in the Han:SPRD (Cy/+) rat model, AICAR combined with 2-Deoxy-D-Glucose facilitated a shift to aerobic metabolism while reducing kidney size and improving renal function (Riwanto et al., 2016).

1.4 Bicaudal-C

1.4.1 Discovery

Bicaudal-C (BicC) was initially identified in 1985 by Wieschaus (Mohler and Wieschaus, 1985), in a mutagenesis screen in Drosophila. While BicC homozygotes females were sterile, BicC heterozygous females produce embryos with "double abdomen" (bicaudal phenotypes) characterized by a mirror-image of 2-4 posterior segments. These phenotypes indicated that BicC has essential roles in oocyte formation and axial patterning.

Owing to its mosaic mode of development, Drosophila localizes specific cell fate determinants asymmetrically already in oocytes to pre-specify the future body axes. For instance, restriction of oskar mRNA to the posterior pole is mandatory to achieve development of posterior structures (Ephrussi and Lehmann, 1992; Gavis and

Lehmann, 1992). BicC restricts oskar mRNA translation to the posterior pole in order to prevent its premature translation, and to thereby allow the formation of anterior structures (Mahone et al., 1995; Saffman et al., 1998). BicC co-immunoprecipitates

Orb (Castagnetti, 2003), the Drosophila homolog of CEPB (Cytoplasmic polyadenylation Element Binding protein). Orb mediates oskar long polyadenylation to reach high levels of translation at the posterior pole. Interestingly, BicC patterning defects are suppressed when the mutation is combined with and orb allele or the aretQB72 allele, encoding Bru, a negative regulator of osk translation. Additionally, BicC mutants mislocalize gurken mRNA between the nucleus and the anterior cortex instead of the dorsal cortex as in the wild-type oocytes (Kugler et al., 2009). Since correct mRNA localization is necessary for Grk protein secretion, these findings established that BicC specifies not only anterior but also dorsal structures.

KH domains are well-known RNA binding domains, for review see (Nicastro et al., 2015). An RNA binding protein containing KH domains, FMRP, the protein mutated in Fragile X syndrome (Darnell et al., 2011), regulates translation of specific mRNAs important for synapse development and function. In C. elegans, GLD-3, which belongs to the Bicaudal-C family of RNA binding proteins and contain a KH domain, influences germline development and functions in sex determination by promoting

35 spermatogenesis (Eckmann et al., 2002). GLD-2 is associated with P granules in early embryos and has cytoplasmic poly(A) polymerase (PAP) activity which is stimulated by GLD-3 in vitro. Interestingly, GLD-2 is not a conventional PAP since it lacks a recognizable RRM (RNA recognition motif)-like domain. GLD-2 association with mRNA is mediated by GLD-3 thereby targeting mRNAs for polyadenylation

(Wang et al., 2002).

BICC1 encodes for an evolutionary conserved RNA-binding protein containing three KH (hnRNP K homology) domains and two KH-like domains, that are distinguished by the absence of GXXG motifs (where G represents glycine and one or both residues X are basic). A long glycine/serine-rich intervening sequence (Gamberi and Lasko, 2012; Hollingworth et al., 2012) follows the KH domains and precedes a SAM (Sterile Alpha Motif) domain that is close to the C-terminal extremity (Kim and Bowie, 2003) (Fig. 11).

Figure 11. MmBicc1 architecture. Scheme depicting the different domains of mBicc1. KH: hnRNP K homology domains, KHL: KH-like domains and SAM: Sterile alpha motif.

The RNA binding properties of Bicc1 KH domains have been confirmed (Bouvrette et al., 2008; Piazzon et al., 2012) and interestingly, a mutant allele of BicC containing a point mutation in one of the KH domains abolish its function in anterior patterning (Mahone et al., 1995).

In C. elegans mutations in the Bicaudal-C homolog bcc-1, which contains a SAM domain, suppresses the lethality associated with a mutation of the essential receptor tyrosine kinase rol-3, suggesting that bcc-1 is a suppressor of rol-3(Jones et al., 2012). Rol-3 functions on the regulation of morphogenesis and development of specialized epidermal tissues (Jones et al., 2013). SAM domains are known to promote protein multimerization in a head-to-tail fashion thanks to two distinct

36 surfaces called EH (End-Helix) and ML (Mid-Loop) (Kim and Bowie, 2003; Knight et al., 2011; Qiao and Bowie, 2005). An in vitro study on the recombinant Bicc1 SAM domain has confirmed its high potential for self-polymerization (Knight et al., 2011)

(Fig. 12).

Figure 12. Bicc1 SAM domain can homopolymerize. Electron microscopy image showing purified negGFP-Bicc1 SAM polymers (arrows). Taken from (Knight et al., 2011).

1.4.2 Role of Bicaudal C in different species

A) In Xenopus laevis

Maternally expressed xBic-C localizes to the vegetal half of the egg (Wessely and De

Robertis, 2000). Overexpression of synthetic xBic-C mRNA leads to ectopic endoderm formation in the absence of mesoderm induction. Furthermore, by deleting all 5 KH domains, it was shown that endoderm specification is dependent on the

RNA-binding properties of xBic-C and not in the SAM domain (Wessely and De Robertis, 2000). xBic-C can bind, and translationally repress, several maternal mRNAs encoding pivotal cell-fate regulators. Analysis of cripto1 mRNA identified in its 3’UTR a Bicc1 RNA-binding element (Zhang et al., 2013). These results pointed to xBic-C as maternal determinant important to promote endoderm germ layer formation during early embryonic development in Xenopus.

37 In addition, xBic-C is express in pronephric epithelial cells, and loss of function by antisense morpholino oligomers results in edematous embryos with impaired osmoregulatory functions and increased pronephric and duct tubule diameters. These characteristics are reminiscent of PKD phenotypes (Tran et al., 2007).

B) In Sea urchin

In Sea urchin, BicC mRNA follows a ubiquitous expression in early developmental stages, suggesting that its function may depend on post-transcriptional regulation or even the localization of its target molecules (Yaguchi et al., 2015). Importantly, BicC morphants develop slower than control embryos, and gastrulation is not achieved. Although BicC morphants still show some but not all aspects of endomesoderm specification, they lose the differentiated anterior neurogenic ectoderm and endomesoderm (Yaguchi et al., 2015). In the neurogenic ectoderm of the embryo,

BicC is important for ankAT-1 (Ankyrin-containing gene specific for Apical Tuft) gene expression which is pivotal for maintaining the length and motility of the apical tuft cilia (Yaguchi et al., 2010).

C) In Mus musculus

In 2000, Wessely et al, identified the cDNA clone encoding the full-length mRNA of murine Bicaudal C (mBicc1), while a high throughput genomic sequences analysis from the Human Genome Program allowed the identification of the human homolog of BicC (Wessely et al., 2001). These Xenopus and Drosophila BicC homologs, named hereafter Bicc1, contain the same domain structure. Mouse and human Bicc1 are 89.8% identical and related to Xenopus they are 78.6% and 81.3% identical, respectively.

By in situ hybridization Bicc1 expression, was first detected at the rostral tip of the primitive streak and later, it delineates the layer of the node from which endoderm and midline mesoderm arise at the late headfold stage (Wessely et al., 2001). Bicc1 is then found at sites of cartilage formation, the diaphragm, the pericardium and the mesenchyme of the emerging lungs. Unlike in Xenopus, mouse Bicc1 is expressed in the development of the kidney at mesonephros and metanephros stages.

Bicc1 contains two transcripts alternatively spliced (Cogswell et al., 2003): transcript A (exons 1-20 and exon 22) encoding for a 977 aa protein and transcript B (exon 1- 22), with a premature stop codon in exon 21, encoding a shorter protein of 951 aa.

Interestingly two mouse models for PKDs arise from spontaneous mutations in Bicc1 gene. The juvenile congenital polycystic kidney disease (jcpk) mutation was recovered in a chlorambucil mutagenesis program at the Wadsworth Center (Flaherty et al., 1995). The mutant Bicc1 protein is truncated after aa 48 (Cogswell et al., 2003) as a result of a point mutation in the splice acceptor site of exon 3. Homozygous jcpk causes severe kidney cystic disease, with cyst emerging from all segments of the nephron. Moreover, the mice exhibit extrarenal manifestations such as enlargement of bile and pancreatic ducts and the gall bladder. They die shortly after birth (P10). Heterozygous +/jcpk show signs of glomerulocystic disease. Due to the extrarenal abnormalities and the heterozygote effects, this model is reminiscent of human

ADPKD.

In comparison, the bpk allele appeared after inbreeding of the BALB/c strain (Guay- Woodford et al., 1996). The bpk mutant allele has a GC insertion in the exon 22 which produces a frame shift and the aberrant elongation of the protein at the C-terminus.

Homozygous mutants express a PKD phenotype that, while less severe than in the jcpk mice, more closely resembles human ARPKD with initially cystic proximal tubules and later appearing in collecting ducts. Death occurs by 4 weeks of age

(Nauta et al., 1993). Like in ARPKD, bpk mutants show apical missorting of EGFR (Sweeney et al., 2001) and increased mTOR signaling (Shillingford et al., 2006).

In our lab, Bicc1 was inactivated by targeted gene disruption using homologous recombination in embryonic stem cells, deleting exons 5 to 11 and introducing a frameshift in exon 12 that resulted in a premature stop of translation. In this mouse model only 50% of the homozygotes developed to term but die shortly after birth, they display complete situs inversions and randomization of asymmetric Nodal signaling concomitant to defects in planar alignment of motile cilia which is required for cilia- driven fluid flow (Maisonneuve et al., 2009).

Homozygous Bicc1-/- are severely affected from renal and pancreatic fluid-filled cyst (Fig. 13), concomitant with ventricular septal heart defects (Maisonneuve et al., 2009). These phenotypes were confirmed by a different gene targeting strategy of the Bicc1 gene, including the hepatic and pancreatic duct dilatations (Tran et al., 2010). Additionally, Bicc1-/- mutant kidneys shows excess accumulation of cAMP (Piazzon et al., 2012), a hallmark of PDK.

Figure 13. Bicc1 expression in mouse kidneys. RNA in situ hybridization of Bicc1 mRNA in WT compared to Bicc1-/- (KO) neonatal kidneys, using a probe against the deleted region. The kidney cystic anatomy reminiscent of PKD is clear in the Bicc1-/- mice. Figure from (Piazzon et al., 2012).

D) In Homo sapiens

More recently, Kraus et al. described two Bicc1 human mutations causing cystic renal dysplasia in pediatric patients. Both patients were heterozygous for these mutations, one nonsense (c.259C>T, p.Gln87*) in the first KH domain and the other missense (c.2795A>G, p.Glu932Gly) in the SAM domain. Furthermore, supported by GWAS data, Bicc1 was identified as a genetic determinant of osteoblastogenesis and bone mineral density, likely by regulating Pkd2 transcription (Mesner et al., 2014).

40 1.4.3 Molecular functions

At the molecular level, numerous studies show that Bicc1 proteins are involved in translational repression of target mRNAs and that KH domains play an essential role for this function.

The importance of the KH domain was initially noticed by the immunoprecipitation of Bicc1 by several KH containing proteins, including Sam68, GLD-1, GRP33, and Qk1 but not FMR1 in HeLa cells lysates (Chen et al., 1997). This study emphasized the fact that a single KH domain present in Sam68 protein can take part in protein-protein interactions beside mediating protein-RNA interactions. Furthermore, a mutation (G296R) in a single KH domain in Drosophila Bic-C mutants reduces the strength of mRNA binding (Saffman et al., 1998). Additional studies demonstrated that the third KH domain of mBicc1 is sufficient for synthetic mRNA binding in vitro (Bouvrette et al., 2008).

Notably, in Drosophila ovaries, it was shown that Bic-C can bind and post transcriptionally repress itself by specific interaction with cis-acting elements in its 5’UTR (including nt 30-184). Bic-C direct association with NOT3/5 component of

CCR4-NOT deadenylase complex in vitro and in vivo (Chicoine et al., 2007). In the same organism, Bicc1 associates with several mRNAs, and in vitro assays demonstrated its role in downregulation of reporters through their 3’UTR. Those validated targets mRNAs, such as cripto-1, which participates in the Nodal pathway, play a key role in development (Zhang et al., 2013).

In mouse, Adenylate Cyclase 6 (AC6) and Protein Kinase A inhibitor α (PKIα) were identified as two mRNA targets of Bicc1. Bicc1 KH domains are sufficient for binding these targets but not to mediate their silencing. Bicc1 loss of function increases the renal protein levels of AC6 which may correlate with increased levels of cAMP in this mouse model of PKD (Piazzon et al., 2012). Additional mRNA targets involved in a different kidney function have been identified, they include Wnk1 and V-ATPase B1, important for ion transport at the distal nephron (Arroyo and Gamba, 2012; Karet et al., 1999).

In vivo, Bicc1 is important to attenuate the canonical Wnt pathway in the posterior notochord. This result was expanded in vitro since in a dose dependent manner, Bicc1 and specifically Bicc1 SAM domain, inhibited Dvl2-mediated induction of a reporter of the canonical Wnt signaling (Maisonneuve et al., 2009). Interestingly, the Bicc1 nonsense mutation identified in patients results in complete abolishment of Bicc1 inhibitory effect over Wnt pathway (Kraus et al., 2012).

Analysis of Bicc1 localization pattern showed that mouse Bicc1 overexpressed in COS-1 cells forms foci around RNA-processing bodies (P-bodies) containing GFP- Dcp1a, the decapping enzyme. Interestingly, this pattern is dependent on the Bicc1 SAM domain but not on KH domains (Maisonneuve et al., 2009). A similar cytoplasmic granular pattern was evidenced when Bic-C colocalize with Trailer Hitch (TraI)-containing ribonucleoprotein complex (mRNP), important for Grk secretion in

Drosophila oocytes (Kugler et al., 2009).

Interestingly, Polycystic kidney disease causal proteins polycystin-1 and polycystin-

2 have been linked to Bicc1. Initially, it was demonstrated that xBic-C regulate the efficient translation of Pkd2 mRNA by antagonizing the repression exerted by miR-

17 family members on pkd2 3’UTR (Tran et al., 2010). On the other hand, loss of PC-

1 protein reduce the Bicc1 protein and transcript levels in vivo and in vitro, and this effect can be rescued by expression of human PC-1 C-terminus in the pkd1-/- cell lines (Lian et al., 2014). These findings suggest Bicc1 acts upstream and downstream of polycystins.

Additional ciliary proteins containing ankyrin repeats have been found in complexes where some of their components can interact with Bicc1. For instance, ANKS6 which also cause renal cyst formation when mutated, contains at its N-terminus 10 tandem ankyrin repeats and a SAM domain at its C-terminus (Brown et al., 2005). ANKS6 was reported to interact with Bicc1 in a KH domain- and, in part RNA-dependent manner (Stagner et al., 2009).

2. Thesis aims

ADPKD is the most common inherited renal pathology and one of the leading causes of the end-stage renal disease. Renal cysts also arise in patients and in mouse models with mutations in the RNA-binding protein Bicc1. Little is known about Bicc1 in renal cyst formation, but given its role in specific mRNA silencing, relying on both its RNA-binding properties and its SAM domain-mediated clustering, Bicc1 appears as an excellent paradigm for studying post-transcriptional regulation of gene expression in a disease-relevant context. Furthermore, the contribution of glycolytic defect in PKD progression was recently uncovered, and studying the underlying molecular mechanisms is of high interest in the field.

To gain knowledge in how Bicc1 is involved in the pathogenesis of polycystic kidneys, my objectives were:

 Studying the ability of Bicc1 to polymerize in cytoplasmic clusters and defining the contribution of this clustering to its translational repression function (Results part I).

 Defining the Bicc1 interactome by using a proteomic approach, validate new interacting partners, evaluate their contribution to the known Bicc1 functions and potentially uncover new roles for the protein (Results part II).

 Analyzing Bicc1 mutant kidneys metabolome and key physiological parameters to describe metabolic alterations in this PKD model (Results part III).

3. Results part I

Manuscript title:

Bicc1 Polymerization Regulates the Localization and Silencing of

Bound mRNA

Authors:

Benjamin Rothé, Lucia Leal-Esteban, Florian Bernet, Séverine Urfer, Nicholas Doerr, Thomas Weimbs, Justyna Iwaszkiewicz, Daniel B. Constam.

Summary of results and contribution:

Bicc1 deficient mice are characterized by renal and pancreatic cysts development and bile duct dilatation (Flaherty et al., 1995; Lemaire et al., 2015; Lian et al., 2014;

Piazzon et al., 2012). In the kidneys, Bicc1 is expressed along the nephron (Bernet F. EPFL TH_6606) and in cultured cells Bicc1 clusters in cytoplasmic foci

(Maisonneuve et al., 2009). However, Bicc1 distribution at the subcellular level in the kidney and liver was not studied before. Here, we showed that, Bicc1 clusters in cytoplasmic puncta in renal proximal tubules and cholangiocytes.

To evaluate whether Bicc1 clustering may influence the distribution of bound target mRNAs, we monitored the localization of a previously described Bicc1 mRNA target (Piazzon et al., 2012). The tool that we designed fused the Luc-AC6-3’UTR proximal reporter, containing the Bicc1 binding region (Piazzon et al., 2012) with 27 MS2 hairpins that form binding sites for the bacteriophage MS2-YFP coat protein (Bertrand et al., 1998). Using this system, we demonstrated that, in presence of wild-type Bicc1, MS2-YFP bound to Luc-AC6-MS2x27 reporter concentrates in cytoplasmic Bicc1 puncta. We concluded that target mRNA is recruited to Bicc1 in cytoplasmic clusters.

To address how the SAM domain regulates Bicc1 clustering, we generated a homology-based 3D model structure. Structure modelling predicted that the Bicc1

SAM domain organizes a central left-handed head-to-tail helix with six SAM units per turn and the RNA binding KH domains are peripheral displayed. By mutating the predicted SAM-SAM interface, we showed that specific residues in this interface are required to polymerize full-length Bicc1 in macromolecular assemblies and to localize its target mRNA to cytoplasmic foci for translation inhibition. SAM polymerization also increases Bicc1 stability, as demonstrated for a Bicc1-polymer mutant or a construct lacking the whole SAM domain. Similar results were also obtained with Bicc1 bpk mouse model.

Importantly, our results provide important structure-function understanding into the Bicc1-mediated translational silencing as they highlight Bicc1 polymerization as essential to recruit a target mRNA into Bicc1 foci and to induce its translational repression. By contrast, Bicc1 function in Wnt signaling pathway inhibition appeared independent of SAM-polymerization.

My contribution to this study comprised the cloning and characterization of the reporter mRNA containing the Bicc1 binding region of the AC6-3’UTR(Piazzon et al., 2012) and 27 repeats of a motif that binds the bacteriophage MS2 coat protein (the

MS2x27 motif) (Bertrand et al., 1998). This reporter was extensively used for validating our hypothesis by in vitro assays.

Bicc1 Polymerization Regulates the Localization and Silencing of Bound mRNA

Benjamin Rothé,a Lucia Leal-Esteban,a Florian Bernet,a Séverine Urfer,a Nicholas Doerr,b Thomas Weimbs,b Justyna Iwaszkiewicz,c Downloaded from Daniel B. Constama Ecole Polytechnique Fédérale de Lausanne (EPFL), SV ISREC, Lausanne, Switzerlanda; Department of Molecular, Cellular, and Developmental Biology and Neuroscience Research Institute, University of California, Santa Barbara, Santa Barbara, California, USAb; Swiss Institute of Bioinformatics, University of Lausanne, Lausanne, Switzerlandc

Loss of the RNA-binding protein Bicaudal-C (Bicc1) provokes renal and pancreatic cysts as well as ectopic Wnt/␤-catenin signaling during visceral left-right patterning. Renal cysts are linked to defective silencing of Bicc1 target mRNAs, including adenylate cyclase 6 (AC6). RNA binding of Bicc1 is mediated by N-terminal KH domains, whereas a C-terminal sterile alpha motif (SAM) self-polymerizes in vitro and localizes Bicc1 in cytoplasmic foci in vivo. To assess a role for multimerization in silencing, we con- http://mcb.asm.org/ ducted structure modeling and then mutated the SAM domain residues which in this model were predicted to polymerize Bicc1 in a left-handed helix. We show that a SAM-SAM interface concentrates Bicc1 in cytoplasmic clusters to speciﬁcally localize and silence bound mRNA. In addition, defective polymerization decreases Bicc1 stability and thus indirectly attenuates inhibition of Dishevelled 2 in the Wnt/␤-catenin pathway. Importantly, aberrant C-terminal extension of the SAM domain in bpk mutant Bicc1 phenocopied these defects. We conclude that polymerization is a novel disease-relevant mechanism both to stabilize Bicc1 and to present associated mRNAs in speciﬁc silencing platforms.

he asymmetric distribution and localized translation of truncated before the first KH domain, whereas a GC insertion in on October 6, 2017 by EPFL Scientific information and libraries TmRNAs control the expression of many proteins in a wide bpk mutant mice changes the reading frame within the last exon range of cells and tissues. Well-known examples include maternal so that 21 amino acids at the C terminus are replaced by 149 determinants of embryo patterning and asymmetric fates in Dro- aberrant residues (17) (see Fig. S1B in the supplemental material). sophila oocytes, such as bicoid, oskar (osk), gurken (grk), and nanos Bicc1jcpk/jcpk mutants develop renal cysts along the entire nephron, mRNAs. Multiple trans-acting factors associate with these combined with dilated pancreatic and liver bile ducts that are mRNAs in dynamic ribonucleoprotein (RNP) complexes to or- reminiscent of autosomal dominant polycystic kidney disease chestrate nuclear processing, translational silencing, transport, (ADPKD). Unlike homozygotes, which die soon after birth, cytoskeletal anchoring, and derepression of translation at the final heterozygotes develop glomerulocystic disease in 25% of the cases destination (1). Defects at any of these steps can have dramatic after ageing (20). In contrast, bpk mutants are a noncongenital consequences. In particular, the failure to activate the translation model of autosomal recessive polycystic kidney disease (ARPKD), of osk mRNA in the posterior pole plasm prevents the formation with renal cysts arising first in proximal tubules and later in col- of abdominal structures and germ cells, whereas precocious osk lecting ducts (21). Compared to the time of cyst formation in jcpk mRNA translation in anterior oocytes blocks head formation and mutants, cyst formation is delayed in bpk mutants, likely because can give rise to a bicaudal phenotype with posterior duplications bpk affects only one of two alternatively spliced transcripts (17, (2–6). 22). ADPKD is caused by mutations in the PKD1 or PKD2 gene, Posterior localization and translational regulation of osk whereas ARPKD results from defects in PKHD1. The correspond- mRNA are mediated by specific cis- and trans-acting elements and by multiple interacting factors and regulatory proteins (7–9), in- ing proteins, polycystin-1, polycystin-2, and fibrocystin, associate cluding Bicaudal-C (Bic-C; reviewed in reference 10). Bicaudal-C comprises three KH domains with characteristic RNA-binding Received 2 April 2015 Returned for modification 2 May 2015 GXXG signatures, where G represents glycine and one or both Accepted 18 July 2015 residues X are basic (11, 12). The three KH and two KH-like do- Accepted manuscript posted online 27 July 2015 mains that are distinguished by the absence of GXXG signatures Citation Rothé B, Leal-Esteban L, Bernet F, Urfer S, Doerr N, Weimbs T, are linked to a C-terminal sterile alpha motif (SAM) by a long Iwaszkiewicz J, Constam DB. 2015. Bicc1 polymerization regulates the localization glycine/serine-rich intervening sequence (see Fig. S1A in the sup- and silencing of bound mRNA. Mol Cell Biol 35:3339–3353. plemental material). The loss of Bic-C in Drosophila oocytes leads doi:10.1128/MCB.00341-15. to the premature derepression of osk translation (13, 14) and ec- Address correspondence to Daniel B. Constam, Daniel.Constam@epfl.ch. topic anterior localization of grk mRNA (15). A biochemical Supplemental material for this article may be found at http://dx.doi.org/10.1128 screen for direct targets revealed binding of Bic-C to stau mRNA /MCB.00341-15. and to several other transcripts (16). In addition, Bic-C has been Copyright © 2015, Rothé et al. This is an open-access article distributed under the terms of the Creative Commons Attribution-Noncommercial-ShareAlike 3.0 shown to recruit CCR4-NOT deadenylase to attenuate its own Unported license, which permits unrestricted noncommercial use, distribution, translation (16). and reproduction in any medium, provided the original author and source are Mutations in mouse and human Bic-C homologs, as well as credited. knockdown of Xenopus Bic-C, revealed an essential role in renal doi:10.1128/MCB.00341-15 tubule morphogenesis (17–19). In jcpk mutant mice, Bicc1 is

October 2015 Volume 35 Number 19 Molecular and Cellular Biology mcb.asm.org 3339 Rothé et al. with each other and regulate calcium flux in response to mechan- MATERIALS AND METHODS ical stimulation by fluid flow (23, 24). In embryonic mouse kid- Plasmids, cloning, and mutagenesis. The luciferase reporter plasmids neys, Pkd1 enhances the expression of Bicc1, which in turn in- pCSϩ::AC6-3=UTRprox and pCSϩ::PKI␣-3=UTRprox have been decreases the levels of Pkd2 mRNA, indicating that Bicc1 acts both scribed previously (27). To construct the AC6-27ϫMS2 reporter plasmid, downstream and upstream of polycystins (25, 26). 24 repeats of a motif that binds the bacteriophage MS2 coat protein were Recent functional analysis identified adenylate cyclase 6 (AC6) amplified by PCR from pCR4-24MS2SL-stable (catalog number 31865; and protein kinase inhibitor ␣ (PKI␣) mRNAs to be the first direct Addgene) and added to 3 existing MS2 sites in plasmid pCSϩ::AC6- targets of mammalian Bicc1 (27). PKI␣ inhibits protein kinase A 3=UTRprox by insertion into a unique PstI site. To subclone the mutant Downloaded from (PKA), whereas AC6 stimulates it by synthesizing cyclic AMP Bicc1 cDNAs encoding the mutation in mutant A (mutA) to mutF and bpk, the plasmid pCMV-SPORT6::HA-Bicc1 (35) was modified by intro- (cAMP), suggesting a dynamic role for Bicc1 in regulating cAMP/ ducing a silent BglII site upstream of the SAM-coding sequence. Mutant PKA signaling. Importantly, cAMP and AC6 promote cystic cDNA fragments were then cloned between this engineered BglII site and growth in Pkd1 mutant mice and in ADPKD patients (28–30). a unique XbaI site. Mutant cDNA fragments encoding the mutations in Bicc1 mutant mice also share other key features with human mutA to mutF were generated by overlap extension PCR. The plasmids ADPKD, including impaired apical-basal sorting of the epidermal carrying Bicc1bpk and the splice variant B of wild-type (WT) Bicc1 were growth factor receptor; hyperactivation of its ligand, transforming obtained by introducing a synthetic cDNA (Proteogenix) into the pCMV- growth factor ␣; and elevated mTOR signaling (31–34). In addi- SPORT6::HA-Bicc1 vector. The sequences of all mutated expression vec- http://mcb.asm.org/ tion, Bicc1 is needed during development for the alignment of tors were verified by Sanger sequencing. The plasmids pTOPFLASH (43), motile node cilia by planar cell polarity (PCP) signals that govern pCDNA3.1::mDvl2 (44), pDCP1a-GFP (45), and pGEX-1␭T::SAM (27) visceral left-right patterning (35, 36). Deregulation of PCP or ca- were previously described. nonical Wnt signaling can also trigger renal cysts (23). Together, Recombinant GST-SAM and custom anti-Bicc1 antibody. The pGEX-1␭T::SAM plasmid was used to produce the recombinant glutathi- these observations highlight the relevance of Bicc1 mutants as a one S-transferase (GST)–SAM protein in the Escherichia coli BL21 strain disease model and the importance of elucidating the molecular (Novagen) as described previously (27). The GST-SAM protein was pu- mechanisms that regulate Bicc1 activities at the crossroads of mul- rified from cell extracts under native conditions, using glutathione-Sep-

tiple signaling pathways. harose 4B, as recommended by the manufacturer (GE Healthcare). Puri- on October 6, 2017 by EPFL Scientific information and libraries Repression of AC6 and PKI␣ mRNAs by Bicc1 depends on fication was carried out in a buffer consisting of 50 mM Tris-HCl, pH 8, specific regions in their proximal 3= untranslated regions 200 mM NaCl, 1 mM dithiothreitol (DTT). When necessary, 20 mM (UTRs) and on cognate microRNAs (miRNAs) (27). While glutathione was added for elution. A custom polyclonal Bicaudal-C anti- Bicc1 binds these RNAs independently of the SAM domain, body against GST-SAM was raised by Proteogenix (Oberhausbergen, deletion of the SAM domain blocks their loading into miRNA- France) in rabbit by development of an antibody against the mouse Bicc1 induced silencing complexes (miRISCs) with Argonaute 2 SAM domain (residues 870 to 984) fused to an N-terminal His tag. The (Ago-2) (27). In addition, the SAM domain increases the po- antibody was affinity purified and tested for its activity against the antigen by enzyme-linked immunosorbent assay. tential of Bicc1 to inhibit the Wnt signaling component Di- Bicc1 mutant mice. Mice heterozygous for a targeted null allele of shevelled 2 (Dvl2) independently of KH domains in TOPflash Bicc1 were maintained on a C57BL/6 mouse genetic background in indi- reporter assays, possibly by localizing Bicc1 in a network of vidually ventilated cages at the Ecole Polytechnique Fédérale de Lausanne cytoplasmic foci that seem to interact with Dvl2 foci (35). (EPFL) animal facility as described previously (35). Mice homozygous for However, to our knowledge, a role for Bicc1 in localizing target the hypomorphic bpk allele of Bicc1 have been described previously (17, mRNAs has not been evaluated. 21, 34). All animal experiments were approved by the Veterinary Service Novel mechanistic insights into mRNA silencing may come of the Swiss canton of Vaud or by the Institutional Animal Care and Use from structure-function analysis of the Bicc1 SAM domain. SAM Committee and adhered to the guidelines in the Guide for the Care and Use domains are found in numerous regulatory proteins and in some of Laboratory Animals (46). instances mediate the formation of homo- or heterooligomers Cell culture and transfections. HEK293T and COS-1 cells were cul- (37–39). The crystal structure of the SAM domain of the tran- tured in Dulbecco modified Eagle medium (Sigma) supplemented with scriptional repressor TEL established that self-association in a 10% fetal bovine serum (FBS; Sigma), glutamine (1%; Invitrogen), and gentamicin (1%; Invitrogen). Plasmids were transfected using the jetPEI head-to-tail configuration via the so-called midloop (ML) and transfection reagent (Polyplus Transfection) according to the manufac- end helix (EH) surfaces can give rise to a helical polymer (40, 41). turer’s instructions. Scrambled and Lsm1 small interfering RNAs were Since a green fluorescent protein (GFP) fusion of the human transfected 24 h prior to plasmid transfection using the INTERFERin BICC1 SAM domain can polymerize in vitro (42), here we asked transfection reagent (Polyplus Transfection). whether self-association controls the localization or function of Indirect immunofluorescence analysis and histological stainings. full-length Bicc1. To address this question, we modeled the three- For immunostaining, COS-1 cells were transfected with hemagglutinin dimensional (3D) structure of the mouse Bicc1 SAM domain and (HA)-tagged Bicc1 (HA-Bicc1) and GFP-Dcp1a (1 ␮g DNA each) in designed point mutations to disrupt a potential SAM-SAM inter- 6-well plates. After 24 h, the cells were split and grown on sterile coverslips face. We show that such mutations inhibit polymerization of full- in 24-well plates. At 48 h after transfection, the cells were fixed for 10 min Ϫ length Bicc1 and its cytoplasmic clustering. Furthermore, we re- at 20°C in methanol and washed with phosphate-buffered saline (PBS). port that polymerization-competent Bicc1 recruits an associated The coverslips were incubated for1hatroom temperature for blocking in PBS containing 1% bovine serum albumin (BSA) and then for2hatroom = AC6-3 UTR reporter mRNA to cytoplasmic foci and that the temperature in blocking buffer containing the anti-HA primary antibody SAM domain promotes mRNA silencing by mediating Bicc1 po- bpk (1/500, rabbit; Sigma). After washes in PBS, the Alexa Fluor 568-conju- lymerization. bpk mutant Bicc1 (Bicc1 ), which is associated gated anti-rabbit secondary antibody was added, and the mixture was with an ARPKD-like disease, phenocopied these defects, indicat- incubated in blocking buffer for1hatroom temperature in the presence ing that SAM polymerization is likely essential for Bicc1 functions of DAPI (4=,6-diamidino-2-phenylindole; 1/10,000). Pictures were ac- in vivo. quired on a Zeiss LSM700 confocal microscope.

3340 mcb.asm.org Molecular and Cellular Biology October 2015 Volume 35 Number 19 Role of Bicc1 SAM Domain Polymerization

For cryosectioning, organs were collected from Bicc1 mutants and mutC, mutD, and mutE and the ⌬SAM mutants, 2 ␮g DNA was used) or control littermates at postnatal day 2, ﬁxed for4hin4%paraformalde- 10-cm dishes (for Bicc1bpk,8␮g DNA was used). Cell extracts were pre- hyde, soaked in 15% sucrose overnight at 4°C, and embedded in opti- pared as mentioned above for the immunoprecipitation assays. Superna- mum-cutting-temperature (OCT) compound on isopentane. Sections (8 tants were incubated for2hat4°Cwith glutathione-Sepharose 4B beads ␮m) were permeabilized for 10 min at room temperature with PBS, 0.2% saturated with GST-SAM or GST for the negative control. After washing Triton X-100 and then incubated for2hatroom temperature for blocking as described above, the beads were resuspended in Laemmli buffer, frac- in PBS containing 1% BSA. An additional blocking step using a strepta- tionated on SDS-polyacrylamide gels, and analyzed by Western blotting.

vidin-biotin blocking kit (catalog number SP-2002; Vector Laboratories) Anti-HA antibodies (1/1,000, rabbit; Sigma) were used for the detection Downloaded from was added speciﬁcally for biotinylated Lotus tetragonolobus lectin (LTL) of HA-Bicc1. The retention of the GST-SAM bait was monitored directly detection. Primary antibodies anti-Bicc1 SAM (1/300, rabbit) (27), anti- by Coomassie staining. Dcp1a (1/100, mouse; Abnova), anti-CK19 (1/50, human; Santa Cruz Sucrose gradient fractionation assays. HEK293T cells were trans- Biotechnology), and biotinylated LTL (1/200; Vector Laboratories) were fected in 10-cm dishes, using 3 plates per condition (2 ␮g DNA/plate). added and incubated overnight at 4°C. Bicc1 antibody was preabsorbed Cell extracts were prepared as mentioned above for the immunoprecipi- with embryo powder (1/100) prior to incubation. After washing in PBS– tation assays. Continuous 15 to 60% sucrose gradients were prepared 0.1% Triton X-100, secondary antibodies (anti-rabbit antibody–Alexa manually by layering and passive diffusion of sucrose solutions prepared Fluor 488, anti-mouse antibody–Alexa Fluor 647, anti-human antibody– in buffer consisting of 20 mM Tris-HCl, pH 7.4, 100 mM NaCl, 0.05%

Alexa Fluor 568, or streptavidin-Alexa Fluor 568) were incubated in PBS– NP-40. Identical volumes of cell extracts were fractionated at 4°C by cen- http://mcb.asm.org/ ϫ 0.1% Triton X-100 for1hatroom temperature. Nuclei were stained with trifugation at 100,000 g for 3 h. Fractions were recovered manually from DAPI (1/10,000) during the ﬁnal washes. Stained sections were mounted the top, fractionated on SDS-polyacrylamide gels, and analyzed by West- in DABCO (1,4-diazabicyclo[2.2.2]octane). Images were acquired on a ern blotting. Anti-HA antibodies (1/1,000, rabbit; Sigma) were used for ␥ Zeiss LSM700 confocal microscope. the detection of HA-Bicc1. Anti- -tubulin (1/2,000, mouse; Sigma) and Molecular modeling and MD simulation. The X-ray structure of the anti-ribosomal protein S6 (anti-RPS6; 1/1,000, mouse; Cell Signaling) ␥ diacylglycerol kinase ␦1 (DGK␦1) SAM dimer (PDB accession number antibodies were used to detect, respectively, -tubulin and RPS6 proteins, 3BQ7)(47) was identiﬁed with tools from the SWISS-MODEL work space which were used as internal controls. (48). The software MODELLER (v9.5) (49) was used to align the Bicc1 Protein half-life measurements. HEK293T cells were transfected in ␦ 6-well plate dishes (1 ␮g DNA). At 24 h after transfection, cycloheximide

and DGK 1 sequences and to construct 100 models of the Bicc1 SAM on October 6, 2017 by EPFL Scientific information and libraries ␮ domain dimer. The best model, according to the discrete optimized pro- (CHX; Sigma) was added at a final concentration of 100 g/ml. Cells were tein energy (DOPE) (50) score, was set up for molecular dynamics (MD) collected before treatment (0 h) and 8, 24, 32, 48, and 56 h after cyclohex- simulation with a GROMACS (v4.5) simulation tool kit (51). The protein imide addition. Cell extracts were prepared as mentioned above for immunoprecipitation assays. Anti-HA antibodies (1/1,000, rabbit; Sigma) was simulated in water and subjected to energy minimization with the ␥ steepest-descent method (52) and state-of-the-art equilibration. The MD were used for the detection of HA-Bicc1, and anti- -tubulin antibodies (1/2,000, mouse; Sigma) were used for normalization. simulation was then carried out using GROMACS (v4.5) software and the Luciferase assays. HEK293T cells were plated in 24-well plates. Af- CHARMM27 force field (53). The system was simulated for 10 ns in the ter 12 h, quadruplicate samples were transfected with the plasmids isobaric-isothermal ensemble, with the pressure being maintained at 1 indicated below (a 1ϫ dose was 0.1 ␮g/well) and with a lacZ expression atm and the temperature being maintained at 300 K. Final models were vector (0.05 ␮g/well) using jetPEI (Polyplus Transfection). At 36 h rendered in the PyMOL program (Schrödinger). after transfection, cells extracts were prepared in buffer consisting of Electron microscopy. Twenty microliters of protein sample at 0.5 25 mM Tris-phosphate, pH 7.8, 2 mM DTT, 2 mM CDTA (1,2-di- ␮g/␮l was deposited on a carbon-Formvar grid (Ted Pella Inc.), incubated aminocyclohexanetetraacetic acid), 10% glycerol, 0.5% Triton X-100. for 2 min, and blotted dry. The grid was then covered with 20 ␮lof2% The measurements of luciferase expression levels were carried out using uranyl acetate, incubated for 2 min, and blotted dry. Samples were visu- 20-fold-diluted extracts, and luminescent counts were normalized to alized on a JEOL JEM-2200FS transmission electron microscope. ␤-galactosidase activity. Results represent mean values from at least 3 Immunoprecipitation and RT-PCR analyses. To monitor protein- independent experiments performed in quadruplicate, and error bars RNA interactions, immunoprecipitations were conducted with HEK293T ␮ show the standard errors of the means (SEMs). Student’s t test was used to cells transfected in 10-cm dishes (for Bicc1 WT, 1 g DNA was used; for calculate P values. Bicc1 mutD, mutant Bicc1 lacking SAM [Bicc1⌬SAM], and the Bicc1 bpk ␮ mutant, 2 g DNA was used). At 36 h after transfection, cells were washed RESULTS with PBS and resuspended in extraction buffer, consisting of 20 mM Tris- Cytoplasmic clustering of Bicc1 in mouse kidney and liver epi- HCl, pH 7.4, 2.5 mM MgCl2, 100 mM NaCl, 5% glycerol, 1 mM DTT, 0.05% NP-40, RNasin (Promega), and protease inhibitors (Roche). After thelial cells. Bicc1-deficient mice are born with cysts in their kid- a brief sonication and centrifugation at 10,000 ϫ g for 10 min, the super- neys and pancreas and dilated liver bile ducts (25, 27, 54). Con- natants were incubated for2hat4°Cwith anti-HA beads (Sigma) on a cordant with a function in renal morphogenesis, the Bicc1 protein wheel. After five washes of 5 min each in washing buffer, consisting of 20 is expressed in the newborn mouse kidney cortex (25, 27). How-

mM Tris-HCl, pH 7.4, 2 mM MgCl2, 200 mM NaCl, 1 mM DTT, and 0.1% ever, the distribution of Bicc1 in this or other tissues has not been NP-40, RNAs were isolated by phenol-chloroform extraction, followed by resolved at a subcellular level. To address this, we stained frozen ethanol precipitation and RQ1 DNase (Promega) treatment. Reverse sections of postnatal kidneys and livers using a novel custom anti- transcription-PCR (RT-PCR) analyses were carried out using SuperScript Bicc1 antibody that speciﬁcally reacts with the tissues of wild-type II reverse transcriptase (Life Technologies) according to the manufactur- but not Bicc1Ϫ/Ϫ mice (see Fig. S2 in the supplemental material). er’s recommendations. The reverse transcription was performed using the Lotus tetragonolobus lectin (LTL), which is expressed in proximal following primer: CAGTGCAGGGTCCGAGGTATTCGGCCTCTGCGC TTTCTC. The PCR was carried out using the forward primer GAAGAT tubules, was used to mark cortical renal structures, whereas the CGGGTTGAACATGGGTCC and the reverse primer CAGTGCAGGGT cholangiocytes of liver bile ducts were marked by cytokeratin-19 CCGAGGTATTC. (CK19). High-resolution imaging revealed a nonhomogeneous GST pulldown assays. To monitor interactions between GST-SAM distribution of endogenous Bicc1 in clusters of cytoplasmic and HA-Bicc1, HEK293T cells were transfected in 6-well plate dishes (for puncta both in renal proximal tubule cells and in cholangiocytes the Bicc1 WT, mutA, mutB, and mutF, 1 ␮g DNA was used; for Bicc1 (Fig. 1). The volumes of these puncta varied from 0.05 to 0.6 ␮m3,

October 2015 Volume 35 Number 19 Molecular and Cellular Biology mcb.asm.org 3341 Rothé et al. Downloaded from http://mcb.asm.org/

FIG 1 Bicc1 protein forms cytoplasmic clusters in mouse kidney cells and bile duct cholangiocytes. (A) Frozen sections of WT mouse kidney labeled with anti-Bicc1 antibodies and with the proximal tubule marker LTL on postnatal day 4. (B) Frozen sections of liver of a WT mouse obtained postnatally labeled with anti-Bicc1 and anti-CK19 antibodies. CK19 is an intermediate filament protein of epithelial tissues. The boxed areas in the left panels are magnified in the three panels to the right. Bars, 10 ␮m (large views) and 2 ␮m (magnified views). on October 6, 2017 by EPFL Scientific information and libraries similar to those of cytoplasmic foci detected by anti-HA staining reporter RNA was sufficient to translocate the MS2-YFP fusion of tagged Bicc1 in transfected HEK293T cells, although the latter protein from the nucleus to cytoplasmic Bicc1-stained foci inde- were larger, on average, possibly due to overexpression rather than pendently of exogenous Bicc1 (see Fig. S4A in the supplemental alternative fixation or the use of anti-HA antibody instead of anti- material). Even though we could not sufficiently deplete endoge- SAM antibody (see Fig. S3 in the supplemental material). These nous Bicc1 foci by RNA interference to more conclusively validate data show that Bicc1 clusters in cytoplasmic puncta and that this their function (not shown), their colocalization with reporter pattern is not restricted to one tissue but could be a general feature RNA strongly corroborates our conclusion that target mRNAs are of Bicc1. recruited to endogenous Bicc1 in cytoplasmic clusters. Cytoplasmic Bicc1 foci contain target mRNA. Bicc1 cluster- 3D model of Bicc1 SAM-SAM interface. Bicc1 is localized in ing may influence the localization of associated target mRNAs or cytoplasmic foci by its SAM domain independently of the RNA- affect RNA binding, or both. To distinguish between these possi- binding KH domains (35). Certain SAM domains can form bilities, we monitored the localization of a reporter mRNA con- dimers or polymeric structures (57), and a polymer of the human taining the Bicc1 binding region of the AC6 3= UTR (27) and 27 Bicc1 SAM domain fused to GFP has been observed in vitro by repeats of a motif that binds the bacteriophage MS2 coat protein electron microscopy (EM) (42). To test whether SAM polymer- (the MS2ϫ27 motif) (55). Coimmunoprecipitation and luciferase ization is responsible for Bicc1 clustering, we searched for muta- assays with HA-Bicc1 from extracts of transfected HEK293T cells tions that specifically disrupt polymerization. Since structure data showed that Bicc1 specifically binds this reporter mRNA and in- for Bicc1 or its SAM domain are currently unavailable, mutations hibits its expression, as expected (Fig. 2A to C). To assess whether were designed on the basis of homology modeling, where the Bicc1 influences mRNA localization, we cotransfected a fusion known structure of a related protein serves as a template. Among protein of MS2 with nuclear yellow fluorescent protein (YFP) into the available templates, we selected the SAM domain dimer of the COS-1 cells. Like HEK293T cells, COS-1 cells show no endoge- diacylglycerol kinase ␦1 (DGK␦1) E35G (PDB accession number nous Bicc1 expression but are more adhesive, have a larger cyto- 3BQ7) because it shares the highest sequence similarity (54%) and plasm, and, thus, are more amenable for imaging. MS2-YFP with- identity (31%) with the Bicc1 SAM domain (Fig. 3A) and can out Luc-AC6-MS2ϫ27 reporter mRNA localized to nuclei both in form head-to-tail polymers (47). A model of dimeric Bicc1 SAM the presence and in the absence of HA-Bicc1 (Fig. 2D, rows 1 and obtained after energy minimization revealed a common globular 2). However, when coexpressed with Luc-AC6-MS2ϫ27 and fold of five ␣ helices, with two SAM subunits being docked to one without Bicc1, MS2-YFP accumulated diffusely throughout the another at characteristic ML and EH surfaces (Fig. 3B)(40). Res- cytoplasm, confirming that MS2-YFP was exported from the nu- idues involved in the dimerization of the DGK␦1 SAM are con- cleus together with the reporter mRNA (Fig. 2D, row 3). Interest- served or replaced by similar amino acids in the Bicc1 SAM do- ingly, the combined expression of the Luc-AC6-MS2ϫ27 reporter main (highlighted in Fig. 3A). At the Bicc1 SAM-SAM interface, 4 with HA-Bicc1 led to the concentration of MS2-YFP in cytoplas- negatively charged amino acids on the ML surface (Glu900, mic Bicc1 puncta, with 89% Ϯ 8% of the cytoplasmic YFP signal Asp902, Asp913, Glu916) and 5 positively charged amino acids colocalizing with Bicc1 foci (Fig. 2D, row 4). We therefore asked from the EH surface (Lys891, Lys915, Arg925, Arg926, Lys927) whether the Luc-AC6-MS2ϫ27 reporter also enters cytoplasmic form strongly polarized electrostatic networks in two independent foci in the mIMCD3 mouse inner medullary collecting duct cell regions of contact (Fig. 3C and D; see also Fig. S5 in the supple- line that expresses endogenous Bicc1 (56). Transfection of the mental material). In addition, residue Phe922 from the EH surface

3342 mcb.asm.org Molecular and Cellular Biology October 2015 Volume 35 Number 19 Role of Bicc1 SAM Domain Polymerization Downloaded from http://mcb.asm.org/ on October 6, 2017 by EPFL Scientific information and libraries

FIG 2 Bicc1 concentrates an associated reporter mRNA in cytoplasmic foci. (A) Principle of the MS2-YFP colocalization assay. The 3= extremity of the Luc-AC6 reporter mRNA was fused to 27 MS2 hairpins, which constitute multiple binding sites for the MS2 protein fused to YFP. CDS, coding sequence; prox, proximal. (B) RNA coimmunoprecipitation. The Luc-AC6-MS2ϫ27 reporter mRNA was expressed in HEK293T cells together with HA-Bicc1 or empty vector. After HA immunoprecipitation (IP), the various fractions were analyzed by Western blotting and by RT-PCR. Five percent of the total extract was used as the input. The ␤-actin mRNA was used as a negative control for the RT-PCR. HA-Bicc1 lacking all KH and KH-like domains (⌬KH) was used as an additional negative control for RNA-binding specificity. (C) Cotransfection of the fluorescent YFP-MS2 fusion protein does not affect the silencing of the Luc-AC6-MS2ϫ27 reporter by HA-Bicc1. ␤-Galactosidase was used as a control for normalization, and the data represent the percent expression relative to that of a mock-treated control. Error bars show SEMs. *, P Ͻ 0.005. (D) Localization by indirect immunofluorescence staining of the Luc-AC6-MS2ϫ27 reporter mRNA and HA-Bicc1 in COS-1 cells. The MS2-tagged mRNA was detected by the relocalization of fluorescent MS2-YFP fusion protein, which binds the MS2 RNA hairpins. Bars, 5 ␮m. reaches into a hydrophobic pocket of the ML surface comprising wide (see Fig. S7 in the supplemental material). This size resem- Phe896, Ile901, Leu909, and Leu917 (Fig. 3E). bles that of human Bicc1 SAM fused to enhanced GFP (EGFP) To assess the stability of this complex, we performed a molec- (42), corresponds to approximately 40 units of the GST-SAM fu- ular dynamics (MD) simulation in water. For 10 ns, the backbone sion protein, and, therefore, likely represents SAM polymers. root mean square deviation (RMSD) was 2.05 Å. Electrostatic and Overall, our results confirm that the Bicc1 SAM interacts with hydrophobic interactions and minimal root mean square fluctu- itself in higher-order structures large enough to be detected in ation (RMSF) of most of the residues persisted, except at the un- vitro. constrained NH2 and COOH extremities, indicating that the com- To obtain a 3D model of a Bicc1 SAM domain polymer, we plex remained stable (see Fig. S6A and B in the supplemental docked multiple dimers to each other via their EH and ML sur- material). The number of intermolecular H bonds increased dur- faces (Fig. 3F). Polymerization of an arbitrary number of SAM ing the dynamic from 5 to an average of 8.9 H bonds per time units in silico formed a left-handed helix with a diameter of 65 Å frame, indicating significant stabilization (see Fig. S6C in the sup- and a pitch of 44 Å encompassing 6 SAM units. As observed in plemental material). This result validates the confidence of our related SAM polymers (40), the N and C termini were positioned dimeric model. To directly confirm the potential to self-polymer- peripherally, indicating that a central SAM polymer in full-length ize, we imaged a GST fusion of recombinant mouse Bicc1 SAM. Bicc1 is surrounded by the N-terminal KH domains, with the Negative-staining EM of GST-Bicc1 SAM revealed ovoid struc- intervening regions forming a linker to the SAM, and by the short tures that were, on average, 28.8 Ϯ 3.9 nm long and 22.8 Ϯ 3.4 nm C termini (see Fig. S1A in the supplemental material). To evaluate

October 2015 Volume 35 Number 19 Molecular and Cellular Biology mcb.asm.org 3343 Downloaded from http://mcb.asm.org/ on October 6, 2017 by EPFL Scientific information and libraries

FIG 3 Molecular modeling of a Bicc1 SAM polymer. (A) ClustalW alignment of mouse (Mus musculus) Bicc1 (mBicc1) and human (Homo sapiens) DGK␦1 (hDGK␦1) SAM domains. The X-ray structure of dimeric DGK␦1 SAM (PDB accession number 3BQ7) served as the template to model the Bicc1 SAM dimer. The two SAM domains share 31% identity and 54% similarity. Predicted ␣ helices are framed. Residues of the ML surface (red), residues from the EH surface (blue), and hydrophobic residues (underlined) are highlighted. Dark and light gray shading corresponds to identical and similar amino acids, respectively, between Bicc1 and the DGK␦1 template. (B) Dimeric Bicc1 SAM model obtained using MODELLER (v9.5) software. The ␣ helix numbers and the side chains of the residues involved in the interface are displayed. (C to E) Magniﬁed views of the main interacting patches in the predicted Bicc1 SAM dimer interface. Acidic and basic residues are displayed in red and blue, respectively. (F) Model of a Bicc1 SAM polymer of 24 units in surface representation. The NH2 terminus and the COOH terminus of each SAM domain are displayed in green and purple, respectively. (G) Model of the Bicc1 KH domain region in surface representation. Models for individual KH domains were obtained by homology modeling using the SWISS-MODEL work space (48) and templates consisting of the structures with PDB accession numbers 1VIG (KH1), 2CTM (KH2), 1WVN (KH3), and 3N89 (KHL1 and -2). Individual KH domains were then superimposed with their homologous domain in the X-ray structure of ceGLD-3 KH domains (PDB accession number 3N89)(58). The KH domains harboring the GXXG signatures for RNA binding are highlighted in color. Their putative RNA-binding surfaces are darkened, and the identity of their GXXG signature sequence is given in parentheses. The KH-like domains (KHL1 and -2) are displayed in gray. (H) Diagram of a transversal section through a polymer of full-length Bicc1. The SAM polymer is located at the center and displays other Bicc1 domains at its periphery. A schematic representation was used for the other domains. C-ter domain, C-terminal domain. (I) Diagram in longitudinal view of Bicc1 KH domains distributed along the surface of the central SAM polymer.

3344 mcb.asm.org Molecular and Cellular Biology October 2015 Volume 35 Number 19 Role of Bicc1 SAM Domain Polymerization the potential impact of SAM polymerization on RNA binding, we ribosomes (Fig. 5A). A similar distribution was observed for en- also modeled the structures of the Bicc1 KH and KHL domains dogenous Bicc1 in extracts of mIMCD3 cells (see Fig. S4B in the using the Caenorhabditis elegans GLD-3 (ceGLD-3) X-ray struc- supplemental material). In contrast, HA-Bicc1 mutant D was con- ture as a template (58)(Fig. 3G). We found that all three RNA- centrated in fractions with significantly lower molecular weights binding pockets are predicted to reside on the same surface of a in three independent experiments. These data suggest that at least large globular domain. For optimal solvent exposure and access to the largest Bicc1 assemblies in such cell extracts likely depend on RNA, the KH domains of polymeric full-length Bicc1 would thus SAM polymerization. However, despite this marked shift to low- have to reside around and peripheral to the central helix organized er-molecular-weight fractions, HA-Bicc1 mutD levels in the first Downloaded from by the SAM domain (Fig. 3H and I), consistent with their N-ter- three fractions did not increase, suggesting that Bicc1 likely fails to minal positioning relative to the SAM domain (Fig. 3F). Each stably accumulate as a free monomer. helical turn might thus present as many as 25 KH domains at the To assess the impact of polymerization on Bicc1 function, we surface of the macromolecular assembly within a distance of only tested whether Bicc1 mutant D forms cytoplasmic puncta. As de- 44 Å. While the exact configuration of such a multimer awaits scribed previously, immunostaining of COS-1 cells and recon- experimental validation, we note that the regulation of its length struction of 3D surfaces detected HA-Bicc1 in a dense network of and helical pitch by SAM polymerization thus likely determines corpuscles outside and around P-bodies marked by GFP-Dcp1a the number of RNA-binding sites as well as their spacing along the (35). In contrast, Bicc1 mutD localized more diffusely throughout http://mcb.asm.org/ longitudinal axis (Fig. 3I). the cytoplasm, similar to the findings for Bicc1⌬SAM (Fig. 5B and Identification of polymerization-defective Bicc1 mutants. C). Importantly, mutant D also failed to localize the Luc-AC6- To test the predictive power of our structural model of dimeric MS2ϫ27 reporter in cytoplasmic foci, even though mRNA bind- Bicc1 SAM, we individually replaced six surface-exposed patches ing was still detected (Fig. 5C; see also Fig. S8 in the supplemental of electrostatic amino acids within or outside the SAM-SAM in- material). Taken together, these results suggest that both Bicc1 terface by alanines (Fig. 4A). Four of these groups of mutations and associated mRNAs are concentrated in cytoplasmic foci by (those in mutB, mutC, mutD, and mutE) should affect intermo- specific residues in the SAM domain that mediate polymerization.

lecular H bonds, while two others (those in mutA and mutF) were SAM polymerization is essential for Bicc1 stabilization and on October 6, 2017 by EPFL Scientific information and libraries deliberately introduced outside the predicted dimerization inter- to silence mRNAs but not to inhibit the Wnt pathway. Since the face (Fig. 4B to D). To screen for polymerization defects, we trans- Bicc1 mutant D apparently did not freely accumulate as a mono- fected HA-tagged Bicc1 mutants A to F into HEK293T cells and mer in density fractionation gradients, we asked whether polym- assessed their retention on glutathione-Sepharose beads coated erization influences the protein half-life. Translation inhibition by with recombinant GST-SAM fusion protein as a bait. We found cycloheximide (CHX) followed by a time course analysis by im- that wild-type HA-Bicc1 efficiently bound to GST-SAM-coated munoblotting showed that wild-type HA-Bicc1 remained stable beads but not to beads coated with GST alone (Fig. 4E). In con- over the entire chase period (56 h). In contrast, the half-lives of the trast, the Bicc1 mutant lacking SAM (HA-Bicc1⌬SAM) failed to HA-Bicc1 mutD and Bicc1⌬SAM proteins were reduced to 34 and bind GST-SAM beads, suggesting that GST-SAM specifically in- 29 h, respectively (Fig. 6A), suggesting that SAM polymerization teracts with the SAM domain of full-length Bicc1. Moreover, a increases Bicc1 stability. Therefore, in all subsequent assays eval- comparison with HA-Bicc1 mutants showed that while the muta- uating their functions, we doubled the dose of transfected DNA tions in mutants C, D, and E abolished the SAM-SAM interaction, for Bicc1 mutants in order to reach expression levels comparable the mutations in mutants A and F did not, a result that concurs to those of wild-type HA-Bicc1 (Fig. 6B). In particular, to deter- with that of our structure model. The mutation in mutant B ex- mine whether polymerization is necessary for mRNA silencing, we hibited an intermediate effect. The mutation in mutant B affected compared the potential of wild-type HA-Bicc1 and mutant D to three amino acids at the periphery of the EH surface, and only one silence the 3= UTR sequences of AC6 and PKI␣ mRNAs in lucif- of them (K891) is predicted to contribute to the SAM-SAM inter- erase reporter assays. HA-Bicc1 significantly repressed the expres- action (Fig. 3C). According to theoretical predictions, this electro- sion of these luciferase reporters, as described previously (27). In static patch is involved in only 1.1 intermolecular H bonds, on contrast, both HA-Bicc1⌬SAM and HA-Bicc1 mutD failed to re- average, explaining why its mutation is insufficient to fully disrupt press these targets even at an elevated dosage (Fig. 6C and D), dimerization (Fig. 4D). Altogether, these results strongly corrob- indicating that the silencing activity of Bicc1 directly depends on orate our structure model and provide new tools to specifically SAM polymerization. probe the consequences of the loss of SAM-SAM interactions in Previously, we have shown that Bicc1 can also block the induc- vivo. tion of the TOPflash reporter of canonical Wnt signaling at the Polymerization governs Bicc1 clustering and localizes asso- level of Dishevelled 2. This inhibition of Dishevelled signaling is ciated mRNA in cytoplasmic puncta. Among our panel of Bicc1 independent of the mRNA-binding KH domains but is potenti- mutations, the mutation in mutant D disrupted the largest num- ated by the Bicc1 SAM domain (35). To distinguish whether inhi- ber of intermolecular H bonds at the SAM-SAM interface (Fig. 4C bition of Dishevelled 2 activity directly or indirectly depends on and D). To distinguish whether this interface mediates Bicc1 Bicc1 polymerization, we cotransfected wild-type or mutant HA- dimerization or the formation of higher-order assemblies, we Bicc1 with TOPflash and a vector expressing Dvl2. As described compared the sizes of wild-type Bicc1 and mutant D in transfected previously (35), HA-Bicc1⌬SAM inhibited the induction of HEK293T cells by sucrose gradient fractionation. Analysis of poly- TOPflash by Dvl2 less efficiently than wild-type Bicc1. Interest- meric HA-Bicc1 complexes in HEK293T cells revealed a broad size ingly, this difference disappeared when the dosage of either HA- distribution, with wild-type HA-Bicc1 extending beyond the frac- Bicc1⌬SAM or mutant D was doubled to reach the level of wild- tions marked by ribosomal protein S6 (RPS6) (59), indicating that type HA-Bicc1 expression (Fig. 6B and E), suggesting that SAM wild-type Bicc1 congregates in molecular assemblies larger than domain polymerization is not essential for Wnt signal inhibition.

October 2015 Volume 35 Number 19 Molecular and Cellular Biology mcb.asm.org 3345 Rothé et al. Downloaded from http://mcb.asm.org/ on October 6, 2017 by EPFL Scientific information and libraries

FIG 4 Screen for Bicc1 SAM polymer mutants. (A) Bicc1 SAM mutant collection. Shading is as for Fig, 3A. Individual electrostatic patches (patches A to F) at the protein surface were replaced by alanines. (B, C) Positions of mutations in the Bicc1 SAM dimer model (B) and on the ML surface and EH surface (C). (D) Table summarizing, for each amino acid patch, the average number of H bonds per time frame during the MD. (E and F) Pulldown of the WT, point mutants, or 8-fold excess bpk mutant HA-tagged Bicc1 from HEK293T cell extracts by glutathione-Sepharose beads coated with a recombinant GST control or GST-Bicc1 SAM. Five percent of total cell extracts were loaded as input.

Inhibition of SAM polymerization and Bicc1 function by the evaluate whether the bpk mutation also inhibits SAM-SAM inter- bpk mutation. Our model of the SAM domain raised the possibil- actions. To compensate for reduced protein stability, the HA- ity that elongation of the C terminus in the Bicc1 bpk mutant Bicc1 bpk mutant expression vector was transfected in large ex- inhibits a polymeric SAM helix by overcrowding the surrounding cess. Nevertheless, the HA-Bicc1 bpk mutant failed to bind space (see Fig. S1B in the supplemental material). To test this recombinant GST-SAM fusion protein in pulldown assays (Fig. hypothesis, we cloned the bpk mutant in a mammalian expression 4F). HA-Bicc1 bpk also failed to cluster around P-bodies (Fig. 5B) vector. Similar to our engineered mutant D, the bpk mutant dis- and to silence the AC6 and PKI␣ 3= UTR luciferase reporters played a drastically reduced half-life compared to that of wild-type (Fig. 6C and D). In contrast, mRNA binding was still detected (see HA-Bicc1 (Fig. 6A and B). We therefore used pulldown assays to Fig. S8 in the supplemental material). In proportion to its expres-

3346 mcb.asm.org Molecular and Cellular Biology October 2015 Volume 35 Number 19 Role of Bicc1 SAM Domain Polymerization Downloaded from http://mcb.asm.org/ on October 6, 2017 by EPFL Scientific information and libraries

FIG 5 SAM polymerization is required for Bicc1 clustering. (A) Density fractionation of WT and polymerization mutant Bicc1 on a sucrose gradient. HEK293T cell extracts containing HA-tagged Bicc1 were fractionated on a continuous 15 to 60% sucrose gradient and analyzed by anti-Bicc1 Western blotting. The migration direction from the top to the bottom of the tube is indicated. RPS6 and ␥-tubulin (␥-Tub) were used as internal controls. The graph below the gels shows the percentage of Bicc1 compared to the total Bicc1 signal for each fraction. Results represent mean values from 3 independent experiments, and error bars show SEMs. (B) Bicc1 polymer mutants fail to accumulate in cytoplasmic foci. The results of indirect immunofluorescence staining of the HA-Bicc1 WT, mutD, the ⌬SAM mutant, or the bpk mutant and the P-body marker GFP-Dcp1a overexpressed in COS-1 cells are shown. Bars, 5 ␮m. (C) Comparative 3D rendering of the HA-Bicc1 WT and mutD by Imaris software. From the original image (center), z-stacks in two directions (z1 and z2, top and right, respectively) and3D reconstruction (bottom) are given. Bars, 2 ␮m. (D) Localization by indirect immunofluorescence staining of the Luc-AC6-MS2ϫ27 reporter mRNA and HA-Bicc1 in COS-1 cells and comparison with that of HA-Bicc1 mutD. Bars, 5 ␮m. sion levels, HA-Bicc1 bpk also still inhibited Dvl2 signaling in both LTL-positive and LTL-negative renal tubule cells, the cystic TOPflash assays (Fig. 6B and E). Overall, these data show that the kidneys of bpk mutant animals showed considerably weaker and C-terminal elongation of the Bicc1 bpk mutant abrogates SAM more diffuse Bicc1 staining that was confined to LTL-positive polymerization and mRNA silencing without disrupting the po- structures (Fig. 7A). This result concurs with the effects of bpk on tential to inhibit Dishevelled 2. Bicc1 protein stability and cytoplasmic clustering observed ex To validate a potential influence of the bpk mutation on poly- vivo. However, despite the abnormally diffuse distribution, some meric Bicc1 in vivo, cryosections of kidneys obtained from bpk Bicc1 foci remained clearly detectable. To evaluate whether these mice postnatally were colabeled using anti-Bicc1 antibody and residual foci could arise from the alternatively spliced transcript B LTL to mark proximal tubules. Compared to the findings for wild- that is unaffected by the bpk mutation (17, 22) (see Fig. S1C in the type Bicc1, where Bicc1 was concentrated in cytoplasmic foci of supplemental material), we overexpressed a synthetic cDNA of

October 2015 Volume 35 Number 19 Molecular and Cellular Biology mcb.asm.org 3347 Rothé et al. Downloaded from http://mcb.asm.org/ on October 6, 2017 by EPFL Scientific information and libraries

FIG 6 SAM polymerization is required for Bicc1 accumulation and silencing activity. (A) Time course analysis of HA-Bicc1 WT and mutants expressed in HEK293T cells after CHX treatment. HA-Bicc1 protein levels were compared with ␥-tubulin levels by Western blotting at 0, 8, 24, 32, 48, and 56 h. The relative level of each protein is presented in the graph at the bottom, and the estimated half-life is given on the right. AU, arbitrary units. (B) Level of the HA-Bicc1 WT and mutants upon transfection with a single dose (ϫ1) or a double dose (ϫ2) of DNA encoding HA-Bicc1. A double dose of transfected DNA encoding HA-Bicc1 mutD, the ⌬SAM mutant, or the bpk mutant is required to obtain a protein level comparable to that of the HA-Bicc1 WT. The relative percentage of WT Bicc1 is indicated for each condition. ␥-Tubulin was used for normalization. (C and D) Silencing of AC6 and PKI␣ 3= UTR luciferase reporters by WT or polymerization mutant Bicc1 in HEK293T cells. ␤-Galactosidase was used as a control for normalization. Error bars show SEMs. *, P Ͻ 0.005. (E) Induction of the TOPflash reporter of Wnt signaling by Dishevelled 2 (Dvl2) in transfected HEK293T cells is inhibited by both WT and polymerization mutant Bicc1. ␤-Galac- tosidase was cotransfected for signal normalization. Error bars show SEMs. *, P Ͻ 0.005. this splice variant in transfected cells. Similar to the findings for attenuated by polymerization-independent Bicc1 functions, e.g., full-length Bicc1 encoded by transcript A, transfected isoform B during Wnt signaling. was readily detected by anti-Bicc1 antibody in cytoplasmic foci, and it efficiently silenced the Luc-AC6 reporter mRNA (Fig. 7B DISCUSSION and C). It is likely, therefore, that the bpk mutant phenotype is By combining structure modeling and mutagenesis with func- caused at least in part by the failure of the affected Bicc1 isoform A tional analysis, we identified a self-polymerization interface in the to polymerize. In addition, these results are consistent with the C-terminal SAM domain of Bicc1 and showed that it stabilizes notion that isoform B remains functional and thus likely accounts full-length Bicc1 in higher-order complexes to regulate its subcel- for residual hypomorphic Bicc1 activity in bpk mutants. Alterna- lular distribution in the cytoplasm. We provide the first evidence tively, or in addition, cystic growth in bpk mutants may also be that Bicc1 also localizes a reporter mRNA that binds the N-termi-

3348 mcb.asm.org Molecular and Cellular Biology October 2015 Volume 35 Number 19 Role of Bicc1 SAM Domain Polymerization Downloaded from http://mcb.asm.org/ on October 6, 2017 by EPFL Scientific information and libraries

FIG 7 Impaired localization of Bicc1 in bpk mutant mouse kidney. (A) Frozen sections of WT and bpk mutant mouse kidneys labeled with anti-Bicc1 antibodies and with the proximal tubule marker LTL on postnatal day 11. Boxed areas are magnified in the panels to the right. Bars, 20 ␮m (large views) and 10 ␮m (magnified views). (B) Bicc1 isoform B (Bicc1-B) accumulates in cytoplasmic foci. The results of indirect immunofluorescence staining of HA-Bicc1 isoforms A and B and the P-body marker GFP-Dcp1a in transfected COS-1 cells are shown. Bars, 5 ␮m. (C) Silencing of AC6 3= UTR luciferase reporter by HA-Bicc1 isoform A or B in HEK293T cells. ␤-Galactosidase was cotransfected for signal normalization. Error bars show SEMs. *, P Ͻ 0.005. nal KH domains and that this process is regulated by SAM clus- ANKS3), 10% can oligomerize (e.g., ANKS6), and 43% are mo- tering. The helical conformation of the polymeric SAM implies nomeric (e.g., Smaug/Vts1p) (42). A purified EGFP fusion protein that RNA-binding sites are hyperconcentrated in orderly arrays in of the human BICC1 SAM was shown by EM to self-polymerize the surrounding space. The loss of SAM polymerization may ex- (42). Here, EM analysis of an analogous mouse Bicc1 SAM fusion plain why the Bicc1 function is impaired in the bpk mouse model with GST revealed the presence of similar structures of approxi- of autosomal recessive polycystic kidney disease. Regulation of mately 40 nm. Extending this observation, we used GST pulldown mRNA localization and silencing by SAM polymerization to our and sucrose gradient fractionation assays to show that the capacity knowledge has not been reported previously. to self-interact in higher-order structures is preserved in full- Structural model of the SAM-SAM interface. SAM domains length Bicc1. The high sequence conservation (31% identity and are found in many eukaryotic proteins with diverse functions. 54% similarity) between the Bicc1 and DGK␦1 SAM domains al- Among 72 known human SAM domains that have previously lowed us to derive a homology-based 3D model of dimeric Bicc1 been analyzed in isolation for their potential to multimerize in SAM. In this model, the EH surface from one Bicc1 SAM domain vitro, 47% form polymers (e.g., TEL, Ph, Shank3, DGK␦1, and interacts with the ML surface of another through a large electro-

October 2015 Volume 35 Number 19 Molecular and Cellular Biology mcb.asm.org 3349 Rothé et al. static network that is stabilized by 8 hydrogen bonds during mo- either bpk or the mutation engineered in mutant D or upon dele- lecular dynamic simulation and by Phe922 interacting with a hy- tion of the entire SAM domain. Imaging of the reporter mRNA drophobic pocket comprising residues L917, L909, I901, and using an MS2-YFP fusion protein revealed extensive colocaliza- F896. Confirming the robustness of our model, alanine scanning tion with wild-type Bicc1 in cytoplasmic foci, whereas in cells mutagenesis of 6 distinct surface-exposed electrostatic clusters lacking Bicc1, it was diffusely distributed throughout the cyto- showed that only the mutations at the predicted SAM-SAM inter- plasm. Interestingly, the reporter mRNA also colocalized with the face impaired polymerization. The determinants mediating this engineered SAM polymerization mutant D, albeit it did so dif- interaction are conserved in homologous SAM domains, consis- fusely throughout the cytoplasm. We conclude that SAM polym- Downloaded from tent with an important biological function (see Fig. S1D in the erization is not essential for RNA binding but is essential to con- supplemental material). centrate associated transcripts in cytoplasmic foci. The SAM domain organizes a helical Bicc1 polymer that is The SAM-SAM interface is required for silencing of Bicc1 inhibited by the bpk mutation. A model of polymerized Bicc1 via target mRNAs. Previous work established that truncated Bicc1 a SAM-SAM interface built in silico led intrinsically to the forma- lacking the SAM domain still binds target mRNAs and miRNAs tion of a left-handed head-to-tail helix with 6 SAM units per turn. yet fails to induce RNA-induced silencing complex loading and Since the N and C termini of each SAM subunit protrude from a silencing (27). Here, we showed that silencing is also abrogated by helical polymer toward the periphery, attached domains occupy mutations that specifically disrupt the SAM-SAM interface (mu- http://mcb.asm.org/ the surrounding space in regular arrays (Fig. 3F to I). The putative tant D) or its potential to polymerize (bpk mutant). These obser- RNA-binding GXXG signatures that distinguish the N-terminal vations provide strong evidence that the SAM domain promotes KH from KH-like domains are thus likely displayed at the periph- silencing by mediating Bicc1 polymerization. In contrast, polym- ery and linked to the central SAM scaffold by the intervening erization was not essential for mRNA binding, indicating that spacer region. This organization suggests that the space near the Bicc1 likely binds target mRNAs before entering polymers. Since center may be too crowded to accommodate bulky additions at mRNA binding to polymerization-deficient Bicc1 was not suffi- the C terminus or other factors in a polymer. Direct evidence in cient for silencing, the transition to a polymeric configuration

support of such a model comes from our analysis of the Bicc1 bpk may be the molecular switch that turns off mRNA translation. on October 6, 2017 by EPFL Scientific information and libraries mutant, where replacement of the short wild-type C terminus by a Such a model predicts that polymerization is likely regulated. In- sequence approximately five times larger was sufficient to disrupt terestingly, a heterozygous mutation in human BICC1 that asso- the ability of Bicc1 to self-interact. The simplest possible explana- ciates with pediatric renal cystic dysplasia replaces the conserved tion is that the C-terminal extension of Bicc1bpk sterically hinders residue E932 in the SAM domain by glycine. Since E932 is outside SAM polymerization. the SAM-SAM interface and dispensable for cytoplasmic cluster- Bicc1 SAM polymerization is not essential to inhibit Dvl ac- ing and Bicc1 silencing activity (19), we speculate that it controls tivity but is necessary to stabilize cytoplasmic foci and to locally access to a steric inhibitor of Bicc1 polymerization. The extent or concentrate target mRNA. A previous analysis suggested that a dynamics of SAM polymerization could also be regulated by other C-terminal fragment encompassing the SAM domain localizes SAM domain proteins, such as ANKS6. ANKS6 binds and colo- Bicc1 in cytoplasmic puncta, but the underlying mechanism re- calizes with Bicc1 (60) and is needed to suppress renal cysts (61, mained elusive (35). Here, immunostaining of cryosections re- 62). Although ANKS6 was initially considered to self-polymerize vealed that endogenous Bicc1 forms similar clusters in renal tu- (42), a more recent study showed that it inhibits polymerization of bule cells and in liver cholangiocytes. We propose that clustering the related SAM domain protein ANKS3, which can also interact in cytoplasmic foci is mediated by SAM polymerization, since with Bicc1 (63, 64). both the engineered polymerization mutant D and Bicc1bpk were Between the KH domains and SAM, Bicc1 contains an inter- diffusely distributed in transfected cells compared to the distribu- vening sequence of 461 residues. Interestingly, the homologous tion of wild-type Bicc1. In good agreement with that finding, clus- sequence of Xenopus Bic-C is sufficient to mediate translational tering of Bicc1 in cytoplasmic foci was also impaired in the kidneys repression when tethered as an MS2 fusion protein to a luciferase of mice with the bpk mutation in vivo. In addition, failure to po- reporter mRNA containing MS2 hairpins, and addition of the lymerize reduced the Bicc1 protein half-life, suggesting that po- SAM domain potentiated this activity (65). In contrast, tethering lymerization may influence some Bicc1 functions indirectly. Con- of KH or SAM domains alone had no effect. These findings sup- firming this prediction, polymerization-deficient Bicc1⌬SAM, port our model that SAM primarily serves to cluster and stabilize Bicc1bpk, and engineered mutant D all inhibited the induction of Bicc1 in polymers. They also concur with our earlier finding that the TOPflash reporter by Dvl2 less potently than wild-type Bicc1, mRNA binding to Bicc1 KH domains alone is not sufficient for but this defect was rescued when the dosage of mutant proteins translational repression (27). Additional factors may include was adapted to restore normal expression levels. Polymeric Bicc1 miRISC components (27) or binding of unknown factors to the therefore likely facilitates the inhibition of Dvl2 activity indirectly region between KH and SAM domains, depending on the context by increasing Bicc1 stability. (65). How these or other factors cooperate with SAM polymeriza- Our findings show that SAM polymerization also influences tion to repress translation will be a fascinating novel area of future the localization or binding of Bicc1-associated RNAs. Previous investigation. experiments established that Bicc1 coimmunoprecipitates AC6 How does polymerization influence Bicc1 silencing activity? and PKI␣ mRNAs and specific miRNAs independently of the One possible mechanism to promote translational repression by SAM domain, whereas deletion of KH domains blocked these in- Bicc1 clustering involves the local enrichment of target mRNAs in teractions (27). In good agreement with those findings, here Bicc1 microdomains that privilege the access to specific silencing factors also bound an AC6-3= UTR luciferase reporter that was tagged through molecular crowding. Crowding influences the thermody- with MS2 hairpins, and this interaction was not overtly altered by namic properties of cellular components, and hyperconcentra-

3350 mcb.asm.org Molecular and Cellular Biology October 2015 Volume 35 Number 19 Role of Bicc1 SAM Domain Polymerization tion, especially of macromolecules, can alter reaction equilibrium tung (to L.L.-E.) and by a Swiss National Science Foundation Sinergia constants by increasing the probability that substrates will meet grant to D.B.C. and by decreasing their conformational entropy (66). In a catalytic model, polymeric Bicc1 may thus thermodynamically favor inter- REFERENCES actions with silencing factors by locally increasing the concentra- 1. Marchand V, Gaspar I, Ephrussi A. 2012. An intracellular transmission tion of associated mRNAs and/or by presenting them in an orderly control protocol: assembly and transport of ribonucleoprotein com- manner, creating, in effect, a catalytic surface for translational plexes. Curr Opin Cell Biol 24:202–210. http://dx.doi.org/10.1016/j.ceb repression. .2011.12.014. Downloaded from 2. Kim-Ha J, Smith JL, Macdonald PM. 1991. oskar mRNA is localized to An alternative, mutually nonexclusive hypothesis is that Bicc1 the posterior pole of the Drosophila oocyte. Cell 66:23–35. http://dx.doi acts as an RNA chaperone. A chaperone model takes into account .org/10.1016/0092-8674(91)90136-M. the structural dimension and striking degree of symmetry of the 3. Ephrussi A, Dickinson LK, Lehmann R. 1991. oskar organizes the germ Bicc1 polymer. The pitch of helical SAM polymers arranges Bicc1 plasm and directs localization of the posterior determinant nanos. Cell 66:37–50. http://dx.doi.org/10.1016/0092-8674(91)90137-N. KH domains such that each individual KH domain n is aligned 4. Ephrussi A, Lehmann R. 1992. Induction of germ cell formation by oskar. with the related n ϩ 6 domain. Six rows of a given KH domain thus Nature 358:387–392. http://dx.doi.org/10.1038/358387a0. cover the polymer along its length. With 3 KH domains per Bicc1 5. Kim-Ha J, Kerr K, Macdonald PM. 1995. Translational regulation of molecule that can bind RNA, a polymer will present these do- oskar mRNA by bruno, an ovarian RNA-binding protein, is essential. Cell http://mcb.asm.org/ mains in 18 regularly spaced rows at the surface. Compared to the 81:403–412. http://dx.doi.org/10.1016/0092-8674(95)90393-3. 6. Markussen FH, Michon AM, Breitwieser W, Ephrussi A. 1995. Trans- binding of monomeric Bicc1, the surface of a polymer is more lational control of oskar generates short OSK, the isoform that induces likely to bind several RNA molecules simultaneously and/or at pole plasma assembly. Development 121:3723–3732. multiple anchoring points and, thus, to favor alternative second- 7. St Johnston D. 2005. Moving messages: the intracellular localization of ary RNA structures that expose or mask speciﬁc binding sites of mRNAs. Nat Rev Mol Cell Biol 6:363–375. http://dx.doi.org/10.1038 /nrm1643. cognate miRNAs or regulatory proteins. The accessibility of 8. Ghosh S, Marchand V, Gaspar I, Ephrussi A. 2012. Control of RNP miRNA seed sequences is essential for silencing and can be regu- motility and localization by a splicing-dependent structure in oskar

lated by specific RNA-binding proteins, such as Pumilio (67–69). mRNA. Nat Struct Mol Biol 19:441–449. http://dx.doi.org/10.1038/nsmb on October 6, 2017 by EPFL Scientific information and libraries In a chaperone model, the conformation of an mRNA that is ini- .2257. 9. Kim G, Pai CI, Sato K, Person MD, Nakamura A, Macdonald PM. 2015. tially captured by monomeric Bicc1 will change its equilibrium Region-specific activation of oskar mRNA translation by inhibition of under the influence of nearby KH domains and possibly other Bruno-mediated repression. PLoS Genet 11:e1004992. http://dx.doi.org RNAs on the surface of multimeric Bicc1 after SAM-SAM polym- /10.1371/journal.pgen.1004992. erization. The identity of RNA recognition motifs, their number, 10. Kugler JM, Lasko P. 2009. Localization, anchoring and translational and their relative positioning in space could thus govern the si- control of oskar, gurken, bicoid and nanos mRNA during Drosophila oogenesis. Fly 3:15–28. http://dx.doi.org/10.4161/fly.3.1.7751. lencing by Bicc1 by a complex code. 11. Bouvrette DJ, Price SJ, Bryda EC. 2008. K homology domains of the Our results demonstrate that the SAM domain mediates the mouse polycystic kidney disease-related protein, Bicaudal-C (Bicc1), me- polymerization of full-length Bicc1. We found that this process is diate RNA binding in vitro. Nephron Exp Nephrol 108:e27–e34. http://dx inhibited in the spontaneously mutated Bicc1 bpk allele and .doi.org/10.1159/000112913. 12. Hollingworth D, Candel AM, Nicastro G, Martin SR, Briata P, Gherzi needed to enrich Bicc1 and associated mRNA in cytoplasmic clus- R, Ramos A. 2012. KH domains with impaired nucleic acid binding as a ters. SAM-dependent polymerization of the unrelated protein tool for functional analysis. Nucleic Acids Res 40:6873–6886. http://dx polyhomeotic-like 2 induces clustering of the polycomb group- .doi.org/10.1093/nar/gks368. repressive complex 1 on chromatin to repress the transcription of 13. Mahone M, Saffman EE, Lasko PF. 1995. Localized Bicaudal-C RNA Hoxb and other target genes (70). SAM-SAM interactions also encodes a protein containing a KH domain, the RNA binding motif of FMR1. EMBO J 14:2043–2055. control the activity of the transcriptional repressor Yan (71). Ex- 14. Saffman EE, Styhler S, Rother K, Li W, Richard S, Lasko P. 1998. tending this portfolio, our data reveal that SAM domain polym- Premature translation of oskar in oocytes lacking the RNA-binding pro- erization can also be used by an RNA-binding protein to localize tein Bicaudal-C. Mol Cell Biol 18:4855–4862. target mRNAs. This feature might become useful to manipulate 15. Kugler JM, Chicoine J, Lasko P. 2009. Bicaudal-C associates with a Trailer Hitch/Me31B complex and is required for efficient Gurken other RNA-binding proteins that normally contain no SAM do- secretion. Dev Biol 328:160–172. http://dx.doi.org/10.1016/j.ydbio mains. Further studies and new experimental approaches are also .2009.01.024. warranted to further investigate the influence of polymerization 16. Chicoine J, Benoit P, Gamberi C, Paliouras M, Simonelig M, Lasko P. on the spatial positioning of the RNA-binding KH domains or on 2007. Bicaudal-C recruits CCR4-NOT deadenylase to target mRNAs and the conformation of bound RNAs and its potential regulation by regulates oogenesis, cytoskeletal organization, and its own expression. Dev Cell 13:691–704. http://dx.doi.org/10.1016/j.devcel.2007.10.002. Bicc1-interacting SAM domain proteins, such as ANKS6 or 17. Cogswell C, Price SJ, Hou X, Guay-Woodford LM, Flaherty L, Bryda ANKS3 (60, 64). EC. 2003. Positional cloning of jcpk/bpk locus of the mouse. Mamm Genome 14:242–249. http://dx.doi.org/10.1007/s00335-002-2241-0. 18. Tran U, Pickney LM, Ozpolat BD, Wessely O. 2007. Xenopus Bicau- ACKNOWLEDGMENTS dal-C is required for the differentiation of the amphibian pronephros. Dev We thank Paul Guichard and Pierre Gönczy Pierre (EPFL) and Xavier Biol 307:152–164. http://dx.doi.org/10.1016/j.ydbio.2007.04.030. Manival (Université Lorraine-Nancy) for their advice on negative-stain- 19. Kraus MR, Clauin S, Pfister Y, Di Maio M, Ulinski T, Constam D, ing EM and for comments on the manuscript and Jessica Dessimoz and Bellanne-Chantelot C, Grapin-Botton A. 2012. Two mutations in human BICC1 resulting in Wnt pathway hyperactivity associated with cystic renal Arne Seitz (EPFL) and their technical staff for assistance with histology dysplasia. Hum Mutat 33:86–90. http://dx.doi.org/10.1002/humu.21610. and imaging. Computations were performed at Vital-IT (http://www.vital 20. Flaherty L, Messer A, Russell LB, Rinchik EM. 1992. Chlorambucil- -it.ch) of the SIB Swiss Institute of Bioinformatics. induced mutations in mice recovered in homozygotes. Proc Natl Acad Sci This work was generously supported by grants from the Stiftung für USA89:2859–2863. http://dx.doi.org/10.1073/pnas.89.7.2859. Wissenschaftliche Forschung Zürich (to B.R.) and the Gebert-Rüf Stif- 21. Nauta J, Ozawa Y, Sweeney WE, Jr, Rutledge JC, Avner ED. 1993. Renal

October 2015 Volume 35 Number 19 Molecular and Cellular Biology mcb.asm.org 3351 Rothé et al.

and biliary abnormalities in a new murine model of autosomal recessive 41. Tran HH, Kim CA, Faham S, Siddall MC, Bowie JU. 2002. Native polycystic kidney disease. Pediatr Nephrol 7:163–172. http://dx.doi.org interface of the SAM domain polymer of TEL. BMC Struct Biol 2:5. http: /10.1007/BF00864387. //dx.doi.org/10.1186/1472-6807-2-5. 22. Ryan S, Verghese S, Cianciola NL, Cotton CU, Carlin CR. 2010. Autosomal 42. Knight MJ, Leettola C, Gingery M, Li H, Bowie JU. 2011. A human recessive polycystic kidney disease epithelial cell model reveals multiple baso- sterile alpha motif domain polymerizome. Protein Sci 20:1697–1706. http: lateral epidermal growth factor receptor sorting pathways. Mol Biol Cell 21: //dx.doi.org/10.1002/pro.703. 2732–2745. http://dx.doi.org/10.1091/mbc.E09-12-1059. 43. Korinek V, Barker N, Morin PJ, van Wichen D, de Weger R, Kinzler 23. Kotsis F, Boehlke C, Kuehn EW. 2013. The ciliary ﬂow sensor and KW, Vogelstein B, Clevers H. 1997. Constitutive transcriptional activa- Ϫ/Ϫ polycystic kidney disease. Nephrol Dial Transplant 28:518–526. http://dx tion by a beta-catenin-Tcf complex in APC colon carcinoma. Science Downloaded from .doi.org/10.1093/ndt/gfs524. 275:1784–1787. http://dx.doi.org/10.1126/science.275.5307.1784. 24. Wang S, Zhang J, Nauli SM, Li X, Starremans PG, Luo Y, Roberts KA, 44. Schwarz-Romond T, Metcalfe C, Bienz M. 2007. Dynamic recruitment Zhou J. 2007. Fibrocystin/polyductin, found in the same protein complex of axin by Dishevelled protein assemblies. J Cell Sci 120:2402–2412. http: with polycystin-2, regulates calcium responses in kidney epithelia. Mol //dx.doi.org/10.1242/jcs.002956. Cell Biol 27:3241–3252. http://dx.doi.org/10.1128/MCB.00072-07. 45. Cougot N, Babajko S, Seraphin B. 2004. Cytoplasmic foci are sites of 25. Lian P, Li A, Li Y, Liu H, Liang D, Hu B, Lin D, Jiang T, Moeckel G, mRNA decay in human cells. J Cell Biol 165:31–40. http://dx.doi.org/10 Qin D, Wu G. 2014. Loss of polycystin-1 inhibits Bicc1 expression during .1083/jcb.200309008. mouse development. PLoS One 9:e88816. http://dx.doi.org/10.1371 46. National Research Council. 2011. Guide for the care and use of laboratory /journal.pone.0088816. animals, 8th ed. National Academies Press, Washington, DC.

26. Tran U, Zakin L, Schweickert A, Agrawal R, Doger R, Blum M, De 47. Harada BT, Knight MJ, Imai S, Qiao F, Ramachander R, Sawaya MR, http://mcb.asm.org/ Robertis EM, Wessely O. 2010. The RNA-binding protein bicaudal C Gingery M, Sakane F, Bowie JU. 2008. Regulation of enzyme localization regulates polycystin 2 in the kidney by antagonizing miR-17 activity. De- by polymerization: polymer formation by the SAM domain of diacylglyc- velopment 137:1107–1116. http://dx.doi.org/10.1242/dev.046045. erol kinase delta1. Structure 16:380–387. http://dx.doi.org/10.1016/j.str 27. Piazzon N, Maisonneuve C, Guilleret I, Rotman S, Constam DB. 2012. .2007.12.017. Bicc1 links the regulation of cAMP signaling in polycystic kidneys to mi- 48. Arnold K, Bordoli L, Kopp J, Schwede T. 2006. The SWISS-MODEL croRNA-induced gene silencing. J Mol Cell Biol 4:398–408. http://dx.doi workspace: a web-based environment for protein structure homology mod- .org/10.1093/jmcb/mjs027. elling. Bioinformatics 22:195–201. http://dx.doi.org/10.1093/bioinformatics 28. Gattone VH, II, Wang X, Harris PC, Torres VE. 2003. Inhibition of renal /bti770. cystic disease development and progression by a vasopressin V2 receptor 49. Eswar N, Webb B, Marti-Renom MA, Madhusudhan MS, Eramian D, antagonist. Nat Med 9:1323–1326. http://dx.doi.org/10.1038/nm935. Shen MY, Pieper U, Sali A. 2006. Comparative protein structure mod- on October 6, 2017 by EPFL Scientific information and libraries 29. Rees S, Kittikulsuth W, Roos K, Strait KA, Van Hoek A, Kohan DE. 2014. eling using Modeller. Curr Protoc Bioinformatics Chapter 5:Unit 5.6. Adenylyl cyclase 6 deficiency ameliorates polycystic kidney disease. J Am Soc http://dx.doi.org/10.1002/0471250953.bi0506s15. Nephrol 25:232–237. http://dx.doi.org/10.1681/ASN.2013010077. 50. Shen MY, Sali A. 2006. Statistical potential for assessment and prediction 30. Harris PC, Torres VE. 2014. Genetic mechanisms and signaling pathways of protein structures. Protein Sci 15:2507–2524. http://dx.doi.org/10.1110 in autosomal dominant polycystic kidney disease. J Clin Invest 124:2315– /ps.062416606. 2324. http://dx.doi.org/10.1172/JCI72272. 51. Pronk S, Pall S, Schulz R, Larsson P, Bjelkmar P, Apostolov R, Shirts 31. Lee DC, Chan KW, Chan SY. 1998. Expression of transforming MR, Smith JC, Kasson PM, van der Spoel D, Hess B, Lindahl E. 2013. growth factor alpha and epidermal growth factor receptor in adult GROMACS 4.5: a high-throughput and highly parallel open source mo- polycystic kidney disease. J Urol 159:291–296. http://dx.doi.org/10 lecular simulation toolkit. Bioinformatics 29:845–854. http://dx.doi.org .1016/S0022-5347(01)64084-9. /10.1093/bioinformatics/btt055. 32. Sweeney WE, Jr, Avner ED. 1998. Functional activity of epidermal 52. Morse PM, Feshbach H. 1953. Asymptotic series; method of steepest growth factor receptors in autosomal recessive polycystic kidney disease. descent, p 434–443. Methods of theoretical physics, part I. McGraw-Hill, Am J Physiol 275:F387–F394. New York, NY. 33. Dell KM, Nemo R, Sweeney WE, Jr, Levin JI, Frost P, Avner ED. 2001. 53. Mackerell AD, Jr, Feig M, Brooks CL, III. 2004. Extending the treatment A novel inhibitor of tumor necrosis factor-alpha converting enzyme ame- of backbone energetics in protein force fields: limitations of gas-phase liorates polycystic kidney disease. Kidney Int 60:1240–1248. http://dx.doi quantum mechanics in reproducing protein conformational distributions .org/10.1046/j.1523-1755.2001.00963.x. in molecular dynamics simulations. J Comput Chem 25:1400–1415. http: 34. Shillingford JM, Murcia NS, Larson CH, Low SH, Hedgepeth R, Brown //dx.doi.org/10.1002/jcc.20065. N, Flask CA, Novick AC, Goldfarb DA, Kramer-Zucker A, Walz G, 54. Flaherty L, Bryda EC, Collins D, Rudofsky U, Montgomery JC. 1995. Piontek KB, Germino GG, Weimbs T. 2006. The mTOR pathway is New mouse model for polycystic kidney disease with both recessive and regulated by polycystin-1, and its inhibition reverses renal cystogenesis in dominant gene effects. Kidney Int 47:552–558. http://dx.doi.org/10.1038 polycystic kidney disease. Proc Natl Acad SciUSA103:5466–5471. http: /ki.1995.69. //dx.doi.org/10.1073/pnas.0509694103. 55. Bertrand E, Chartrand P, Schaefer M, Shenoy SM, Singer RH, Long 35. Maisonneuve C, Guilleret I, Vick P, Weber T, Andre P, Beyer T, RM. 1998. Localization of ASH1 mRNA particles in living yeast. Mol Cell Blum M, Constam DB. 2009. Bicaudal C, a novel regulator of Dvl 2:437–445. http://dx.doi.org/10.1016/S1097-2765(00)80143-4. signaling abutting RNA-processing bodies, controls cilia orientation 56. Fu Y, Kim I, Lian P, Li A, Zhou L, Li C, Liang D, Coffey RJ, Ma J, Zhao and leftward flow. Development 136:3019–3030. http://dx.doi.org/10 P, Zhan Q, Wu G. 2010. Loss of Bicc1 impairs tubulomorphogenesis of .1242/dev.038174. cultured IMCD cells by disrupting E-cadherin-based cell-cell adhesion. 36. Pennekamp P, Menchen T, Dworniczak B, Hamada H. 2015. Situs Eur J Cell Biol 89:428–436. http://dx.doi.org/10.1016/j.ejcb.2010.01.002. inversus and ciliary abnormalities: 20 years later, what is the connection? 57. Qiao F, Bowie JU. 2005. The many faces of SAM. Sci STKE 2005:re7. Cilia 4:1. http://dx.doi.org/10.1186/s13630-014-0010-9. 58. Nakel K, Hartung SA, Bonneau F, Eckmann CR, Conti E. 2010. Four 37. Ponting CP. 1995. SAM: a novel motif in yeast sterile and Drosophila KH domains of the C. elegans Bicaudal-C ortholog GLD-3 form a globular polyhomeotic proteins. Protein Sci 4:1928–1930. http://dx.doi.org/10 structural platform. RNA 16:2058–2067. http://dx.doi.org/10.1261/rna .1002/pro.5560040927. .2315010. 38. Carroll M, Tomasson MH, Barker GF, Golub TR, Gilliland DG. 1996. 59. Hofmann I, Casella M, Schnolzer M, Schlechter T, Spring H, Franke The TEL/platelet-derived growth factor beta receptor (PDGF beta R) fu- WW. 2006. Identification of the junctional plaque protein plakophilin 3 in sion in chronic myelomonocytic leukemia is a transforming protein that cytoplasmic particles containing RNA-binding proteins and the recruit- self-associates and activates PDGF beta R kinase-dependent signaling ment of plakophilins 1 and 3 to stress granules. Mol Biol Cell 17:1388– pathways. Proc Natl Acad SciUSA93:14845–14850. http://dx.doi.org/10 1398. .1073/pnas.93.25.14845. 60. Stagner EE, Bouvrette DJ, Cheng J, Bryda EC. 2009. The polycystic 39. Meruelo AD, Bowie JU. 2009. Identifying polymer-forming SAM do- kidney disease-related proteins Bicc1 and SamCystin interact. Biochem mains. Proteins 74:1–5. http://dx.doi.org/10.1002/prot.22232. Biophys Res Commun 383:16–21. http://dx.doi.org/10.1016/j.bbrc.2009 40. Kim CA, Gingery M, Pilpa RM, Bowie JU. 2002. The SAM domain of .03.113. polyhomeotic forms a helical polymer. Nat Struct Biol 9:453–457. 61. Hoff S, Halbritter J, Epting D, Frank V, Nguyen TM, van Reeuwijk J,

3352 mcb.asm.org Molecular and Cellular Biology October 2015 Volume 35 Number 19 Role of Bicc1 SAM Domain Polymerization

Boehlke C, Schell C, Yasunaga T, Helmstadter M, Mergen M, Filhol E, Bicaudal-C spatially controls translation of vertebrate maternal mRNAs. Boldt K, Horn N, Uefﬁng M, Otto EA, Eisenberger T, Elting MW, van RNA 19:1575–1582. http://dx.doi.org/10.1261/rna.041665.113. Wijk JA, Bockenhauer D, Sebire NJ, Rittig S, Vyberg M, Ring T, Pohl 66. Zhou HX, Rivas G, Minton AP. 2008. Macromolecular crowding and M, Pape L, Neuhaus TJ, Elshakhs NA, Koon SJ, Harris PC, Grahammer conﬁnement: biochemical, biophysical, and potential physiological con- F, Huber TB, Kuehn EW, Kramer-Zucker A, Bolz HJ, Roepman R, sequences. Annu Rev Biophys 37:375–397. http://dx.doi.org/10.1146 Saunier S, Walz G, Hildebrandt F, Bergmann C, Lienkamp SS. 2013. /annurev.biophys.37.032807.125817. ANKS6 is a central component of a nephronophthisis module linking 67. Kertesz M, Iovino N, Unnerstall U, Gaul U, Segal E. 2007. The role of NEK8 to INVS and NPHP3. Nat Genet 45:951–956. http://dx.doi.org/10 site accessibility in microRNA target recognition. Nat Genet 39:1278– .1038/ng.2681. 1284. http://dx.doi.org/10.1038/ng2135. Downloaded from 62. Taskiran EZ, Korkmaz E, Gucer S, Kosukcu C, Kaymaz F, Koyunlar C, 68. Kedde M, van Kouwenhove M, Zwart W, Oude Vrielink JA, Elkon R, Agami R. 2010. A Pumilio-induced RNA structure switch in p27-3= UTR Bryda EC, Chaki M, Lu D, Vadnagara K, Candan C, Topaloglu R, controls miR-221 and miR-222 accessibility. Nat Cell Biol 12:1014–1020. Schaefer F, Attanasio M, Bergmann C, Ozaltin F. 2014. Mutations in http://dx.doi.org/10.1038/ncb2105. ANKS6 cause a nephronophthisis-like phenotype with ESRD. J Am Soc 69. Miles WO, Tschop K, Herr A, Ji JY, Dyson NJ. 2012. Pumilio facilitates Nephrol 25:1653–1661. http://dx.doi.org/10.1681/ASN.2013060646. miRNA regulation of the E2F3 oncogene. Genes Dev 26:356–368. http: 63. Leettola CN, Knight MJ, Cascio D, Hoffman S, Bowie JU. 2014. Char- //dx.doi.org/10.1101/gad.182568.111. acterization of the SAM domain of the PKD-related protein ANKS6 and 70. Isono K, Endo TA, Ku M, Yamada D, Suzuki R, Sharif J, Ishikura T, its interaction with ANKS3. BMC Struct Biol 14:17. http://dx.doi.org/10 Toyoda T, Bernstein BE, Koseki H. 2013. SAM domain polymerization

.1186/1472-6807-14-17. links subnuclear clustering of PRC1 to gene silencing. Dev Cell 26:565– http://mcb.asm.org/ 64. Yakulov TA, Yasunaga T, Ramachandran H, Engel C, Muller B, Hoff S, 577. http://dx.doi.org/10.1016/j.devcel.2013.08.016. Dengjel J, Lienkamp SS, Walz G. 2015. Anks3 interacts with 71. Qiao F, Song H, Kim CA, Sawaya MR, Hunter JB, Gingery M, Rebay I, nephronophthisis proteins and is required for normal renal development. Courey AJ, Bowie JU. 2004. Derepression by depolymerization; struc- Kidney Int 87:1191–1200. http://dx.doi.org/10.1038/ki.2015.17. tural insights into the regulation of Yan by Mae. Cell 118:163–173. http: 65. Zhang Y, Cooke A, Park S, Dewey CN, Wickens M, Sheets MD. 2013. //dx.doi.org/10.1016/j.cell.2004.07.010. on October 6, 2017 by EPFL Scientific information and libraries

October 2015 Volume 35 Number 19 Molecular and Cellular Biology mcb.asm.org 3353 4. Results part II: Biochemical and functional analysis of Bicc1 interactome

Part of the results of this chapter are contained in the manuscript under revision in

PLOS genetics:

Role of Bicaudal C1 in renal gluconeogenesis and its novel interaction with the CTLH complex.

Lucia Leal-Esteban, Benjamin Rothé, Simon Fortier, Manuela Isenschmid, and

Daniel B. Constam.

The data in this chapter mostly comes from my own work. Contributions of the co- authors of this manuscript are acknowledged where indicated.

4.1. Bicc1 interactome screening

To identify novel Bicc1-dependent processes, I conducted a protein interactome screen in a Flp-In™ T-REx™ 293 knock-in cell line that I engineered to express tandem affinity-tagged Bicc1-HA-StrepII-tagged as a doxycycline-inducible transgene. To generate the inducible stable cell line, I co-transfected Flp-In™T- REx™ 293 cells with pOG44 Flp recombinase expression plasmid and with pcDNA™5/FRT/TO vector containing MmBicc1-HA-StrepII (Fig. 14A). Correct integration by Flp-mediated homologous recombination gave rise to several hygromycin-resistant clones that were independently isolated and evaluated for protein induction efficiency. In one representative clone examined, MmBicc1-HA- StrepII expression relative to γ-tubulin in cells treated with doxycycline (Dox) reached levels comparable to those of endogenous Bicc1 in mIMCD3 cells (Fig. 14B). To validate that MmBicc1-HA-StrepII was functional, I compared its post-transcriptional silencing activity to that of transiently expressed HA-Bicc1 using luciferase mRNA reporters containing target 3'UTR fragments of AC6 or PKIα (Piazzon et al., 2012). I

61 observed a similar reduction of the luciferase activity, confirming that the silencing activity is equivalent in both systems (Fig. 14C). These results confirmed that the evaluated clone expressed Bicc1-HA-StrepII in a functional manner.

Figure 14. Doxycycline-inducible Bicc1-HA-StrepII expression and function. (A) MmBicc1 was tagged with 4 internal and 2 C-terminal HA and 2 Strep-II epitopes. Positions of 3 KH (K-Homology domain), 2 KH-like domains (KHL) and one sterile alpha motif (SAM) are indicated. (B) Western blot of MmBicc1-HA-StrepII in Flp-In T-rex cell extracts treated with or without doxycycline for 24 hrs. Expression relative to endogenous Bicc1 levels in mIMCD3 cells (arrow) was quantified by densitometric analysis. Traces of MmBicc1-HA-StrepII (arrowhead) leaked from the preceding well. (C) Expression of AC6 and PKIα 3'UTR luciferase reporters in Flp-In T-rex HEK293T cells transfected with MmBicc1-HA or empty vector (mock) or induced with doxycycline to express MmBicc1-HA- StrepII. Luciferase values are normalized to co-transfected β-galactosidase. Bars represent mean±S.E.M. n= 3. Student’s t-test was used to find significance p-value *p <0.05; **p <0.01

With this approach MmBicc1-HA-StrepII protein, along with the associated protein partners, are recovered in two sequential steps by tandem affinity purification (TAP) (Gavin et al., 2011). In our standard protocol, cells were cultured for one day in the presence or absence (control) of doxycycline, and Bicc1 was purified by Strep-tag pull-down immediately followed by anti-HA immunoprecipitation. Eluates were analyzed by Liquid Chromatography-Mass Spectrometry (LC/MS) in two independent experiments. This analysis found 243 co-purifying proteins that were enriched at least 2-fold compared to the uninduced negative control sample (Table 1).

62 Table 1. Bicc1 interactors enriched at least 2-fold

63 64 65 4.2 Bicc1 co-purified ciliopathies related proteins

The protein that was most enriched by co-purification with Bicc1 was CNOT1 and, to a lesser extent, several other subunits of the CCR4-NOT complex. CCR4-NOT is involved in mRNA deadenylation and previously has been shown that CNOT3/5 binds and genetically interact with Drosophila Bicaudal-C (Chicoine et al., 2007), suggesting that our in vitro screening identified interactors that have a biological significance.

Among novel candidate Bicc1-interacting factors, we observed that ANKS3 was 32- fold enriched in Bicc1 IP conditions. This interaction was recently found by overexpressing tagged proteins in HEK293T cells (Yakulov et al., 2015). ANKS3 has been identified in patients with autosomal recessive laterality defects, which is a shared feature of several ciliopathies (Shamseldin et al., 2016). Moreover a link with kidney development was demonstrated in Zebrafish since ANKS3 knockdown promoted pronephric cystogenesis (Yakulov et al., 2015). I confirmed this interaction by co-immunoprecipitation assays in mIMCD3 cell, where endogenous Bicc1 efficiently bound ANKS3, while the IgG negative control failed to do so (Fig. 15A).

Since Bicc1 is an RNA-binding protein, we wondered whether ANKS3 binding is modulated by RNA. Endogenous ANKS3 co-immunoprecipitated with doxycycline- induced MmBicc1-HA-StrepII from Flp-In™ T-REx™ 293 cell extracts irrespective of whether they were treated with or without RNaseA, indicating that RNA is not essential to mediate this interaction (Fig. 15B).

Since Bicc1 and ANKS3 both contain a SAM domain, one hypothesis was that they might interact through heterooligomerization of their SAM domains. To map ANKS3- interacting domains in Bicc1, I performed pull-down experiments using as baits recombinant GST-fusions of full-length Bicc1 or individual KH, the intervenient sequence (INV) or SAM domains (purified by my colleague Dr. B. Rothé). Glutathione beads coated with full length GST-Bicc1 or GST-Bicc1 fragments were incubated with a cell extract expressing ANKS3-Flag. Only full-length GST-Bicc1 and GST- Bicc1SAM domain efficiently pulled down ANKS3-Flag protein. This result suggests that the Bicc1 SAM domain is sufficient to bind ANKS3 in vitro (Fig. 15C).

66 Figure 15. Bicc1 interacts with ANKS3 through its SAM domain independently of RNA. (A) Endogenous ANKS3 and Bicc1 co-immunoprecipitation in mIMCD3 (n=2). (B) Endogenous ANKS3 co-immunoprecipitation by STREP-tagged Bicc1 in inducible T-Rex 293 cells. Induction was carried out by tetracycline and cell extracts were treated with or without RNase A before anti-STREP immunoprecipitation (n=1). (C) Pull-down of ANKS3- Flag in HEK293T cell extracts using GST fusions of full-length Bicc1, or KH, IVS or SAM domains bound to glutathione beads. GST alone was used as negative control. Input was 5% of total cell extracts (n=1). Figure cropped from a different exposured film.

A complex network of interactions between Bicc1, ANKS3 and ANKS6 was further characterized by B. Rothé and S. Fortier, in collaboration with C. Leettola and J. Bowie from UCLA. They found that ANKS3 and ANKS6 can regulate the topology of cytoplasmic Bicc1 clusters. Specifically, neither ANKS3 nor ANKS6 formed macroscopic homopolymers on their own. However, binding to ANKS3 dispersed foci of Bicc1 polymers, whereas co-recruitment of ANKS6 by ANKS3 enlarged them. ANKS6 induced the formation of large ANKS3/ANKS6 heteropolymers that can

67 recruit Bicc1 even independently of its SAM domain, likely through a novel interaction of Bicc1 KH domains with ANK repeats of ANKS3. Surprisingly, dispersion of Bicc1 foci by ANKS3 did not alter the silencing of the reporter mRNA. Indeed, cotransfection of the reporter mRNA rescued the assembly of Bicc1/ANKS3 heterooligomers into giant foci, with an spatial re-arrangement of ANKS3 at their periphery. Thus, strongly suggesting that mRNA can compete with ANKS3 for KH binding and therefore silencing ability is retained. Although limited by heterologous expression, this study allowed to delineate with first approximation some rules how the assembly of this SAM-mediated macromolecular scaffold is regulated. See in annex 1 the manuscript.

Our results suggest that the spatial configuration of macromolecular Bicc1 protein complexes can be significantly modulated both by specific interacting proteins and by RNA binding. Indicating that RNA binding can influence Bicc1 interacting partners, since it impacts the conformation and/or accessibility of Bicc1 to bind other proteins. Future studies will address how inhibition of Bicc1 self-polymerization modifies Bicc1 interactome.

Bicc1, ANKS3, and ANKS6 complexes suggests that they may function in similar signaling pathways containing NPHPs proteins (Hoff et al., 2013; Yakulov et al., 2015). In keeping with this idea, other players that appear in the interacting network of the ANKS and NPHPs proteins are the Never In Mitosis A (NIMA) kinases. Nek7, Nek6, and Nek9 participate in the same signaling cascade, where Nek6 and Nek7 are phosphorylated and activated by Nercc1/Nek9. In this cascade, Nek7 phosphorylates the mitotic kinesin Eg5 to control early centrosome separation and mitotic spindle formation (Bertran et al., 2011; Sdelci et al., 2011). Additionally, Nek7 was recently described as a target of ANKS3 for its cytoplasmic retention, yet, its function in the cytoplasm remains elusive (Ramachandran et al., 2015).

Using co-immunoprecipitation assay, I validated the interaction of Bicc1 and Nek7 at endogenous levels in mIMCD3 cells, (Fig. 16A). Curiously, we did not find ANKS6 or Nek8 in Bicc1 interactome, in agreement with the literature, suggesting that ANKS3 but not ANKS6 interacts with Nek7 (Ramachandran et al., 2015).

68 Prompted by this result and knowing that ANKS3 retains Nek7 in the cytoplasm, we wondered whether knock-down of Nek7 would affect Bicc1 function in post- transcriptional silencing. For this purpose, I used Nek7 RNAi. HEK293T cells were treated with two increasing doses of siNek7 showing a good knock-down efficiency

(Fig. 16B). However, I did not observe changes in Bicc1-mediated repression of the AC6 3'UTR luciferase reporter expression. These results confirm that Nek7 is not essential for Bicc1 translational silence of target mRNA.

Figure 16. Bicc1 interacts with NEK7 but its silencing activity is unaffected by NEK7 knock-down. (A) Co-immunoprecipitation of endogenous Bicc1 with NEK7 in mIMCD3 cells. IgG served as negative control (n=2). (B) Expression of AC6 3'UTR luciferase reporter in HEK293T cells transfected with MmBicc1-HA or empty vector (mock) after siNEK7 transfection during 48hrs. Luciferase values are normalized to co-transfected β-galactosidase. Bars represent mean±S.E.M. n= 3. WB shows the efficiency of siNEK7 doses in cells used in the luciferase assay. Protein levels were normalized to γ-Tubulin expression.

69 The kat and jck mouse models harbor spontaneous mutations in Nek1 and Nek8, respectively. These models are not known to exhibit a clear cell cycle disorder but instead, polycystic kidney disease is the principal phenotype (Liu et al., 2002; Upadhya et al., 2000).

To study if a regulatory interaction between Bicc1 and Nek7 is occurring in vivo, I monitored Nek7 protein levels in wild-type compared to Bicc1-/- newborn kidneys. WB analysis showed that Nek7 protein is 1.6-fold increase in Bicc1-/- P2 kidneys (Fig. 17).

Figure 17. Bicc1 lost increases the levels of Nek7 in mice kidneys. Western blot analysis of Nek7 in Bicc1-/- kidney extracts. n=5 per genotype. Bars represent mean ± SEM. **p <0.01

Taken together, these results indicate that Bicc1 participates in NPH signaling pathways by networking with key players such as ANKS6, ANKS3, and Nek7. Moreover, these data suggest that the interaction between Bicc1 and ANKS6 is mediated by ANKS3, which is in agreement with a previous model where ANKS6 and Bicc1 interaction was likely indirect (Stagner et al., 2009).

4.3 Bicc1 co-purified the CTLH complex

Among the novel interactions, we observed 11- to 48-fold enrichment of multiple subunits of the C-terminal to Lissencephaly Homology (CTLH) complex (Kobayashi et al., 2007), including WD repeat-containing protein WDR26, Ran-binding protein-9 (RanBP9), Macrophage Erythroblast Attacher (MAEA), Two-hybrid-Associated protein With RanBPM 1 (TWA1), Required for Meiotic Nuclear Division 5 homolog A (RMND5a), and 2.5-fold enrichment of Armadillo repeat containing 8 (ARMC8). The

70 only known CTLH subunits that were not enriched by Bicc1-HA-StrepII by MS were Muskelin-1 (MKLN1) and GID4 (formerly C17orf39) (Table 2).

Table 2. Mass spectrometric identification of CTLH complex subunits co-purified with TAP- tagged Bicc1. CTLH/GID Unique peptide counts Rank subunit Gene name Name ctrl IP1 ctrl IP2 Δ 4 GID7 WDR26 WD repeat-containing protein 26 0 28 0 20 48 22 GID1 RanBP9 Ran-binding protein 9 0 13 0 8 21 43 GID9 MAEA1 Macrophage erythroblast attacher 0 11 0 4 15 46 GID8 TWA1 Glucose-induced degradation protein 8 homolog 0 7 0 6 13 64 GID2 RMND5a Required for meiotic nuclear division 5 homologA 0 8 0 3 11 90 GID5 ARMc8 Armadillo repeat-containing protein 8 0 5 0 1 6

The mammalian CTLH complex and the orthologous Gid complex in S. cerevisiae are conserved multi-protein complexes enriched in multiple subunits for

Lissencephaly Homology (LisH), CTLH and C-terminal CT11-RanBP9 (CRA) domains, and for transducin/WD40 and armadillo repeats (Francis et al., 2013; Kobayashi et al., 2007). When yeast growing on ethanol are switched to a glucose containing-medium, the Gid complex consisting of 9 subunits becomes activated to degrade gluconeogenic enzymes via the proteasomal Pro/N-end rule pathway or via lysosomes (Chen et al., 2017b; Hämmerle et al., 1998; Santt et al., 2008). The orthologous mammalian CTLH complex has no known function and contained no GID4 under the conditions examined in HEK293 cells (Kobayashi et al., 2007).

To validate binding to Bicc1, we performed co-immunoprecipitation assays. In HEK293T cells which do not express endogenous Bicc1, anti-Flag IP of DDK-tagged WDR26 co-immunoprecipitated RanBP9, TWA1 and exogenous HA-Bicc1 (Fig. 18A). WDR26 also co-precipitated MKLN1, whereas anti-MKLN1 IP pulled down HA-Bicc1 together with RanBP9 (Fig. 18B). Moreover, immunoprecipitation of endogenous Bicc1 in extracts of mouse inner medullary collecting duct (mIMCD3) cells enriched endogenous RanBP9, WDR26, and Twa1 (Fig. 18C). These results

71 confirm that multiple subunits of the mammalian CTLH complex bind Bicc1 in transfected HEK293T cells and at endogenous expression levels in mIMCD3 cells.

Figure 18. Bicc1 interaction with the mammalian CTLH complex subunits. (A) Anti-Flag (DDK) coimmunoprecipitation of HA-Bicc1 and endogenous RanBP9, Twa1 and MKLN1 in HEK293T extracts. Mock-transfected cells served as negative control. (B) Anti-MKLN1 coimmunoprecipitation of HA-Bicc1 and endogenous RanBP9 in HEK293T cells. Pre-immune IgG served as negative control. Figure cropped from different exposure time films. (C) Coimmunoprecipitation of RanBP9, WDR26, and Twa1 with endogenous Bicc1 in mIMCD3 cells extract using anti-Bicc1 or pre-immune IgG (negative control). Figures representative of two experiments.

4.3.1 Loss of Bicc1 increases WDR26 protein in newborn kidneys

To evaluate potential interactions of the CTLH subunits with Bicc1 in vivo, we

compared the expression levels of CTLH subunits in wild-type (WT) and Bicc1-/- mice.

Western blot analysis revealed significant upregulation of WDR26 protein levels in

the knockout (P2 n=5, P≤0.01) and a similar trend for GID4 and RanBP9, but not for

TWA1 proteins (Fig. 19A-D). By contrast, no changes were observed in WDR26 or

72 GID4 mRNAs (Fig. 19E). These results indicate that Bicc1 attenuates the levels of some but not all CTLH subunits in vivo.

Figure 19. Elevated protein levels of CTLH subunits in Bicc1-/- kidneys. (A-D) Western blot analysis of (A) RanBP9, (B) Twa1, (C) WDR26 and (D) GID4 in WT and Bicc1-/- kidney extracts. Bars represent mean ± SEM. **p <0.01 (n=5 per genotype) (E) RT-qPCR analysis of WDR26 and GID4 mRNAs in WT and Bicc1-/- kidneys (n=3 per genotype). GAPDH mRNA was used for normalization.

4.3.2 Bicc1 maintains normoglycemia and normal renal expression of several gluconeogenic enzymes

In S. cerevisiae grown in high glucose, the Gid complex polyubiquitinates the gluconeogenic enzymes Fructose-1,6-biphosphatase (FBP1) and

Phosphoenolpyruvate carboxykinase-1 (PEPCK) in a necessary step for catabolite inactivation that will lead to their degradation to facilitate a metabolic switch to glycolysis (Regelmann et al., 2003; Santt et al., 2008).

To consider the hypothesis that Bicc1 and the CTLH interacting complex regulate gluconeogenesis in the mammalian kidney, we started by assessing the blood glucose levels of newborn mice. Interestingly, Bicc1-/- pups at postnatal day P2 have significantly reduced glycemia (Fig. 20A). To validate that Bicc1 promotes gluconeogenic gene expression, and to assess potential post-transcriptional regulation, we compared relative levels of PEPCK and the rate-limiting enzyme FBP1

73 by RT-qPCR and Western blot. Compared to WT, Bicc1 mutants down-regulated PEPCK mRNA and protein 3.4- and 10-fold, respectively, whereas FBP1 only decreased 2-fold specifically at the protein level and only in kidneys (Fig. 20B, C) but not in livers (Fig. 20D). These results show that Bicc1 maintains the expression of both PEPCK and FBP1 proteins specifically in kidneys.

Figure 20. Reduced glycemia and downregulation of FBP1 and PEPCK expression in Bicc1-/- kidneys. (A) Blood glucose measurements in WT and Bicc1-/- mice at postnatal day P2 (n ≥ 6 per genotype). Bars represent mean ± SEM. 1-way ANOVA: 0.0001 followed by Bonferroni’s multiple comparisons test **p <0.01; **p <0.001 (B) RT-qPCR analysis of FBP1 and PEPCK1 mRNA in WT (n=3) and Bicc1-/- (n=3) kidneys. GAPDH mRNA was used for normalization. Bars represent mean ± SEM. *p<0.05 (C) Western blot analysis of FBP1 and PEPCK1 in WT (n=3) and Bicc1-/- kidneys (n=3) at P2. γ-tubulin was used as loading control. (D) Western blot analysis of FBP1 and PEPCK in WT (n=3) and Bicc1-/- livers (n=3) at P2. γ-Tubulin was used as loading control.

74 4.3.3. Bicc1 upregulates FBP1 in mIMCD3 cells concomitant with its inhibitory effect on WDR26 (GID7) expression and other CTLH subunits examined

To in vitro validate our hypothesis that Bicc1 and the CTLH interacting complex regulate gluconeogenesis and since FBP1 transcription was unaffected by Bicc1 in the kidneys, we investigated whether Bicc1 and its interaction with CTLH complex regulate FBP1 at the protein level in two cell lines. Ectopic expression of HA-Bicc1 in HEK293T cells increased the accumulation of FBP1 (Fig. 21A), whereas depletion of Bicc1 by a specific siRNA in mIMCD3 cells (Fu et al., 2010; Piazzon et al., 2012) decreased it (Fig. 21B).

Figure 21. Endogenous Bicc1 in mIMCD3 cells increases FBP1 protein levels (A) Western blot analysis of FBP1 in HEK293T cells transfected with the indicated dose of HA-Bicc1. (n=3). Bars represent mean ± SEM. *p <0.05. (B) Western blot of FBP1 in mIMCD3 48hrs after transfection of Scramble or Bicc1 siRNA (n=10). Below: Bars represent mean ± SEM. ***p <0.001.

Alternatively, my colleague Dr. S. Fortier, inactivated Bicc1 by CRISPR/Cas9 editing using single guide RNA (Fig. 22A). Western blot revealed the loss of Bicc1 protein in several independent clones and that FBP1 levels among such sgBicc1 clones were variable (Fig. 22B).

75 sgBicc1 CRISPR clones IMCD3 wt IMCD3 #4 #5 #12 #14 #2 #3

Bicc1

FBP1

Gapdh

Figure 22. CRISPR-editing of Bicc1 in mIMCD3 cells. (A) Scheme of Bicc1 targeting by CRISPR/Cas9 genome editing. (B) Western blot of Bicc1 and FBP1 expression in CRISPR/Cas9 treated mIMCD3 cell clones.

I selected sgBicc1 clone #14 for further analysis because it constantly showed very reduced levels of FBP1. Moreover, I observed that FBP1 RNA levels did not differ between sgBicc1 clone and WT (Fig. 23).

Figure 23. Bicc1 inactivation by CRISPR/Cas9 decreases FBP1 protein but no mRNA levels. Western blot and RT-qPCR analysis (right panel) of FBP1 in mIMCD3 cells (WT) and in a representative CRISPR-edited clone (sgBicc) (n=3 per genotype). For RT-qPCR GAPDH mRNA was used for normalization. Bars represent mean ± SEM. **p <0.01

If Bicc1 promotes FBP1 accumulation by inhibiting its proteasomal targeting by

CTLH complex, proteasome inhibitor treatment should increase FBP1 levels in

76 Bicc1 deficient cells more potently than in WT. Indeed, when treated with the proteasomal inhibitor MG132, only sgBicc cells tended to increase FBP1, whereas WT cells decreased it (Fig. 24A). These results indicate that Bicc1 stabilizes FBP1 at least in part by attenuating its degradation, but that proteasomes in parallel likely degrade also a positive regulator of FBP1 expression.

To evaluate whether FBP1 is regulated by the CTLH complex, I depleted WDR26 or Twa1 or both by RNAi. WDR26 and Twa1 were depleted within 48 hrs after siRNA treatment. Contrary to our prediction, FBP1 levels were not upregulated but rather decreased by the time of analysis, and only in sgBicc but not in WT cells (Fig. 24B). These data suggest that Bicc1 does not upregulate FBP1 by directly inhibiting the CTLH complex, in mIMCD3 cells.

Figure 24. Bicc1 regulate FBP1 levels independently of its interaction with the CTLH complex. (A) Western blot of FBP1 in mIMCD3 cells (WT) and a representative CRISPR-edited clone (sgBicc1) treated with Bafilomycin A1 (BafA1, 100 nM), MG132 [5 µM] or both for 4 hrs. The mean ± SEM fold changes in three experiments relative to untreated WT cells are shown below. **p <0.01. (B) Western blot of FBP1, WDR26, and Twa1 in mIMCD3 transfected with siScr or siWDR26 or siTWA1 siRNAs for 48 hrs. Error bars represent mean ± SEM of two experiments. *p <0.05.

77 4.3.4 The CTLH complex and proteasomal degradation attenuate Bicc1 levels in mIMCD3 cells Alternative targets of the CTLH complex may include Bicc1 itself. In keeping with this idea, depletion of CTLH subunits in mIMCD3 cells significantly increased endogenous Bicc1 protein but not mRNA levels (Fig. 25A). To assess how Bicc1 is degraded, we treated mIMCD3 cells for 4 hours with MG132 or Bafilomycin A1 or empty vehicle. Treatment with either of these compounds significantly increased

Bicc1 protein levels, albeit without inducing detectable levels of slower migrating polyubiquitinated degradation intermediates (Fig. 25B). These results suggest that

CTLH complex, proteasomes and lysosomes all contribute to curbing Bicc1 levels, likely independently of polyubiquitination.

We have shown that Bicc1 is stabilized in large cytoplasmic foci by self- polymerization of its SAM domain (results part I). To distinguish whether the CTLH complex interacts with Bicc1 in cytoplasmic foci or with a non-polymerized pool, or both, I performed co-immunostainings of HA-Bicc1 and endogenous RanBP9 in HEK293T cells. I observed no co-localization of RanBP9 with cytoplasmic Bicc1 foci (Fig. 25C). Therefore, we investigated in more detail whether self-polymerization of Bicc1 and binding to the CTLH complex are mutually exclusive.

78 Figure 25. Bicc1 is a target of the CTLH complex. (A) Western blot and RT-qPCR analysis of endogenous Bicc1 in WT mIMCD3 cells treated for 48 hrs. with scrambled or WDR26 and Twa1 siRNAs. Quantifications show the fold change in Bicc1 protein (middle) and mRNA levels (right) relative to scramble-treated cells in three independent experiments. TATA-box binding protein (TBP) mRNA was used for normalization. (B) Bicc1 protein levels in mIMCD3 cells after treatment with Bafilomycin A1 (BafA1, 100 nM), MG132 [5 µM] or both for 4 hrs. Bars represent mean ± SEM fold changes in Bicc1 protein levels relative to untreated cells in two experiments. (C) Coimmunostaining of HA-Bicc1 and endogenous RanBP9 in HEK293T (n=1). Bars 10 µm.

4.3.5 The SAM domain independently of its self-polymerization mediates

Bicc1 recruitment to the CTLH complex To assess how Bicc1 binds the CTLH complex, I conducted co-immunoprecipitation

experiments with truncated HA-Bicc1 in transfected HEK293T cells. Deletion of the

Bicc1 SAM domain abolished binding to CTLH subunits, whereas deletion of the KH

domains did not (Fig. 26A). To confirm whether Bicc1 binding is mediated by the SAM

domain, I used GST-SAM fusion protein in the pull-down assays in HEK293T cell extracts. Irrespective of the presence or absence of exogenous HA-Bicc1, GST-SAM

79 pulled down endogenous TWA1. GST alone failed to do so, indicating that the SAM domain is sufficient to mediate this interaction (Fig. 26B).

Figure 26. Bicc1 SAM domain mediates binding with the CTLH complex. (A) Anti-HA co-immunoprecipitation of CTLH subunits with HA-Bicc1 or its truncated derivatives Bicc1ΔKH and Bicc1ΔSAM in HEK293T cells. Inputs correspond to 5% of the total extract. Mock-transfected cells were used as negative control. (n=1). (B) Pull down of endogenous Twa1 by a GST fusion protein of Bicc1 SAM domain or GST alone (control) in HEK293T cells transfected with or without HA-Bicc1 (n=1). Figure cropped from different exposure time films.

During previous studies for Bicc1 self-polymerization characterization, B. Rothé created a series of Bicc1 polymerization mutants used here for studying the interaction with the CTLH complex (mut-D and mut-I) (Fig. 27A-D). Mutant-D (mut-

D) characterization is described in the first results chapter.

80 Figure 27. In vitro characterization of the novel Bicc1 self-polymerization mutant mut-I. (A) SAM domain of MmBicc1 and positions of charged residues (red and blue) and their conservation (gray shading, determined by ClustalW alignment) in end helix (EL) and mid loop (ML) surfaces of the dimerization interface. Lysine substitutions in electrostatic patches of mutant Bicc1 mut-I (E900D902/KK) or alanine substitutions in Bicc1 mut-D (D913K915E916/AAA) are indicated. (B) Pull-down of full-length HA-Bicc1 (WT) or its mutant E900D902/KK derivative (mut-I) by glutathione-Sepharose beads coated with GST alone or with a GST fusion of the wild-type Bicc1 SAM domain. Input represents 5% of cell extracts. (C) Anti-HA immunostaining of HA-Bicc1 or its polymerization mutant derivative mut I in Cos- 1 cells. Note that the mut-I mutation suppresses polymerization of full-length HA-Bicc1 in cytoplasmic foci. (D) Expression of Luc-AC6-3’UTRprox reporter normalized to β- galactosidase in HEK293T cells co-transfected with WT HA-Bicc1 or with a single (x1) or double dose (x2) of HA-Bicc1 mut-I. Bars represent mean±SEM of 3 experiments. *p<0.05. Anti-HA Western blot shows the expression of the HA-Bicc1 proteins in the cells analyzed in the luciferase assay. All data in this figure was provided by B. Rothé.

81 To distinguish whether the SAM domain binds CTLH complexes indirectly by forming homo- or heterooligomers, or directly, I monitored pull down by GST fusions of full- length Bicc1 and its mutant derivatives mut-D and mut-I where self-polymerization is inhibited by specific mutations in mut-D and mut-I (D913;K915;E916/AAA and E900;D902/KK, respectively). Interestingly, inhibition of SAM polymerization by either of these mutations did not diminish Bicc1-mediated pull-down of TWA1, RanBP9, and WDR26 but rather increased it (Fig. 28), indicating that Bicc1 polymerization may curb the interaction with the CTLH complex subunits.

Figure 28. Unpolymerized Bicc1 preferentially binds the CTLH subunits Pull-down of endogenous Twa1, RanBP9, and WDR26 in HEK293T cell extracts by GST alone or by GST fusions of WT Bicc1 or its polymerization mutant forms mut-D or mut-I. 5% of total cell extracts were loaded as input. Coomassie staining of proteins that resisted transfer to the nitrocellulose membrane confirms equal loading of the GST fusions (n=1).

To validate that Bicc1 binds CTLH complex independently of polymerization, I fractionated HA-tagged Bicc1 on sucrose density gradients. I found that RanBP9, WDR26, GID4, TWA1, ARMC8, and MKLN1 all co-fractionated with HA-Bicc1 in low molecular weight (LMW) fractions 3 to 9, but not with HA-Bicc1 high molecular weight

82 (HMW) fractions 11-17 (Fig. 29). Thus, these results suggest that the CTLH complex specifically associates with LMW forms of Bicc1.

Figure 29. The CTLH complex co-fractionates with LMW Bicc1. Western blot analysis of density-fractionated HEK293T cell extracts on a continuous 15 to 60% sucrose gradient. Arrows indicate the order for collecting the fractions (top to bottom). Density fractionation in extracts of mock- (empty vector) (left side) or HA-Bicc1-transfected cells (right side). Graphs below show percentages of the total of each subunit per fraction. Gray areas denote LMW fractions 1 to 9. Results represent mean ± SEM from 2 experiments.

83 4.3.6 Yeast two-hybrid mapping reveals that ARMC8 mediates binding of the

CTLH complex to non-polymeric Bicc1 To comprehensively map direct interactions of the CTLH subunits with Bicc1 and among themselves, we did yeast two-hybrid assays. I cloned each CTLH subunit as bait and prey and B. Rothé and M. Isenschmid conducted the methodology in which each bait was compared to empty prey plasmid, and the strength of each interaction was evaluated by adding 3-Amino-1,2,4-triazol (3AT), a competitive inhibitor of the HIS3 reporter gene product. CTLH subunits were alternately tested as a bait and prey. The strongest interactions among CTLH subunits were those of TWA1 with ARMC8, RMND5A, and RANBP9 (Fig. 30). Strong interactions were also observed between RanBP9 and WDR26, and between MAEA and RMND5A, similar to those among the corresponding yeast proteins (Menssen et al., 2012). However, in contrast to the Gid complex topology described in yeast, we observed only a very weak interaction between GID4 and ARMC8 (GID5).

84 Figure 30. Binding of CTLH subunits to each other in yeast two-hybrid assays. All CTLH subunits were tested between each other at increasing doses of 3‐Amino‐1,2,4‐ triazole (3AT), to assess the strength of each interaction. Empty Gal4-AD was included to determine the level of growth induced by non-specific binding (background). Data are representative of 2 experiments with similar results.

Interestingly, no significant interactions were observed between CTLH subunits and WT Bicc1. However, the polymerization mutant Bicc1 mut-D strongly interacted with

ARMC8 (GID5) (Fig. 31A, B). These data indicate that Bicc1 self-polymerization and binding to the CTLH complex are mutually exclusive and that unpolymerized full- length Bicc1 directly connects to the CTLH complex through ARMC8.

85 - - - - - A LT LTH L-T- L-T-H- 3AT Gal4-AD 3AT Gal4-AD X 0 0 0.5 1 2.5 5 10 20 (mM) X 0 0 0.5 1 2.5 5 10 20 (mM) ) ) 1 AD(-) 2 AD(-) BD BD - - GID GID ( ( 4 4 9 Bicc1-WT a Bicc1-WT 5 Gal Gal Bicc1-mut D Bicc1-mut D RanBP Rmnd

------LT LTH 3AT LT LTH 3AT X Gal4-AD X Gal4-AD 0 0 0.5 1 2.5 5 10 20 (mM) 0 0 0.5 1 2.5 5 10 20 (mM) ) ) 4 3 AD(-) AD(-) BD BD - GID - GID ( ( 4 Bicc1-WT 4 Bicc1-WT 39 H 2 orf Gal Gal

Bicc1-mut D 17 Bicc1-mut D UBE C

L-T- L-T-H- L-T- L-T-H- Gal4-AD 3AT Gal4-AD 3AT X 0 0 0.5 1 2.5 5 10 20 (mM) X 0 0 0.5 1 2.5 5 10 20 (mM) ) ) 7

5 AD(-) AD(-) BD BD GID - - ( GID ( 4 Bicc1-WT 4 Bicc1-WT 8 26 Gal Gal

Armc Bicc1-mut D

WDR Bicc1-mut D

L-T- L-T-H- 3AT L-T- L-T-H- 3AT Gal4-AD Gal4-AD X 0 0 0.5 1 2.5 5 10 20 (mM) X 0 0 0.5 1 2.5 5 10 20 (mM) ) ) 9

8 AD(-) AD(-) BD BD - - GID 4 GID ( 4 ( Bicc1-WT Bicc1-WT 1 Gal Gal

Bicc1-mut D MAEA Bicc1-mut D TWA

B UBE2H Bicc1 (GID3) mutD

MAEA (GID9)

TWA1 (GID8)

C17orf39 (GID4)

Interaction strength RANBP9 (Bait Prey) (GID1) 1-2 WDR26 3-4 (GID7) 5-6 7-8

Figure 31. Bicc1 binds the CTLH complex via ARMc8 and independently of SAM polymerization. (A) Binding of CTLH subunits to each other and to Bicc1 WT or mut-D (D913K915E916/AAA) in yeast two-hybrid assays. For each CTLH subunit, Bicc1 WT or mut-D interactions were tested at increasing doses of 3‐Amino‐1,2,4‐triazole (3AT), to assess the strength of each interaction. Empty Gal4-AD was included to determine the level of growth induced by non-specific binding (background). Data are representative of 2 experiments with similar results. (B) Human CTLH complex subunit interactions and binding to Bicc1 mut-D mapped by yeast two-hybrid assays. Arrows depict the relative strength of binding in two independent experiments.

86 4.4 Discussion

Bicc1-HA-StrepII purification identified some ciliopathies-related proteins

We validated at endogenous levels ANKS3 and Nek7 as Bicc1 interacting partners. This result led us to investigate the interaction between Bicc1 and ANKS3, as well as the related SAM protein ANKS6 whose association with Bicc1 and ANKS3 has previously been reported (Bakey et al., 2015; Delestré et al., 2015; Leettola et al.,

2014; Stagner et al., 2009; Yakulov et al., 2015). ANKS6 is part of the NPHP network and connects NEK8 (NPHP9) to INVS (NPHP2) and NPHP3 (Hoff et al., 2013). ANKS6 SAM domain mutation (R823W) in the cy/+ rat model of PKD cause a defective interaction with ANKS3, while in a different model ANKS6 (I747N) mice increased ANKS6 and Bicc1 interaction, that otherwise in WT conditions is very weak (Bakey et al., 2015). They hypothesized that Bicc1 is expressed in proximal tubules where cysts are absent in the ANKS6 (I747N) mice suggesting that Bicc1 and ANKS6 likely interact through a protein complex to suppress cyst formation. Their hypothesis is supported by our results that indicate Bicc1 binding to ANKS3 bridge ANKS6 interaction. These results highlight the roles of SAM domain-containing proteins in a network that is relevant for kidney functions.

Furthermore, our studies focus on explore the regulation of Bicc1, ANKS3 and ANKS6 assembly in homo- and hetero-oligomers by a Bicc1 target mRNA. We have shown Bicc1-KH domains binds target mRNAs while SAM-domain polymerization mediates recruitment in large foci for translational silencing (Results part I and (Piazzon et al., 2012)). Our results indicate that binding an mRNA target influences the cytoplasmic assembly into large foci overcoming the fact that Bicc1 polymerization can be halted by Bicc1-ANKS3 hetero-oligomers in absence of target mRNA. Our biochemical analysis identified the interaction between Bicc1 and ANKS3, specifically involving their respective KH and ANK domains. Thus, a possibility is that ANKS3 and the target mRNA compete both for KH domains binding, but that, under the conditions examined, the RNA take over as primary partner and significantly alters the protein complex topology.

87 Whether or how Bicc1-ANKS3-ANKS6 assembly can regulate specific ciliary or cytoplasmic signaling pathways, remains to be determined. In the cAMP/PKA signaling pathway, AC6 and PKIa mRNAs are translationally repressed by Bicc1 SAM-mediated polymerization (Results part I and (Piazzon et al., 2012)). However, we have not observed an effect of ANKS3 or ANKS6 in regulating Bicc1-mediated mRNA silencing.

Importantly, protein-protein interactions can also determine the recruitment of a protein in a specific compartment. Bicc1 cytoplasmic localization has been extensively described (Results part I and (Lian et al., 2014; Maisonneuve et al., 2009; Stagner et al., 2009)) in vitro and in vivo. More recently, Bicc1 ciliary localization was demonstrated in LLCPK renal cells (Mohieldin et al., 2015) and at centrosomes of dividing mIMCD3 cells where it seems to stimulate transcription of specific target mRNA, rather than inhibiting it (Iaconis et al., 2017). ANKS6 is recruited by Inversin into cilia leading to activation of NEK8 kinase and phosphorylation of ANKS6 (Czarnecki et al., 2015; Hoff et al., 2013). NEK8 was suggested to indirectly phosphorylate polycystin-2 (Sohara et al., 2008) and to bind ANKS3 (Czarnecki et al., 2015; Yakulov et al., 2015), but when mutated, NEK8 hyperactivates YAP signaling and is a cause of polycystic kidneys (Grampa et al., 2016; Liu et al., 2002).

Our results found Bicc1 interaction with NEK7 kinase. Whether NEK7 phosphorylates Bicc1 or ANKS proteins is still not determined. Nevertheless, NEK7 knockdown did not alter Bicc1 translational repression of target mRNAs. Interestingly, Bicc1 curb NEK7 protein levels in vivo. Classically, NEK7 has been linked to mitotic progression downstream for NEK9 (Belham et al., 2003; Bertran et al., 2011). Recent evidence places NEK7 as a switch between mitosis and inflammasome assembly since Nek7 binds directly to the innate immunity receptor NLRP3 (Shi et al., 2015) and this event seems to be mutually exclusive from mitosis, at least in macrophages. Remarkably, the NLRP3 receptor is expressed in tubular epithelial cells and it is increased in a renal inflammation and fibrosis rat model (Romero et al., 2017).

It would be interesting to elucidate whether Bicc1 interaction and increased NEK7 protein levels in the kidneys, have direct effects over ANKS3 mediated cytoplasmic

88 retention of NEK7, either to regulate cell cycle progression or inflammatory responses or both.

Bicc1 and CTLH complex novel interaction

A proteomic screen and validation by coimmunoprecipitation, yeast two-hybrid and density gradient fractionation assays revealed that Bicc1 in its non-polymerized form interacts with the CTLH complex. We showed that in sharp contrast to orthologous Gid proteins in lower eukaryotes, the endogenous CTLH complex in mammalian mIMCD3 cells does not decrease the rate-limiting gluconeogenic enzyme FBP1, but rather increases it, at least in cells depleted of Bicc1. Nevertheless, deletion of Bicc1 in mice led to reduced glycemia and diminished expression of the gluconeogenic enzymes PEPCK and FBP1 specifically in kidneys but not in liver, correlating with increased accumulation of CTLH complex. Analysis in cultured mIMCD3 cells indicates that CTLH complex in turn curbs the levels of Bicc1. These results reveal multilayered regulation of FBP1 expression by Bicc1 and CTLH complexes (Fig. 32): Besides mediating a net increase in FBP1, Bicc1 can also inhibit a stimulatory input of CTLH subunits, whereas reciprocal inhibition of Bicc1 in its non-polymerized form by the CTLH complex will facilitate negative feedback regulation.

Figure 32. Regulatory interactions between CTLH complex and low molecular weight (LMW) Bicc1. An epistasis analysis using RNAi and CRISPR in the model cell line mIMCD3 shows that Bicc1 stimulates FBP1 expression ex vivo, although independently of its interaction with the CTLH complex. Instead, the CTLH complex and Bicc1 mutually downregulate each other Stippled arrow: Effect of CHLT in Bicc1-deficient mIMCD3 cells.

Among the proteins that co-purified with inducible TAP tag-purified Bicc1, we found CNOT1 and several other subunits of the CCR4-NOT deadenylase, excluding NOT- 4, -5, -6/6L or -9. The absence of certain subunits may explain why Bicc1 does not

89 overtly alter the length of poly(A) tails of target mRNAs analyzed in HEK293T cells or in mouse kidneys (Piazzon et al., 2012; Tran et al., 2010).

Bicc1 interaction with the mammalian CTLH complex subunits.

In this study, we focused our attention on a novel interaction of Bicc1 with the CTLH complex. In S. cerevisiae, the core of the homologous Gid complex is formed by Gid1 and Gid8, whereas Gid5 and Gid9 recruit Gid2 and binds to the core by the interaction between Gid9 and Gid8 (Menssen et al., 2012; Pitre et al., 2006). Gid2 and Gid9 are RING domain E3 ubiquitin ligases. GID7 is another direct partner of GID1. They each contain a CTLH motif that is also found in Gid1 and Gid8 and in mammalian RanBP9 (GID1), RMND5a (GID2), TWA1 (GID8), MAEA (GID9), WDR26 (GID7) and in MKLN1 (Kobayashi et al., 2007; Menssen et al., 2012).

To our knowledge, our yeast two-hybrid assays of mammalian CTLH complex establish for the first time the strength of interaction between its components. We show that the core interactions occur between TWA1 and ARMC8, RMND5A and RANBP9, which differs from the homologous Gid complex. Additional strong interactions were observed between RANBP9 and WDR26, and between MAEA and RMND5A, which was also described in the yeast homologs.

Of all CTLH subunits, WDR26 was most enriched by TAP tag-purified Bicc1. WDR26 is a cytoplasmic protein that can inhibit MAPK-induced activation of serum response transcription factors (Zhu et al., 2004). WDR26 and other WD40 repeat proteins such as the UV-damaged DNA protein (DDB1) recruit substrates to the E3 Ub ligase CUL4 (Higa et al., 2006). TAP tag purification of Bicc1 also enriched DDB1, but not CUL4. Possibly, WDR26 and/or DDB1 are substrate-specific adaptors for more than one E3 Ub ligase.

RanBP9 has been mainly hypothesized to be a scaffolding protein with diverse functions according to those protein interactions (Salemi et al., 2017). RanBP9 also binds the RNA-binding KH domain protein Fragile Mental Retardation (FMRP) (Menon et al., 2004). However, our two-hybrid assay detected no interaction with Bicc1 KH domains (not shown). Of known RanBP9 partners (Salemi et al., 2017)

90 TAP-tagged Bicc1 also enriched Yippee-like-5 (YPEL5) and low levels of Ran itself and Ran-binding proteins RanBP10, RanBP2, RanGAP1, but not Polo-like kinase (Plk1). RanBP9 localizes in the cytoplasm and nucleus and, in a complex with other CTLH subunit inhibits the deacetylase HDAC6, a known activator of cilia disassembly (Pugacheva et al., 2007; Salemi et al., 2015).

Muskelin1 and other CTLH components were recovered in a proteomic screening using ciliary proteins (including some IFT proteins) as baits in HEK293T cells (Boldt et al., 2016). Future studies should investigate whether Bicc1-CTLH complexes may influence mitotic spindle assembly or cilia disassembly.

Bicc1 attenuates renal expression of CTLH complex subunits

Besides physically interacting with the CTLH complex, Bicc1 diminished the accumulation of some of its subunits. For WDR26, this effect reached significance, and we noted a similar trend for RanBP9 and GID4. In S. cerevisiae, the Gid complex becomes active when it binds Gid4, a regulatory subunit that accumulates within minutes after addition of glucose to degrade gluconeogenic enzymes (Chen et al., 2017b; Santt et al., 2008). A function for human GID4 is elusive, and binding to CTLH complex to our knowledge has not been demonstrated previously. We found that GID4 was present in HEK293T cells and partitioned with other CTLH subunits in sucrose density gradients. GID4 also interacted with RanBP9 and RMND5a in two- hybrid assays. Although this topology differs from S. cerevisiae where Gid4 is recruited by Gid5 (Menssen et al., 2012), these data suggest that human GID4 can bind the mammalian CTLH complex. Interestingly, the CTLH complex that was enriched by TAP-tagged Bicc1 contained no detectable GID4. Possibly, GID4 recruitment to mammalian CTLH complex is inhibited by regulatory factors despite the presence of high glucose. In addition, Bicc1 may repress GID4 expression. However, testing whether Bicc1 prevents GID4 access was impossible under the conditions examined because GID4 binding to CTLH complexes in HEK293T cells is below detection already in the absence of Bicc1 (Kobayashi et al., 2007). Thus, other factors likely restrict access of mammalian GID4 to the CTLH complex. Alternatively,

91 CTLH complexes containing GID4 may be rapidly degraded, as described in

S. cerevisiae (Santt et al., 2008).

The CTLH complex preferentially binds unpolymerized Bicc1 via its SAM domain, thereby limiting Bicc1 accumulation

We found that depletion of CTLH complex subunits by RNAi increased endogenous Bicc1 levels in mIMCD3 cells. In the simplest possible model, the CTLH complex accelerates Bicc1 turnover by direct binding. In keeping with this idea, Bicc1 directly interacted with the CTLH subunit ARMC8 in two-hybrid assays. ARMC8 has previously been shown to reduce the accumulation of E-cadherin and of α- and β- catenin to promote cell invasiveness in various cancers (Jiang et al., 2015; Zhao et al., 2016). Despite associated E3 ligases of the CTLH complex, ARMC8 targets α- catenin for proteasomal degradation independently of polyubiquitination (Menssen et al., 2012; Suzuki et al., 2008). Interestingly, only the non-polymerizing mutant Mut- D, but not WT Bicc1 interacted with ARMC8, suggesting self-polymerization protects Bicc1 against the CTLH complex. Compatible with this idea, cytoplasmic foci of polymerized Bicc1 did not detect amounts of CTLH complexes, as determined by anti-RanBP9 co-immunostaining. Together, these findings suggest that binding to CTLH complexes depletes a pool of free Bicc1 that is not stabilized by SAM domain- mediated self-polymerization (Fig. 32).

While the CTLH complex is not the only regulator of Bicc1 degradation, our finding that it binds Bicc1 via its unpolymerized SAM domain suggests that Bicc1 has distinct functions depending on its state of polymerization. These results also show that in our TAP strategy for mass spectrometry screening, molecules of monomeric and polymeric Bicc1 are purified, suggesting that the validation of Bicc1 interactors has to distinguish between non-polymeric or polymeric Bicc1 to better attribute Bicc1 functions. It will be interesting to determine whether a pool of unpolymerized Bicc1 outside cytoplasmic Bicc1 foci specifically associates with cilia or centrosomes, or whether its availability is regulated by cell metabolism or vice versa.

92 Role of Bicc1 in renal gluconeogenesis

A switch of glucose metabolism from glycolysis to gluconeogenesis is important during fasting and at birth when newborns feed on milk that largely lacks glucose. During prolonged fasting, up to 25% of gluconeogenesis derives from extrahepatic sources such as kidney and intestine (Gerich, 2010). Our findings demonstrate an essential role for Bicc1 in promoting blood glucose homeostasis and renal expression of gluconeogenic enzymes PEPCK and FBP1 in newborn mice.

We did not study the transcriptional downregulation of PEPCK1 in the Bicc1-/- kidneys or other causes of reduced FBP1 protein levels. Future studies should address this topic because PEPCK is known to be extensively controlled at the transcription level since it has several responsive elements in its promoter including those for insulin, glucocorticoid receptor, thyroid hormone or cAMP (Jitrapakdee, 2012). Increased cAMP dissociates the catalytic and regulatory subunits of PKA. The catalytic subunit can translocate to the nucleus and phosphorylates CREB (pCREB) which in turn can bind cAMP-responsive element (CRE) together with its co-activator CRTC2/TORC2, enhancing the transcription of gluconeogenic genes such as pyruvate carboxylase (PC), PEPCK, and Glucose-6-phosphatase (G6Pase) (Jitrapakdee, 2012).

FBP1 is allosterically regulated by cAMP/PKA signaling. cAMP activation of PKA can phosphorylate PFK2/FBPase2 that catalyzes the conversion of fructose-2,6- phosphate into fructose-6-phosphate, thus reducing the availability of this key FBP1 allosteric inhibitor, thereby increasing FBP1 activity (Murray et al., 2003).

In Bicc1-/- kidneys, increased cAMP levels at early postnatal levels (P4) (Piazzon et al., 2012). Concurrently, Bicc1-/- cystic lining cells showed increased pCREB staining (Bernet F. EPFL TH_6606). Despite these findings, Bicc1-/- kidneys drastically reduced PEPCK1 transcription, one possible explanation would be that activity of other transcription factors and coactivators, including FoxO and PGC1a (Quinn and Yeagley, 2005), may be deficient in absence of Bicc1 thus limiting gluconeogenic genes transcription. Assessing Bicc1 role in their regulation of transcription would provide further insights into disease-relevant mechanisms.

93 FBP1 regulation by the CTLH complex

In S. cerevisiae the Gid complex inhibits gluconeogenesis by targeting FBP1 and PEPCK for proteasomal degradation by the proline N-end rule, in a process where the assembly of the full complex is required for its activity (Chen et al., 2017b; Santt et al., 2008). FBP1 levels decreased in Bicc1 mutant kidneys without a corresponding decrease at the mRNA level. Depletion of Bicc1 by RNAi also decreased FBP1 protein in cultured IMCD3 cells, whereas ectopic expression of Bicc1 in HEK293T cells increased it. To test whether endogenous Bicc1 maintain steady levels of FBP1 protein by inhibiting the CTLH complex, we depleted the subunits TWA1 or WDR26 in cultured mIMCD3 cells. Knockdown of the CTLH complex by RNAi did not stabilize FBP1 but rather diminished it, at least in Bicc1 mutant cells. We also observed no accumulation of polyubiquitinated FBP1 intermediates in cells treated with MG132. While these observations do not rule out that Bicc1 may stabilize FBP1 protein by inhibiting CTLH complexes in renal proximal tubules, this is clearly not the case in proliferating mIMCD3 cells. In basal-like breast cancer, FBP1 expression can be transcriptionally repressed by Snail1 to stimulate glycolysis and confer a metabolic advantage to cancer stem cells (Dong et al., 2013). FBP1 is also depleted in nearly all renal cell carcinomas to promote their aggressiveness (Li et al., 2014). It would be interesting to investigate whether this process involves Bicc1 or its interaction with CTLH complexes.

94 5. Results part III: Role of Bicc1 in cell metabolism

5.1 Analysis of Bicc1 metabolic interactors and Bicc1-/- mouse model metabolome

Mounting evidence from mouse and rat models suggests that PKD progression relies on important metabolic alterations, including a Warburg-like phenotype (Klawitter et al., 2013; Menezes et al., 2016; Rowe et al., 2013). After validating Bicc1 binding to selected metabolic enzymes found in the interactome screen, this chapter focuses on the analysis of the metabolome and its correlation with other metabolic alterations in Bicc1-/- kidneys and cell lines.

5.1.1 The Bicc1 interactome is enriched for metabolic proteins

A) GO-term analysis shows enrichment of proteins related to metabolic processes.

To obtain an overview of the biological processes mediated by Bicc1-interacting proteins, I used PANTHER™ Version 12.0 (released 2017-07-10) for GO-term analysis (Mi et al., 2017). Based on the full list of Bicc1 interactors with more than 2- fold exclusive unique peptide counts enrichment in Bicc1 IP compared to negative control (Table 1 section 4.1), the analysis revealed metabolic processes (GO:0008152) as the second-most enriched GO term, with 48% of the hits belonging to this class of proteins (Fig. 33).

95 Figure 33. “GO biological process” terms enriched for Bicc1 interactors. Pie chart of the 10 GO terms that were significantly enriched for proteins co-purified by TAP-tagged Bicc1 (Table 1). After the broad GO term Cellular Processes, Metabolic processes (red) is the second most frequent GO term linked to Bicc1-enriched proteins.

For example, enzymes in important metabolic processes that co-purified with Bicc1 included Acetyl-CoA Carboxylase Alpha (ACACA) and Fatty acid synthase (FASN) which are rate-limiting for fatty acid synthesis by catalyzing the carboxylation of acetyl CoA to malonyl CoA and the synthesis of long-chain fatty acid from acetyl-CoA and malonyl-CoA, respectively (Mashima et al., 2009). Among mitochondrial enzymes, we found the mitochondrial ADP/ATP translocase 2 (SLC25A5), which acts as a antiporter for ADP translocation from the cytoplasm into the mitochondrial matrix and conversely, for ATP translocation from mitochondrial matrix to the cytoplasm (Clémençon et al., 2013). We also identified the translocase of inner mitochondrial membrane 50 (TIMM50), a subunit of the TIM23 inner mitochondrial membrane translocase complex, which recognizes the mitochondrial targeting signal on cargo proteins designated to the inner membrane of mitochondrial matrix (Endo et al., 2003). Another candidate hit associated with the Metabolic Processes GO term was the multifunctional enzyme CAD (carbamoyl-phosphate synthetase 2, aspartate transcarbamylase, and dihydroorotase), which catalyzes the synthesis of pyrimidines

(Huang and Graves, 2003). These results support the hypothesis of a novel role of Bicc1 as a regulator of disease-relevant metabolic processes.

96 B) In vitro validation of Bicc1 interactor candidates and analysis of their protein levels in Bicc1-/- kidneys.

To validate candidate hits, I assessed their binding to endogenous Bicc1 using co- immunoprecipitation assays. Endogenous FASN bound Bicc1 in cultured mIMCD3 cells while it failed to do so in the IgG negative control. I also observed efficient binding of endogenous FASN, ACACA and CAD to Bicc1-HA-StrepII-tagged in doxycycline-induced T-Rex HEK293 cells. As negative control, no binding was observed in non-induced cells (Fig. 34). These results indicate that these metabolic enzymes bind to Bicc1.

Figure 34. Bicc1 interacts in vitro with FASN, ACACA and CAD proteins. Co-immunoprecipitation of endogenous Bicc1 with FASN in mIMCD3 cells (top left) and Bicc1-HA-StrepII with FASN, ACACA, and CAD in T-Rex HEK293. IgG or non-induced cells served as negative controls, respectively (n=1).

Serendipitously, I observed during these co-IPs that induction of Bicc1-HA-StrepII- tagged consistently increased the levels of FASN, CAD, and ACACA, in the input samples. To validate such a regulatory interaction in vivo, I monitored protein levels of FASN, ACACA, and CAD in wild-type compared to Bicc1-/- newborn kidneys. WB analysis showed that FASN levels are significantly decreased, and the same trend is observed for ACACA and CAD levels in Bicc1-/- kidneys (Fig. 35).

97 Figure 35. Bicc1 maintains protein levels of FASN, ACACA and CAD in the mice kidneys. Western blot analysis of FASN, ACACA, and CAD in P2 stage kidneys of the indicated genotypes (n≥4 per genotype). Protein levels were normalized to γ-Tubulin expression. Error bars represent SEM. Student t-test, *p <0.05.

These results suggest a novel role for Bicc1 binding and maintaining the protein levels of several rate-limiting metabolic enzymes in vivo and in vitro. In the future, it would be interesting to identify molecular mechanisms underlying this relationship between Bicc1 and enzymes involved in fatty acid and pyrimidine synthesis, as well as a potential effect of Bicc1 on their enzymatic activities. Given the novel role of Bicc1 in maintaining the protein levels of these enzymes in kidneys and the enrichment of these and several other 'metabolic proteins' in our interactome screen, it is tempting to speculate that Bicc1 may primarily function as a regulator of cell metabolism. In such a scenario, metabolic defects may directly contribute to disease progression in the Bicc1-/- model, possibly already at early postnatal stages.

98 5.1.2 Metabolomic analysis of Bicc1-/- kidneys.

Metabolic changes that promote cyst formation in ADPKD include a Warburg-like glucose metabolism and fatty acid oxidation defects (Magistroni and Boletta, 2017; Menezes et al., 2016; Rowe et al., 2013). To systematically characterize metabolic alterations in Bicc1-/- mice kidneys, we conducted an untargeted metabolomic screen on freshly frozen kidneys of 8 WT and 8 Bicc1-/- mice at post-natal day 2 in collaboration with Dr J. Ivanisevic at the metabolomics facility of the University of Lausanne. Compared to a targeted approach, an untargeted screen reduces bias in metabolite detection, since the output is a global overview of the relative changes. However, the data analysis starts from a large number of features that must be manually curated to finally achieve the validation of the significant identified metabolites. Using HILIC ESI Mass spectrometry acquisition in positive and negative ionization modes to produce positive or negative ions, combined with tandem mass spectrometry matching, we validated identities of detected features (or peaks). Data processing including signal intensity drift corrections and quality control filtering (described in materials and methods).

An unsupervised principal component analysis (PCA) from all metabolite peaks (3482 and 3767 peaks in positive and negative mode) showed a clear segregation of the tested samples, indicating that mutant kidneys clearly have a distinct metabolomic profile (Fig. 36A). Visualization of the significantly different features and fold changes in Bicc1-/- compared to WT kidneys are represented in volcano plots (Fig. 36B).

99 A Positive acquisition Negative acquisition

B 10 10 ) ) 8 ttest 8 . ttest . e e alu 6 6 alu v v p p ( ( 10 4 4 10 Log Log − − 2 2 0 0

−2 −1 0 1 2 −2 −1 0 1 2 Log2FC Log2FC

Figure 36. Renal metabolomic analysis unambiguously distinguishes Bicc1-/- mice. (A) Principal component analysis (PCA) and loading biplots (Pareto scaling) of the kidney metabolome from wild-type (WT), Bicc1-/- (KO) mice and pooled samples (QC) obtained a satisfactory validation according to clear discrimination between groups. All detected features driving sample grouping (loadings) are projected in the background in light gray. (B) Volcano plots from all detected features summarizing fold-change (x axis) and t-test criteria (y axis) for metabolites that are up- (positive values in red) or down-regulated (negative values in green) in Bicc1-/- kidneys. Vertical dotted lines represent fold change of 1.5 or -1.5, and horizontal dotted line represent the threshold of statistical significance (p = 0.05). Plots on the left and the right side represent data from the positive and negative acquisition modes, respectively.

After removing the chemical and bioinformatic noise and redundant isotope, adduct, and ion-source fragment features, and after manual curation based on the biological relevance of each feature (i.e. endogenous vs. exogenous nature of the metabolite), putative relevant metabolites were validated using the accurate mass, retention time and MS/MS data. Accurate masses were searched against databases METLIN and HMDB (Tautenhahn et al., 2012; Wishart et al., 2013) reducing the list of putative identified features to 468 and 535 in positive and negative mode, respectively. Final

100 identification were performed by matching acquired MS/MS data for altered metabolites with standard MS/MS data developed and online available at METLIN database (Benton et al., 2008; Tautenhahn et al., 2012).

Overall, 34 significantly altered metabolites were identified (with taurine being found in both positive and negative ionization modes). The fold changes, calculated as ratios of ion counts (mean values) in KO compared to WT kidneys, and p-values of each identified altered metabolite are shown in Table 3. 20 out of 34 (58%) validated metabolites significantly increased in Bicc1-/- compared to WT kidneys, while 14 (42%) significantly decreased.

The pattern of variation of identified metabolites in WT and Bicc1-/- kidneys is represented in a Heat map combined with Hierarchical Cluster Analysis in “R” software based on altered metabolite pattern similarity between WT and Bicc1-/- kidneys (Fig. 37). In agreement with the PCA analysis, alterations in the identified metabolites clearly segregated the samples according to their genotypes. Thus, several features that likely represent specific metabolites accumulate at significantly different levels in Bicc1-/- compared to WT kidneys.

101 Figure 37. Heat map and hierarchical cluster analysis of metabolites validated by MS/MS pattern matching as significant de-regulated in Bicc1-/- kidneys. Metabolites listed on the right characterize the metabolic alterations specifically related to Bicc1 lost. A z-transformation was performed on all peak areas to scale the data. Significant decreases in Bicc1-/- kidneys compared to WT are shown in red, while significant increases are in green. Color intensities represent the magnitude of the difference relative to the average level of a given metabolite across all samples.

102 To have a better idea of the metabolic changes resulting from Bicc1 loss-of-function, I performed a pathway analysis using the online tool MetaboAnalyst 3.0 (http://www.metaboanalyst.ca/) (Xia and Wishart, 2016).

Graphs reporting pathways decreased and/or increased in Bicc1-/- kidneys are shown in (Fig. 38). Metabolic pathways labeled and localized towards the top right quadrant represent those having undergone the biggest changes. Their detailed analysis is presented in the following section.

103 Figure 38. Pathway analysis of all metabolites that were validated as de-regulated in Bicc1-/- kidneys. Each figure displays all matched pathways as circles. The sizes and colors of each circle represent pathway impact values (x-axis) and p-values (y-axis), respectively, obtained by enrichment analysis using the KEGG database. Graphs representing the pathway analysis were generated using MetaboAnalysist 3.0 (Xia and Wishart, 2016).

104 5.1.3 Metabolic pathways altered in Bicc1-/- kidneys

A) Purine and pyrimidine metabolism

Validated metabolites that were most deregulated in the Bicc1-/- kidneys included 14.7-fold increase in allantoin, the final degradation product of uric acid in mice (Oda et al., 2002) and 0.3-fold decrease in inosine monophosphate (IMP), the second- most strongly depleted metabolite. Both correlate with an increase purine catabolism which is a major source of uric acid (Maiuolo et al., 2016) (Fig. 39). Furthermore, hypoxanthine which is oxidized into xanthine and once again into uric acid significantly decreased 0.6-fold (Table 3). These data imply that Bicc1-/- kidneys catabolize purine in large excess.

Figure 39. Purine biosynthesis Purine metabolism. In the de novo pathway, 5-phospho-α-D-ribosyl 1-pyrophosphate (PRPP) is metabolized to IMP which can be used to generate GMP and AMP. Catabolism of these nucleotides generate guanine, hypoxanthine or adenine that can be used in the salvage pathway to regenerate purines. Red and green squares represent metabolites down- or up-regulated, respectively, in Bicc1-/- kidneys. Figure adapted from (Ichida et al., 2012).

105 Levels of GMP (guanosine monophosphate, a derivative of IMP) were reduced accordingly (0.8-fold) in Bicc1-/- kidneys. Generation of IMP utilizes the amino acids glycine, aspartate, and glutamine. Aspartate significantly decreased (0.7-fold) whereas glutamine increased (1.3-fold), in Bicc1-/- mutants.

Deficits in specific metabolites tend to be compensated in part by parallel pathways. In the case of nucleotides, the purine salvage pathway uses intermediate metabolites originating from catabolic reactions like adenine and a co-substrate 5-phospho-α-D- ribosyl 1-pyrophosphate (PRPP), to generate AMP. In this context, the concomitant increased levels of AMP and adenosine in Bicc1-/- mutants (1.3 and 1.5-fold, respectively) may be explained by such mechanism (Ichida et al., 2012).

In the pyrimidine synthesis pathway, both UMP (uridine monophosphate) and CMP

(cytidine monophosphate) were significantly reduced (0.8-fold) in Bicc1-/- kidneys.

Interestingly, I found that CAD, the rate limiting enzyme for the three first steps in the de novo synthesis of pyrimidines, binds to Bicc1 in vitro and that Bicc1 is important to maintain CAD protein levels in the kidneys (Fig 34-35). These results suggest that

Bicc1 may have a direct role in regulating pyrimidine nucleotide metabolism (Fig. 40).

Figure 40. Pyrimidine de novo synthesis. Steps 1-3 (red circles) are catalyzed by the trifunctional enzyme CAD (carbamoyl phosphate synthetase 2+ aspartate transcarbamoylase + dihydroorotase). Figure adapted from (Löffler et al., 2016). Red and green squares represent metabolites down- or upregulated, respectively, in Bicc1-/- kidneys. See acronyms in text.

106 B) Arginine/proline metabolism

The second-most enriched metabolite in Bicc1-/- kidneys is dimethylarginine (7.8- fold). This metabolite is recognized as a toxic non-proteinogenic amino acid in kidney and lung chronic diseases (Leone et al., 1992; Vallance and Leiper, 2004; Zakrzewicz and Eickelberg, 2009). Asymmetric dimethylarginine (ADMA) is generated within polypeptides by post-translational arginine methylation (Morales et al., 2016). Symmetric dimethylarginine (SDMA) is an isomer of ADMA and is released together with ADMA from polypeptides by proteolysis, and their elimination directly depends on renal excretion (Tain and Hsu, 2017). The accumulation of ADMA and SMDA correlates with the pathology of human ADPKD and with chronic kidney disease likely because it inhibits nitric oxide (NO) synthesis and thereby contributing to inflammation and oxidative stress (Klawitter et al., 2014; Rysz et al., 2017) (Fig.41).

While ADMA can be demethylated to form Arginine, we did not observe increased arginine levels in the Bicc1-/- kidneys. However, two metabolites involved in arginine metabolism, citrulline and ornithine significantly increased (1.6 and 2.1-fold). In the mitochondria, ornithine can form citrulline by binding carbamoyl-phosphate and subsequently in the cytoplasm, arginine can be synthesized from citrulline by Argininosuccinate synthetase (ASS) and Argininosuccinate lyase (ASL) (Curis et al., 2005). Citrulline and ornithine also participate in the urea cycle. Thus, in the Bicc1-/- mutants this mechanism of toxic waste product conversion from ammonia into urea seems to increase. A significant increase in hydroxypyruvic acid in Bicc1-/- kidneys (2.3-fold) suggests increased oxidative deamination of serine that may contribute to subsequent release of ammonia (Chung et al., 2010).

Alternatively, arginine, ornithine, glutamate/glutamine can serve as substrates for proline synthesis in specific tissues. Among these, the kidney is one of the most active organs in terms of proline synthesis (Wu et al., 2008). Proline and its metabolite hydroxyproline are enriched (2.4- and 2.2-fold, respectively) in Bicc1-/- mutants. They constitute one-third of the amino acids in collagens, which correspond to nearly 30% of body proteins. Proline has antioxidant properties (Kaul et al., 2008), and the accumulation of hydroxy-DL-proline-2-14C has been associated with reduced

107 hydroxyproline metabolism in patients with renal insufficiency (Hart et al., 1976) (Fig. 41).

Figure 41. Arginine metabolism Arising as products of arginine catabolism, SMDA and ADMA negatively regulate nitric oxide synthase (NOS), whereas Ornithine can generate proline in a reaction catabolized by ornithine aminotransferase (OAT) or generate polyamines by ornithine decarboxylase (ODC). Green squares represent metabolites accumulated in Bicc1-/- kidneys. Figure adapted from (Scott et al., 2014).

C) Citrate (TCA) cycle intermediates

The TCA cycle plays a central role to produce both energy and biosynthetic precursors (Anderson et al., 2017). Only two intermediates were detected and significantly altered in the Bicc1-/- kidneys. Citrate and succinate increased 6.8- and 3.3-fold respectively, compared to WT kidneys. Importantly, in the cpk mouse model of PKD, a non-targeted renal metabolomic analysis at P14 revealed significantly increased citrate and acetyl-CoA levels and concomitant downregulation of downstream TCA cycle metabolites revealing a shunt towards the synthesis of the oncometabolite 2-hydroxyglutarate at the expense of glutamine utilization. These

108 findings were further validated in ADPKD human samples (Hwang et al., 2015). By contrast, Bicc1 mutants did not show elevated acetyl-CoA or 2-hydroxyglutarate levels, or downregulation of TCA intermediates.

A critical pathway to replenish the TCA cycle with α-ketoglutarate (αKG), in proliferating cells is glutaminolysis (Anderson et al., 2017). Glutamine was not depleted in Bicc1-/- kidneys but rather increased (1.3-fold). Thus, a shortage of αKG and Acetyl-CoA to fuel the cycle and accumulate citrate seems unlikely, even though they remained unidentified in our analysis. Citrate accumulation may reflect enhanced production of cytosolic acetyl-CoA for the synthesis of fatty acids, probably to meet the demand for lipids that are required for membrane formation (López et al.,

2005) (Fig. 42).

Figure 42. Tricarboxylic acid (TCA) cycle Starting from Acetyl-CoA, derived for example from pyruvate, arrows indicate the direction of the metabolites through the cycle. Three molecules of NADH, one molecule of FADH2, two molecules of CO2 and one ATP are produced per cycle. Green squares represent metabolites enriched in Bicc1-/- compared to WT kidneys. Figure adapted (O’Neill et al., 2016).

109 D) Glutathione metabolism

Cystine, the most depleted metabolite (0.2-fold) in Bicc1-/- kidneys, is the oxidized product of cysteine, which is the limiting amino acid for Glutathione synthesis in humans (Lyons et al., 2000) and other species (Chung et al., 1990). Importantly, both reduced (GSH) and oxidized (GSSG) forms of glutathione comprise the major antioxidant system in aerobic cells, are increased in Bicc1-/- kidneys 7.1 and 4.5-fold, respectively. Bicc1-/- kidneys likely synthesize these metabolites in elevated amounts to cope with oxidative stress. In glutathione metabolism, methionine and taurine are involved as a precursor and the end product, respectively. Both were reduced in

Bicc1-/- kidneys (0.7- and 0.8-fold, respectively). Interestingly, taurine has renoprotective, antioxidant and anti-inflammatory properties (Karbalay-Doust et al., 2012; Latchoumycandane et al., 2014), and it dampens uric acid production in rats with kidney injury (Feng et al., 2017) (Fig. 43).

Figure 43. Glutamate and cystine transporters cooperate to regulate glutathione synthesis The cystine antiporter incorporates cystine inside the cell in exchange for glutamate. The glutamate transporter takes up glutamate from extracellular milieu to restore the intracellular balance. Oxidative stress can reduce the capacity of glutamate transporter. Red and green squares represent metabolites that are diminished or enriched, respectively, in Bicc1-/- kidneys. Adapted from (Porcheray et al., 2006).

E) Ketogenesis

Bicc1-/- kidneys accumulate 3-hydroxy-3-methyl-Glutaric acid (HMG) and the ketone

3-hydroxybutyric acid (7.1- and 2.9-fold, respectively). Besides the liver, where

110 ketogenesis is mainly occurring, the kidney has been also recognized as a ketogenic organ. HMG is an important intermediate metabolite in the synthesis of ketone bodies from β-oxidation of fatty acids (Fig. 44).

Figure 44. Synthesis of 3-hydroxybutyrate from fatty acids and ketogenic amino acids. Ketogenesis begins with the conversion of Acyl-CoA to acetyl-CoA (Ac-CoA) via β-oxidation in mitochondria. Ac-CoA is further metabolized into ketone bodies (acetoacetate, 3HB, and acetone). Green ovals represent metabolites accumulated in Bicc1-/- kidneys. FFA: Free fatty acids, ILE: isoleucine, LEU: leucine, LYS: lysine, THR: threonine, TRP: tryptophan, TYR: tyrosine, 3HB: β-hydroxybutyrate, 3HBDH: 3HB dehydrogenase, AcetoAc-CoA: acetoacetyl- CoA, HMG-CoA: 3-hydroxy-3-methyl glutaryl-CoA, HMGCL: HMG-CoA lyase, HMGCS: HMG-CoA synthase. Adapted from (Miyazaki et al., 2015).

111 F) Carnitine pathway

For β-oxidation in mitochondria, long chain fatty acids are activated by CoA. The resulting acyl-CoA is attached to carnitine by Carnitine palmitoyl transferase I (CPTI) to produce acylcarnitine and deliver it into the mitochondrial intermembrane space. From there, acylcarnitine enters the mitochondrial matrix, where Carnitine palmitoyl transferase II (CPTII) liberates the carnitine by re-attaching the acyl moiety of acylcarnitine to CoA to produce the acyl-CoA that undergoes β-oxidation (McCoin et al., 2015). Free carnitine, or acetylcarnitine after acetylation by carnitine acetyl transferase (CAT), can be transported back from the mitochondrial matrix to the cytoplasm. Bicc1-/- kidneys accumulate acetylcarnitine (1.6-fold) and have reduced levels of carnitine (0.7-fold) (Fig. 45).

Figure 45. Carnitine shuttle In contrast to medium-chain fatty acids, long-chain fatty acids rely on the carnitine shuttle to cross inside the mitochondrial matrix. Red and green squares represent metabolites downregulated and increased, respectively, in Bicc1-/- kidneys. CPT I: carnitine palmitoyltransferase type 1. CPT II: carnitine palmitoyltransferase type 2. CACT: carnitine- acylcarnitine translocase. Figure adapted from (Wong and Crawford, 2011).

112 G) Glycerophospholipid metabolism

The third-most depleted metabolite in Bicc1-/- kidneys was choline (0.5-fold). Choline is the precursor of phosphatidylcholine (PC), a phospholipid essential for maintaining the cell membrane integrity for adequate flow of substances across the plasma membrane (Jacobs et al., 2010). Choline is also the main source of methyl groups (in the S-adenosylmethionine synthesis pathways) important for DNA, RNA, lipids and protein methylation (Glier et al., 2014). In line with this result, LysoPC (16:0) consisting of one chain of palmitic acid at the C-1 position was also diminished (0.7- fold) in Bicc1-/- kidneys. LysoPCs can mediate cell differentiation, proliferation, survival, adhesion, migration, and invasion by binding to lysophospholipid receptors (LPL-R) (Ishii et al., 2004).

H) Intermediates in inflammatory pathways

Urocanic acid is a metabolite of histidine deamination. Increased levels of urocanic acid in Bicc1-/- kidneys (2.9-fold), suggest a disturbance in the histidine pathway, which has been associated with systemic inflammation, oxidative stress and increased mortality in CKD patients (Watanabe et al., 2008). Histidine has anti- inflammatory and anti-oxidant properties due to the ability of the imidazole ring to scavenge reactive oxygen species that are generated during inflammation (Peterson et al., 1998). In addition, Bicc1-/- kidneys accumulated myoinositol (1.4-fold), a known second messenger in inflammatory cells that also increases in plasma of CKD patients (Chen et al., 2017a). These findings may reflect an associated inflammatory status in Bicc1-/- kidneys.

I) Uridine diphosphate glucose

Bicc1-/- kidneys also showed 0.8-fold reduced levels of UDP-glucose (Uridine diphosphate glucose. UDP-glucose is used as a lipid head group in glycosylated lipids such as, glucosylceramide (GlcCer), cholesterylglucoside (ChlGlc), and phosphatidylglucoside (PtdGlc) which are components of cellular membranes (Ishibashi et al., 2013). Alternatively, UDP-glucose can donate glucose for glycogen synthesis primarily in the liver, and to a lesser extent in kidneys (Adeva-Andany et

113 al., 2016). Since Bicc1-/- animals have reduced glycemia, an increase in renal glycogen synthesis is highly unlikely. A more plausible scenario is that UDP-glucose is diminished by cystic growth and the associated increased demand for membrane synthesis.

5.2 Additional metabolic alterations observed in Bicc1-/- mice

5.2.1 Bicc1-/- kidneys accumulate lipid droplets

In the lipid metabolism, lipid droplets (LD) are specialized storage organelles in cells, containing a core of neutral lipids, triglycerides and cholesterol. They originate in the endoplasmic reticulum, where the terminal enzymes catalyzing neutral lipid synthesis are localized. LDs protects from fatty acids induced lipotoxicity and their breakdown classically attributed to cytosolic lipases on the LD monolayer, releases free fatty acids into the cytoplasm to feed mitochondrial β-oxidation (Hashemi and Goodman, 2015). Studies in rodent models of ADPKD suggest that cystic growth also associates with perturbations in lipid metabolism (Klawitter et al., 2013; Menezes et al., 2016). Therefore, and because of my observation that Bicc1 interacts with FASN and

ACACA, I investigated whether Bicc1-/- kidneys show signs of altered lipid metabolism.

To detect lipid droplets, I stained kidney sections with Oil Red O dye. To our surprise, the number of lipid droplets significantly increased in Bicc1-/- kidneys compared to WT (Fig. 46A). By contrast, no such increase was observed at embryonic stages (E18.5), suggesting that Bicc1 inhibits lipid droplet accumulation specifically after birth (Fig. 46B).

To test whether Bicc1 may regulate lipid uptake, I cultured WT or CRISPR/Cas9 (sgBicc1) mIMCD3 cells during 24 hrs. in presence or absence of Oleic acid as an exogenous source of lipids. After staining with BODIPY 493/503, a marker of neutral lipids, cells were analyzed by flow cytometry. I did not observe significant changes in fluorescence intensity between the two cell lines before or after Oleic acid treatment compared to vehicle conditions (Fig. 46C). This result indicates that, under the

114 conditions analyzed, loss of Bicc1 does not alter lipid droplet content or uptake in cultured mIMCD3 cells.

Figure 46. Bicc1 prevents lipid droplet accumulation in kidneys, but not in cultured mIMCD3 cells. (A) Cryo-sections of WT and Bicc1-/- kidneys at postnatal day P2 stage stained with Oil red oil for lipid droplets (arrows). The average number of lipid droplets and nuclei counted in at least 4 different fields in two kidneys per genotype are shown to the right. Student t-test (** p-value <0.01). (B) Cryo-sections of WT and Bicc1-/- kidneys at embryonic day E18.5 were stained with Oil Red O. No lipid droplets were detected at this stage. (C) In vitro staining of lipid droplets in WT or sgBicc1 mIMCD3 cells using Bodipy 493/503. Representative images after Oleic acid (O.A) treatment as an external source of lipids. Quantification represents the fluorescent signal per number of cells per field (n=5) in each condition tested.

115 5.2.2 Bicc1-/- kidneys show increase autophagic flux, rather than decreasing it

Triglycerides and cholesterol stored in lipid droplets can also be broken down by lipophagy, which relies on the lysosomal degradative autophagic pathway (Liu and Czaja, 2013). To monitor autophagy, I analyzed the expression levels of the lipidated form of microtubule associated protein 1 light chain 3 (LC3-II), a marker of the maturation of autophagosomal membranes, and of the LC3-interacting protein p62/SQSTM1 that is degraded together with cargo of autophagosomes as they mature into autophagolysosomes (Klionsky et al., 2016). Western blot analysis revealed that while LC3-II protein levels increased, p62/SQSTM1 decreased in P2

Bicc1-/- kidneys compared to WT (Fig. 47A). These results suggest that Bicc1-/- kidneys significantly increase autophagy above baseline, likely because of the metabolic alterations described before. Thus, we can rule out defective autophagy as the cause of the accumulation of lipid droplets.

To assess how the loss of Bicc1 stimulates autophagy, I investigated two signaling pathways that regulate autophagy initiation, AMPK and mTOR (Kim et al., 2011) in Bicc1-/- kidneys. AMPK promotes autophagy by phosphorylation of autophagy- initiating kinase Ulk1. I observed decreased levels of total AMPK in Bicc1-/- kidneys compared to WT littermates (Fig. 47B). Activated pospho-AMPK negatively regulates mTOR (Huang and Manning, 2008). This result should be further confirmed using phospho-AMPK antibodies. A reduction of AMPK indicates that the Bicc1-/- mouse model resembles other models of PKD in altered cellular metabolic sensors (Chiaravalli et al., 2016; Takiar et al., 2011). mTOR normally inhibits autophagy. To assess mTOR activation I monitored two of its targets, phospho-4E-BP1 and phospho-p70 S6K (Jung et al., 2010). WB analysis showed that Bicc1-/- kidneys upregulated phospho-4E-BP1 and phospho-p70 S6K levels, rather than diminishing them. Total protein levels of 4E-BP1 also increased (Fig. 47C). These results agree with earlier studies that showed mTOR hyperactivation in Bicc1bpk/bpk mice and in ADPKD patients (Shillingford et al., 2006; Stephanis et al., 2017). Overall, these results show that Bicc1 loss-of-function

116 significantly increases basal autophagy in the kidney, despite concomitant upregulation of the mTORC1 pathway.

Figure 47. Bicc1 reduces autophagy while increasing mTORC1 levels. (A) WB analysis of LC3 I-II and p62/SQSTM1 protein levels in WT and Bicc1-/- kidneys. The levels of LC3-II were normalized to γ-tubulin. n≥4 per genotype. (B) WB analysis of total AMPK protein levels in P2 WT and Bicc1-/- kidneys. The levels of AMPK were normalized to γ-tubulin. N=3 per genotype. (C) WB analysis of the mTORC1 targets phospho- and total 4EBP1 and phospho-p70 S6K in P2 WT and Bicc1-/- kidneys. γ-tubulin or GAPDH were used as loading controls. n=3 per genotype. All bar graphs represent mean ± S.E.M. Student’s t-test was used to determine significant differences p-value *<0.05 **<0.01

To address whether Bicc1 loss-of-function similarly affects autophagy, I analyzed mIMCD3 clones that were depleted of endogenous Bicc1 by CRISPR/Cas9-editing 117 (sgBicc1), followed by doxycycline induction of a lentiviral tetO::HA-Bicc1; rtTA3 transgene (engineered by my colleague Dr. S. Fortier). Bafilomycin A (BafA1), an inhibitor of autophagosome and lysosome fusion and a neutralizer of lysosomal pH, was used to assess its impact on autophagic flux. Increased LC3-II levels after BafA1 treatment are indicative of enhanced autophagosomes formation. I found by WB analysis that cells depleted of Bicc1 (sgBicc) accumulated excess LC3-II after Bafilomycin A1 treatment, whereas re-introducing Bicc1 expression reverted this effect (Fig. 48A). These data indicate that Bicc1 may inhibit autophagic flux at least when cultured in high glucose DMEM.

For comparison, we cultured the same cells in glucose-free DMEM. In the absence of glucose, Bicc1 loss-of-function seemed to recapitulate similar effects as those observed in high glucose (even with a clearer tendency), where sgBicc1 cells showed increased levels of LC3-II and the add-back clone reverted this effect (Fig. 48B). Thus, Bicc1 may contribute to metabolic homeostasis by keeping basal levels of autophagy low. Although, additional experiments will be needed and sufficiently repeated to rule out clonal effects, including analysis of p62/SQSTM1, to ascertain that Bicc1 curbs autophagic flux in vitro similar to what I observed in vivo.

118 Figure 48. Bicc1 reduces autophagic flux in mIMCD3 cells. (A) WB analysis of LC3 I-II levels in WT, sgBicc1 and add-back inducible Bicc1 cells treated with Bafilomycin A1 (100 nM) or vehicle (EtOH) during 4 hours in high glucose or (B) glucose-free DMEM. For quantifications, LC3-II was normalized to γ-tubulin. Bars represent changes relative to WT vehicle treated cells (n=1).

5.2.3. Effect of Bicc1 loss on cell energy phenotype

Prompted by the metabolic disturbances observed, and to shed light on direct effect of Bicc1 loss-of-function in cellular metabolism, we evaluated metabolic phenotypes of sgBicc1 mIMCD3 cells using the XF extracellular flux analyzer (Seahorse). This technology allows to simultaneously measure two important bioenergetic pathways producing ATP, i.e. mitochondrial respiration and glycolysis. Glycolysis converts glucose to pyruvate, which in the cytosol can be reduced to lactate (independently of oxygen). This reaction releases protons into the extracellular medium through various proton pumps, leading to an extracellular acidification (Brooks, 2010). Glucose, fatty acids, and glutamate can be completely oxidized into CO2 and H2O through the TCA cycle, and the electrons carried by

NADH and FADH2 can flow down in the electron

119 transport chain (ETC) to form a transmembrane proton gradient across the inner mitochondrial membrane. At the end of the ETC, electrons serve to reduce O2, giving rise to H2O. Thus, the proton gradient drives phosphorylation of ADP to ATP as protons return to the mitochondrial matrix through the ATP synthase complex, coupling respiration with ATP production (Winer and Wu, 2014). The XF analyzer performs a kinetic measurement of respiration by assessing the oxygen consumption rate (OCR), defined by the decrease of oxygen in the culture medium over time. Concomitantly, this assay can determine the extracellular acidification rate (ECAR) by measuring how quickly the medium accumulates free protons (Fig. 49).

Figure 49. Overview of the pathways assessed by Seahorse technology. The orange panel represents the entrance of glycose in the glycolytic pathway. Catabolism of glucose into pyruvate which can be converted into lactate and excreted by monocarboxylate transporters (not shown). The citric acid cycle generates energy equivalents to feed the electron transport chain for cellular respiration. The electron transport chain can be blocked at different steps by adding specific inhibitors during analysis in the XF analyzer. The purple area represents the glycerol-3P cycle which links glycolysis to the electron transport chain by providing dihydroxyacetone-phosphate. The seahorse icons represent the parameters assessed by the XF analyzer. Figure adapted from (Zhang et al., 2012).

120 Simultaneously, I evaluated the effect of mitochondrial stressors which challenge the respiratory capacity of the cells to uncover potential defects. The addition of Oligomycin blocks the proton channel in the F0 subunit of ATP synthase (complex V), producing a massive drop in mitochondrial ATP production and consequently a compensatory increase in cellular glycolytic flux to produce more ATP and maintain cellular homeostasis (Wu et al., 2007). On the other hand, the maximal respiratory capacity can be revealed by treating cells with carbonylcyanide-4 (trifluoromethoxy) phenylhydrazone (FCCP). By uncoupling respiration from oxidative phosphorylation, FCCP allows the oxidation of available substrates which accelerate the electron transport chain (Winer and Wu, 2014).

To estimate glycolytic capacity, we analyzed ECAR both before (basal levels) and after addition of oligomycin. I did not observe significant differences in basal ECAR between WT and sgBicc1 cells, and addition of oligomycin increased ECAR production in both genotypes (Fig. 50A). However, the ECAR induced by oligomycin treatment in Bicc1 mutant cells (sgBicc1) was significantly lower (8.7 mpH/min) than in control mIMCD3 cells (10.4 mpH/min). Our finding that sgBicc1 mIMCD3 cells did not increase lactate production (as determined by baseline ECAR levels) does not support a model that endogenous Bicc1 acts as an inhibitor of glycolysis. However, how the loss of Bicc1 dampened glycolytic capacity in response to oligomycin is unclear.

To assess the mitochondrial respiration of these cells, I estimated the OCR levels before and after addition of mitochondrial stressors. I found out that loss of Bicc1 significantly decreased the basal mitochondrial respiration by 35 pMol/min (27% reduction of total) (Fig. 50B). The ATP-coupled mitochondrial respiration detected after challenge by oligomycin treatment was reduced proportionately in sgBicc1 compared to WT cells (by 18 pMol/min; 28% of total). Furthermore, while the maximal respiratory capacity determined by FCCP-stimulated OCR was 1.3-fold above basal levels in WT cells, no increase in OCR was seen in sgBicc1 cells. This result shows that the maximal respiratory capacity in sgBicc1 cells is reached already in the absence of FCCP, with zero reserve capacity left. Thus, under the conditions

121 examined, Bicc1 is critical to maintain normal cellular respiration. In addition, these results imply that Bicc1 mutant mIMCD3 cells meet their demand for ATP differently than cancerous cells as they do not rely on a Warburg-like glucose metabolism marked by increased glycolytic flux (Fig. 50C) (Vander Heiden et al., 2009).

Figure 50. Glycolysis and mitochondrial respiration profiles of sgBicc1 mIMCD3 cells. (A) Calculated ECAR in basal conditions (before) and after addition of Oligomycin to assess glycolytic capacity in the indicated cell lines. (B) Calculated OCR in basal conditions (before), and after addition of Oligomycin and FCCP to assess ATP-coupled mitochondrial respiration and maximal respiratory capacity, respectively. (C) Metabolic switching diagram depicts the relative metabolic state at basal conditions between WT (black circle) and sgBicc1 (white circle) mIMCD3 cells. Bars represent the mean±S.E.M of n=6 during time. Student’s t-test was used to determine significant differences between two cell lines. *p<0.05 ***p<0.001

122 5.2.4 Systemic alterations in Bicc1-/- mice glucose and ketone metabolism

The most striking metabolic alteration that we observed in Bicc1-/- neonates was their reduced glucose levels. Blood glucose levels normally are regulated by insulin and glucagon that are released by beta and alpha pancreatic cells, respectively. These hormones are secreted in response to high or low glucose levels, respectively, and they are regulated by several factors including the incretin hormone glucagon-like peptide 1 (GLP-1). Additionally, other hormones such as cortisol can regulate blood glucose levels (Gosmanov et al., 2016).

To assess whether blood glucose in Bicc1 mutant mice decrease due to hormonal changes or due to downregulation of renal gluconeogenic enzymes, I first measured insulin levels in plasma. Plasma from newborns at P2 showed no increase but rather a significant decrease in the levels of insulin in Bicc1-/- mutants compared to WT littermates (Fig. 51A). This result indicates that hypoglycemia is not caused by hyperinsulinemia. Instead, a role for impaired renal gluconeogenesis seems more plausible. Glucagon like peptide-1 (GLP-1) is a gut hormone that induces β-cells to release insulin (Baggio and Drucker, 2007). To evaluate whether it may contribute to lower insulin levels, we monitored GLP-1 in plasma of Bicc1-/- mutant mice by immunoassay kit. We observed a tendency to reduced levels, although it did not reach statistical significance (Fig. 51B). These results indicate that blood glucose levels in Bicc1-/- mutant mice (Fig. 20A section 4.3.2) unlikely result from hormonal alterations.

Besides a role in gluconeogenesis, kidneys are essential to reabsorb glucose from glomerular filtrate in proximal tubules. To test whether glucose reabsorption is defective we collected urine from the bladder of P2 WT and Bicc1-/- mice after decapitation. Bicc1-/- mutants showed on average 2-fold higher glucose concentrations in the urine than their WT littermates (Fig. 51C). This result indicates that besides downregulating gluconeogenic enzymes, Bicc1 mutant kidneys fail to efficiently reabsorb glucose. Overall, we conclude that Bicc1-/- mutant neonates have decreased renal gluconeogenic capacity and glucose reabsorption, and that both

123 mechanisms contribute to reduce blood glucose levels. By contrast, a defect in hepatic gluconeogenesis is unlikely (Fig 20D. section 4.3.2).

During hypoglycemia or starvation, the organism must meet energy demands by activating alternative metabolic pathways. These include glycogenolysis to breakdown glycogen into glucose, and ketogenesis, the utilization of metabolites from the catabolism of fatty acids and ketogenic amino acids to produce ketone bodies. As a source of energy during fasting, ketone bodies are particularly important for the brain. Kidneys can generate ketone bodies, such as β-hydroxybutyrate since they express HMGCS2, the enzyme necessary for ketogenesis (Zhang et al., 2011). To interrogate the status of alternative energy sources that may help Bicc1-/- mutant mice to cope with reduced blood glucose levels, I measured β-hydroxybutyrate in plasma collected from P2 WT and Bicc1-/- mice. A 37% increase in β-hydroxybutyrate levels observed in Bicc1-/-mice suggests that the Bicc1 mutants hyperactivate ketogenesis (Fig. 51D). Importantly, this increase correlates with the increased level of β-hydroxybutyrate (3-hydroxybutyric acid) in mutant kidneys observed by metabolomics, supporting the hypothesis that kidneys contribute to increased ketogenesis to cope with reduced blood glucose levels in Bicc1-/- neonates.

124 Figure 51. Plasma insulin and GLP-1 levels, ketone bodies, and urinary glucose concentrations in Bicc1-/- at postnatal day P2. (A) Insulin and (B) GLP-1 concentrations in plasma collected from the indicated genotypes at P2 stage. (C) Urine glucose levels from bladders of P2 mice. n=7 Bicc1-/-, n=15 Bicc1+/+. (D) Plasma concentration of β-hydroxybutyrate from the indicates genotypes at P2 stage. n=4 Bicc1-/-, n=8 Bicc1+/+. *p<0.05, Student’s t-test.

5.3 Discussion

Inhibition of autophagy by mTORC1 and other pathways, together with alterations in glucose and lipid metabolism have recently been proposed to be important drivers of ADPKD progression (Rowe et al., 2013; Rowe and Boletta, 2014; Klawitter et al.,

2013; Menezes et al., 2016). Since the Bicc1 protein interactome was enriched for metabolism related enzymes, I investigated possible links between Bicc1 and kidney metabolism and autophagy. Here, I showed that Bicc1 inhibits autophagy rather than stimulating it, and that it coimmunoprecipitates with rate-limiting metabolic enzymes and contributes to maintain their protein levels, and that deletion of Bicc1 significantly alters the renal metabolome. These results implicate Bicc1 as an important metabolic regulator in the etiology of polycystic kidneys (Fig. 52).

Figure 52. Metabolic alterations in Bicc1-/- newborn mice found in this study By combining protein interactome, kidney metabolomics and in vitro analysis of metabolic parameters, we found nucleotides, lipids, amino acids (aa) metabolism and the gluconeogenesis pathway significantly de-regulated in Bicc1-/- mice at early post-natal stage. In particular, energy demand of cystic growth appears to be fueled primarily by fatty acid oxidation. Increased ketones production may lead to acidosis further aggravated by impaired renal PEPCK expression that halts glutaminolysis mediated effect on HCO3- secretion into blood and NH4+ excretion into urine. See below in the text.

125 Bicc1-/- kidneys show increased autophagy

Autophagic responses are generally impaired in ADPKD models (Belibi et al., 2011b; Rowe et al., 2013). This defect contributes to the imbalance between proliferation and apoptosis in cystic cells driving disease progression. Furthermore, several compounds used to inhibit cystogenesis such as rapamycin, metformin, triptolide and curcumin, are also autophagy inducers (De Rechter et al., 2016). Here I found that, Bicc1-/- kidneys increase basal autophagy evidenced by increased LC3-II protein levels and a parallel decrease of p62/SQSTM1.

To explain this result, I assessed autophagy regulatory pathways such as mTORC1 and AMPK. I observed reduced levels of total AMPK concomitant with an hyperactivation of mTORC1. Although phospho-AMPK was not assessed in this study, one possibility is that mTORC1 is hyperactivated in Bicc1-/- kidneys by decreased levels of AMPK and likely phospho-AMPK. Another possibility is that the activation of Wnt signaling in Bicc1-/- kidneys (Maisonneuve et al., 2009) indirectly upregulates mTOR as described in a Wnt gain-of-function mouse model (Inoki et al., 2006). In PKD studies phopho-AMPK has been shown to be reduced concomitant with enhanced ERK signaling and are likely to be the cause of higher anaerobic glycolysis described in PKD1 knockout cells and kidneys (Boletta, 2009). Irrespectively of the mechanism, our results indicate that Bicc1-/- kidneys circumvent mTORC1 autophagy inhibition, to enhance autophagic responses that play critical roles for the cell. How Bicc1 inhibits autophagy in parallel to mTORC1 activation, is unknown. One possibility is that Bicc1 regulates CREB and/or farnesoid X receptor (FXR) which rule the transcription of autophagic genes during fasting/feeding periods, respectively (Seok et al., 2014).

Novel protein interactions strongly suggest a role for Bicc1 in coordinating multiple metabolic pathways

An unbiased proteomic screen for Bicc1-interacting factors identified multiple novel partners whose biological processes are referenced in cellular metabolism. Among these we validated the interaction with two key enzymes in the de novo fatty acid

126 synthesis: ACACA, the rate limiting enzyme that carboxylates acetyl-coA to form malonyl-CoA and FASN, the enzyme that generates long-chain fatty acids from acetyl-CoA and malonyl-CoA in the presence of NADPH (Mashima et al., 2009). In addition, Bicc1 interacted with the rate-limiting trifunctional enzyme CAD which catalyzes the first three reactions in the de novo pyrimidine synthesis pathway (Buel et al., 2013). Importantly, Bicc1 was also required to maintain normal protein expression levels of these enzymes in the kidney of newborn mice. Whether Bicc1 stimulates the expression of these enzymes by promoting their translation or stability and how this might influence their activities should be investigated in future studies.

Bicc1-/- kidneys reflects alterations in pyrimidine and purine metabolism

Reduced levels of UMP and CMP might be explained by our observation that Bicc1 binds CAD, the rate-limiting enzyme in de novo pyrimidine synthesis, and by the decrease in CAD protein levels that we observed in Bicc1-/- kidneys. Further studies should directly test whether Bicc1 binding regulate CAD activity.

In addition, we observed marked alterations in purine metabolism, including increased accumulation of allantoin and a reduction of IMP and GMP. Purine degradation is a major source of uric acid (allantoin in rodents) which appears concomitantly with high levels of adenosine (Taylor et al., 2010). Accumulation of allantoin and adenosine and several different metabolites in the purine metabolism have been observed in the cpk model (Taylor et al., 2010), the Lewis polycystic kidney rat (Abbiss et al., 2012) and in CDK patients (Chen et al., 2017a). Moreover, adenosine is involved in the purinergic signaling that has an unfavorable role in the pathophysiology of PKD, since adenosine stimulates cAMP and CFTR driving ion and water secretion and thereby contributing to cyst expansion (Hovater et al., 2008).

Newborn Bicc1-/- kidneys do not increase glycolytic flux

In contrast to previous studies in orthologous (Rowe and Boletta, 2014) and non- orthologous models of ADPKD (Riwanto et al., 2016), we detected no increase in the levels of lactate or other glycolytic intermediates in Bicc1-/- compared to wild-type kidneys. ECAR measurements also revealed no change in aerobic glycolysis upon

127 CRISPR/Cas9 editing of Bicc1 in mIMCD3 cells. Although some studies on orthologous ADPKD models have failed to find a prominent glycolytic response (Menezes et al., 2016; Warner et al., 2016), ADPKD patient-derived cells increase expression of multiple glycolytic enzymes including HK1, ALDOA, ALDOC, LDHA, PKM2 at expense of gluconeogenic enzymes, such as G6PC, FBP1, BPGM or GALM. These changes are accompanied a switch from aerobic to anaerobic glycolysis characteristic of many cancer cells (Warburg effect) (Rowe et al., 2013). However, the data cannot distinguish which defect appears first. Our finding that gluconeogenic enzymes decreased independently of any apparent change in glycolysis raise the interesting possibility that downregulation of gluconeogenic enzymes is an early determinant of PKD progression.

Impaired PEPCK expression may contribute to lipid droplet accumulation

In addition to accumulation of lipid droplets, Bicc1-/- kidneys accumulate citrate and succinate, two TCA cycle intermediates, suggesting impaired TCA flux. TCA flux is dependent on continuous efficient removal of TCA anions in a cataplerotic process mediated by PEPCK to prevent accumulation of anions (Hakimi et al., 2005). Interestingly, targeted deletion of PEPCK specifically in liver also results in increased triglyceride levels and provokes fatty liver (steatosis), despite increased protein levels of several free-fatty acid oxidizing enzymes (She et al., 2000). Thus, Bicc1 knockout kidneys likely accumulate TCA intermediates and lipid droplets likely due to the drastic inhibition of PEPCK1 mRNA and protein expression. In keeping with this hypothesis, lipid droplets also accumulate in renal tubule cells of PEPCK1 null mice, and reconstitution of PEPCK1 expression specifically in liver was not sufficient to suppress this phenotype (Semakova et al., 2017). Why PEPCK1 transcription is inhibited despite increased cAMP accumulation and PKA signaling in cystic Bicc1 kidneys is unknown. Renal PEPCK1 expression does not decrease in a rat model of chronic renal failure (Choukroun et al., 1994), and general acidosis should increase it in proximal tubules rather than decreasing it (Curthoys et al., 2007). Therefore, a key question for future studies will be how loss of Bicc1 blocks PEPCK accumulation.

128 Besides impaired PEPCK1 expression and ectopic lipid droplets, newborn Bicc1-/- kidneys also showed diminished protein levels of FASN and ACACA, the enzymes that mediate de novo lipid synthesis. If Bicc1 protein binding inhibits their activities, even reduced levels of ACACA and FASN together with the observed 6.8-fold increase in citrate may be sufficient to accelerate the biogenesis and storage of neutral lipids and triglycerides in lipid droplets. However, inhibition of these enzyme activities by Bicc1 binding has not been demonstrated, and depletion of their expression levels in Bicc1 null kidneys alone would more likely inhibit LD accumulation than stimulating it.

Bicc1-/- mice enhance ketogenesis

The perinatal period is a stressful moment for the pup until it adapts its metabolism to a high-fat low carbohydrate milk diet. Here, I showed that Bicc1-/- mice must cope with abnormal reduced glucose levels in the postnatal period. I observed increased levels of plasma β-hydroxybutyrate in Bicc1-/- compared to WT littermates, indicating that Bicc1-/- mice activate ketogenesis to meet their demand. Our untargeted renal metabolomic screen also revealed 7.1-fold accumulation of 3-Hydroxy-3- methylglutaric acid (HMG) and 2.9-fold accumulation of β-hydroxybutyrate confirming that Bicc1-/- newborn kidneys activate ketogenesis. Furthermore, these results imply that Bicc1-/- kidneys undergo β-oxidation to generate ketone bodies. Bicc1-/- kidneys accumulated 1.6-fold levels of acetylcarnitine, suggesting an increase of long chain fatty acid transport into the mitochondrial matrix for β-oxidation (Wong and Crawford, 2011). In line with this result, Bicc1 was described recently as promoter of mRNA translation (including Cpt1a) in cooperation with OFD1 (Iaconis et al., 2017). Future analysis of mitochondrial enzymes controlling the carnitine pathway, such as Carnitine palmitoyltransferase 1A and 2 (CPT1 and CPT2), will be important to test our prediction that increased acetylcarnitine and decreased carnitine reflect increased flux into mitochondria to sustain high levels of beta oxidation.

We assessed in vitro OCR levels in Bicc1 loss-of-function mIMCD3 cells. These cells displayed significant impairment of basal and challenged mitochondrial respiration.

The most striking phenotype observed in Bicc1 loss-of-function mIMCD3 cells is the

129 abrogated spare respiratory capacity (i.e the increment in OCR after FCCP addition compared to basal OCR), indicating that Bicc1 helps to maintain the extra respiratory capacity to produce energy under stress which plays a critical role in cellular long- term survival (Choi et al., 2009; Nicholls, 2009). Depletion of the spare respiratory capacity has been involved in different cardiac and neurodegenerative pathologies and in aging (Desler et al., 2012). It will be important to determine in future studies the mechanism by which Bicc1 regulates in vitro mitochondrial function. The finding of several mitochondrial proteins, such as the SLC25A5 and SLC25A6, in the proteomic screen leads us to speculate that Bicc1 may directly regulate their activity. Additionally, the in vitro analysis of fatty acid oxidation using exogenous sources of fatty acids would address more specifically the mitochondrial functionality in Bicc1 loss-of-function cells.

Role of PEPCK in acid-base balance and disease progression

Bicc1-/- mutant mice face the neonatal stage having impaired renal gluconeogenesis together with glucose lost by urinary excretion which may directly contribute to significant reduced glycemia. To cope with this situation, mutant mice increase the production of ketone bodies and its release into the bloodstream. Oxidation of ketone bodies provides Acetyl-CoA to fuel the TCA cycle. β-hydroxybutyrate is an organic acid that when accumulated in the plasma may lead to ketoacidosis. In case of acidosis, the proximal tubule cells increase the uptake and catabolism of glutamine, which in the mitochondria is the main source for ammoniagenesis (Curthoys and Moe, 2014).

In mammals, the α-ketoglutarate derived from renal glutaminolysis is converted into glucose if cytoplasmic PEPCK is present divert oxaloacetate from the TCA cycle into the gluconeogenesis pathway (Gerich et al., 2001; Pagliara and Goodman, 1970). Coupling of renal gluconeogenesis and ammoniagenesis (from glutaminolysis) in

+ proximal tubules is key to enable each secreted NH4 ion to drive the production of

- one new HCO3 from CO2 and water, and hence to maintain the acid-base balance

- through the net production of HCO3 that can cross the basolateral membrane and reach the bloodstream (Curthoys and Moe, 2014; Hamm et al., 2015).

130 In Bicc1-/- mutant kidneys we observed increased glutamine and drastically reduced PEPCK expression (Table 3 and Fig. 20C section 4.3.2), indicating that the source of ammoniagenesis is enlarged. However, the contribution of gluconeogenesis to generate bicarbonate, and thus the capacity of the kidney to buffer acidosis is impaired. We also observed increased levels of ornithine and citrulline. These compounds are intermediate metabolites of the urea cycle, a metabolic process that

+ coverts the excess of nitrogen (such as NH3 and NH4 ) into a less toxic urea, which can then be excreted in the urine (Klein et al., 2011). Together, these results suggest Bicc1-/- mutants likely accumulate excess of ammonium and may succumb to systemic acidosis or hypoglycemia.

The hypothesis that ammonium contributes to the pathogenesis of polycystic kidney disease was first considered in the association of chronic hypokalemia and renal cyst acquisition (Alpern and Toto, 1990). Follow up studies demonstrated that the metabolism of renal ammoniagenesis plays a role in the development of the disease and that excess ammonium can promote inflammation (by activation of the complement) or stimulate cell growth (Torres et al., 1994a). In the Han:SPRD Cy/+ rat model of PKD, rats fed with ammonium chloride aggravated the cystic disease, while the opposite was achieved after alkali supplementation in the diet (Torres et al., 1994a). In the same model marked downregulation of PEPCK and several transporters involved in ammoniagenesis argued that it is through the impairment to produce and excrete ammonia that the bicarbonate biosynthesis and its buffering capacity are inhibited (Bürki et al., 2015).

Prior to alterations in glomerular filtration rates (GFR), ADPKD patients develop a urine concentration defect revealed by impaired urinary excretion of ammonium and impaired urinary acidification, likely due to defective trapping of ammonium in the renal medulla despite of enhanced ammoniagenesis (Torres et al., 1994b). In Bicc1- /- mutant kidneys, impaired gluconeogenesis thus is predicted to impair the balance in ammonium secretion thereby perpetuating the disturbances in acid-base homeostasis which may contribute to disease progression (Fig 53). It would be interesting to investigate in future studies, the mechanisms leading to renal PEPCK1

131 depletion in Bicc1-/- mice, as a central hub of the metabolic phenotype in this mouse model.

Figure 53. Acidosis enhances glutaminolysis to restore acid-base balance. (A) Increase expression of transporters and mitochondrial enzymes together with cytosolic PEPCK ensures glutaminolysis and gluconeogenesis to produce the adequate amounts of NH+4 that acidifies the luminal fluid and net synthesis of HCO3- that is transported into the blood B) Bicc1-/- mutant kidneys may fail in this mechanism of acid-base balance regulation since they have impaired gluconeogenesis thus inhibiting the adequate handling of acidosis by the renal cells. Adapted from (Curthoys and Moe, 2014).

Importance of metabolomic analysis in Bicc1-/- model

While renal metabolomics is an approach commonly applied to investigate biomarkers in kidney diseases including CDK and renal cell carcinoma (Hakimi et al., 2016; Rysz et al., 2017), some studies have determined the renal metabolome of mouse or rat models in polycystic kidney diseases (Hwang et al., 2015; Menezes et al., 2016; Rowe et al., 2013; Taylor et al., 2010). Previous studies in kidney cancer and in the cpk model of PKD concluded that, compared to urine and plasma, tissues are the most sensitive bio-sample to uncover biological changes by metabolomics (Ganti et al., 2012; Hwang et al., 2015). Renal metabolomic studies are important since kidney function has a broad impact on circulating metabolites which may have important functions on their own in the etiology of chronic kidney diseases (Rhee, 2015). To our knowledge this is the first metabolomic analysis in Bicc1-/- mutants. To

132 provide a global view of metabolic alterations, we decided to use untargeted metabolomics. However, this approach has the disadvantage to be exhaustive since we may not detect all the metabolites of a single pathway, which was the case in this study. Nonetheless, identification and validation after tandem mass spectrometry of more than 30 metabolites gave a good overview of the deregulated metabolic pathways in Bicc1-/- kidneys, which can be analyzed in a more targeted manner in future studies.

Early disease markers can provide more biological insight than alterations occurring at later stages. A caveat in our study is the aggressive cystic disease of the Bicc1-/- mice that already exist at early neonatal stages when all the analyses in this study were done. Future studies may consider the experimental design in earlier developmental stages or in a conditional Bicc1-/- model. This chapter mainly described the complex metabolic alterations in the polycystic kidney Bicc1-/- model. The results discussed here strongly argue that further and more focused analysis must consider that metabolic flux patterns are branched, connected and shunted into parallel pathways, and that synergically they may contribute to the disease development.

133 6. Concluding remarks

This study shows for the first time Bicc1 interactome. From this new dataset, we specifically focused on the analysis of some ciliopathy-related proteins (ANKS3, ANKS6), the mammalian CTLH complex and some metabolic enzymes that led us to describe the renal metabolome in Bicc1 deficient mice. Our results highlight the critical role of Bicc1 in maintaining renal gluconeogenic enzymes and glycemia and highlight several metabolic alterations that may contribute to disease progression and to the early lethality of Bicc1 deficient mice.

First, we were interested in the role of SAM domain in Bicc1 functions. In a first article, we demonstrated that Bicc1 self-polymerizes through its SAM domain to form cytoplasmic foci that concentrate and silence target mRNAs (Results part I). The identification of the SAM protein ANKS3 in the Bicc1 interactome, as well as their interaction previously described with ANKS6, led us to study the network of interaction between these three SAM domain-containing proteins. Our results, compiled in an article currently under revision, show how ANKS3 and ANKS6 interact with Bicc1 to regulate the topology of cytoplasmic Bicc1 clusters.

Our proteomic screen identified the mammalian CTLH complex which prompted us to hypothesize that Bicc1 may regulate the degradation of gluconeogenic enzymes through this interaction. We extensively characterized the binding of Bicc1 and the CTLH subunits and our results show that Bicc1 in its non-polymerized form interacts with the CTLH complex. However, the endogenous CTLH complex in mammalian mIMCD3 cells does not decrease the rate-limiting gluconeogenic enzyme FBP1. Instead, this interaction seems to enable Bicc1 and CTLH complex to mutually downregulate each other. Nevertheless, deletion of Bicc1 in mice led to reduced glycemia and diminished expression of the gluconeogenic enzymes PEPCK and FBP1 specifically in kidneys. These results are presented in an article currently under revision.

134 This novel characteristic of Bicc1-/- mice phenotype and our results that several rate- limiting enzymes in lipid and pyrimidine syntheses are enriched in Bicc1 interactome were complemented by renal metabolomic analysis. We found that nucleotides, lipids, amino acids metabolites are significantly de-regulated in Bicc1-/- kidneys at early postnatal stages without an evident increase in glycolytic flux. The energy demand of cystic growth appears to be fueled primarily by fatty acid oxidation. Increased ketones production may lead to acidosis further aggravated by impaired renal PEPCK expression since renal gluconeogenesis cooperates with glutaminolysis to secrete and excrete ions that regulate acid-base balance. Altogether, these results implicate Bicc1 as an important metabolic regulator in the etiology of polycystic kidneys.

135 7. Materials and methods

Plasmids

To generate TAP-tagged MmBicc1-HA-StrepII, a previously described HA-tagged mouse Bicc1 cDNA(Maisonneuve et al., 2009) was amplified by PCR and fused in- frame as a Kpn I restriction fragment to 2xHA-2xStrep-II (WSHPQFEK) tag in the inducible expression vector pcDNA5™/FRT/TO (courtesy of Aebersold Lab). Alanine substitutions of residues D913, K915, and E916 in HA-Bicc1 have been described previously (Results part I). Briefly, the mutated DNA fragment was amplified by overlap extension PCR and then subcloned between BglII and XbaI sites of pCMV- SPORT6::HA-Bicc1. The mutant Bicc1-mutI cDNA carrying Lys substitutions in residues E900 and D902 was generated analogously. DNA fragments for Y2H assays were cloned by PCR in plasmid pACT2 distal to the GAL4 transactivation domain coding sequence, or in plasmid pGBKT7 distal to the GAL4 DNA-binding domain coding sequence using primers with unique restriction sites. To derive plasmid pACT2::Bicc1-D913K915E916/AAA, the mouse SAM-D913K915E916/AAA encoding fragment was excised from pCMV-SPORT6::HA-Bicc1- D913K915E916/AAA (Bicc1-mutD) by Stu I and Xho 1 and subcloned between the corresponding sites of human Bicc1 in plasmid pACT2::hBICC1. The luciferase reporter plasmids Luc-AC6 3’UTRprox and Luc-PKIα 3’UTRprox in the vector pCS2+ have been described(Piazzon et al., 2012).

Cell culture and transfections

HEK293T and mIMCD3 cells were cultured in DMEM medium (Sigma) supplemented with 10% fetal bovine serum (FBS; Sigma), glutamine (1%; Invitrogen), and gentamicin (1%; Invitrogen). These and all other cell lines in this study were negative for mycoplasma as determined by an ELISA-based assay (Roche). Expression vectors were transfected using jetPEI transfection reagent (Polyplus) according to the manufacturer's instructions. Small interfering RNAs against WDR26 (TAAAGGCTTTAGCTCATTCAGGTCA) Twa1 (CCGACTCATCATGAACTAC) and Bicc1 (CCAACCACGUAUCCUAUAATT) or scrambled control (Microsynth) were

136 transfected during 48 hrs using INTERFERin® transfection reagent (Polyplus). CRISPR/Cas9 editing of Bicc1 in mIMCD3 was carried out using the guide sequence 5’-GCGAGCGCAGCACCGACTCGCCGG-3’ cloned into the expression vector PX458 containing GFP-tagged Cas9 (Ran et al., 2013). The resulting sgRNA/Cas9 expression vector was transfected and after 24h, the cells were trypsinized, washed with PBS and resuspended in PBS/1% FBS for single cell sorting for GFP by FACS into 96-well plate containing complete medium. Clonal cell lines were expanded and screened by anti-Bicc1 by Western blot.

Indirect immunofluorescence

HEK293T cells were transfected with 1 µg of HA-Bicc1 in 6 well plates. After 24 hrs, the cells were plated on coverslips. At 48 hrs post-transfection, cells were washed with phosphate-buffered saline (PBS) and methanol-fixed during 10 min at -20°C. Cells were washed again with 1X PBS and incubated with blocking solution containing 1% BSA for 1 hr at room temperature. Mouse anti-HA antibody (Sigma) and rabbit anti-RanBP9 (Abcam) were diluted 1:500 and 1:100, respectively, in blocking solution and added to the cells during 2 hrs at room temperature. After washing with 1X PBS, cells were incubated with Alexa 647-conjugated secondary donkey anti-mouse and Alexa 488-conjugated donkey anti-rabbit IgG during 1hr at room temperature in the presence of 4’,6-diamidino-2-phenylindole (DAPI). Pictures were acquired by confocal microscopy using a Zeiss LSM700 microscope.

Generation of stable inducible cell line and TAP-tag purification

In the Flp-In™ system, three different vectors are used to generate an isogenic stable cell line. The first major component is the pFRT/lac Zeo target site vector that has been used to generate the Flp-In™ host cell line. The vector contains a lacZ-Zeocin™ fusion gene allowing the selection of Flp-In™ host cell line containing an integrated FRT site. The FRT site serves as the binding and cleavage site for the Flp recombinase. The second major component of the Flp-In™ system is the pcDNA™5/FRT expression vector into which the gene of interest is cloned downstream of the human CMV promoter. The vector also contains the hygromycin

137 resistance gene with an FRT site embedded in the 5′ coding region. The hygromycin resistance gene lacks a promoter and the ATG initiation codon. The third major component of the Flp-In™ System is the pOG44 plasmid which constitutively expresses the Flp recombinase (Broach et al., 1982; Broach & Hicks, 1980; Buchholz et al., 1996). Upon cotransfection, the Flp recombinase expressed from pOG44 mediates a homologous recombination between the FRT sites (integrated into the genome and on pcDNA™5/FRT) such that the pcDNA™5/FRT construct is inserted into the genome at the integrated FRT site. Insertion of pcDNA™5/FRT into the genome at the FRT site brings the SV40 promoter and the ATG initiation codon (from pFRT/lacZeo) into proximity and frame with the hygromycin resistance gene and inactivates the lacZ-Zeocin™ fusion gene. Thus, stable Flp-In™expression cell lines can be selected for hygromycin resistance, Zeocin™sensitivity, lack of β- galactosidase activity, and expression of the recombinant protein of interest. Several clones were isolated and validated for MmBicc1-HA-StrepII expression after 24 hrs of doxycycline induction.

For tandem affinity purification (Gavin et al., 2011), 3 dishes of Bicc1-HA-StrepII

T-Rex cells were treated without (control) or with doxycycline (1 µg/ml) during 24 hours. Cells were rinsed and scraped from dishes in Tris-buffered saline (TBS) containing 150 mM NaCl in 50 mM Tris.HCl pH 7.6, followed by extraction with

TBS containing 1 mM DTT, 0.05% NP-40, phosphatase inhibitor cocktail 3 (Sigma),

RNAse inhibitors (Promega) and protease inhibitor cocktail (Roche). After sonication and centrifugation at 10'000 × g for 10 min, supernatants were incubated on a rotating wheel for 2 hrs at 4°C with Strep-Tactin Sepharose resin (IBA). After incubation, resins were loaded in Mobicol columns (MoBiTec) and washed with 10 ml 0.05% NP-40 in TBS, and eluted on ice using desthiobiotin during 15 min as described (Gloeckner et al., 2009). For the second purification, eluates were incubated with anti-HA agarose beads (Sigma, A2095) during 2 hrs at 4°C. After incubation, beads were loaded on Mobicol columns, rinsed with 10 ml Tris.HCl pH 7.4 (10 mM) containing 100 mM NaCl, 2.5 mM

MgCl2, 1 mM DTT, and 0.02% NP-40, and eluted using 125 mM HCl into a vial containing 50 µl of Triethylammonium bicarbonate (TAEB) neutralization buffer

(Sigma). Eluates were concentrated in 40 µl using

138 Amicon Ultra 0.5 ml 3K (Millipore). Laemmli buffer was added to the eluates and boiled samples were loaded in a 4-15% Mini-PROTEAN® TGX™ precast Gel (BioRad).

LC-MS/MS analysis

LC-MS/MS analysis was done at the Proteomic core facility (PCF) at Ecole Polytechnique Fédérale de Lausanne (EPFL). Peptides were extracted from Coomassie Brilliant blue R-250 stained gel slices and subjected to tryptic digestion. Reverse phase separations were performed on a Dionex Ultimate 3000 RSLC nano- UPLC system (Thermo Fisher Scientific) connected online an Orbitrap Elite Mass Spectrometer (Thermo Fisher Scientific) piloted with Xcalibur (version 2.1) and Tune (version 2.5.5), as described in (Chopra et al., 2014). Samples were analyzed using Mascot (version 2.6, Matrix Science, Boston, MA, USA) set up to search the human subset of the UniProt database (Release 2017_02). For visualization and validation, MS/MS data was entered in Scaffold 4 (Proteome Software Inc., Portland, OR). The peptide identification threshold and the protein identification threshold based on matches with at least 2 identified peptides were set at 1% FDR. The MS analysis was conducted in duplicates from two biological replicates.

Luciferase reporter assays

To measure luciferase expression, HEK293T cells were plated in 24-well plates and transfected on the following day with the indicated plasmids (100 ng/well) and with a lacZ expression vector (50 ng/well) in triplicate samples using jetPEI (Polyplus Transfection). 24 hrs after transfection, cells were lysed in buffer containing 25 mM Tris-phosphate, pH 7.8, 2 mM DTT, 2 mM 1,2-diaminocyclohexanetetraacetic acid (CDTA), 10% glycerol, and 0.5% Triton X-100. Cell extracts were diluted 20-fold, and luminescent counts detected in a Centro XS3 LB960 microplate luminometer (Berthold Technologies). Values were normalized to β-galactosidase activity measured by spectrophotometry in a Safire2 microplate reader. Results represent mean values from at least three independent experiments. Error bars show standard errors of the mean (SEM). Student's t-test was used to calculate P values.

139 Co-immunoprecipitation

HEK293T cells were transfected with the indicated expression plasmids in 10 cm dishes (for Bicc1-HA and DDK-WDR26, we used 2 µg/dish). 24 hrs after transfection, cells were washed with PBS and proteins were extracted with lysis buffer containing 1X TBS, 1mM DTT, 0.05% NP-40, 1x protease inhibitor cocktail (Roche) and phosphatase inhibitor cocktail 3 (Sigma). Endogenous Bicc1 was immunoprecipitated from mIMCD3 cell extracts. In brief, confluent cells in 10 cm dishes were washed with 1X PBS and resuspended in extraction buffer as described above. After sonication and centrifugation at 10'000 × g for 10 min, supernatants were incubated on a rotating wheel for 2 hrs at 4°C with anti-HA agarose antibody or anti-FLAG® M2 affinity gel (Sigma), or with protein G Sepharose beads (GE Healthcare) precoated with Bicc1 custom antibody (Results part I) or pre-immune IgG (R&D Systems). Beads loaded on Mobicol columns were washed with 10 ml of

. Tris HCl pH 7.4, 100 mM NaCl, 2.5 mM MgCl2, 1 mM DTT, and 0.02% NP-40 washing buffer and resuspended in Laemmli buffer. Elutions were fractionated on SDS-PAGE gels and analyzed by Western blotting.

GST pull-down

Recombinant GST-Bicc1 WT, GST-Bicc1 MutD, and GST-Bicc1 MutI proteins were produced in the E. coli BL21 strain (Novagen) as previously described using pGEX- 1λT expression vector (Piazzon et al., 2012). GST fusion proteins were purified from cell extract under native conditions, using Glutathione Sepharose 4B (GE Healthcare) in 50 mM Tris.HCl pH 8, 200 mM NaCl, 1 mM DTT according to manufacturer's instructions, rinsed twice with 20 mM Tris.HCl pH 7.4; 750 mM NaCl; 1 mM DTT washing buffer followed by Tris.HCl pH 7.4, 20 mM; NaCl 200 mM; DTT 1 mM washing buffer and eluted by addition of 20 mM glutathione. Confluent HEK293T cells in 10 cm dishes were lysed in TBS buffer containing 1 mM DTT, 0.05% NP-40, Phosphatase inhibitor cocktail 3 (Sigma), RNAse inhibitors (Promega) and Protease inhibitors cocktail (Roche). After sonication and centrifugation at 10'000 × g for 10 min, supernatants were incubated on a rotating wheel for 2 hrs at 4°C with glutathione-Sepharose 4B beads saturated with GST-Bicc1 WT, GST-MutD, and

140 GST-MutI or GST alone (negative control). After washing on Mobicol columns with 5

. ml of 20 mM Tris HCl pH 7.4 containing 200 mM NaCl, 2 mM MgCl2, 1 mM DTT, and 0.05% NP-40, the beads were resuspended in Laemmli buffer. Eluates were fractionated on SDS-PAGE gels and analyzed by Western blotting. The loading of GST fusion proteins was validated indirectly by Coomassie staining of residual proteins retained in the gel.

Sucrose density gradient fractionation

Subconfluent HEK293T cells were transfected with Bicc1-HA or empty vector in 10 cm dishesg (2 μ DNA/dish) using 3 plates per condition. Cell extracts were prepared after 24 hrs as described above for the GST pull-down assay. Continuous 15 to 60% sucrose gradients were prepared as described previously (Results part I). Fractions were recovered manually starting from the top, fractionated on SDS-PAGE gels, and analyzed by Western blotting. γ-tubulin was used as a control.

Reverse transcription qPCR

Total RNA from mIMCD3 cells and kidneys was isolated using TRIzol (Life Technologies) and RNeasy plus mini kit (Qiagen) following manufacturer's instructions. Total RNA was digested with RQ1 DNase (Promega). Reverse transcription of cDNA was carried out using SuperScript III reverse transcriptase and Oligo dT (Life Technologies) according to the manufacturer's recommendations. The qPCR was performed in a QuantStudio™ 6 Flex real-time PCR systems (Applied Biosystems) using goTaq qPCR 2x Master Mix (Promega). Samples were analyzed as triplicates, and expression levels were calculated with the manufacturer’s software using the ΔΔCt method. The qRT-PCR primers are described in the following table.

141 Yeast two-hybrid assay

To assess binding of each CTLH complex subunit to WT Bicc1 or Bicc1- D913K915E916/AAA, they were each fused to the DNA-binding domain of the GAL4 transcription factor (GAL4‐BD) in the pGBKT7 plasmid (Clontech) as bait proteins. In parallel, they were also fused to the activation domain of the GAL4 transcription factor (GAL4‐AD) in the pACTII plasmid (Clontech) as prey proteins. The reporter gene used in this study is the HIS3 gene required for histidine biosynthesis. To monitor bait and prey interactions, appropriate pACTII (LEU2) and pGBKT7 (TRP1) plasmids were transformed into haploid cells from strain CG1945 (mat a; ura3‐52, his3‐200, ade2‐101, lys2‐801, trp1‐901, leu2‐3, 112, gal4‐542, gal80‐538, cyhr2, LYS2::GAL1UAS‐GAL1TATA‐HIS3, URA3::GAL417‐mers(x3)‐CYC1TATALacZ) and strain Y187 (mat α; gal4, gal80, ade2‐101, his3‐200, leu2‐3,112, lys2‐801, trp1‐ 901, ura3‐52, URA3::Gal1UAS GAL1TATA‐LacZ), respectively, using the lithium acetate method (Gietz and Schiestl, 2007). After crossing on YPD medium, diploid cells were selected on media suitable for double selection (Leu‐, Trp‐) and then

142 plated on media suitable for triple selection (Leu‐, Trp‐, His‐). Where indicated, 3‐ Amino‐1, 2, 4‐triazole (3‐AT) was added as a competitive inhibitor of histidine synthesis to evaluate the strength of the interactions. Growth was assessed after three days of incubation at 30°C. The interactions were confirmed in two independent experiments.

Antibodies

For Western blot analysis, the following antibodies were used according to manufacturer's instructions: Anti-PEPCK 1:1000 (Abgent), anti-MAEA 1:1000 (R&D systems), anti-C17orf39 1:500 (Aviva), anti-WDR26 1:1000 (Bethyl Lab), anti-Twa1 1:1000 (Proteintech), anti-FBP1 1:800 (Sigma), anti-RanBP9 1:1000 (Abcam), anti- Armc8 1:500 (Santacruz), anti-ACACA 1:1000 (Cell signaling), anti-FASN 1:1000 (Cell signaling), anti-CAD 1:1000 (Bethyl Lab), anti-LC3 I/II 1:1000 (Cell signaling), anti-p62/ SQSTM1 1:1000 (ProteinTech), anti-4E-BP1 1:100 (Cell signaling), anti- Phospho 4E-BP1 (Thr37/46) 1:1000 (Cell signaling), anti-Phospho-p70 S6 Kinase (Ser371) 1:1000 (Cell signaling), anti-AMPKα 1:1000 (Cell signaling), anti-Bicc1 1:1000 (raised by Proteogenix), anti-HA 1:1000 (Sigma), anti-γ-Tubulin 1:1000 (Sigma).

Analysis of kidneys lysates, plasma and urine measurements in Bicc1 mutant mice

Bicc1+/- heterozygous mice on a C57Bl/6 mouse genetic background were maintained in ventilated cages at the animal facility of the Ecole Polytechnique Fédérale de Lausanne (EPFL) (Maisonneuve et al., 2009). All animal experiments were approved by the Veterinary Service of the Swiss canton of Vaud or by the Institutional Animal Care and Use Committee and adhered to the guidelines in the Guide for the Care and Use of Laboratory Animals (National Research Council. 2011. Guide for the care and use of laboratory animals, 8th ed. National Academies Press, Washington, DC). Newborn mice were decapitated at postnatal day P2, and total blood glucose was measured using the Accu-Check Aviva Nano glucometer (Roche). For insulin and GLP-1 measurements, plasma was collected in EDTA tubes and frozen until analyzed by Meso Scale Discovery immunoassay kits in the Phenotyping

143 unit, CPG at EPFL. β-Hydroxybutyrate was measured in plasma using the β- Hydroxybutyrate (Ketone Body) Colorimetric Assay Kit from Cayman Chemical. Urine was collected after decapitation directly from the bladder by syringe aspiration, tubes and frozen until analyzed by Phenotyping unit, CPG at EPFL. Kidney and livers were immediately snap frozen in liquid nitrogen after dissection. For WB, the tissues were grinded with a pestle in lysis buffer and sonicated using a Bioruptor device (Diagenode). After centrifugation, protein concentration was quantified using Bradford assay (Bio-Rad).

Sample preparation and normalization for metabolomics

Kidney samples were pre-extracted and homogenized by the addition of 150 µL of MeOH:H2O (4:1) per 10 mg of frozen tissue weight, in the Cryolys Precellys 24 sample Homogenizer (2 x 20 seconds at 10000 rpm, Bertin Technologies, Rockville,

MD, US) with ceramic beads. The bead beater was air-cooled down at a flow rate of 110 L/min at 6 bar. Homogenized extracts were centrifuged for 15 minutes at 4000 g at 4°C (Hermle, Gosheim, Germany). The resulting supernatant was collected and evaporated to dryness in a vacuum concentrator (LabConco, Missouri, US). The protein pellets were evaporated and lysed in 20 mM Tris-HCl (pH 7.5), 4M guanidine hydrochloride, 150 mM NaCl, 1 mM Na2EDTA, 1 mM EGTA, 1% Triton, 2.5 mM sodium pyrophosphate, 1 mM beta-glycerophosphate, 1 mM Na3VO4, 1 µg/ml leupeptin using brief probe-sonication (5 pulses x 5 sec). BCA Protein Assay Kit (Thermo Scientific, Massachusetts, US) was used to measure (A562nm) total protein concentration (Hidex, Turku, Finland). The dry extracts were reconstituted using ACN:H2O (1:1, v/v) and the volume of re-suspension was adjusted based on measured protein content in each pellet (Ivanisevic et al., 2013).

Data acquisition and LC-MS analysis for metabolomics

Kidney extracts were analyzed by HILIC HESI (±) HRMS in full scan acquisition mode on a ThermoFisher Scientific LC-MS Q Exactive™ Hybrid Quadrupole-Orbitrap™ system interfaced with Thermo Accela 1250 UPLC pump and CTC PAL Analytics autosampler (Thermo Scientific, Massachusetts, US). Samples were analyzed using

144 a BEH Amide, 1.7 μm, 100 mm × 2.1 mm I.D. column (Waters, Massachusetts, US) in positive and negative ionization modes. The mobile phase was composed of A = 20 mM ammonium formate and 0.1 % FA in water and B = 0.1 % FA in 100% acetonitrile for positive mode and A = 25 mM ammonium acetate and 25 mM ammonium hydroxide in water and B = 100% acetonitrile for the negative mode. In positive mode, the linear elution gradient from 95% B (0-1.5’) down to 30% B was applied (1.5’-10.50’). A 4.5 min post run was applied for HILIC, to ensure the column re-equilibration and maintain the reproducibility. In negative mode, the linear elution gradient from 85% B (0-1.5’) down to 30% B was applied (1.5’-10.50’) followed by 4.5 min post run. The flow rate was 400 μL/min and the sample injection volume was 10µl. HESI source conditions were set as follows: sheath gas flow at 60, Aux gas flow rate at 20, Sweep gas flow rate at 0, spray voltage at 3kV, capillary temperature at 300°C, s-lens RF level at 60 and aux gas heater temperature at 50°C. The instrument was set to acquire over the m/z range 60-900, with the MS acquisition parameters set as follows: resolution at 70’000, micro scans at 1, AGC target at 1e6 and maximum inject time at 100. Following the data analysis, targeted MS/MS analysis of selected dysregulated metabolite features was performed using the inclusion list of ions of interest, isolation windows set at 4.0 m/z with stepped collision energy at 10, 30 and 50 eV and AGC target at 1e5.

Pooled Quality Control (QC) samples (representative of the entire sample set) were analyzed periodically (every 4 samples) throughout the overall analytical run to assess the quality of the data, correct the signal intensity drift (attenuation in most cases, that is inherent to LC-MS technique and MS detector due to sample interaction with the instrument over time) and remove the peaks with poor reproducibility (CV > 30%) that can be considered as a chemical or a bioinformatic noise. The resulting output, including a table of metabolite features, extracted ion chromatograms and multigroup cloud plot, were exported directly from the online package called XCMS (various forms (X) of chromatography mass spectrometry).

145 Statistical analysis for metabolomics

Parametric Welch t-test (or non-parametric Mann-Whitney) was used to extract the significantly altered metabolite features between WT and KO with an arbitrary level of significance, p-value = 0.05 (and q-value as an adjusted p-value using an optimized False Discovery Rate approach). Data were log transformed prior to t-test application. Multivariate statistical analysis was achieved with SIMCA software (Version 14.1, Umetrics, Sweden) and used for data quality assessment and exploration. The unsupervised Principal Component Analysis (PCA) was applied as an exploratory diagnostic tool to evaluate the data quality before and after the signal intensity drift correction, as well as for the global overview of sample grouping.

For comparison between different pathways, MetaboAnalyst 3.0 consider the total/maximum importance of each pathway being 1, the importance of each metabolite node is the percentage regarding the total pathway importance and the pathway impact value is the cumulative percentage from the matched metabolite nodes. This tool combines results from pathway enrichment with pathway topology analysis to facilitate the identification of the most relevant pathways dysregulated. The pathway analysis module integrates the KEGG (Kyoto Encyclopedia of Genes and Genomes) metabolic pathway database as the matching database and Mus musculus as the supported organism pathway library (Xia and Wishart, 2016). Importantly, the matched pathways generated are arranged by p-values of pathway enrichment (i.e matched with the KEGG database) rather than the p-values shown in Table 3. To estimate the impact of a given pathway and the importance of specific nodes, I included a pathway topology analysis. This comprises so-called node centrality measures (i.e betweenness centrality) since it is known that changes in key positions or “hubs” of a network will produce a bigger impact on the pathway than changes occurring in relatively isolated positions. Thus, the topological analysis gives us the pathway impact that allow us to know whether the metabolites whose levels significantly change in Bicc1-/- kidneys are placed in key positions for the identified enriched pathway.

146 Oil Red O staining

Cryo-sections of WT or Bicc1-/- kidneys at E.18.5 or P2 stage were subject to Oil red O staining and Hematoxylin counterstaining, at the EPFL histology facility following their standard protocol.

BODIPY™ staining

WT and CRISPR/Cas9 sgBicc targeted cells were cultured in normal media until treated with a mix of Oleic acid-BSA in warm DMEM without FBS in a final concentration of 400 µM of Oleic Acid (Sigma), during 24 hrs incubation. Then, cells were washed with PBS and trypsinized and washed and resuspended in PBS containing 1µg/mL of BODIPY™ 493/503 (4,4-Difluoro-1,3,5,7,8-Pentamethyl-4- Bora-3a,4a-Diaza-s-Indacene) from ThermoFisher. Staining took place following manufacturer’s instructions. After incubation at 37°C in the dark, cells were washed and resuspended in PBS containing DAPI for cytometry analysis. For quantification, signal intensity (A.U) was analyzed using Image J. For each condition, at least 3 fields were analyzed, A.U were measured and normalized by number of cells in each field.

OCR and ECAR measurements

Oxygen consumption rate (OCR) and extracellular acidification rate (ECAR) were measured using a Seahorse XF96 extracellular flux analyzer (Seahorse Bioscience, North Billerica, MA). 24 hours before the experiment, CRISPR/Cas9 Bicc1 targeted (sgBicc1) and wild-type mIMCD3cells were plated into on Seahorse XF-96 plates (Seahorse Bioscience) at a density of 8.000 cells per well. Cells were cultured in normal high glucose DMEM (Sigma) in a humidified incubator at 37°C with 5% CO2. Next day before running the experiment, the medium was replaced by XF base medium (supplemented with 10 mM glucose, 1 mM sodium pyruvate, 2mM glutamine pH 7.4) and incubated at 37 °C in a non-CO2 incubator for 1 hour. Machine protocol was set up at 37°C with baseline rates measured 5 times every 3 minutes followed by injection of inhibitors: oligomycin (2µg/ml) and FCCP (5 μM). Readings were taken 5 times every 3 minutes in between and after injections. After running the assay, the

147 protein content in each well was quantified by Bradford assay (BioRad). Bars represent the mean of six wells over 15 minutes of reading ± SEM. The graph is representative of two independent experiments.

Statistical analysis

Error bars represent the standard error or the mean (SEM). Two-tail student t-test was used to compare the differences in mean between 2 conditions to calculate p- values. 1-way ANOVA and Bonferroni’s multiple comparisons test were used to compare groups of unpaired values and determine the significance (p-value) of every mean compared to every other mean.

148 8. Annex

The following manuscript is in press in Structure.

149 Manuscript

Crystal structure of Bicc1 SAM polymer and mapping of physical interactions 1 2 between the ciliopathy-associated proteins Bicc1, ANKS3 and ANKS6 3 4 5 Benjamin Rothé1, Catherine N. Leettola2, Lucia Leal-Esteban1, Duilio Cascio2, Simon Fortier1, 6 7 Manuela Isenschmid1, James U. Bowie2, and Daniel B. Constam1. 8 9 10 11 1 Ecole Polytechnique Fédérale de Lausanne (EPFL) SV ISREC, Station 19, CH-1015 12 13 Lausanne, Switzerland 14 2 15 Department of Chemistry and Biochemistry, UCLA-DOE Institute of Genomics and 16 Molecular Biology Institute, University of California, Los Angeles, Boyer Hall 611 Charles 17 Proteomics, 18 E. Young Dr. E, Los Angeles, California 90095-1570, USA 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 1 64 65 Summary 1 2 3 4 Head-to-tail polymers of sterile alpha motifs (SAM) can scaffold large macromolecular 5 6 complexes. Several SAM domain proteins that bind each other are mutated in patients with 7 cystic kidneys or laterality defects, including the Ankyrin (ANK) and SAM domain-containing 8 9 proteins ANKS6 and ANKS3, and the RNA-binding protein Bicc1. To address how their interactions 10 11 are regulated, we first determined a high-resolution crystal structure of a Bicc1-SAM polymer, 12 13 revealing a canonical SAM polymer with a high degree of flexibility in the subunit interface 14 orientations. We further mapped interactions between full-length and distinct domains of Bicc1, 15 16 ANKS3 and ANKS6. Neither ANKS3 nor ANKS6 alone formed macroscopic homopolymers in vivo. 17 18 However, ANKS3 recruited ANKS6 to Bicc1, and the three proteins together cooperatively generated 19 20 giant macromolecular complexes. Thus, the giant assemblies are shaped both by SAM domains, their 21 22 flanking sequences, and SAM-independent protein-protein and protein-mRNA interactions. 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 2 64 65 Introduction 1 2 3 4 A growing number of disease-associated mutations are found in genes encoding proteins 5 6 involved in primary cilia biogenesis and function. These so-called ciliopathies include developmental 7 or later-onset chronic disorders such as nephronophthisis (NPH) and polycystic kidney diseases (PKD) 8 9 (for review, (Hildebrandt et al., 2011)). At the apical surface of epithelial cells, primary cilia integrate 10 11 mechanical, osmotic and chemical cues to control proliferation, differentiation and cell polarity 12 13 through Wnt, cAMP/PKA, Hippo, mTOR and other signaling pathways (for review, (Walz, 2017)). How 14 such diverse stimuli are integrated and transduced by ciliary proteins to control organogenesis and 15 16 tissue homeostasis is an active area of investigation. Limited space within cilia and specific scaffolds 17 18 formed by intraflagellar transport proteins may be important to facilitate rapid signal propagation by 19 20 concentrating specific molecules in close proximity to each other (for review, (Nachury, 2014)). 21 22 Signaling molecules may also be locally enriched in specific compartments by ‘polymerizers’ that 23 assemble dynamic platforms called signalosomes (for review, (Bienz, 2014)). A widespread 24 25 polymerizer is the sterile alpha motif (SAM), consisting of α-helical bundles of approximately 70 26 27 amino acids that can form dimers, closed oligomers or helical polymers in head-to-tail configuration 28 29 via so-called mid-loop (ML) and end-helix (EH) surfaces (for review, (Qiao and Bowie, 2005)). 30 Regularly arrayed macromolecular complexes scaffolded by SAM domain polymers can 31 32 accommodate associated proteins or RNAs, but little is known how the dynamics or signal amplitude 33 34 of such complexes are regulated (for review, (Bienz, 2014)). 35 36 37 38 Mutations in the SAM domains of ANKS6 and Bicc1 are linked to polycystic kidneys (Cogswell 39 et al., 2003; Kraus et al., 2012; Taskiran et al., 2014), whereas ANKS3 mutations have been identified 40 41 in human patients with autosomal recessive laterality defects, another trait common in several 42 43 ciliopathies (Shamseldin et al., 2016). ANKS6 contains 11 ANK repeats at its N-terminus and a SAM 44 45 domain close to its C-terminus, while ANKS3 consists of 6 N-terminal ANK repeats and a SAM domain 46 followed by 168 amino acids at the C-terminus. ANK repeats pack next to each other with a slight 47 48 torsion angle in rigid platforms to accommodate specific interacting proteins (for review, (Mosavi et 49 50 al., 2004)). The functional relevance of ANKS6 in kidney development is highlighted by missense 51 52 mutations in its SAM domain that cause cystic disease in the PKD cy/+ rat model (Brown, 2005), in 53 54 mouse (Bakey et al., 2015) and in human patients with a nephronophthisis-like disease (Taskiran et 55 al., 2014). In mouse kidneys, ANKS3 immunostaining has been reported in cilia (Delestré et al., 2015), 56 57 whereas in multi-ciliated Xenopus epidermal cells, GFP-tagged ANKS3 localized at basal bodies and in 58 59 large cytoplasmic foci (Yakulov et al., 2015). A role in kidney development was demonstrated by 60 61 62 63 3 64 65 ANKS3 knock-down in zebrafish that increased cysts already in the pronephros, the first stage 1 2 of kidney development (Yakulov et al., 2015). 3 4 5 ANKS3 and ANKS6 interact both with each other and with Bicc1 (Bakey et al., 2015; Delestré et al., 6 7 2015; Leettola et al., 2014; Ramachandran et al., 2015; Stagner et al., 2009; Yakulov et al., 2015). A 8 9 link of Bicc1 to cystic kidneys was discovered in bpk mutant mice, a non-orthologous model of 10 autosomal recessive PKD (ARPKD), where a GC insertion changes the reading frame so that the last 11 12 21 amino acids are replaced by 149 aberrant residues (Cogswell et al., 2003). Bicc1 is an RNA-binding 13 14 protein composed of three KH and two KH-like domains that are linked by an intervening sequence 15 16 to a C-terminal SAM domain. Targeted deletion of Bicc1 revealed additional roles in the planar 17 18 positioning of motile node cilia that specify visceral left-right asymmetry and in repressing the 19 translation of Adenylate Cyclase 6 (AC6) and PKA inhibitor α (PKIα) mRNAs in kidneys of newborn 20 21 mice (Maisonneuve et al., 2009; Piazzon et al., 2012). Bicc1 binds the 3'UTRs of these target mRNAs 22 23 via KH domains, whereas self-polymerization of its SAM domain mediates recruitment to large 24 25 cytoplasmic foci for translational silencing (Rothé et al., 2015). Recently, Bicc1 has also been 26 detected in primary cilia of porcine renal epithelial LLC-PK cells (Mohieldin et al., 2015) and at 27 28 centrosomes of dividing mIMCD3 cells, where it appears to stimulate the translation of specific 29 30 target mRNAs, rather than inhibiting it (Iaconis et al., 2017). Despite 41 to 57% sequence identity, 31 32 the SAM domains of Bicc1, ANKS6 and ANKS3 differ in their potential to self polymerize. The ANKS6 33 34 SAM domain is intrinsically unable to self-interact, whereas recombinant SAM domains of Bicc1 and 35 ANKS3 spontaneously form homopolymers (Knight et al., 2011; Leettola et al., 2014). Self-association 36 37 via the SAM domain has been confirmed in vivo for full-length Bicc1 (Rothé et al., 2015). A 38 39 homology-based 3D model of Bicc1-SAM polymers suggests a helical conformation with a pitch of 6 40 41 SAM domains (Rothé et al., 2015), in contrast to the crystal structure of the ANKS3-SAM polymer 42 which revealed a pitch of 8 SAM domains per turn (Leettola et al., 2014). How ANKS3 or ANKS6 may 43 44 affect the conformation or function of Bicc1 homo- or heterooligomers is unknown. 45 46 47 48 Here, we use structural and biochemical approaches combined with yeast two-hybrid assays to 49 50 comprehensively map interacting domains of Bicc1, ANKS3 and ANKS6 and to assess their 51 potential to heterooligomerize and to regulate the assembly or activity of cytoplasmic 52 53 Bicc1. Determination of the crystal structure of homopolymeric Bicc1 SAM domain and 54 55 measurements of its affinity for recombinant ANKS3 SAM revealed that the ML and EH surfaces 56 57 of each isolated SAM domain can mediate homo- or heterooligomer formation in vitro, 58 consistent with a model of co-polymerization. However, analysis in a cellular context shows 59 60 that full-length ANKS3 instead disperses large Bicc1 assemblies through its long C-terminal 61 62 domain, suggesting this flanking 63 4 64 65 sequence inhibits the elongation of SAM domain co-polymers. Furthermore, we show that full-length 1 2 ANKS6 associates with Bicc1 indirectly through ANKS3, thereby rescuing the formation of Bicc1- 3 ANKS3 complexes in large cytoplasmic foci. Our findings suggest that the assembly or disassembly of 4 5 diverse heteromeric macromolecular SAM-mediated scaffolds is regulated by distinct biophysical 6 7 properties of both the SAM domains and their relative positioning and specific flanking sequences. 8 9 10 11 12 13 Results 14 Crystal structure of the Bicc1 SAM domain polymer 15 16 The sequences of Bicc1 and ANKS3 SAM domains are closely related, especially in their ML and 17 18 EH surfaces (63% and 67% identity, respectively) (Fig. 1A). To characterize Bicc1 homo- 19 20 oligomers, we solved a high resolution crystal structure using a mutant SAM domain of human Bicc1 21 22 where residue R924 was substituted by glutamic acid. This mutation destabilizes the self-association 23 24 of Bicc1 SAM (see below) enough to solubilize the SAM domain sufficiently for crystallization. Details 25 of the structure determination and refinement are provided in Table 1 and in Methods. We 26 27 solved the structure by multi-wavelength anomalous dispersion (MAD) phasing using a 28 29 selenomethionine (SeMet)-replaced Bicc1 SAM R924E mutant. The SeMet structure was then 30 31 used to solve the structure of a native R294E mutant. Since the SeMet structure could be refined 32 to a higher resolution (1.75 Å) than the native structure (2.00 Å), we employed the SeMet structure 33 34 for analysis of the SAM domain interactions. There were three similar but not identical molecules 35 36 in the asymmetric unit (RMSD on all atoms is 0.430- 0.631 Å) that assemble to form a helical 37 38 polymer with 6 SAM domains per turn, much like many other SAM domains (Fig. 1B). Like 39 40 other SAM polymers, the subunit interfaces are formed between mid-loop (ML) and end-helix (EH) 41 interfaces, but there are unusually large differences in orientation between the non-identical 42 43 subunits in the crystal, suggesting a high degree of structural flexibility in the Bicc1-SAM polymer 44 45 (Fig. 1B, C). The ML surface of Bicc1 consists of negatively charged residues around a shallow 46 47 hydrophobic patch (Fig. 1D, left) that packs against a Phe protrusion surrounded by complementary 48 positively charged residues in the EH surface (Fig. 1D, right). Interface residues involved in ionic 49 50 interactions (D900, D911, K913), hydrogen bonds (E898, K913, K925), hydrophobic packing (F920), 51 52 and shape complementarity (packing of T903 against the helical backbone of the EH surface) 53 54 stabilize Bicc1 SAM domain self-interactions (Fig. 1D). As electrostatic interactions are 55 56 adaptable, it is possible that the electrostatically stabilized interface leads to its apparent 57 flexibility. To test whether flexibility may allow SAM polymer architectures to vary, we modeled 58 59 polymers using each of the three interfaces separately. We found that polymers constructed with 60 61 only the BA interface have a helical repeat of ~4 SAM domains and a helical pitch of ~48 Å. When 62 63 constructed with only the CB interface, there were ~11 SAM domains per helical repeat 5 64 65 with a pitch of 158 Å, whereas polymers with only the AC interface harbored ~7 SAM domains 1 2 per helical repeat with a pitch of 54 Å. Thus, it is conceivable that the flexibility to switch interfaces 3 may enable Bicc1 SAM polymers to adjust their conformations to environmental conditions. 4 5 6 7 To validate that the structure seen in the crystal corresponds to the polymer in solution, we 8 9 performed mutational analysis using a native gel assay (Knight et al., 2011). In this assay we fuse the 10 SAM domain to a negatively charged GFP (negGFP), which provides consistent migration toward the 11 12 cathode. Polymeric SAM-negGFP fusions migrate more slowly on a native gel than monomers. As 13 14 shown in figure 1E, ML surface mutations E898K, D900K, T903E, D911K and the EH surface mutations 15 16 K913E, F920E, R924E, K925E each increased the mobility of negGFP fusion proteins on native gels, 17 18 indicating that they destabilized Bicc1 homopolymers as expected from the crystal structure, even 19 though not all the residues make contacts in every orientation in the polymer. 20 21 22 23 Bicc1 and ANKS3 SAM domains associate in a manner compatible with co-polymerization 24 25 The structure of the Bicc1 SAM domain strongly resembles that of ANKS3, showing an 26 27 average root-mean-square deviation (RMSD) across backbone atoms of 0.754 Å, and of 0.814 28 29 Å across all atoms (Leettola et al., 2014). Since Bicc1 and ANKS3 have recently been shown to 30 31 interact (Ramachandran et al., 2015; Stagner et al., 2009; Yakulov et al., 2015), we tested 32 whether negGFP fusions of their SAM domains bind each other in gel shift assays. We found that 33 34 negGFP-tagged Bicc1 SAM and negGFP-tagged ANKS3 SAM exhibited diminished electrophoretic 35 36 mobility upon 1:1 mixing of the two proteins, indicating robust binding (Fig. 1E). To identify the 37 38 protein surfaces mediating this interaction, we mixed ANKS3 SAM 1:1 with Bicc1 SAM domains 39 40 containing specific point mutations in either the ML or EH surface. Irrespective of whether the Bicc1 41 SAM domain was mutated in its ML or EH surface, it still shifted ANKS3 (Fig. 1E), suggesting that 42 43 either of these surfaces alone are sufficient to bind the SAM domain of ANKS3. In other words, 44 45 ANKS3-SAM can bind to both surfaces. To validate this conclusion, we mixed the Bicc1 ML mutants 46 47 E898K and D911K 1:1 with ANKS3 EH mutant F472E for analysis by size exclusion chromatography. 48 A definite elution upshift confirmed heterodimer formation (Fig. 1F). We also tested 49 50 functionality of the alternate interface by mixing the Bicc1 EH mutants F920E and R924E with the 51 52 ANKS3 ML mutant I455E. Analysis of the resulting complexes by Size Exclusion Chromatography 53 54 with Multi-Angle Light Scattering analysis (SEC-MALS) revealed in each case a single peak, eluting 55 56 at molecular weights of 17.4 ± 0.6 kDa (Bicc1 R924E/ANKS3 I455E) or 18.4 ± 0.1 kDa (Bicc1 F920E/ 57 ANKS3 I455E), respectively, consistent with the predicted size of heterodimers (16.1 kDa) (Fig. 58 59 1G). These results show that the Bicc1 and ANKS3 SAM domains can bind each other via two 60 61 distinct interfaces: Bicc1-EH + ANKS3-ML and Bicc1-ML + ANKS3-EH. 62 63 6 64 65 To measure the strength of these interactions by SPR, we first immobilized the Bicc1 EH mutant 1 2 R924E and added increasing concentrations of the ANKS3 ML mutant I455E in the mobile phase, 3 allowing only their non-mutated native surfaces to interact. The binding affinity of this Bicc1-ML/ 4 5 ANKS3-EH interface measured by SPR was 2.2 ± 0.1 μM (Fig. S1A). Conversely, SPR analysis of 6 7 immobilized ANKS3 EH mutant F472E with increasing concentrations of the Bicc1 ML mutants E898K 8 9 or D911K revealed a binding affinity of 1.35 ± 0.1 μM for the Bicc1-EH/ANKS3-ML interface (Fig. S1B, 10 C). This interaction was specific since ANKS3 EH mutant F472E did not bind Bicc1 E898K/F920E or 11 12 D911K/F920E where both ML and EH surfaces are disrupted (Fig. S1D). Since the affinities of both 13 14 possible interfaces are in the same order of magnitude, and are similar to the measured affinity of 15 16 Bicc1 SAM for itself (1.35 ± 0.1 μM) (Fig. S1E-G), the Bicc1 and ANKS3 SAM domains may form 17 18 alternating co-polymers. Alternatively, the EH surface of ANKS3 can strongly bind the ML surface of 19 ANKS6 (Kd = 249 ± 8 nM) (Leettola et al., 2014). Given its similarity, the EH surface of Bicc1 20 21 should bind ANKS6 as well. We tested this hypothesis using negGFP fusion proteins. When the 22 23 wild-type SAM domains of Bicc1 and ANKS6 were mixed 1:1, there was no gel shift, indicating that 24 25 they failed to interact under the conditions examined (Fig. S1H). Therefore, ANKS6 seems less likely 26 than ANKS3 to form SAM domain co-polymers with Bicc1. 27 28 29 30 Semi-quantitative analysis of intramolecular self-interactions by yeast two-hybrid assays 31 32 To map intra- and intermolecular interactions of Bicc1, ANKS3, ANKS6 and their respective 33 34 domains, we conducted yeast two-hybrid (Y2H) assays in S. cerevisiae where neither Bicc1 nor ANKS3 35 36 or ANKS6 are conserved (Figs. 2A-I, S2). In yeast, interactions between a bait (fused to the 37 DNA binding domain of Gal4, Gal4‐BD) and a prey (fused to the Gal4 activation domain, Gal4‐ 38 39 AD) is revealed by selective growth conferred by induction of the HIS3 reporter gene, while 40 41 titration of 3‐Amino‐1,2,4‐triazol as a competitive inhibitor of the HIS3 gene product serves to 42 43 estimate the strength of each interaction. Each full-length protein (FL) and individual domains were 44 tested as bait and prey to reduce false negative results. Empty prey plasmid was used to detect false 45 46 positives and the threshold above which interactions were considered specific. Y2H assays 47 48 unexpectedly revealed significant binding of Bicc1-FL to its own KH domains (Fig. 2A, B), and it 49 50 confirmed that the Bicc1 SAM domain strongly interacts with itself (Fig. 2C). By contrast, self- 51 52 interactions of ANKS3-SAM or ANKS6-SAM were weak or undetectable, respectively (Fig. 2F, I). 53 These results are consistent with prior results indicating that the SAM domain of Bicc1 is more 54 55 prone to self-polymerize than that of ANKS3, and that the ANKS6 SAM is monomeric (Knight et al., 56 57 2011; Leettola et al., 2014). 58 59 60 61 62 63 7 64 65 Y2H reveals an intricate interaction network between Bicc1, ANKS3 and ANKS6 1 2 Y2H analysis of Bicc1, ANKS3 and ANKS6 in pairs revealed that full length Bicc1 and ANKS3 and 3 4 their respective SAM domains bind each other (Fig. 2A, C, D, F). Strong interactions were also 5 detected between ANKS3 and ANKS6 and between their SAM domains (Fig. 2D, F, G, I). In 6 7 sharp contrast, only minimal binding was detected between isolated SAM domains of Bicc1 and 8 9 ANKS6, (Fig. 2C, I), or the corresponding full-length proteins (Fig. 2A, G). These data confirm the 10 11 conclusion of our gel shift assay and extend it to full-length proteins that the Bicc1 SAM domain 12 preferably binds the SAM domain of ANKS3 and not ANKS6. Additional robust interactions between 13 14 Bicc1 and ANKS3 involved their ANK and KH domains. Especially the ANK repeats of ANKS3 15 16 strongly bound Bicc1-FL both as bait and as prey (Fig. 2A, E) and, conversely, the KH domain of 17 18 Bicc1 bound ANKS3-FL (Fig. 2B, D). However, Bicc1-KH and ANKS3-ANK domains alone failed to 19 20 directly interact (Fig. 2B, E), suggesting that additional determinants are required for their stable 21 association. Equally unexpected, ANKS3-FL strongly interacted with ANKS6-ANK (Fig. 2D). This 22 23 interaction was confirmed using ANKS6-ANK as a bait despite increased background and, conversely, 24 25 by significant binding of ANKS6-FL as a bait with ANKS3-ANK (Fig. 2G). Together, these observations 26 27 suggest that Bicc1-ANKS3 and ANKS3-ANKS6 complexes rely both on SAM-SAM interaction, as well 28 as additional binding surfaces involving ANK repeats and KH domains (Fig. 2J). 29 30 31 32 ANKS3 recruits ANKS6 to Bicc1 33 34 Since ANKS6 associated with ANKS3 but not with Bicc1 in Y2H assays, it may be recruited to Bicc1 by 35 36 ANKS3. To test this hypothesis, we used HEK293T cells that present the advantage of not expressing 37 endogenous Bicc1, and we performed co-immunoprecipitation (coIP) assays by co-expressing HA- 38 39 Bicc1 or v5-ANKS6 or both with ANKS3-Flag. Irrespective of whether HA-Bicc1 and v5-ANKS6 were 40 41 co-expressed jointly or individually, they each co-immunoprecipated with ANKS3-Flag (Fig. 3A). Thus, 42 43 Bicc1 and ANKS6 do not appear to compete with each other for binding to ANKS3. Conversely, a HA- 44 45 Bicc1 pull down efficiently enriched ANKS3-Flag as well as an ANKS3-Flag truncation mutant lacking 46 the C-terminal 168 amino acids (ANKS3ΔCter-Flag) (Fig. 3B). These results confirm Bicc1-ANKS3 47 48 binding and show that it is independent of the ANKS3 C-terminal domain. In sharp contrast to 49 50 ANKS3-Flag, v5-ANKS6 barely co-immunoprecipitated with HA-Bicc1, in line with two-hybrid assays 51 52 showing that these ANKS6 and Bicc1 only weakly interact. Possibly, small amounts of endogenous 53 ANKS3 stabilize this interaction, because in the presence of ANKS3-Flag, v5-ANKS6 was greatly 54 55 enriched in HA-Bicc1 immunoprecipitates, indicating that ANKS3 acts as a bridge to stabilize binding 56 57 of Bicc1 to ANKS6. 58 59 60 61 62 63 8 64 65 Analysis of Bicc1 homo- and heterooligomers by sucrose density gradient fractionation confirms 1 ANKS3-mediated ANKS6 recruitment in high molecular weight complexes 2 3 To confirm whether ANKS6 recruitment to Bicc1 is mediated by ANKS3, we analyzed oligomeric 4 5 complexes by sucrose density gradient fractionation and their re-partitioning to different density 6 fractions after co-transfection in HEK293T cells. This approach allows to discriminate complexes by 7 8 their size and to evaluate how each subunit influences the protein network extension. Co- 9 10 transfection of Bicc1 with ANKS3 shifted their respective curves from two peaks to a single one with 11 12 a maximum in the intermediate molecular weight (IMW) range around fraction 13 (Fig. 4A-C). 13 14 Interestingly, deletion of the C-terminal domain of ANKS3 shifted ANKS3 and Bicc1 complexes even 15 further to the high molecular weight (HMW) range and in amounts that far exceeded those observed 16 17 without exogenous ANKS3 (43% versus 11%). The same HMW fractions 19-23 also enriched ANKS3 18 19 and ANKS6 if both proteins were co-expressed with or without Bicc1, but not if either of them was 20 21 transfected alone (Fig. 4A-D). In sharp contrast, no interference was observed between Bicc1 and 22 ANKS6 (Fig. 4A, B, D), arguing that the effects of ANKS3 were highly specific. As an additional control, 23 24 the distributionof γ-tubulin remained unchanged in all conditions examined, confirming correct 25 26 sucrose fractionation (Fig. 4E). Thus, Bicc1, ANKS6 and ANKS3 form complexes of dramatically 27 28 diverse sizes, depending on their combination and on the C-terminal domain of ANKS3, which 29 30 inhibits the recruitment of Bicc1 into heterooligomeric HMW complexes. 31 32 33 Full length ANKS3 via its C-terminal domain can inhibit cytoplasmic clustering of Bicc1 34 35 Since HA-Bicc1 and ANKS3-Flag co-fractionated in sucrose gradient, and because recruitment of 36 37 ANKS6 increased the apparent average size of HA-Bicc1 and ANKS3-Flag in HMW complexes, we 38 39 asked whether these effects correlate with increased Bicc1 clustering in cytoplasmic foci. To address 40 this question, we analyzed combinations of HA-Bicc1, ANKS3-Flag and v5-ANKS6 in HeLa cells 41 42 by indirect immunofluorescent staining. As shown previously, HA-Bicc1 formed large 43 44 cytoplasmic clusters, reflecting its strong capacity to polymerize (Maisonneuve et al., 2009; Rothé et 45 46 al., 2015). By contrast, ANKS3-Flag alone localized diffusely throughout the cytoplasm, and v5- 47 48 ANKS6 alone was enriched at cell protrusions and at the cortex beneath the plasma membrane 49 (Fig. 5A). To evaluate effects of full-length proteins on each other’s localization, we co- 50 51 expressed ANKS3-Flag with v5-ANKS6, followed by immunostaining with mouse anti-Flag and 52 53 rabbit anti-ANKS6 antibodies in combination (Fig. 5B) or with mouse anti-Flag or mouse anti-v5 54 55 individually (Fig. 5C). ANKS3 and ANKS6 drastically changed their distinct individual distributions 56 and co-localized in large cytoplasmic foci only if they were co-expressed. ANKS3-Flag also 57 58 translocated to cytoplasmic foci when co-expressed with HA-Bicc1, although these were 59 60 extremely small compared to foci of Bicc1 alone or of ANKS3-ANKS6 complexes (Fig. 5D). To assess 61 62 how ANKS3 disperses large Bicc1 foci, we compared it 63 9 64 65 to mutant ANKS3ΔCter-Flag. Contrary to wild-type ANKS3, truncated ANKS3ΔCter-Flag co-localized 1 2 with HA-Bicc1 exclusively in large or medium size cytoplasmic foci in 90% and 10% of the cells, 3 respectively (n=60, Fig. 5E). Only the smaller of these percentiles showed significant diffuse staining 4 5 of both proteins also outside foci throughout the cytoplasm (bottom panel). This result shows that 6 7 ANKS3 disperses Bicc1 foci specifically through its C-terminal domain. In sharp contrast to ANKS3- 8 9 Flag, v5-ANKS6 did not translocate to HA-Bicc1 foci (Fig. 5D). This result confirms our earlier 10 conclusion that ANKS6 and Bicc1 do not directly interact, and that co-localization of Bicc1 with ANKS3 11 12 is not merely a non-specific consequence of any SAM domain or ANK repeat protein overexpression. 13 14 15 16 To assess whether Bicc1-ANKS3 foci can be influenced by ANKS6 or vice versa, we co- 17 18 transfected all three proteins. Double immunostaining of HA-Bicc1 and ANKS3-Flag or HA-Bicc1 19 and v5-ANKS6 revealed that the presence of v5-ANKS6 suppressed the scattering of Bicc1 foci by 20 21 ANKS3-Flag and instead induced the formation of even larger cytoplasmic domains harboring 22 23 all three proteins (Fig. 5F). Interestingly, ANKS3-Flag was specifically enriched at the periphery but 24 25 not in the center of such foci, whereas HA-Bicc1 and v5-ANKS6 were homogeneously 26 distributed. Besides confirming the conclusion of the coimmunopreciptation experiments showing 27 28 that ANKS6 is linked to Bicc1 by ANKS3, this result shows that ANKS6 recruitment alters the 29 30 topology of macromolecular Bicc1-ANKS3 complexes in a non-random consistent fashion. 31 32 33 34 The Bicc1-ANKS3-ANKS6 network does not rely exclusively on SAM-SAM interfaces 35 36 To evaluate the contribution of the Bicc1 SAM domain to ANKS3 or ANKS6 interactions, we 37 conducted co-immunoprecipitation assays using truncated Bicc1ΔSAM or the polymerization mutant 38 39 Bicc1 mut-D where residues D913, K915 and E916 (mouse Bicc1 numbering) are substituted by alanines 40 41 to disrupt the ML and EH surfaces of the SAM-SAM interface (Rothé et al., 2015) (Fig. S3A). 42 43 Surprisingly, both mutants still coimmunoprecipitated ANKS3 alone or together with ANKS6 (Fig. 44 S3B). In parallel, we evaluated the potential of ANKS3 and ANKS6 to cluster HA-tagged Bicc1 mut-D 45 46 or Bicc1ΔSAM in cytoplasmic foci. Both Bicc1 mut-D and Bicc1ΔSAM are unable to self-polymerize in 47 48 large cytoplasmic clusters on their own (Maisonneuve et al., 2009; Rothé et al., 2015). Their 49 50 cytoplasmic localization also remained diffuse in the presence of ANKS3. However, when co- 51 52 expressed together with ANKS6, ANKS3 significantly rescued the localization of HA-Bicc1ΔSAM or 53 HA-Bicc1 mut-D in cytoplasmic foci (Fig. S3C, D). These results agree with our Y2H data that Bicc1 54 55 binds ANKS3 in part independently of a SAM-SAM interaction through its KH domains, and that the 56 57 ANKS3 SAM domain interfaces with the SAM domain of ANKS6 (Figs. 2B, D, F, I, S2). Thus, ANKS3- 58 59 ANKS6 hetero-oligomers can recruit Bicc1 in part independently of its own SAM domain, whereas 60 Bicc1 SAM 61 62 63 10 64 65 domain polymerization reinforces this network and remodels the shape of the resulting 1 2 macromolecular complexes. 3 4 5 While ANKS3 and ANKS6 do not affect Bicc1-mediated silencing of a reporter mRNA, mRNA binding 6 instead changes the topology of protein complexes 7 8 The molecular function of polymeric Bicc1 involves the cytoplasmic localization and translational 9 10 repression of bound mRNA (Rothé et al., 2015). Prompted by our finding that ANKS3 and ANKS6 can 11 12 modulate the topology of cytoplasmic Bicc1 clusters, we hypothesized that they regulate Bicc1- 13 14 mediated silencing of the target 3′ UTR of AC6 mRNA (Piazzon et al., 2012). Surprisingly, Bicc1 15 repressed the Luc-AC6 3'UTR reporter irrespective of the presence or absence of exogenous ANKS3 16 17 or ANKS6 (Fig. 6A). To assess whether the mRNA itself formed a scaffold to restore large Bicc1 foci, 18 19 we imaged the cytoplasmic distribution of a Luc-AC6-MS2x27 reporter mRNA by tethering 20 21 fluorescent MS2-YFP fusion protein which normally is retained in the nucleus by a nuclear 22 localization signal (NLS) (Fig. 6B) (Rothé et al., 2015). As described previously, co-expression with HA- 23 24 Bicc1 alone efficiently recruited the Luc-AC6-MS2x27 reporter and associated MS2-YFP to 25 26 cytoplasmic HA-Bicc1 foci (Fig. 6C). In the absence of Luc-AC6-MS2x27 mRNA, ANKS3-Flag with or 27 28 without HA-Bicc1 as a control did not alter the nuclear localization of MS2-YFP (Fig. 6D, rows 1 and 29 30 3). Co-expression with ANKS3-Flag alone also did not influence the localization of Luc-AC6-MS2x27 31 mRNA or vice versa (Fig. 6D, row 2). However, when co-expressed together with both HA-Bicc1 and 32 33 ANKS3-Flag, the reporter mRNA rescued large Bicc1 foci despite the presence of Flag-ANKS3 (Fig. 6D, 34 35 row 4). Moreover, Flag-ANKS3 staining was conspicuously absent at the center of such rescued Bicc1 36 37 foci and instead decorated their periphery (Figs. 6D, S4), similar to what we observed when ANKS3 38 and Bicc1 are jointly coexpressed with ANKS6 (Fig. 5F). These results suggest that RNA binding can 39 40 dramatically alter the spatial configuration of macromolecular Bicc1 protein complexes. 41 42 43 44 The bpk mutation does not block the association of Bicc1 with ANKS3 or ANKS6 45 46 The bpk mouse model is a noncongenital model of ARPKD characterized by an aberrant C- 47 48 terminal extension that inhibits Bicc1 polymerization and stabilization in large cytoplasmic 49 foci (Cogswell et al., 2003; Rothé et al., 2015). To evaluate whether the bpk mutation affects 50 51 interactions of Bicc1 with ANKS3 or ANKS6, we co-expressed ANKS3-Flag and v5-ANKS6 alone or in 52 53 combination with bpk mutant HA-Bicc1. Similar to wild-type HA-Bicc1, the bpk mutant co- 54 55 immunoprecipitated with full-length and C-terminally truncated ANKS3-Flag, and it did not 56 impair the recruitment of ANKS6 (Fig. 7A). By contrast, when co-expressed with ANKS3 and ANKS6 57 58 together in HeLa cells, Bicc1 bpk only partly accumulated in ANKS3-ANKS6 foci and partly remained 59 60 diffusely localized throughout the cytoplasm (Fig. 7B). Besides confirming that the ANKS6- 61 62 mediated scaffolding of Bicc1-ANKS3 63 11 64 65 complexes into giant clusters is partly independent of SAM domain polymerization, these data show 1 2 that the bpk mutant phenotype is not simply caused by a lack of ANKS3 or ANKS6 binding, but 3 possibly by failure to correctly connect such complexes in a mesh-like macromolecular network 4 5 (Fig. S5). 6 7 8 9 Global analysis revealed that SAM domains mainly localize in protein C-terminal regions 10 11 Since SAM domain-mediated scaffolding of large Bicc1 foci was inhibited both by the 12 aberrant C-terminal elongation of the bpk frame-shift mutation and by ANKS3 via its long C-terminal 13 14 region after the SAM domain, we queried the InterPro database (Finn et al., 2017) 15 16 (https://www.ebi.ac.uk/interpro/) to systematically address whether the positioning of SAM domains 17 18 is skewed in favor of long or short flanking sequences. In 178 proteins larger than 280 amino acids 19 20 and containing a single SAM domain, we determined whether the center of the SAM domain resides 21 in the N-terminal or C-terminal quartiles of the host protein, or in the middle section (from 25 to 75% 22 23 of the sequence). Indeed, 68% of the SAM domains analyzed resided in the C-terminal quartile, 24 25 compared to only 17% and 14% that were found in the N-terminal or middle regions (Table S1). 26 27 Furthermore, in proteins where the SAM domain is positioned near the C-terminus, the average 28 number of residues after the SAM domain is only 29. This bias for SAM domains to localize near the 29 30 C-terminus may reduce the risk of self-aggregation of unfinished polypeptide chains during mRNA 31 32 translation. Alternatively, overcrowding of the space around the SAM polymer helix by bulky N- and 33 34 C-termini together may sterically hinder head-to-tail polymerization. 35 36 37 38 39 Discussion 40 41 SAM domains often form polymers in vitro, but how this process or their 42 43 heterooligmerization are regulated is unclear. Since complexes of Bicc1, ANKS3 and ANKS6 have 44 45 previously been linked to polycystic kidneys, we explored the regulation of their assembly into homo- 46 47 and heterooligomers by SAM domains and their flanking sequences and by a Bicc1 target RNA. 48 Crystal structure determination and biophysical analysis revealed that the SAM domain of Bicc1 49 50 polymerizes in a flexible manner with itself. Alternatively, the ML and EH surfaces can also mediate 51 52 co-polymerization with the SAM domain of ANKS3, but not ANKS6. Unexpectedly, reconstitution 53 54 experiments in heterologous cells showed that neither ANKS3 nor ANKS6 formed macroscopic 55 56 homopolymers on their own, in sharp contrast to Bicc1. However, binding to ANKS3 dispersed foci of 57 Bicc1 polymers, whereas co-recruitment of ANKS6 by ANKS3 enlarged them. ANKS6 induced large 58 59 ANKS3-ANKS6 heteropolymers that recruited Bicc1 partly independently of its SAM domain, likely 60 61 through additional interfaces involving Bicc1 KH domains and ANK repeats of ANKS3. Interestingly, 62 63 12 64 65 binding of a synthetic mRNA to Bicc1 KH domains similarly rescued the assembly of Bicc1-ANKS3 1 2 heterooligomers into giant foci, with a stereotypic spatial re-arrangement of ANKS3 at their 3 periphery. The potential of individual SAM domains to form homo- or heteropolymers was 4 5 modulated by flanking sequences, suggesting it cannot be directly inferred from analyzing only their 6 7 interactions in isolation. Together, our observations allow us to propose biochemical rules for the 8 9 self-assembly of these macromolecular RNP complexes as a first approximation. Stoichiometries and 10 the localization of heterooligomeric complexes likely vary in physiological contexts, but given the vast 11 12 number of SAM domain proteins, similar rules may govern the assembly of other SAM domain- 13 14 mediated heterooligomeric protein scaffolds. 15 16 17 Structure of the Bicc1 SAM domain and mutational analysis of its interaction with ANKS3 SAM 18 19 The high-resolution crystal structure revealed that the helical SAM domain polymer lacks radial 20 21 symmetry because the ML and EH surfaces of each SAM subunit interfaced with each other in 3 22 23 different orientations. Interestingly, all residues validated by point mutations to be involved in 24 polymerization made contacts in at least one of the three interfaces, but no binding interface 25 26 involved all of these residues. We cannot formally rule out that such variability is an artifact linked 27 28 e.g. to the R924E mutation that we used to increase protein solubility before crystallization. 29 30 However, flexible binding has previously been observed between the SAM domains of EphA2 and 31 32 SHIP2, where native ML-EH interactions populate two different orientations (Lee et al., 2012). 33 Flexibility resulting from the electrostatic nature of the ML-EH binding interface may increase the 34 35 versatility of macromolecular Bicc1 scaffolds by facilitating malleable docking of ANKS3 or other 36 37 interacting proteins. Our data identify the SAM domain of ANKS3 as the first direct binding partner of 38 39 the Bicc1-SAM. Affinity measurements and point mutations showed that the SAM:SAM interfaces in 40 heteromeric ANKS3/Bicc1 SAM complexes have similar binding affinities as those in Bicc1-SAM (this 41 42 study) and ANKS3-SAM homopolymers (Leettola et al., 2014), suggesting they likely consist of 43 44 alternating copolymers rather than blocks of two homooligomers. A report that a GFP fusion of 45 46 ANKS3 forms large cytoplasmic aggregates in multiciliated epidermis of Xenopus embryos was 47 48 consistent with both scenarios but did not distinguish between them (Yakulov et al., 2015). 49 50 51 ANKS3 can disperse or enlarge polymeric Bicc1 scaffolds, depending on its association with 52 ANKS6 53 54 55 In contrast to Bicc1- and ANKS3-SAM domains, the ANKS6-SAM previously has been shown to only 56 57 interact with the EH surface of the ANKS3-SAM but not with itself (Knight et al., 2011; Leettola et al., 58 2014). However, interactions between full-length proteins remained to be mapped. Here, 59 60 reconstitution in heterologous cells showed that neither ANKS6 nor ANKS3 alone accumulated in 61 62 63 13 64 65 HMW fractions of sucrose density gradients or as macroscopic higher order complexes in cytoplasmic 1 2 foci, suggesting they do not form long homopolymers in vivo. Thus, large cytoplasmic foci of ANKS3 3 described in zebrafish embryos (Yakulov et al., 2015) are likely assembled by interacting factors. 4 5 Possible candidates include ANKS6 but not Bicc1, because our reconstitution analysis in HeLa cells 6 7 showed that ANKS3 and ANKS6 congregate in cytoplasmic foci only if they are co-expressed (with or 8 9 without Bicc1), whereas complexes of ANKS3 with Bicc1 alone were dispersed throughout the 10 cytoplasm. In good agreement, our Y2H and coimmunoprecipitation assays showed that ANKS3 11 12 independently binds Bicc1 and ANKS6. In sharp contrast, binding of Bicc1 to ANKS6 was minimal, 13 14 except in the presence of ANKS3. These results strongly suggest that ANKS6-Bicc1 complexes 15 16 previously described in kidneys (Bakey et al., 2015) are stabilized by ANKS3. Previous analysis of Bicc1 17 18 and ANKS6 deletion constructs predicted that they depend on a third partner to mediate their 19 interaction between the ANKS6-SAM and Bicc1-KH domains (Stagner et al., 2009). We propose that 20 21 this partner is ANKS3. Overall, our findings suggest that only Bicc1 forms self-polymers on its own 22 23 and that ANKS3 limits their enlargement, whereas co-recruitment of ANKS6 promotes it, possibly by 24 25 capping the ANKS3-SAM domain so that segments of homopolymeric Bicc1 have more chance to 26 form inside the macromolecular complexes (Fig. S5). 27 28 29 30 Giant modular protein complexes of Bicc1, ANKS3 and ANKS6 form through an intricately 31 interwoven network of protein interactions 32 33 The estimated size of endogenous Bicc1 foci in kidney cells and of homopolymeric Bicc1 HMW 34 35 complexes in sucrose gradients is larger than ribosomes (Rothé et al., 2015). Bicc1-ANKS3-ANKS6 36 37 complexes in cytoplasmic foci and on density fractionation gradients were even larger, with no 38 39 obvious constraints limiting their growth. What regulates the size of foci in renal tissue is unknown. 40 Both bpk mutant Bicc1 and two other polymerization mutants, Bicc1 mut-D and Bicc1ΔSAM, still 41 42 formed cytoplasmic foci with ANKS3 and ANKS6. However, compared to wild-type Bicc1 foci 43 44 scaffolded by SAM-SAM interactions, they were smaller and surrounded by abnormally diffuse Bicc1, 45 46 suggesting that complete assembly of large Bicc1-ANKS3-ANKS6 complexes requires both the 47 48 scaffolding by Bicc1 SAM polymerization and additional interactions. In good agreement, our two- 49 hybrid experiments suggest that ANKS3 and Bicc1 interconnect both via their SAM domains and by 50 51 their ANK repeats and KH domains. In addition, the SAM domain and ANK repeats of ANKS3 strongly 52 53 bound those of ANKS6, and Bicc1-KH domains also formed direct contacts with full-length Bicc1. 54 55 Thus, the assembly of Bicc1 heterooligomers clearly involves not only SAM domains. A mesh-like 56 structure encompassing SAM homo- and/or hetero-polymers linked through additional interactions 57 58 is more plausible (Fig. S5G, H). In addition, the regulation of the relative concentrations 59 60 61 62 63 14 64 65 of individual components of these heterooligomeric complexes and their post-translational 1 2 modifications or their sequestration in cilia or other compartments may also play a role. 3 4 5 Effect of a target mRNA on the size and topology of Bicc1 protein complexes 6 7 To our surprise, dispersal of large Bicc1 foci by ANKS3 into small ones was also suppressed in cells 8 9 overexpressing an mRNA that contains the Bicc1-interacting region of the AC6 3'UTR. The 10 11 simplest explanation might be that mRNA can link small Bicc1-ANKS3 foci to each other into big ones 12 (Fig. S5I). Such a model is consistent with the fact that Bicc1 and ANKS6 are linked by ANKS3 and co- 13 14 immunoprecipitate more efficiently before than after RNase A treatment (Stagner et al., 2009). The 15 16 pivot used by mRNA to enlarge Bicc1-ANKS3 complexes likely involves the Bicc1 KH domains. 17 18 Since Bicc1-KH domains can also directly contact ANKS3 and full-length Bicc1, future studies 19 20 should investigate whether such contacts affect mRNA binding or vice versa. Protein scaffolding 21 functions have been reported for long non-coding RNAs which brings molecules in close proximity to 22 23 enhance their interactions and functions in crowded cellular environments. However, to our 24 25 knowledge, no scaffolding functions have been described for mRNAs. Thus, besides being a target, 26 27 the mRNA might serve to regulate associated protein complexes. 28 29 30 Modulation of SAM polymerization by flanking sequences 31 32 Previously, we have shown that Bicc1 SAM domain self-polymerization was blocked by a frame- 33 34 shift mutation in Bicc1bpk that aberrantly elongates the C-terminus by 149 amino acids (Rothé et al., 35 36 2015). Similarly, clustering by a SAM domain is inhibited in the unrelated protein Polyhomeotic upon 37 fusion to a C-terminal GFP tag (Isono et al., 2013). Here, the deletion of the C-terminal domain that 38 39 flanks the SAM domain of ANKS3 was not sufficient to induce its self-polymerization, but 40 41 enabled it to form HMW complexes with Bicc1. According to structure models, helical SAM polymers 42 43 are expected to display N- and C-terminal flanking sequences at their periphery (for review, 44 45 (Qiao and Bowie, 2005)), suggesting that they may impose important steric constraints if their size or 46 shape forces them to clash with each other. Here, our systematic analysis of 180 human proteins 47 48 revealed that their SAM domains reside in the C-terminal quartile in 68% of all cases, 49 50 consistent with a possible link between C-terminal positioning and the ability of a given SAM 51 52 domain to polymerize. Although few SAM domains have been confirmed to polymerize in the 53 context of their full-length proteins, some form self-polymers despite their positioning close to 54 55 the N-terminus, e.g. in SLP-76 (Liu et al., 2013). Nevertheless, the distribution bias in SAM 56 57 location inside the protein sequence points to a selection mechanism favoring SAM domains close 58 59 to C-termini. 60 61 62 63 15 64 65 Among structural domains frequently associated with SAM domains in 178 proteins analyzed (Table 1 2 S1), ANK repeats are found for 18 entries (10%), whereas KH domains are only found in Bicc1. Of 3 note, in every case ANK repeats are upstream of the SAM domain and localize in the N-terminal 4 5 half of the protein, suggesting this combination is particularly suited for a molecular 6 7 scaffolding function. Indeed, while the SAM domain is a powerful domain for polymerizing and 8 9 concentrating proteins, ANK repeats are versatile platforms for the recruitment of other 10 interacting partners (for review, (Mosavi et al., 2004)). A recent example is the poly(ADP-ribose) 11 12 polymerase Tankyrase and its role in Wnt signaling, where ANK repeats serve as a docking site 13 14 for target proteins and SAM-dependent polymerization potentiates their modification by ADP- 15 16 ribose (Mariotti et al., 2016; Riccio et al., 2016). 17 18 19 Are Bicc1-ANKS3-ANKS6 scaffolds acting as modular signaling platforms? 20 21 Recent findings reported that three SAM proteins, namely Sfmbt, Scm and Ph, exhibit 22 23 different polymerization properties and interact together in a hierarchic manner to scaffold 24 25 the assembly of Polycomb (PcG) protein complexes on the chromatin and to induce 26 27 transcriptional repression (Frey et al., 2016). Whether macromolecular Bicc1-ANKS3-ANKS6 28 complexes act as modular scaffolds to similarly regulate specific signaling effectors remains to be 29 30 determined. In the cAMP/PKA signaling pathway, SAM-mediated polymerization is required for 31 32 translational repression of AC6 and PKIa mRNAs, but a role for ANKS3 or ANKS6 in regulating Bicc1- 33 34 mediated mRNA silencing has not been found. Polymerization also increases Bicc1 stability 35 36 (Rothé et al., 2015), which is important for curbing mTORC1 and canonical Wnt signaling 37 (Maisonneuve et al., 2009; Shillingford et al., 2012). However, even a polymerization-defective 38 39 mutant Bicc1 inhibits canonical Wnt signals if it is expressed at normal levels (Rothé et al., 2015). It 40 41 will be interesting to investigate in future studies whether a scaffolding function impacts the 42 43 activation of the kinase NEK8 by ANKS6 and its effect on YAP/TAZ signaling (Czarnecki et al., 2015; 44 Hoff et al., 2013), or the cytoplasmic retention of NEK7 by ANKS3 (Ramachandran et al., 2015). 45 46 47 48 49 Materials and Methods 50 51 52 Plasmids and cDNA cloning. SAM domain fusions of human Bicc1 (residues 870-939), ANKS3 53 54 (residues 421-490), and ANKS6 (residues 771-840) with negGFP or hexahistidine small ubiquitin-like 55 56 modifier (SUMO) tags were cloned as described previously (Knight et al., 2011; Leettola et al., 2014). 57 58 The Quickchange method (Agilent) was used for site-directed mutagenesis. Mammalian expression 59 60 plasmids pCMV-SPORT6::HA-Bicc1, pCS+::AC6-3’UTRprox-MS2x27 and pcDNA6::V5-ANKS6 (Hoff et 61 62 63 16 64 65 al., 2013; Rothé et al., 2015) have been described previously. ANKS3 cDNA (NM_133450), expressed 1 2 from pCMV6-Entry and fused to C-terminal Flag tag, was from OriGene (clone ID: RC223862). For 3 pCMV6-Entry::ANKS3ΔCter-Flag, the coding sequence for amino acids 1 to 490 was amplified by PCR 4 5 and subcloned between BamHI and MluI sites of pCMV6-Entry. For Y2H, Bicc1, ANKS3 and ANKS6 6 7 coding fragments were amplified by PCR and subcloned in pACT2 and pGBKT7 plasmids using BamHI- 8 9 XhoI and BamHI-SalI restriction sites, respectively. Primers used for PCR amplification are listed in 10 Supplemental Experimental Procedures. 11 12 13 14 Protein expression and purification. Details for recombinant proteins expression and purification are 15 16 given in Supplemental Experimental Procedures. 17 18 19 Crystallization and structure determination. We used multi-wavelength anomalous dispersion 20 21 (MAD) to solve a selenomethionine (SeMet)-replaced Bicc1 SAM R924E mutant to 1.75 Å resolution. 22 23 We also crystallized a native R924E mutant which had a slightly more compact unit cell and solved 24 25 this structure to 2.00 Å resolution by molecular replacement. Both the native and SeMet constructs 26 crystallized in space group P212121 with 3 molecules in the asymmetric unit. The coordinates have 27 28 been deposited in the PDB with accession codes 4RQM and 4RQN. The detailed protocol and 29 30 crystallographic statistics are given in Supplemental Experimental Procedures and Table 1. 31 32 33 34 negGFP Native Gel Binding Assays. Native gel binding assays were performed as previously 35 described (Leettola et al., 2014). See Supplemental Experimental Procedures for more details. 36 37 38 39 Analytical Size-Exclusion Chromatography and SEC-MALS. The interaction between the recombinant 40 41 Bicc1 and ANKS3 SAM domains was assessed by size-exclusion chromatography. Size Exclusion 42 Chromatography with Multi-Angle Light Scattering analyses were performed as previously described 43 44 (Leettola et al., 2014). The detailed protocol is given in Supplemental Experimental Procedures. 45 46 47 48 Cell culture and transfections. HEK293T, HeLa and Flp-In T-Rex 293 cells were cultured in DMEM 49 50 (Sigma) supplemented with 10% fetal bovine serum (FBS, Sigma), glutamine 1% (Invitrogen) and 51 gentamycine 1% (Invitrogen). Plasmids were transfected using jetPEITM (Polyplus Transfection) 52 53 according to the manufacturer instructions. 54 55 56 57 Yeast two-hybrid and co-immunoprecipitation assays. Protein co-immunoprecipitation from 58 transfected HEK293T cells, and protein-protein interaction by yeast two-hybrid assays were 59 60 61 62 63 17 64 65 performed as described (Rothé et al., 2014, 2015). See Supplemental Experimental Procedures for 1 2 more details. 3 4 5 Sucrose gradient fractionation assays. HEK293T cell extracts were fractionated on continuous 15- 6 7 60% sucrose gradients as described in (Rothé et al., 2015). See Supplemental Experimental 8 9 Procedures for more details. 10 11 12 Indirect immunofluorescence analyses. Immunostaining in HeLa cells were performed as described 13 14 in (Rothé et al., 2015). See Supplemental Experimental Procedures for more details. 15 16 17 18 19 20 Acknowledgments 21 22 We are grateful to Dr Arne Seitz (EPFL) and their technical staff for assistance with imaging. We thank 23 24 Brendan Amer for assistance with SEC-MALS. We thank Mike Collazo and Mike Sawaya at the UCLA- 25 26 DOE X-ray Crystallization and Crystallography Core Facilities, which are supported by DOE Grant DE- 27 28 FC02-02ER63421. We thank M. Capel, K. Rajashankar, N.Sukumar, J.Schuermann, I. Kourinov and F. 29 30 Murphy at NECAT beamlines 24-ID at APS, which are funded by the National Institute of General 31 Medical Sciences from the National Institutes of Health (P41 GM103403). The Pilatus 6M detector on 32 33 24-ID-C beam line is funded by a NIH-ORIP HEI grant (S10 RR029205). Use of the APS is supported by 34 35 DOE under Contract No. DE-AC02-06CH11357. This work was supported by grants from Gebert-Rüf 36 37 Stiftung (Rare Diseases grant GRS-051/13) and the Stiftung für Wissenschaftliche Forschung (Zürich) 38 to D.B.C, NIH Grant 5R01DK100482 to J.U.B, and a Ruth L. Kirschstein National Research Service 39 40 Award GM007185 to C.N.L. 41 42 43 44 45 References 46 47 Bakey, Z., Bihoreau, M.-T., Piedagnel, R., Delestré, L., Arnould, C., de Villiers, A. d’Hotman, Devuyst, 48 O., Hoffmann, S., Ronco, P., Gauguier, D., et al. (2015). The SAM domain of ANKS6 has different 49 interacting partners and mutations can induce different cystic phenotypes. Kidney Int. 88, 299–310. 50 51 52 Bienz, M. (2014). Signalosome assembly by domains undergoing dynamic head-to-tail 53 polymerization. Trends Biochem. Sci. 39, 487–495. 54 55 Brown, J.H. (2005). Missense Mutation in Sterile Motif of Novel Protein SamCystin is Associated with 56 Polycystic Kidney Disease in (cy/+) Rat. J. Am. Soc. Nephrol. 16, 3517–3526. 57 58 59 Cogswell, C., Price, S.J., Hou, X., Guay-Woodford, L.M., Flaherty, L., and Bryda, E.C. (2003). Positional 60 cloning of jcpk/bpk locus of the mouse. Mamm. Genome 14, 242–249. 61 62 63 18 64 65 Czarnecki, P.G., Gabriel, G.C., Manning, D.K., Sergeev, M., Lemke, K., Klena, N.T., Liu, X., Chen, Y., Li, 1 Y., San Agustin, J.T., et al. (2015). ANKS6 is the critical activator of NEK8 kinase in embryonic situs 2 determination and organ patterning. Nat. Commun. 6, 6023. 3 4 Delestré, L., Bakey, Z., Prado, C., Hoffmann, S., Bihoreau, M.-T., Lelongt, B., and Gauguier, D. (2015). 5 6 ANKS3 Co-Localises with ANKS6 in Mouse Renal Cilia and Is Associated with Vasopressin Signaling 7 and Apoptosis In Vivo in Mice. PloS One 10, e0136781. 8 9 Finn, R.D., Attwood, T.K., Babbitt, P.C., Bateman, A., Bork, P., Bridge, A.J., Chang, H.-Y., Dosztányi, Z., 10 El-Gebali, S., Fraser, M., et al. (2017). InterPro in 2017—beyond protein family and domain 11 annotations. Nucleic Acids Res. 45, D190–D199. 12 13 14 Frey, F., Sheahan, T., Finkl, K., Stoehr, G., Mann, M., Benda, C., and Müller, J. (2016). Molecular basis 15 of PRC1 targeting to Polycomb response elements by PhoRC. Genes Dev. 30, 1116–1127. 16 17 Hildebrandt, F., Benzing, T., and Katsanis, N. (2011). Ciliopathies. N. Engl. J. Med. 364, 1533–1543. 18 19 Hoff, S., Halbritter, J., Epting, D., Frank, V., Nguyen, T.-M.T., van Reeuwijk, J., Boehlke, C., Schell, C., 20 Yasunaga, T., Helmstädter, M., et al. (2013). ANKS6 is a central component of a nephronophthisis 21 22 module linking NEK8 to INVS and NPHP3. Nat. Genet. 45, 951–956. 23 24 Iaconis, D., Monti, M., Renda, M., van Koppen, A., Tammaro, R., Chiaravalli, M., Cozzolino, F., Pignata, 25 P., Crina, C., Pucci, P., et al. (2017). The centrosomal OFD1 protein interacts with the translation 26 machinery and regulates the synthesis of specific targets. Sci. Rep. 7. 27 28 29 Isono, K., Endo, T.A., Ku, M., Yamada, D., Suzuki, R., Sharif, J., Ishikura, T., Toyoda, T., Bernstein, B.E., 30 and Koseki, H. (2013). SAM Domain Polymerization Links Subnuclear Clustering of PRC1 to Gene 31 Silencing. Dev. Cell 26, 565–577. 32 33 Karplus, P.A., and Diederichs, K. (2012). Linking Crystallographic Model and Data Quality. Science 336, 34 1030–1033. 35 36 37 Knight, M.J., Leettola, C., Gingery, M., Li, H., and Bowie, J.U. (2011). A human sterile alpha motif 38 domain polymerizome. Protein Sci. Publ. Protein Soc. 20, 1697–1706. 39 40 Kraus, M.R.-C., Clauin, S., Pfister, Y., Di Maïo, M., Ulinski, T., Constam, D., Bellanné-Chantelot, C., and 41 Grapin-Botton, A. (2012). Two mutations in human BICC1 resulting in Wnt pathway hyperactivity 42 associated with cystic renal dysplasia. Hum. Mutat. 33, 86–90. 43 44 45 Lee, H.J., Hota, P.K., Chugha, P., Guo, H., Miao, H., Zhang, L., Kim, S.-J., Stetzik, L., Wang, B.-C., and 46 Buck, M. (2012). NMR Structure of a Heterodimeric SAM:SAM Complex: Characterization and 47 Manipulation of EphA2 Binding Reveal New Cellular Functions of SHIP2. Structure 20, 41–55. 48 49 Leettola, C.N., Knight, M.J., Cascio, D., Hoffman, S., and Bowie, J.U. (2014). Characterization of the 50 SAM domain of the PKD-related protein ANKS6 and its interaction with ANKS3. BMC Struct. Biol. 14, 51 52 17. 53 54 Liu, H., Thaker, Y.R., Stagg, L., Schneider, H., Ladbury, J.E., and Rudd, C.E. (2013). SLP-76 Sterile ? 55 Motif (SAM) and Individual H5 ? Helix Mediate Oligomer Formation for Microclusters and T-cell 56 Activation. J. Biol. Chem. 288, 29539–29549. 57 58 Maisonneuve, C., Guilleret, I., Vick, P., Weber, T., Andre, P., Beyer, T., Blum, M., and Constam, D.B. 59 60 (2009). Bicaudal C, a novel regulator of Dvl signaling abutting RNA-processing bodies, controls cilia 61 orientation and leftward flow. Development 136, 3019–3030. 62 63 19 64 65 Mariotti, L., Templeton, C.M., Ranes, M., Paracuellos, P., Cronin, N., Beuron, F., Morris, E., and 1 Guettler, S. (2016). Tankyrase Requires SAM Domain-Dependent Polymerization to Support Wnt-β- 2 Catenin Signaling. Mol. Cell 63, 498–513. 3 4 Mohieldin, A.M., Haymour, H.S., Lo, S.T., AbouAlaiwi, W.A., Atkinson, K.F., Ward, C.J., Gao, M., 5 6 Wessely, O., and Nauli, S.M. (2015). Protein composition and movements of membrane swellings 7 associated with primary cilia. Cell. Mol. Life Sci. 72, 2415–2429. 8 9 Mosavi, L.K., Cammett, T.J., Desrosiers, D.C., and Peng, Z. (2004). The ankyrin repeat as molecular 10 architecture for protein recognition. Protein Sci. 13, 1435–1448. 11 12 Nachury, M.V. (2014). How do cilia organize signalling cascades? Philos. Trans. R. Soc. B Biol. Sci. 369, 13 14 20130465–20130465. 15 16 Piazzon, N., Maisonneuve, C., Guilleret, I., Rotman, S., and Constam, D.B. (2012). Bicc1 links the 17 regulation of cAMP signaling in polycystic kidneys to microRNA-induced gene silencing. J. Mol. Cell 18 Biol. 4, 398–408. 19 20 Qiao, F., and Bowie, J.U. (2005). The Many Faces of SAM. Sci. Signal. 2005, re7-re7. 21 22 23 Ramachandran, H., Engel, C., Müller, B., Dengjel, J., Walz, G., and Yakulov, T.A. (2015). Anks3 alters 24 the sub-cellular localization of the Nek7 kinase. Biochem. Biophys. Res. Commun. 464, 901–907. 25 26 Riccio, A., McCauley, M., Langelier, M.-F., and Pascal, J. (2016). Tankyrase Sterile ? Motif Domain 27 Polymerization Is Required for Its Role in Wnt Signaling. Structure 24, 1573–1581. 28 29 30 Rothé, B., Back, R., Quinternet, M., Bizarro, J., Robert, M.-C., Blaud, M., Romier, C., Manival, X., 31 Charpentier, B., Bertrand, E., et al. (2014). Characterization of the interaction between protein 32 Snu13p/15.5K and the Rsa1p/NUFIP factor and demonstration of its functional importance for 33 snoRNP assembly. Nucleic Acids Res. 42, 2015–2036. 34 35 Rothé, B., Leal-Esteban, L., Bernet, F., Urfer, S., Doerr, N., Weimbs, T., Iwaszkiewicz, J., and Constam, 36 37 D.B. (2015). Bicc1 Polymerization Regulates the Localization and Silencing of Bound mRNA. Mol. Cell. 38 Biol. 35, 3339–3353. 39 40 Shamseldin, H.E., Yakulov, T.A., Hashem, A., Walz, G., and Alkuraya, F.S. (2016). ANKS3 is mutated in 41 a family with autosomal recessive laterality defect. Hum. Genet. 135, 1233–1239. 42 43 Shillingford, J.M., Leamon, C.P., Vlahov, I.R., and Weimbs, T. (2012). Folate-Conjugated Rapamycin 44 45 Slows Progression of Polycystic Kidney Disease. J. Am. Soc. Nephrol. 23, 1674–1681. 46 47 Stagner, E.E., Bouvrette, D.J., Cheng, J., and Bryda, E.C. (2009). The polycystic kidney disease-related 48 proteins Bicc1 and SamCystin interact. Biochem. Biophys. Res. Commun. 383, 16–21. 49 50 Taskiran, E.Z., Korkmaz, E., Gucer, S., Kosukcu, C., Kaymaz, F., Koyunlar, C., Bryda, E.C., Chaki, M., Lu, 51 D., Vadnagara, K., et al. (2014). Mutations in ANKS6 Cause a Nephronophthisis-Like Phenotype with 52 53 ESRD. J. Am. Soc. Nephrol. 25, 1653–1661. 54 55 Walz, G. (2017). Role of primary cilia in non-dividing and post-mitotic cells. Cell Tissue Res. 56 57 Yakulov, T.A., Yasunaga, T., Ramachandran, H., Engel, C., Müller, B., Hoff, S., Dengjel, J., Lienkamp, 58 S.S., and Walz, G. (2015). Anks3 interacts with nephronophthisis proteins and is required for normal 59 renal development. Kidney Int. 87, 1191–1200. 60 61 62 63 20 64 65 Table 1. Crystallographic data collection and refinement statistics 1 2 BICC1 R924E BICC1 R924E 3 SeMet Native 4 PDB Accession # 4RQM 4RQN 5 Data collection 6 APS 24-ID-C APS 24-ID-C 7 Location 8 Space group P21•2121 P212121 9 Peak High remote Inflection 10 Wavelength (Å) 0.9791 0.9714 0.9793 0.9795 11 12 Cell dimensions 13 a, b, c (Å) 46.81, 64.43, 70.64 46.26, 64.03, 70.05 46.25, 64.04, 70.07 41.68, 59.22, 68.66 14 α, β, γ (°) 90.0, 90.0, 90.0 90.0, 90.0, 90.0 90.0, 90.0, 90.0 90.0, 90.0, 90.0 15 Resolution (Å) 1.75 2.40 2.40 2.00 16 17 Rsym 0.068 (.550) 0.094 (.784) 0.096 (.866) 0.097 (0.505) 18 I/σI 10.29 (2.05) 8.13 (1.93) 7.78 (1.74) 14.94 (4.38) 19 CC1/2 99.5 (67.7) 99.3 (77.1) 99.4 (75.0) 99.6 (95.2) 20 Completeness (%) 93.0 (92.3) 97.2 (94.6) 97.3 (93.7) 98.5 (98.7) 21 22 Redundancy 2.27 (2.15) 2.78 (2.76) 2.77 (2.65) 12.65 (13.40) 23 Phasing Statistics 24 Number of sites 3 25 Mean figure of 26 merit 27 0.602/0.778 28 MAD/after density 29 modification 30 MapCC (SHELXE) 0.822 31 CC (%) 65.46 32 33 Refinement 34 Resolution (Å) 1.75 2.00 35 No. reflections 22189 11813 36 Rwork/ Rfree 0.2297/0.2672 0.2124/0.2652 37 38 No. atoms 39 Protein 1461 1397 40 Water 40 16 41 Zinc 2 1 42 2 43 B-factor (Å ) 43.0 54.1 44 R.m.s deviations 45 Bond lengths (Å) 0.008 0.008 46 Bond angles (º) 1.12 1.08 47 48 Highest resolution shell is shown in parenthesis. Rsym = ∑ |I − < I > |/∑ < I >, where I is the observed intensity 49 and is the average intensity from observations of symmetry-related reflections. CC1/2 = correlation 50 coefficient between two halves of the data (Karplus and Diederichs, 2012). Rwork = ∑ |Fobs – Fcalc|/∑Fobs, where 51 Fobs and Fcalc are the observed and calculated structure factor amplitudes, respectively. Rfree is calculated for a 52 53 set of reflections (10%) that were not included in atomic refinement. 54 55 56 57 58 59 60 61 62 63 22 64 65 FIGURE LEGENDS 1 2 3 4 Figure 1: Crystal structure of the Bicc1 SAM domain and binding to ANKS3 SAM. (A) Sequence 5 alignment of the SAM domains of human Bicc1 (residues 877-938), ANKS3 (residues 429-490), and 6 7 ANKS6 (residues 771-840). α-helices are numbered above. Residues encompassing ML and EH 8 9 surfaces are shaded in blue or red, respectively. (B) Ribbon model of Bicc1 SAM polymer crystal 10 11 structure. The helical polymer containing 6 subunits per turn is not symmetrical. Each asymmetric 12 13 unit contains 3 subunits and 2 zinc ions. The zinc ions each have half occupancy and are modeled as 14 grey spheres. A single polymer is shown in the context of the unit cell. The chain ID of each subunit is 15 16 labeled. (C) Bicc1 SAMs associate in three different orientations within the polymer. Each homodimer 17 18 has been aligned using the subunit containing the EH surface, allowing the varying angles of ML 19 20 surface interaction to be apparent. Surface area buried at each interface was calculated using the 21 PISA server. (D) Residues critical for Bicc1 SAM interaction are highlighted on the AC interface. 22 23 Surface electrostatics calculated using APBS in Pymol and contoured at ±1 kT/e show the charge 24 25 complementarity between the negatively charged ML surface and positively charged EH surface. The 26 27 dimer structure (middle) contains the R924E mutation. In the EH surface residue 924 is modeled as 28 29 the wild-type Arg to more clearly show the positive charge of this surface. (E) negGFP native gel 30 analysis of Bicc1 SAM and ANKS3 SAM alone (first two lanes) or together (third lane). Residues 31 32 colored red or blue in (D) were validated as being crucial for the ML or EH surfaces, respectively. 33 34 (F) Analytical size exclusion chromatography analysis of the Bicc1-EH/ANKS3-ML interface. Compared 35 36 to monomeric ML-mutant Bicc1 E898K and D911K and monomeric EH-mutant ANKS3 F472E alone, 37 1:1 mixtures of the two proteins elute earlier, consistent with heterodimer formation. (G) SEC-MALS 38 39 analysis of complexes of ML-mutant ANKS3 I455E with EH-mutant Bicc1 F920E or R924E, 40 41 respectively, revealed molecular weights of 17.4 ± 0.6 or 18.4 ± 0.1 kDa, corresponding in size to 42 43 heterodimers (expected MW 16.1 kDa) formed by the wild-type Bicc1-ML/ANKS3-EH interface. 44 45 Figure 2: Y2H mapping of the Bicc1-ANKS3-ANKS6 interaction network. Full-length (FL) human 46 47 Bicc1, truncated KH (residues 1 to 414), IVS (415-872) and SAM domain (873-974) fragments, and 48 49 full-length ANKS3 and ANKS6 and their respective ANK (34-220 and 8-423, respectively) and SAM 50 51 (425-488 and 773-836, respectively) domains were used as baits and preys in yeast-two-hybrid (Y2H) 52 53 assays. A growth control in non-selective medium without leucine (L) and tryptophane (T) is shown in 54 the first column (L-T-). Interactions are revealed in triple selective medium (L-T-H-) lacking histidine, 55 56 supplemented with the indicated increasing amounts of 3-aminotriazol (3-AT). (A-I) Each panel shows 57 58 one bait protein fused to the DNA-binding domain of Gal4 (Gal4-BD) and the strength of its 59 60 interactions with all candidate preys fused to Gal4 activation domain (Gal4-AD). (J) Summary of 61 62 63 23 64 65 protein interactions observed in Y2H assays. 3-AT resistance values determined for each interaction 1 2 were used to define an interaction strength scale (box) from the weakest (thin dotted black line) to 3 the strongest (thick solid red line). 4 5 6 Figure 3: Bicc1-ANKS6 association is stabilized by ANKS3. (A) HA-Bicc1 and v5-ANKS6 co- 7 8 immunoprecipitation with ANKS3-Flag in HEK293T cell extracts by anti-Flag beads. (B) ANKS3-Flag, 9 10 ANKS3ΔCter-Flag and v5-ANKS6 co-immunoprecipitation with HA-Bicc1 by anti-HA beads. Input was 11 5% of total cell extracts. γ-tubulin was used as negative control. 12 13 14 Figure 4: Density gradient fractionation reveals complexes of diverse sizes and composition. (A) 15 16 Combinations of HA-Bicc1, ANKS3-Flag, ANKS3ΔCter-Flag and v5-ANKS6 were transiently expressed in 17 18 HEK293T cells. Cell extracts were fractionated on a continuous 15 to 60% sucrose gradient and 19 analyzed by Western-blotting. The migration from the top to the bottom is indicated. γ-tubulin was 20 21 used as internal control. (B-E) Graphs showing the percentage of protein for each fraction compared 22 23 to the cumulated signal in the whole gradient. Error bars show SEMs. 24 25 26 Figure 5: Cytoplamic clustering of Bicc1 is inhibited by ANKS3 C-terminal domain and rescued by 27 ANKS6. (A) Immunfluorescent staining of individually transfected HA-Bicc1, ANKS3-Flag, ANKS3ΔCter- 28 29 Flag or v5-ANKS6. (B, C) Immunfluorescent staining of co-transfected ANKS3-Flag and v5-ANKS6 30 31 stained (B) together with anti-Flag and anti-ANKS6 antibodies or (C) only one at a time with anti-Flag 32 33 or anti-v5 antibodies. (D) Co-immunostaining of HA-Bicc1 with v5-ANKS6 or ANKS3-Flag or with (E) C- 34 terminally truncated ANKS3ΔCter-Flag. (F) Co-localization of HA-Bicc1 together with ANKS3-Flag and 35 36 v5-ANKS6. Bottom rows in (D) shows enlargement of the boxed area in the merge panel. Bars, 5 μm. 37 38 39 Figure 6: Bicc1-mediated repression of a reporter mRNA is not affected by exogenous ANKS3 and 40 41 ANKS6. (A) Silencing of AC6 3′ UTR luciferase reporter by HA-Bicc1 in absence or in presence of 42 ANKS3-Flag or v5-ANKS6 in transfected HEK293T cells. β-Galactosidase was co-transfected for signal 43 44 normalization. Data represent mean ± SEM of 3 experiments. (B) Localization of the Luc-AC6-MS2×27 45 46 reporter mRNA, HA-Bicc1 and ANKS3-Flag in transfected HeLa cells. The MS2-tagged mRNA was 47 48 imaged by nuclear MS2-YFP fusion protein which is retained in the cytoplasm when bound to the 49 50 MS2 RNA hairpins. HA-Bicc1 and ANKS3-Flag were detected by immunofluorescent staining using 51 anti-HA or anti-Flag antibodies. Bars, 5 μm. 52 53 54 Figure 7: ANKS3 and ANKS6 partially rescue the recruitment of bpk mutant Bicc1 to cytoplasmic 55 56 foci despite inhibition of its self-polymerization by an aberrently elongated C-terminus. (A) Co- 57 immunoprecipitation of ANKS3-Flag or ANKS3ΔCter-Flag and v5-ANKS6 with WT or bpk mutant HA- 58 59 Bicc1 by anti-HA antibodies. Five percent of total HEK293T cell extracts were loaded as input. γ- 60 61 62 63 24 64 65 tubulin was used as negative control. (B) Co-localization of bpk mutant HA-Bicc1 with ANKS3-Flag 1 2 transfected with or without v5-ANKS6 in HeLa cells and stained with anti-HA and anti-Flag antibodies. 3 v5-ANKS6 was unstained because anti-v5 and anti-Flag antibodies are both from mouse. Bars, 5 μm. 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 25 64 65 Figure 1 Click here to download Figure Figure 1.tif Figure 2 Click here to download Figure Figure 2_revised.tif Figure 3 Click here to download Figure Figure 3.tif Figure 4 Click here to download Figure Figure 4.tif Figure 5 Click here to download Figure Figure 5.tif Figure 6 Click here to download Figure Figure 6.tif Figure 7 Click here to download Figure Figure 7.tif Supplemental Text and Figures

Supplemental Data

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 Figure S1: Affinity of Bicc1 SAM domain homotypic and heterotypic interactions. (A) SPR analysis of the binding 41 affinity of soluble ML-mutant ANKS3 SAM I455E for immobilized EH-mutant Bicc1 SAM R924E. (B, C) SPR 42 43 analysis of the binding affinities of soluble ML mutant Bicc1 E898K (B) or Bicc1 D911K (C) for immobilized EH- 44 mutant ANKS3 F472E. (D) Raw Biacore data and relative response units of the Bicc1 SAM double mutants 45 46 E898K/F920E and D911K/F920E flowed over an ANKS3 F472E SAM-conjugated Biacore chip. ANKS3 binding to 47 BICC1 double mutants with mutations in both the ML and EH surfaces was close to background, demonstrating that 48 49 ANKS3 and Bicc1 SAM domains associate via their ML and EH surfaces. (E, F) SPR analysis of the binding affinity of 50 51 the native Bicc1 SAM interface was measured by equilibrium binding of the Bicc1 ML mutants E898K (E) or D911K 52 (F) to immobilized Bicc1 EH mutant R924E. The Bicc1 SAM homotypic interaction has an average Kd = 1.35 ± 0.1 53 54 µM. (G) Raw Biacore data and relative response units of the Bicc1 double mutants E898K/F920E and D911K/F920E 55 flowed over a Bicc1 R924E conjugated Biacore chip. Bicc1 double mutants defective in both the EH and ML surfaces 56 57 exhibit minimal binding, demonstrating that the measured binding affinity is between a native ML and EH surface and 58 not an alternate binding site. (H) negGFP native gel analysis of Bicc1 SAM and ANKS6 SAM alone (first two lanes) or 59 60 together (third lane). For SPR, error bars of triplicate measurements that were fit to a 1:1 steady-state model are smaller 61 than the data points. 62 63 64 65

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 Figure S2: CIRCOS representation of the Y2H mapping. (A) CIRCOS representations of the relative strengths of 54 55 self-interactions among specific regions within Bicc1, ANKS3 and ANKS6. (B) Details of interactions between the 56 57 indicated pairs of proteins and their specific domains tested in both ways as bait (BD) and prey (AD). (C) Summary of 58 the Bicc1-ANKS3-ANKS6 cross-interactions. The scale of interaction strengths (bottom right) was defined based on 59 60 relative resistance to increasing concentrations of 3-AT, a competitive inhibitor of the reporter gene product. 61 62 63 64 65 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37

38 39 40 Figure S3: Bicc1 polymerization is not essential to bind ANKS3 and to form cytoplasmic foci with ANKS3 and 41 42 ANKS6. (A) Graphical representation of the main amino acids involved in the Bicc1 SAM-SAM interface according 43 44 the Bicc1-SAM crystal structure shown in Figure 1. (B) ANKS3-Flag, and v5-ANKS6 co-immunoprecipitation by HA- 45 Bicc1 mut-D and ΔSAM in HEK293T cells. Five percent of total cell extracts were loaded as input. γ-tubulin was used 46 47 as negative control. (C, D) Immunostaining of HA-Bicc1 mut-D and ΔSAM together with ANKS3-Flag with or without 48 v5-ANKS6 in HeLa cells using anti-HA and anti-Flag antibodies. v5-ANKS6 was left unstained due to expected cross- 49 50 reactivity of secondary antibodies. Bars, 5 µm. 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 Figure S4: 3D rendering of Bicc1-ANKS3-mRNA scaffolds. (A) Z-projection of the Luc-AC6-MS2×27 reporter 53 54 mRNA and immunostained HA-Bicc1 and ANKS3-Flag in HeLa cells. The distribution of MS2-tagged mRNA was 55 detected by monitoring the retention of nuclear MS2-YFP fusion protein in the cytoplasm. Bar, 4 µm. (B) 3D rendering 56 57 of Bicc1-ANKS3-mRNA foci in panel A using Imaris software. Bar, 4 µm. (C) Enlarged and superimposed views of 58 ANKS3-Flag and reporter RNA (left panel), or ANKS3-Flag, reporter RNA and HA-Bicc1 (right panel) after 3D 59 60 rendering. Yellow color marking the reporter mRNA volume is transparent to not obscure the distribution of HA-Bicc1 61 62 (magenta) and ANKS3-Flag (blue) inside the foci. Bar, 4 µm. 63 64 65 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 Figure S5: Possible topologies of heterooligomeric scaffolds formed by Bicc1, ANKS3 and ANKS6. (A) Long 47 48 Bicc1 homopolymer. (B) ANKS3 homooligomer, depicted here as a dimer because of its inability to form macroscopic 49 cytoplasmic foci, although sucrose density fractionation suggests that SAM domains and ANK may interlink oligomer 50 51 repeats. (C) ANKS6 monomer. (D) ANKS3-ANKS6 complex interfaced through SAM-SAM and ANK-ANK 52 interactions. (E) Short Bicc1-ANKS3 heterooligomers capped on either side by ANKS3. (F) Model of how the deletion 53 54 of the long C-terminal domain of ANKS3 may allow Bicc1-ANKS3ΔCter co-polymers to elongate further. (G- 55 H) Models how ANKS6 co-recruitment may connect small Bicc1-ANKS3 heterooligomers into large Bicc1-ANKS3- 56 57 ANKS6 scaffolds by co-polymerization or meshing, respectively. Bicc1 still associates with ANKS3-ANKS6 58 complexes even when polymerization of its own SAM domain is blocked by specific mutations, (panel D), consistent 59 60 with KH domain binding to ANKS3 (Figs. 2B, S3B). (I) Bicc1-ANKS3 heterooligomers and Bicc1 homopolymers can 61 be scaffolded into large ribonucleoprotein complexes by a Bicc1-bound mRNA. 62 63 64 65 Table S1: Localization of SAM domains in human protein sequences. Human SAM protein sequences were retrieved from InterPro database (https://www.ebi.ac.uk/interpro/). SAM, ANK and KH domains were localized in the 1 N-terminal part (first quartile of the sequence), Middle part (second and third quartile) and C-terminal part (fourth 2 3 quartile) according the positioning of their median amino acid. The domain positioning is given in % of the total protein 4 5 length. 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 References

Abbiss, H., Maker, G.L., Gummer, J., Sharman, M.J., Phillips, J.K., Boyce, M., and Trengove, R.D. (2012). Development of a non-targeted metabolomics method to investigate urine in a rat model of polycystic kidney disease. Nephrology 17, 104– 110.

Adeva-Andany, M.M., González-Lucán, M., Donapetry-García, C., Fernández- Fernández, C., and Ameneiros-Rodríguez, E. (2016). Glycogen metabolism in humans. BBA Clin. 5, 85–100.

Aguiari, G., Varani, K., Bogo, M., Mangolini, A., Vincenzi, F., Durante, C., Gessi, S., Sacchetto, V., Catizone, L., Harris, P., et al. (2009). Deficiency of polycystic kidney disease-1 gene (PKD1) expression increases A3 adenosine receptors in human renal cells: Implications for cAMP-dependent signalling and proliferation of PKD1- mutated cystic cells. Biochim. Biophys. Acta BBA - Mol. Basis Dis. 1792, 531–540.

Alpern, R.J., and Toto, R.D. (1990). Hypokalemic Nephropathy — A Clue to Cystogenesis? N. Engl. J. Med. 322, 398–399.

Amerongen, R. van, and Nusse, R. (2009). Towards an integrated view of Wnt signaling in development. Development 136, 3205–3214.

Anderson, N.M., Mucka, P., Kern, J.G., and Feng, H. (2017). The emerging role and targetability of the TCA cycle in cancer metabolism. Protein Cell 1–22.

Arroyo, J.P., and Gamba, G. (2012). Advances in WNK Signaling of Salt and Potassium Metabolism: Clinical Implications. Am. J. Nephrol. 35, 379–386.

Baert, L. (1978). Hereditary polycystic kidney disease (adult form): a microdissection study of two cases at an early stage of the disease. Kidney Int. 13, 519–525.

Baggio, L.L., and Drucker, D.J. (2007). Biology of Incretins: GLP-1 and GIP. Gastroenterology 132, 2131–2157.

Bakey, Z., Bihoreau, M.-T., Piedagnel, R., Delestré, L., Arnould, C., de Villiers, A. d’Hotman, Devuyst, O., Hoffmann, S., Ronco, P., Gauguier, D., et al. (2015). The SAM domain of ANKS6 has different interacting partners and mutations can induce different cystic phenotypes. Kidney Int. 88, 299–310.

187 Belham, C., Roig, J., Caldwell, J.A., Aoyama, Y., Kemp, B.E., Comb, M., and Avruch, J. (2003). A Mitotic Cascade of NIMA Family Kinases: Nercc1/Nek9 ACTIVATES THE Nek6 AND Nek7 KINASES. J. Biol. Chem. 278, 34897–34909.

Belibi, F., Zafar, I., Ravichandran, K., Segvic, A.B., Jani, A., Ljubanovic, D.G., and Edelstein, C.L. (2011a). Hypoxia-inducible factor-1α (HIF-1α) and autophagy in polycystic kidney disease (PKD). Am. J. Physiol. - Ren. Physiol. 300, F1235–F1243.

Belibi, F., Ravichandran, K., Zafar, I., He, Z., and Edelstein, C.L. (2011b). mTORC1/2 and rapamycin in female Han:SPRD rats with polycystic kidney disease. Am. J. Physiol. - Ren. Physiol. 300, F236–F244.

Benton, H.P., Wong, D.M., Trauger, S.A., and Siuzdak, G. (2008). XCMS2: Processing Tandem Mass Spectrometry Data for Metabolite Identification and Structural Characterization. Anal. Chem. 80, 6382–6389.

Benzing, T., and Schermer, B. (2012). Clinical spectrum and pathogenesis of nephronophthisis: Curr. Opin. Nephrol. Hypertens. 21, 272–278.

Benzing, T., Simons, M., and Walz, G. (2007). Wnt Signaling in Polycystic Kidney Disease. J. Am. Soc. Nephrol. 18, 1389–1398.

Bergman, H., and Drury, D.R. (1938). The relationship of kidney function to the glucose utilization of the extra abdominal tissues. Am. J. Physiol. Content 124, 279– 284.

Bergmann, C. (2015). ARPKD and early manifestations of ADPKD: the original polycystic kidney disease and phenocopies. Pediatr. Nephrol. Berl. Ger. 30, 15–30.

Bergmann, C., von Bothmer, J., Ortiz Brüchle, N., Venghaus, A., Frank, V., Fehrenbach, H., Hampel, T., Pape, L., Buske, A., Jonsson, J., et al. (2011). Mutations in Multiple PKD Genes May Explain Early and Severe Polycystic Kidney Disease. J. Am. Soc. Nephrol. JASN 22, 2047–2056.

Bertran, M.T., Sdelci, S., Regué, L., Avruch, J., Caelles, C., and Roig, J. (2011). Nek9 is a Plk1-activated kinase that controls early centrosome separation through Nek6/7 and Eg5. EMBO J. 30, 2634–2647.

188 Bertrand, E., Chartrand, P., Schaefer, M., Shenoy, S.M., Singer, R.H., and Long, R.M. (1998). Localization of ASH1 mRNA Particles in Living Yeast. Mol. Cell 2, 437– 445.

Blanco, G., and Wallace, D.P. (2013). Novel role of ouabain as a cystogenic factor in autosomal dominant polycystic kidney disease. Am. J. Physiol. - Ren. Physiol. 305, F797–F812.

Boehlke, C., Kotsis, F., Patel, V., Braeg, S., Voelker, H., Bredt, S., Beyer, T., Janusch, H., Hamann, C., Gödel, M., et al. (2010). Primary cilia regulate mTORC1 activity and cell size through Lkb1. Nat. Cell Biol. 12, 1115–1122.

Boldt, K., van Reeuwijk, J., Lu, Q., Koutroumpas, K., Nguyen, T.-M.T., Texier, Y., van Beersum, S.E.C., Horn, N., Willer, J.R., Mans, D.A., et al. (2016). An organelle- specific protein landscape identifies novel diseases and molecular mechanisms. Nat. Commun. 7.

Boletta, A. (2009). Emerging evidence of a link between the polycystins and the mTOR pathways. PathoGenetics 2, 6.

Boletta, A., Qian, F., Onuchic, L.F., Bragonzi, A., Cortese, M., Deen, P.M., Courtoy, P.J., Soria, M.R., Devuyst, O., Monaco, L., et al. (2001). Biochemical characterization of bona fide polycystin-1 in vitro and in vivo. Am. J. Kidney Dis. 38, 1421–1429.

Boron, W.F., and Boulpaep, E.L. (2017). Organization of the urinary system: Functional anatomy of the kidney. In Medical Physiology, (Elsevier), pp. 722–729.

Bouvrette, D.J., Price, S.J., and Bryda, E.C. (2008). K Homology Domains of the Mouse Polycystic Kidney Disease-Related Protein, Bicaudal-C (Bicc1), Mediate RNA Binding in vitro. Nephron Exp. Nephrol. 108, e27–e34.

Brant, S.R., Yun, C.H., Donowitz, M., and Tse, C.-M. (1995). Cloning, tissue distribution, and functional analysis of the human Na+/N+ exchanger isoform, NHE3. Am. J. Physiol.-Cell Physiol. 269, C198–C206.

Braun, W.E. (2014). Advances in autosomal dominant polycystic kidney disease-- 2014 and beyond. Cleve. Clin. J. Med. 81, 545–556.

Brooks, G.A. (2010). What does glycolysis make and why is it important? J. Appl. Physiol. Bethesda Md 1985 108, 1450–1451. 189 Brown, J.H., Bihoreau, M.-T., Hoffmann, S., Kränzlin, B., Tychinskaya, I., Obermüller, N., Podlich, D., Boehn, S.N., Kaisaki, P.J., Megel, N., et al. (2005). Missense Mutation in Sterile α Motif of Novel Protein SamCystin is Associated with Polycystic Kidney Disease in (cy/+) Rat. J. Am. Soc. Nephrol. 16, 3517–3526.

Buel, G.R., Kim, S.G., and Blenis, J. (2013). mTORC1 Signaling Aids in CADalyzing Pyrimidine Biosynthesis. Cell Metab. 17, 633–635.

Burcklé, C., Gaudé, H.-M., Vesque, C., Silbermann, F., Salomon, R., Jeanpierre, C., Antignac, C., Saunier, S., and Schneider-Maunoury, S. (2011). Control of the Wnt pathways by nephrocystin-4 is required for morphogenesis of the zebrafish pronephros. Hum. Mol. Genet. 20, 2611–2627.

Bürki, R., Mohebbi, N., Bettoni, C., Wang, X., Serra, A.L., and Wagner, C.A. (2015). Impaired expression of key molecules of ammoniagenesis underlies renal acidosis in a rat model of chronic kidney disease. Nephrol. Dial. Transplant. 30, 770–781.

Bycroft, M., Bateman, A., Clarke, J., Hamill, S.J., Sandford, R., Thomas, R.L., and Chothia, C. (1999). The structure of a PKD domain from polycystin‐1: implications for polycystic kidney disease. EMBO J. 18, 297–305.

Carroll, T.J., and Das, A. (2011). Planar cell polarity in kidney development and disease. Organogenesis 7, 180–190.

Castagnetti, S. (2003). Orb and a long poly(A) tail are required for efficient oskar translation at the posterior pole of the Drosophila oocyte. Development 130, 835– 843.

Chapman, A.B., Bost, J.E., Torres, V.E., Guay-Woodford, L., Bae, K.T., Landsittel, D., Li, J., King, B.F., Martin, D., Wetzel, L.H., et al. (2012). Kidney volume and functional outcomes in autosomal dominant polycystic kidney disease. Clin. J. Am. Soc. Nephrol. CJASN 7, 479–486.

Chen, D.-Q., Cao, G., Chen, H., Liu, D., Su, W., Yu, X.-Y., Vaziri, N.D., Liu, X.-H., Bai, X., Zhang, L., et al. (2017a). Gene and protein expressions and metabolomics exhibit activated redox signaling and wnt/β-catenin pathway are associated with metabolite dysfunction in patients with chronic kidney disease. Redox Biol. 12, 505– 521.

190 Chen, S.-J., Wu, X., Wadas, B., Oh, J.-H., and Varshavsky, A. (2017b). An N-end rule pathway that recognizes proline and destroys gluconeogenic enzymes. Science 355, eaal3655.

Chen, T., Damaj, B.B., Herrera, C., Lasko, P., and Richard, S. (1997). Self- association of the single-KH-domain family members Sam68, GRP33, GLD-1, and Qk1: role of the KH domain. Mol. Cell. Biol. 17, 5707–5718.

Chiaravalli, M., Rowe, I., Mannella, V., Quilici, G., Canu, T., Bianchi, V., Gurgone, A., Antunes, S., D’Adamo, P., Esposito, A., et al. (2016). 2-Deoxy-d-Glucose Ameliorates PKD Progression. J. Am. Soc. Nephrol. 27, 1958–1969.

Chicoine, J., Benoit, P., Gamberi, C., Paliouras, M., Simonelig, M., and Lasko, P. (2007). Bicaudal-C Recruits CCR4-NOT Deadenylase to Target mRNAs and Regulates Oogenesis, Cytoskeletal Organization, and Its Own Expression. Dev. Cell 13, 691–704.

Choi, S.W., Gerencser, A.A., and Nicholls, D.G. (2009). Bioenergetic analysis of isolated cerebrocortical nerve terminals on a microgram scale: spare respiratory capacity and stochastic mitochondrial failure. J. Neurochem. 109, 1179–1191.

Choi, Y.-H., Suzuki, A., Hajarnis, S., Ma, Z., Chapin, H.C., Caplan, M.J., Pontoglio, M., Somlo, S., and Igarashi, P. (2011). Polycystin-2 and phosphodiesterase 4C are components of a ciliary A-kinase anchoring protein complex that is disrupted in cystic kidney diseases. Proc. Natl. Acad. Sci. 108, 10679–10684.

Chopra, T., Hamelin, R., Armand, F., Chiappe, D., Moniatte, M., and McKinney, J.D. (2014). Quantitative Mass Spectrometry Reveals Plasticity of Metabolic Networks in Mycobacterium smegmatis. Mol. Cell. Proteomics MCP 13, 3014–3028.

Choukroun, G., Bankir, L., and Trinh-Trang-Tan, M.M. (1994). Effect of chronic renal failure on the abundance of mRNA for enzymes of intermediary metabolism in kidney and liver. Exp. Nephrol. 2, 358–363.

Chung, S.P., Sogabe, K., Park, H.K., Song, Y., Ono, K., El-Magd, A., M, R., Shishido, Y., Yorita, K., Sakai, T., et al. (2010). Potential cytotoxic effect of hydroxypyruvate produced from d-serine by astroglial d-amino acid oxidase. J. Biochem. (Tokyo) 148, 743–753.

191 Chung, S.T., Chacko, S.K., Sunehag, A.L., and Haymond, M.W. (2015). Measurements of Gluconeogenesis and Glycogenolysis: A Methodological Review. Diabetes 64, 3996–4010.

Chung, T.K., Funk, M.A., and Baker, D.H. (1990). L-2-oxothiazolidine-4-carboxylate as a cysteine precursor: efficacy for growth and hepatic glutathione synthesis in chicks and rats. J. Nutr. 120, 158–165.

Clémençon, B., Babot, M., and Trézéguet, V. (2013). The mitochondrial ADP/ATP carrier (SLC25 family): Pathological implications of its dysfunction. Mol. Aspects Med. 34, 485–493.

Cogswell, C., Price, S.J., Hou, X., Guay-Woodford, L.M., Flaherty, L., and Bryda, E.C. (2003). Positional cloning of jcpk/bpk locus of the mouse. Mamm. Genome 14, 242–249.

Conduit, P.T., Wainman, A., and Raff, J.W. (2015). Centrosome function and assembly in animal cells. Nat. Rev. Mol. Cell Biol. 16, 611–624.

Consortium, I.P.K.D., and others (1995). Polycystic kidney disease: the complete structure of the PKD1 gene and its protein. Cell 81, 289–298.

Corbit, K.C., Shyer, A.E., Dowdle, W.E., Gaulden, J., Singla, V., and Reiter, J.F. (2008). Kif3a constrains β-catenin-dependent Wnt signalling through dual ciliary and non-ciliary mechanisms. Nat. Cell Biol. 10, 70–76.

Cornec-Le Gall, E., Audrezet, M.-P., Chen, J.-M., Hourmant, M., Morin, M.-P., Perrichot, R., Charasse, C., Whebe, B., Renaudineau, E., Jousset, P., et al. (2013). Type of PKD1 Mutation Influences Renal Outcome in ADPKD. J. Am. Soc. Nephrol. 24, 1006–1013.

Curis, E., Nicolis, I., Moinard, C., Osowska, S., Zerrouk, N., Bénazeth, S., and Cynober, L. (2005). Almost all about citrulline in mammals. Amino Acids 29, 177.

Curthoys, N.P., and Moe, O.W. (2014). Proximal Tubule Function and Response to Acidosis. Clin. J. Am. Soc. Nephrol. CJASN 9, 1627–1638.

Curthoys, N.P., Taylor, L., Hoffert, J.D., and Knepper, M.A. (2007). Proteomic analysis of the adaptive response of rat renal proximal tubules to metabolic acidosis. Am. J. Physiol. - Ren. Physiol. 292, F140–F147. 192 Czarnecki, P.G., Gabriel, G.C., Manning, D.K., Sergeev, M., Lemke, K., Klena, N.T., Liu, X., Chen, Y., Li, Y., San Agustin, J.T., et al. (2015). ANKS6 is the critical activator of NEK8 kinase in embryonic situs determination and organ patterning. Nat. Commun. 6, 6023.

Darnell, J.C., Van Driesche, S.J., Zhang, C., Hung, K.Y.S., Mele, A., Fraser, C.E., Stone, E.F., Chen, C., Fak, J.J., Chi, S.W., et al. (2011). FMRP Stalls Ribosomal Translocation on mRNAs Linked to Synaptic Function and Autism. Cell 146, 247– 261.

Davidow, C.J., Maser, R.L., Rome, L.A., Calvet, J.P., and Grantham, J.J. (1996). The cystic fibrosis transmembrane conductance regulator mediates transepithelial fluid secretion by human autosomal dominant polycystic kidney disease epithelium in vitro. Kidney Int. 50, 208–218.

De Rechter, S., Decuypere, J.-P., Ivanova, E., van den Heuvel, L.P., De Smedt, H., Levtchenko, E., and Mekahli, D. (2016). Autophagy in renal diseases. Pediatr. Nephrol. 31, 737–752.

Delestré, L., Bakey, Z., Prado, C., Hoffmann, S., Bihoreau, M.-T., Lelongt, B., and Gauguier, D. (2015). ANKS3 Co-Localises with ANKS6 in Mouse Renal Cilia and Is Associated with Vasopressin Signaling and Apoptosis In Vivo in Mice. PLoS ONE 10.

Delling, M., Indzhykulian, A.A., Liu, X., Liu, Y., Xie, T., Corey, D.P., and Clapham, D.E. (2016). Primary Cilia Are Not Calcium-Responsive Mechanosensors. Nature 531, 656–660.

Desler, C., Hansen, T.L., Frederiksen, J.B., Marcker, M.L., Singh, K.K., and Juel Rasmussen, L. (2012). Is There a Link between Mitochondrial Reserve Respiratory Capacity and Aging?

Distefano, G., Boca, M., Rowe, I., Wodarczyk, C., Ma, L., Piontek, K.B., Germino, G.G., Pandolfi, P.P., and Boletta, A. (2009). Polycystin-1 Regulates Extracellular Signal-Regulated Kinase-Dependent Phosphorylation of Tuberin To Control Cell Size through mTOR and Its Downstream Effectors S6K and 4EBP1. Mol. Cell. Biol. 29, 2359–2371.

193 Dong, C., Yuan, T., Wu, Y., Wang, Y., Fan, T.W.M., Miriyala, S., Lin, Y., Yao, J., Shi, J., Kang, T., et al. (2013). Loss of FBP1 by Snail-Mediated Repression Provides Metabolic Advantages in Basal-like Breast Cancer. Cancer Cell 23, 316–331.

Dusso, A.S. (2011). Kidney disease and vitamin D levels: 25-hydroxyvitamin D, 1,25- dihydroxyvitamin D, and VDR activation. Kidney Int. Suppl. 1, 136–141.

Düvel, K., Yecies, J.L., Menon, S., Raman, P., Lipovsky, A.I., Souza, A.L., Triantafellow, E., Ma, Q., Gorski, R., Cleaver, S., et al. (2010). Activation of a Metabolic Gene Regulatory Network Downstream of mTOR Complex 1. Mol. Cell 39, 171–183.

Eckmann, C.R., Kraemer, B., Wickens, M., and Kimble, J. (2002). GLD-3, a bicaudal- C homolog that inhibits FBF to control germline sex determination in C. elegans. Dev. Cell 3, 697–710.

Endo, T., Yamamoto, H., and Esaki, M. (2003). Functional cooperation and separation of translocators in protein import into mitochondria, the double-membrane bounded organelles. J. Cell Sci. 116, 3259–3267.

Ephrussi, A., and Lehmann, R. (1992). Induction of germ cell formation by oskar. Nature 358, 387–392.

Fantus, D., Rogers, N.M., Grahammer, F., Huber, T.B., and Thomson, A.W. (2016). Roles of mTOR complexes in the kidney: implications for renal disease and transplantation. Nat. Rev. Nephrol. 12, 587–609.

Feng, Y., Sun, F., Gao, Y., Yang, J., Wu, G., Lin, S., and Hu, J. (2017). Taurine decreased uric acid levels in hyperuricemic rats and alleviated kidney injury. Biochem. Biophys. Res. Commun. 489, 312–318.

Finetti, F., Paccani, S.R., Riparbelli, M.G., Giacomello, E., Perinetti, G., Pazour, G.J., Rosenbaum, J.L., and Baldari, C.T. (2009). Intraflagellar transport is required for polarized recycling of the TCR/CD3 complex to the immune synapse. Nat. Cell Biol. 11, 1332–1339.

Flaherty, L., Bryda, E.C., Collins, D., Rudofsky, U., and Montgomery, J.C. (1995). New mouse model for polycystic kidney disease with both recessive and dominant gene effects. Kidney Int. 47, 552–558.

194 Francis, O., Han, F., and Adams, J.C. (2013). Molecular Phylogeny of a RING E3 Ubiquitin Ligase, Conserved in Eukaryotic Cells and Dominated by Homologous Components, the Muskelin/RanBPM/CTLH Complex. PLoS ONE 8, e75217.

Fu, Y., Kim, I., Lian, P., Li, A., Zhou, L., Li, C., Liang, D., Coffey, R.J., Ma, J., and Zhao, P. (2010). Loss of Bicc1 impairs tubulomorphogenesis of cultured IMCD cells by disrupting E-cadherin-based cell-cell adhesion. Eur. J. Cell Biol. 89, 428–436.

Fukao, T., Mitchell, G., Sass, J.O., Hori, T., Orii, K., and Aoyama, Y. (2014). Ketone body metabolism and its defects. J. Inherit. Metab. Dis. 37, 541–551.

Gainullin, V.G., Hopp, K., Ward, C.J., Hommerding, C.J., and Harris, P.C. (2015). Polycystin-1 maturation requires polycystin-2 in a dose-dependent manner. J. Clin. Invest. 125, 607–620.

Gamberi, C., and Lasko, P. (2012). The Bic-C Family of Developmental Translational Regulators. Comp. Funct. Genomics 2012, 1–23.

Ganti, S., Taylor, S.L., Kim, K., Hoppel, C.L., Guo, L., Yang, J., Evans, C., and Weiss, R.H. (2012). Urinary acylcarnitines are altered in human kidney cancer. Int. J. Cancer J. Int. Cancer 130, 2791–2800.

Garcia-Gonzalo, F.R., Corbit, K.C., Sirerol-Piquer, M.S., Ramaswami, G., Otto, E.A., Noriega, T.R., Seol, A.D., Robinson, J.F., Bennett, C.L., Josifova, D.J., et al. (2011). A Transition Zone Complex Regulates Mammalian Ciliogenesis and Ciliary Membrane Composition. Nat. Genet. 43, 776–784.

Gastel, M.D.A. van, and Torres, V.E. (2017). Polycystic Kidney Disease and the Vasopressin Pathway. Ann. Nutr. Metab. 70, 43–50.

Gavin, A.-C., Maeda, K., and Kühner, S. (2011). Recent advances in charting protein–protein interaction: mass spectrometry-based approaches. Curr. Opin. Biotechnol. 22, 42–49.

Gavis, E.R., and Lehmann, R. (1992). Localization of nanos RNA controls embryonic polarity. Cell 71, 301–313.

Gerich, J.E. (2010). Role of the kidney in normal glucose homeostasis and in the hyperglycaemia of diabetes mellitus: therapeutic implications. Diabet. Med. 27, 136– 142. 195 Gerich, J.E., Meyer, C., Woerle, H.J., and Stumvoll, M. (2001). Renal Gluconeogenesis: Its importance in human glucose homeostasis. Diabetes Care 24, 382–391.

Germino, G.G. (2005). Linking cilia to Wnts. Nat. Genet. 37, 455–457.

Gherman, A., Davis, E.E., and Katsanis, N. (2006). The ciliary proteome database: an integrated community resource for the genetic and functional dissection of cilia. Nat. Genet. 38, 961–962.

Gietz, R.D., and Schiestl, R.H. (2007). Large-scale high-efficiency yeast transformation using the LiAc/SS carrier DNA/PEG method. Nat. Protoc. 2, 38–41.

Glier, M.B., Green, T.J., and Devlin, A.M. (2014). Methyl nutrients, DNA methylation, and cardiovascular disease. Mol. Nutr. Food Res. 58, 172–182.

Goggolidou, P. (2014). Wnt and planar cell polarity signaling in cystic renal disease. Organogenesis 10, 86–95.

González-Perrett, S., Kim, K., Ibarra, C., Damiano, A.E., Zotta, E., Batelli, M., Harris, P.C., Reisin, I.L., Arnaout, M.A., and Cantiello, H.F. (2001). Polycystin-2, the protein mutated in autosomal dominant polycystic kidney disease (ADPKD), is a Ca2+- permeable nonselective cation channel. Proc. Natl. Acad. Sci. 98, 1182–1187.

Gosmanov, A.R., Gosmanova, E.O., and Kovesdy, C.P. (2016). Evaluation and management of diabetic and non-diabetic hypoglycemia in end-stage renal disease. Nephrol. Dial. Transplant. 31, 8–15.

Grampa, V., Delous, M., Zaidan, M., Odye, G., Thomas, S., Elkhartoufi, N., Filhol, E., Niel, O., Silbermann, F., Lebreton, C., et al. (2016). Novel NEK8 Mutations Cause Severe Syndromic Renal Cystic Dysplasia through YAP Dysregulation. PLOS Genet. 12, e1005894.

Grantham, J.J., Cook, L.T., Wetzel, L.H., Cadnapaphornchai, M.A., and Bae, K.T. (2010). Evidence of Extraordinary Growth in the Progressive Enlargement of Renal Cysts. Clin. J. Am. Soc. Nephrol. 5, 889–896.

Greka, A., and Mundel, P. (2012). Cell Biology and Pathology of Podocytes. Annu. Rev. Physiol. 74, 299–323.

196 Guay-Woodford, L.M. (2006). Renal cystic diseases: diverse phenotypes converge on the cilium/centrosome complex. Pediatr. Nephrol. 21, 1369–1376.

Guay-Woodford, L.M., and Desmond, R.A. (2003). Autosomal Recessive Polycystic Kidney Disease: The Clinical Experience in North America. Pediatrics 111, 1072– 1080.

Guay-Woodford, L.M., Bryda, E.C., Christine, B., Lindsey, J.R., Collier, W.R., Avner, E.D., D’Eustachio, P., and Flaherty, L. (1996). Evidence that two phenotypically distinct mouse PKD mutations, bpk and jcpk, are allelic. Kidney Int. 50, 1158–1165.

Gwinn, D.M., Shackelford, D.B., Egan, D.F., Mihaylova, M.M., Mery, A., Vasquez, D.S., Turk, B.E., and Shaw, R.J. (2008). AMPK phosphorylation of raptor mediates a metabolic checkpoint. Mol. Cell 30, 214–226.

Habbig, S., Bartram, M.P., Müller, R.U., Schwarz, R., Andriopoulos, N., Chen, S., Sägmüller, J.G., Hoehne, M., Burst, V., Liebau, M.C., et al. (2011). NPHP4, a cilia- associated protein, negatively regulates the Hippo pathway. J. Cell Biol. 193, 633– 642.

Hackenthal, E., Paul, M., Ganten, D., and Taugner, R. (1990). Morphology, physiology, and molecular biology of renin secretion. Physiol. Rev. 70, 1067–1116.

Hakimi, A.A., Reznik, E., Lee, C.-H., Creighton, C.J., Brannon, A.R., Luna, A., Aksoy, B.A., Liu, E.M., Shen, R., Lee, W., et al. (2016). An Integrated Metabolic Atlas of Clear Cell Renal Cell Carcinoma. Cancer Cell 29, 104–116.

Hakimi, P., Johnson, M.T., Yang, J., Lepage, D.F., Conlon, R.A., Kalhan, S.C., Reshef, L., Tilghman, S.M., and Hanson, R.W. (2005). Phosphoenolpyruvate carboxykinase and the critical role of cataplerosis in the control of hepatic metabolism. Nutr. Metab. 2, 33.

Hallows, K.R., Raghuram, V., Kemp, B.E., Witters, L.A., and Foskett, J.K. (2000). Inhibition of cystic fibrosis transmembrane conductance regulator by novel interaction with the metabolic sensor AMP-activated protein kinase. J. Clin. Invest. 105, 1711– 1721.

Hamm, L.L., Nakhoul, N., and Hering-Smith, K.S. (2015). Acid-Base Homeostasis. Clin. J. Am. Soc. Nephrol. 10, 2232–2242.

197 Hämmerle, M., Bauer, J., Rose, M., Szallies, A., Thumm, M., Düsterhus, S., Mecke, D., Entian, K.-D., and Wolf, D.H. (1998). Proteins of Newly Isolated Mutants and the Amino-terminal Proline Are Essential for Ubiquitin-Proteasome-catalyzed Catabolite Degradation of Fructose-1,6-bisphosphatase of Saccharomyces cerevisiae. J. Biol. Chem. 273, 25000–25005.

Hanaoka, K., Qian, F., Boletta, A., Bhunia, A.K., Piontek, K., Tsiokas, L., Sukhatme, V.P., Guggino, W.B., and Germino, G.G. (2000). Co-assembly of polycystin-1 and - 2 produces unique cation-permeable currents. Nature 408, 990–994.

Harris, P.C., and Torres, V.E. (2014). Genetic mechanisms and signaling pathways in autosomal dominant polycystic kidney disease. J. Clin. Invest. 124, 2315–2324.

Hart, W., van den Hamer, C.J., and van der Sluys Veer, J. (1976). The use of hydroxy-DL-proline-2-(14)C in the investigation of hydroxyproline metabolism in normal subjects and in patients with renal insufficiency. Clin. Nephrol. 6, 379–387.

Hashemi, H.F., and Goodman, J.M. (2015). The life cycle of lipid droplets. Curr. Opin. Cell Biol. 33, 119–124.

Hatting, M., Tavares, C.D.J., Sharabi, K., Rines, A.K., and Puigserver, P. (2017). Insulin regulation of gluconeogenesis: Insulin regulation of gluconeogenesis. Ann. N. Y. Acad. Sci.

Hiesberger, T., Gourley, E., Erickson, A., Koulen, P., Ward, C.J., Masyuk, T.V., Larusso, N.F., Harris, P.C., and Igarashi, P. (2006). Proteolytic Cleavage and Nuclear Translocation of Fibrocystin Is Regulated by Intracellular Ca2+ and Activation of Protein Kinase C. J. Biol. Chem. 281, 34357–34364.

Higa, L.A., Wu, M., Ye, T., Kobayashi, R., Sun, H., and Zhang, H. (2006). CUL4- DDB1 ubiquitin ligase interacts with multiple WD40-repeat proteins and regulates histone methylation. Nat. Cell Biol. 8, 1277–1283.

Hoff, S., Halbritter, J., Epting, D., Frank, V., Nguyen, T.-M.T., van Reeuwijk, J., Boehlke, C., Schell, C., Yasunaga, T., Helmstädter, M., et al. (2013). ANKS6 is a central component of a nephronophthisis module linking NEK8 to INVS and NPHP3. Nat. Genet. 45, 951–956.

198 Hollingworth, D., Candel, A.M., Nicastro, G., Martin, S.R., Briata, P., Gherzi, R., and Ramos, A. (2012). KH domains with impaired nucleic acid binding as a tool for functional analysis. Nucleic Acids Res. 40, 6873–6886.

Hopp, K., Ward, C.J., Hommerding, C.J., Nasr, S.H., Tuan, H.-F., Gainullin, V.G., Rossetti, S., Torres, V.E., and Harris, P.C. (2012). Functional polycystin-1 dosage governs autosomal dominant polycystic kidney disease severity. J. Clin. Invest. 122, 4257–4273.

Hovater, M.B., Olteanu, D., Welty, E.A., and Schwiebert, E.M. (2008). Purinergic signaling in the lumen of a normal nephron and in remodeled PKD encapsulated cysts. Purinergic Signal. 4, 109–124.

Huan, Y., and Adelsberg, J. van (1999). Polycystin-1, the PKD1 gene product, is in a complex containing E-cadherin and the catenins. J. Clin. Invest. 104, 1459–1468.

Huang, J., and Manning, B.D. (2008). The TSC1–TSC2 complex: a molecular switchboard controlling cell growth. Biochem. J. 412, 179–190.

Huang, M., and Graves, L.M. (2003). De novo synthesis of pyrimidine nucleotides; emerging interfaces with signal transduction pathways. Cell. Mol. Life Sci. CMLS 60, 321–336.

Huang, L., Szymanska, K., Jensen, V.L., Janecke, A.R., Innes, A.M., Davis, E.E., Frosk, P., Li, C., Willer, J.R., Chodirker, B.N., et al. (2011). TMEM237 Is Mutated in Individuals with a Joubert Syndrome Related Disorder and Expands the Role of the TMEM Family at the Ciliary Transition Zone. Am. J. Hum. Genet. 89, 713–730.

Hwang, V.J., Kim, J., Rand, A., Yang, C., Sturdivant, S., Hammock, B., Bell, P.D., Guay-Woodford, L.M., and Weiss, R.H. (2015). The cpk model of recessive PKD shows glutamine dependence associated with the production of the oncometabolite 2-hydroxyglutarate. Am. J. Physiol. - Ren. Physiol. 309, F492–F498.

Iaconis, D., Monti, M., Renda, M., van Koppen, A., Tammaro, R., Chiaravalli, M., Cozzolino, F., Pignata, P., Crina, C., Pucci, P., et al. (2017). The centrosomal OFD1 protein interacts with the translation machinery and regulates the synthesis of specific targets. Sci. Rep. 7.

199 Ibraghimov-Beskrovnaya, O., and Bukanov, N. (2008). Polycystic kidney diseases: from molecular discoveries to targeted therapeutic strategies. Cell. Mol. Life Sci. CMLS 65, 605–619.

Ichida, K., Amaya, Y., Okamoto, K., and Nishino, T. (2012). Mutations Associated with Functional Disorder of Xanthine Oxidoreductase and Hereditary Xanthinuria in Humans. Int. J. Mol. Sci. 13, 15475–15495.

Inoki, K., Ouyang, H., Zhu, T., Lindvall, C., Wang, Y., Zhang, X., Yang, Q., Bennett, C., Harada, Y., Stankunas, K., et al. (2006). TSC2 Integrates Wnt and Energy Signals via a Coordinated Phosphorylation by AMPK and GSK3 to Regulate Cell Growth. Cell 126, 955–968.

Ishibashi, Y., Kohyama-Koganeya, A., and Hirabayashi, Y. (2013). New insights on glucosylated lipids: Metabolism and functions. Biochim. Biophys. Acta BBA - Mol. Cell Biol. Lipids 1831, 1475–1485.

Ishii, I., Fukushima, N., Ye, X., and Chun, J. (2004). Lysophospholipid receptors: signaling and biology. Annu. Rev. Biochem. 73, 321–354.

Ivanisevic, J., Zhu, Z.-J., Plate, L., Tautenhahn, R., Chen, S., O’Brien, P.J., Johnson, C.H., Marletta, M.A., Patti, G.J., and Siuzdak, G. (2013). Toward ‘Omic Scale Metabolite Profiling: A Dual Separation–Mass Spectrometry Approach for Coverage of Lipid and Central Carbon Metabolism. Anal. Chem. 85, 6876–6884.

Jacobs, R.L., Zhao, Y., Koonen, D.P.Y., Sletten, T., Su, B., Lingrell, S., Cao, G., Peake, D.A., Kuo, M.-S., Proctor, S.D., et al. (2010). Impaired de Novo Choline Synthesis Explains Why Phosphatidylethanolamine N -Methyltransferase-deficient Mice Are Protected from Diet-induced Obesity. J. Biol. Chem. 285, 22403–22413.

Janssens, P., Weydert, C., Rechter, S.D., Wissing, K.M., Liebau, M.C., and Mekahli, D. (2017). Expanding the role of vasopressin antagonism in polycystic kidney diseases: From adults to children? Pediatr. Nephrol. 1–14.

Jelkmann, W. (2011). Regulation of erythropoietin production: Erythropoietin production. J. Physiol. 589, 1251–1258.

200 Jensen, V.L., and Leroux, M.R. (2017). Gates for soluble and membrane proteins, and two trafficking systems (IFT and LIFT), establish a dynamic ciliary signaling compartment. Curr. Opin. Cell Biol. 47, 83–91.

Jiang, G., Yang, D., Wang, L., Zhang, X., Xu, H., Miao, Y., Wang, E., and Zhang, Y. (2015). A novel biomarker ARMc8 promotes the malignant progression of ovarian cancer. Hum. Pathol. 46, 1471–1479.

Jitrapakdee, S. (2012). Transcription factors and coactivators controlling nutrient and hormonal regulation of hepatic gluconeogenesis. Int. J. Biochem. Cell Biol. 44, 33– 45.

Jones, M.R., Rose, A.M., and Baillie, D.L. (2012). Oligoarray Comparative Genomic Hybridization-Mediated Mapping of Suppressor Mutations Generated in a Deletion- Biased Mutagenesis Screen. G3 GenesGenomesGenetics 2, 657–663.

Jones, M.R., Rose, A.M., and Baillie, D.L. (2013). The ortholog of the human proto- oncogene ROS1 is required for epithelial development in C. elegans. Genes. N. Y. N 2000 51, 545–561.

Jung, C.H., Ro, S.-H., Cao, J., Otto, N.M., and Kim, D.-H. (2010). mTOR regulation of autophagy. FEBS Lett. 584, 1287–1295.

Karbalay-Doust, S., Noorafshan, A., and Pourshahid, S.-M. (2012). Taxol and Taurine Protect the Renal Tissue of Rats after Unilateral Ureteral Obstruction: A Stereological Survey. Korean J. Urol. 53, 360.

Karet, F.E., Finberg, K.E., Nelson, R.D., Nayir, A., Mocan, H., Sanjad, S.A., Rodriguez-Soriano, J., Santos, F., Cremers, C.W.R.J., Pietro, A.D., et al. (1999). Mutations in the gene encoding B1 subunit of H+-ATPase cause renal tubular acidosis with sensorineural deafness. Nat. Genet. 21, 84–90.

Karihaloo, A., Koraishy, F., Huen, S.C., Lee, Y., Merrick, D., Caplan, M.J., Somlo, S., and Cantley, L.G. (2011). Macrophages Promote Cyst Growth in Polycystic Kidney Disease. J. Am. Soc. Nephrol. 22, 1809–1814.

Kathem, S.H., Mohieldin, A.M., and Nauli, S.M. (2014). The Roles of Primary cilia in Polycystic Kidney Disease. AIMS Mol. Sci. 1, 27–46.

201 Kaul, S., Sharma, S.S., and Mehta, I.K. (2008). Free radical scavenging potential of L-proline: evidence from in vitro assays. Amino Acids 34, 315–320.

Kim, C.A., and Bowie, J.U. (2003). SAM domains: uniform structure, diversity of function. Trends Biochem. Sci. 28, 625–628.

Kim, J., Kundu, M., Viollet, B., and Guan, K.-L. (2011). AMPK and mTOR regulate autophagy through direct phosphorylation of Ulk1. Nat. Cell Biol. 13, 132–141.

Kim, M., Kim, M., Lee, M.-S., Kim, C.-H., and Lim, D.-S. (2014). The MST1/2-SAV1 complex of the Hippo pathway promotes ciliogenesis. Nat. Commun. 5, 5370.

Kim, S., Nie, H., Nesin, V., Tran, U., Outeda, P., Bai, C.-X., Keeling, J., Maskey, D., Watnick, T., Wessely, O., et al. (2016). The polycystin complex mediates Wnt/Ca2+ signalling. Nat. Cell Biol.

Klawitter, J., Zafar, I., Klawitter, J., Pennington, A.T., Klepacki, J., Gitomer, B.Y., Schrier, R.W., Christians, U., and Edelstein, C.L. (2013). Effects of lovastatin treatment on the metabolic distributions in the Han: SPRD rat model of polycystic kidney disease. BMC Nephrol. 14, 165.

Klawitter, J., Reed-Gitomer, B.Y., McFann, K., Pennington, A., Klawitter, J., Abebe, K.Z., Klepacki, J., Cadnapaphornchai, M.A., Brosnahan, G., Chonchol, M., et al. (2014). Endothelial dysfunction and oxidative stress in polycystic kidney disease. Am. J. Physiol. - Ren. Physiol. 307, F1198–F1206.

Klein, J.D., Blount, M.A., and Sands, J.M. (2011). Urea Transport in the Kidney. In Comprehensive Physiology, R. Terjung, ed. (Hoboken, NJ, USA: John Wiley & Sons, Inc.), p.

Klionsky, D.J., Abdelmohsen, K., Abe, A., Abedin, M.J., Abeliovich, H., Acevedo Arozena, A., Adachi, H., Adams, C.M., Adams, P.D., Adeli, K., et al. (2016). Guidelines for the use and interpretation of assays for monitoring autophagy (3rd edition). Autophagy 12, 1–222.

Knight, M.J., Leettola, C., Gingery, M., Li, H., and Bowie, J.U. (2011). A human sterile alpha motif domain polymerizome. Protein Sci. Publ. Protein Soc. 20, 1697–1706.

Kobayashi, N., Yang, J., Ueda, A., Suzuki, T., Tomaru, K., Takeno, M., Okuda, K., and Ishigatsubo, Y. (2007). RanBPM, Muskelin, p48EMLP, p44CTLH, and the 202 armadillo-repeat proteins ARMC8α and ARMC8β are components of the CTLH complex. Gene 396, 236–247.

Koulen, P., Cai, Y., Geng, L., Maeda, Y., Nishimura, S., Witzgall, R., Ehrlich, B.E., and Somlo, S. (2002). Polycystin-2 is an intracellular calcium release channel. Nat. Cell Biol. 4, 191–197.

Kraus, M.R.-C., Clauin, S., Pfister, Y., Di Maïo, M., Ulinski, T., Constam, D., Bellanné- Chantelot, C., and Grapin-Botton, A. (2012). Two mutations in human BICC1 resulting in Wnt pathway hyperactivity associated with cystic renal dysplasia. Hum. Mutat. 33, 86–90.

Kugler, J.-M., Chicoine, J., and Lasko, P. (2009). Bicaudal-C Associates with a Trailer Hitch/Me31B Complex and is Required for Efficient Gurken Secretion. Dev. Biol. 328, 160–172.

Kühl, M., Geis, K., Sheldahl, L.C., Pukrop, T., Moon, R.T., and Wedlich, D. (2001). Antagonistic regulation of convergent extension movements in Xenopus by Wnt/beta- catenin and Wnt/Ca2+ signaling. Mech. Dev. 106, 61–76.

Lal, M., Song, X., Pluznick, J.L., Di Giovanni, V., Merrick, D.M., Rosenblum, N.D., Chauvet, V., Gottardi, C.J., Pei, Y., and Caplan, M.J. (2008). Polycystin-1 C-terminal tail associates with β-catenin and inhibits canonical Wnt signaling. Hum. Mol. Genet. 17, 3105–3117.

Latchoumycandane, C., Nagy, L.E., and McIntyre, T.M. (2014). Chronic ethanol ingestion induces oxidative kidney injury through taurine-inhibitable inflammation. Free Radic. Biol. Med. 69, 403–416.

Leettola, C.N., Knight, M.J., Cascio, D., Hoffman, S., and Bowie, J.U. (2014). Characterization of the SAM domain of the PKD-related protein ANKS6 and its interaction with ANKS3. BMC Struct. Biol. 14, 17.

Lemaire, L.A., Goulley, J., Kim, Y.H., Carat, S., Jacquemin, P., Rougemont, J., Constam, D.B., and Grapin-Botton, A. (2015). Bicaudal C1 promotes pancreatic NEUROG3+ endocrine progenitor differentiation and ductal morphogenesis. Development 142, 858–870.

203 Leone, A., Moncada, S., Vallance, P., Calver, A., and Collier, J. (1992). Accumulation of an endogenous inhibitor of nitric oxide synthesis in chronic renal failure. The Lancet 339, 572–575.

Li, B., Qiu, B., Lee, D.S.M., Walton, Z.E., Ochocki, J.D., Mathew, L.K., Mancuso, A., Gade, T.P.F., Keith, B., Nissim, I., et al. (2014). Fructose-1,6-bisphosphatase opposes renal carcinoma progression. Nature 513, 251–255.

Li, M., Wang, X., Meintzer, M.K., Laessig, T., Birnbaum, M.J., and Heidenreich, K.A. (2000). Cyclic AMP Promotes Neuronal Survival by Phosphorylation of Glycogen Synthase Kinase 3β. Mol. Cell. Biol. 20, 9356–9363.

Li, Q., Montalbetti, N., Wu, Y., Ramos, A., Raychowdhury, M.K., Chen, X.-Z., and Cantiello, H.F. (2006). Polycystin-2 Cation Channel Function Is under the Control of Microtubular Structures in Primary Cilia of Renal Epithelial Cells. J. Biol. Chem. 281, 37566–37575.

Lian, P., Li, A., Li, Y., Liu, H., Liang, D., Hu, B., Lin, D., Jiang, T., Moeckel, G., Qin, D., et al. (2014). Loss of Polycystin-1 Inhibits Bicc1 Expression during Mouse Development. PLoS ONE 9, e88816.

Lienkamp, S., Ganner, A., Boehlke, C., Schmidt, T., Arnold, S.J., Schäfer, T., Romaker, D., Schuler, J., Hoff, S., Powelske, C., et al. (2010). Inversin relays Frizzled-8 signals to promote proximal pronephros development. Proc. Natl. Acad. Sci. U. S. A. 107, 20388–20393.

Liu, K., and Czaja, M.J. (2013). Regulation of lipid stores and metabolism by lipophagy. Cell Death Differ. 20, 3–11.

Liu, S., Lu, W., Obara, T., Kuida, S., Lehoczky, J., Dewar, K., Drummond, I.A., and Beier, D.R. (2002). A defect in a novel Nek-family kinase causes cystic kidney disease in the mouse and in zebrafish. Development 129, 5839–5846.

Löffler, M., Carrey, E.A., and Zameitat, E. (2016). Orotate (orotic acid): An essential and versatile molecule. Nucleosides Nucleotides Nucleic Acids 35, 566–577.

López, M., Tovar, S., Vázquez, M.J., Nogueiras, R., Señarís, R., and Diéguez, C. (2005). Sensing the fat: Fatty acid metabolism in the hypothalamus and the melanocortin system. Peptides 26, 1753–1758.

204 Low, S.H., Vasanth, S., Larson, C.H., Mukherjee, S., Sharma, N., Kinter, M.T., Kane, M.E., Obara, T., and Weimbs, T. (2006). Polycystin-1, STAT6, and P100 Function in a Pathway that Transduces Ciliary Mechanosensation and Is Activated in Polycystic Kidney Disease. Dev. Cell 10, 57–69.

Lu, H., Galeano, M.C.R., Ott, E., Kaeslin, G., Kausalya, P.J., Kramer, C., Ortiz- Brüchle, N., Hilger, N., Metzis, V., Hiersche, M., et al. (2017). Mutations in DZIP1L, which encodes a ciliary-transition-zone protein, cause autosomal recessive polycystic kidney disease. Nat. Genet. 49, 1025–1034.

Lu, W., Peissel, B., Babakhanlou, H., Pavlova, A., Geng, L., Fan, X., Larson, C., Brent, G., and Zhou, J. (1997). Perinatal lethality with kidney and pancreas defects in mice with a targetted Pkd1 mutation. Nat. Genet. 17, 179–181.

Lyons, J., Rauh-Pfeiffer, A., Yu, Y.M., Lu, X.-M., Zurakowski, D., Tompkins, R.G., Ajami, A.M., Young, V.R., and Castillo, L. (2000). Blood glutathione synthesis rates in healthy adults receiving a sulfur amino acid-free diet. Proc. Natl. Acad. Sci. 97, 5071–5076.

MacDonald, B.T., Tamai, K., and He, X. (2009). Wnt/β-catenin signaling: components, mechanisms, and diseases. Dev. Cell 17, 9–26.

Magistroni, R., and Boletta, A. (2017). Defective glycolysis and the use of 2-deoxy- d-glucose in polycystic kidney disease: from animal models to humans. J. Nephrol. 1–9.

Mahone, M., Saffman, E.E., and Lasko, P.F. (1995). Localized Bicaudal-C RNA encodes a protein containing a KH domain, the RNA binding motif of FMR1. EMBO J. 14, 2043.

Maisonneuve, C., Guilleret, I., Vick, P., Weber, T., Andre, P., Beyer, T., Blum, M., and Constam, D.B. (2009). Bicaudal C, a novel regulator of Dvl signaling abutting RNA-processing bodies, controls cilia orientation and leftward flow. Development 136, 3019–3030.

Maiuolo, J., Oppedisano, F., Gratteri, S., Muscoli, C., and Mollace, V. (2016). Regulation of uric acid metabolism and excretion. Int. J. Cardiol. 213, 8–14.

205 Markowitz, G.S., Cai, Y., Li, L., Wu, G., Ward, L.C., Somlo, S., and D’Agati, V.D. (1999). Polycystin-2 expression is developmentally regulated. Am. J. Physiol. - Ren. Physiol. 277, F17–F25.

Mashima, T., Seimiya, H., and Tsuruo, T. (2009). De novo fatty-acid synthesis and related pathways as molecular targets for cancer therapy. Br. J. Cancer 100, 1369– 1372.

Massinen, S., Hokkanen, M.-E., Matsson, H., Tammimies, K., Tapia-Páez, I., Dahlström-Heuser, V., Kuja-Panula, J., Burghoorn, J., Jeppsson, K.E., Swoboda, P., et al. (2011). Increased Expression of the Dyslexia Candidate Gene DCDC2 Affects Length and Signaling of Primary Cilia in Neurons. PLoS ONE 6.

Mather, A., and Pollock, C. (2011). Glucose handling by the kidney. Kidney Int. 79, S1–S6.

McCoin, C.S., Knotts, T.A., and Adams, S.H. (2015). Acylcarnitines[mdash]old actors auditioning for new roles in metabolic physiology. Nat. Rev. Endocrinol. 11, 617–625.

McMahon, A.P. (2016). Development of the Mammalian Kidney. Curr. Top. Dev. Biol. 117, 31–64.

Mehta, L., and Jim, B. (2017). Hereditary Renal Diseases. Semin. Nephrol. 37, 354– 361.

Menezes, L.F., Zhou, F., Patterson, A.D., Piontek, K.B., Krausz, K.W., Gonzalez, F.J., and Germino, G.G. (2012). Network Analysis of a Pkd1-Mouse Model of Autosomal Dominant Polycystic Kidney Disease Identifies HNF4α as a Disease Modifier. PLoS Genet. 8.

Menezes, L.F., Lin, C.-C., Zhou, F., and Germino, G.G. (2016). Fatty Acid Oxidation is Impaired in An Orthologous Mouse Model of Autosomal Dominant Polycystic Kidney Disease. EBioMedicine 5, 183–192.

Menon, R.P., Gibson, T.J., and Pastore, A. (2004). The C terminus of fragile X mental retardation protein interacts with the multi-domain Ran-binding protein in the microtubule-organising centre. J. Mol. Biol. 343, 43–53.

Menssen, R., Schweiggert, J., Schreiner, J., Kušević, D., Reuther, J., Braun, B., and Wolf, D.H. (2012). Exploring the Topology of the Gid Complex, the E3 Ubiquitin 206 Ligase Involved in Catabolite-induced Degradation of Gluconeogenic Enzymes. J. Biol. Chem. 287, 25602–25614.

Mesner, L.D., Ray, B., Hsu, Y.-H., Manichaikul, A., Lum, E., Bryda, E.C., Rich, S.S., Rosen, C.J., Criqui, M.H., Allison, M., et al. (2014). Bicc1 is a genetic determinant of osteoblastogenesis and bone mineral density. J. Clin. Invest. 124, 2736–2749.

Meyer, C., Dostou, J.M., Welle, S.L., and Gerich, J.E. (2002). Role of human liver, kidney, and skeletal muscle in postprandial glucose homeostasis. Am. J. Physiol. - Endocrinol. Metab. 282, E419–E427.

Mi, H., Huang, X., Muruganujan, A., Tang, H., Mills, C., Kang, D., and Thomas, P.D. (2017). PANTHER version 11: expanded annotation data from Gene Ontology and Reactome pathways, and data analysis tool enhancements. Nucleic Acids Res. 45, D183–D189.

Miyazaki, T., Honda, A., Ikegami, T., Iwamoto, J., Monma, T., Hirayama, T., Saito, Y., Yamashita, K., and Matsuzaki, Y. (2015). Simultaneous quantification of salivary 3-hydroxybutyrate, 3-hydroxyisobutyrate, 3-hydroxy-3-methylbutyrate, and 2- hydroxybutyrate as possible markers of amino acid and fatty acid catabolic pathways by LC–ESI–MS/MS. SpringerPlus 4.

Mohieldin, A.M., Haymour, H.S., Lo, S.T., AbouAlaiwi, W.A., Atkinson, K.F., Ward, C.J., Gao, M., Wessely, O., and Nauli, S.M. (2015). Protein composition and movements of membrane swellings associated with primary cilia. Cell. Mol. Life Sci. CMLS 72, 2415–2429.

Mohler, J., and Wieschaus, E.F. (1985). Dominant maternal-effect mutations of Drosophila melanogaster causing the production of double-abdomen embryos. Genetics 112, 803–822.

Morales, J., and Schneider, D. (2014). Hypoglycemia. Am. J. Med. 127, S17-24.

Morales, Y., Cáceres, T., May, K., and Hevel, J.M. (2016). Biochemistry and regulation of the protein arginine methyltransferases (PRMTs). Arch. Biochem. Biophys. 590, 138–152.

Moser, M., Matthiesen, S., Kirfel, J., Schorle, H., Bergmann, C., Senderek, J., Rudnik-Schöneborn, S., Zerres, K., and Buettner, R. (2005). A mouse model for

207 cystic biliary dysgenesis in autosomal recessive polycystic kidney disease (ARPKD). Hepatology 41, 1113–1121.

Mount, D.B. (2014). Thick Ascending Limb of the Loop of Henle. Clin. J. Am. Soc. Nephrol. CJASN 9, 1974–1986.

Muff, M.A., Masyuk, T.V., Stroope, A.J., Huang, B.Q., Splinter, P.L., Lee, S.-O., and LaRusso, N.F. (2006). Development and characterization of a cholangiocyte cell line from the PCK rat, an animal model of Autosomal Recessive Polycystic Kidney Disease. Lab. Invest. 86, 940–950.

Mullins, L.J., Conway, B.R., Menzies, R.I., Denby, L., and Mullins, J.J. (2016). Renal disease pathophysiology and treatment: contributions from the rat. Dis. Model. Mech. 9, 1419–1433.

Murray, R.K., Granner, D.K., Mayes, P.A., and Rodwell, V.W. (2003). Harper’s Illustrated Biochemistry (McGraw-Hill).

Nauli, S.M., Alenghat, F.J., Luo, Y., Williams, E., Vassilev, P., Li, X., Elia, A.E.H., Lu, W., Brown, E.M., Quinn, S.J., et al. (2003). Polycystins 1 and 2 mediate mechanosensation in the primary cilium of kidney cells. Nat. Genet. 33, 129–137.

Nauta, J., Ozawa, Y., Sweeney, W.E., Rutledge, J.C., and Avner, E.D. (1993). Renal and biliary abnormalities in a new murine model of autosomal recessive polycystic kidney disease. Pediatr. Nephrol. 7, 163–172.

Nicastro, G., Taylor, I.A., and Ramos, A. (2015). KH–RNA interactions: back in the groove. Curr. Opin. Struct. Biol. 30, 63–70.

Nicholls, D.G. (2009). Spare respiratory capacity, oxidative stress and excitotoxicity. Biochem. Soc. Trans. 37, 1385–1388.

Oda, M., Satta, Y., Takenaka, O., and Takahata, N. (2002). Loss of urate oxidase activity in hominoids and its evolutionary implications. Mol. Biol. Evol. 19, 640–653.

Omran, H. (2010). NPHP proteins: gatekeepers of the ciliary compartment. J. Cell Biol. 190, 715–717.

O’Neill, L.A.J., Kishton, R.J., and Rathmell, J. (2016). A guide to immunometabolism for immunologists. Nat. Rev. Immunol. 16, 553–565.

208 Ong, A.C.M., Devuyst, O., Knebelmann, B., and Walz, G. (2015). Autosomal dominant polycystic kidney disease: the changing face of clinical management. The Lancet 385, 1993–2002.

Oud, M., Lamers, I., and Arts, H. (2016). Ciliopathies: Genetics in Pediatric Medicine. J. Pediatr. Genet. 06, 018–029.

Outeda, P., Menezes, L., Hartung, E.A., Bridges, S., Zhou, F., Zhu, X., Xu, H., Huang, Q., Yao, Q., Qian, F., et al. (2017). A novel model of autosomal recessive polycystic kidney questions the role of the fibrocystin C-terminus in disease mechanism. Kidney Int.

Pagliara, A.S., and Goodman, A.D. (1970). Relation of renal cortical gluconeogenesis, glutamate content, and production of ammonia. J. Clin. Invest. 49, 1967–1974.

Palsson, R., Sharma, C.P., Kim, K., McLaughlin, M., Brown, D., and Arnaout, M.A. (1996). Characterization and cell distribution of polycystin, the product of autosomal dominant polycystic kidney disease gene 1. Mol. Med. 2, 702–711.

Pan, J., and Snell, W. (2007). The Primary Cilium: Keeper of the Key to Cell Division. Cell 129, 1255–1257.

PD, W., L, G., X, L., and Cr, B. (1999). The PKD1 gene product, “polycystin-1,” is a tyrosine-phosphorylated protein that colocalizes with alpha2beta1-integrin in focal clusters in adherent renal epithelia. Lab. Investig. J. Tech. Methods Pathol. 79, 1311– 1323.

Pema, M., Drusian, L., Chiaravalli, M., Castelli, M., Yao, Q., Ricciardi, S., Somlo, S., Qian, F., Biffo, S., and Boletta, A. (2016). mTORC1-mediated inhibition of polycystin- 1 expression drives renal cyst formation in tuberous sclerosis complex. Nat. Commun. 7, 10786.

Peters, D.J.M., van de Wal, A., Spruit, L., Saris, J.J., Breuning, M.H., Bruijn, J.A., and de Heer, E. (1999). Cellular localization and tissue distribution of polycystin-1. J. Pathol. 188, 439–446.

209 Peterson, J.W., Boldogh, I., Popov, V.L., Saini, S.S., and Chopra, A.K. (1998). Anti- inflammatory and antisecretory potential of histidine in Salmonella-challenged mouse small intestine. Lab. Investig. J. Tech. Methods Pathol. 78, 523–534.

Piazzon, N., Maisonneuve, C., Guilleret, I., Rotman, S., and Constam, D.B. (2012). Bicc1 links the regulation of cAMP signaling in polycystic kidneys to microRNA- induced gene silencing. J. Mol. Cell Biol. 4, 398–408.

Piontek, K., Menezes, L.F., Garcia-Gonzalez, M.A., Huso, D.L., and Germino, G.G. (2007). A critical developmental switch defines the kinetics of kidney cyst formation after loss of Pkd1. Nat. Med. 13, 1490–1495.

Pitre, S., Dehne, F., Chan, A., Cheetham, J., Duong, A., Emili, A., Gebbia, M., Greenblatt, J., Jessulat, M., Krogan, N., et al. (2006). PIPE: a protein-protein interaction prediction engine based on the re-occurring short polypeptide sequences between known interacting protein pairs. BMC Bioinformatics 7, 365.

Plotnikova, O.V., Nikonova, A.S., Loskutov, Y.V., Kozyulina, P.Y., Pugacheva, E.N., and Golemis, E.A. (2012). Calmodulin activation of Aurora-A kinase (AURKA) is required during ciliary disassembly and in mitosis. Mol. Biol. Cell 23, 2658–2670.

Porcheray, F., Léone, C., Samah, B., Rimaniol, A.-C., Dereuddre-Bosquet, N., and Gras, G. (2006). Glutamate metabolism in HIV-infected macrophages: implications for the CNS. Am. J. Physiol. - Cell Physiol. 291, C618–C626.

Praetorius, H.A., and Spring, K.R. (2001). Bending the MDCK cell primary cilium increases intracellular calcium. J. Membr. Biol. 184, 71–79.

Promeneur, D., Kwon, T.-H., Frøkiær, J., Knepper, M.A., and Nielsen, S. (2000). Vasopressin V2-receptor-dependent regulation of AQP2 expression in Brattleboro rats. Am. J. Physiol. - Ren. Physiol. 279, F370–F382.

Pugacheva, E.N., Jablonski, S.A., Hartman, T.R., Henske, E.P., and Golemis, E.A. (2007). HEF1-dependent Aurora A activation induces disassembly of the primary cilium. Cell 129, 1351–1363.

Qian, C.-N., Knol, J., Igarashi, P., Lin, F., Zylstra, U., Teh, B.T., and Williams, B.O. (2005). Cystic Renal Neoplasia Following Conditional Inactivation of Apc in Mouse Renal Tubular Epithelium. J. Biol. Chem. 280, 3938–3945.

210 Qian, F., Watnick, T.J., Onuchic, L.F., and Germino, G.G. (1996). The molecular basis of focal cyst formation in human autosomal dominant polycystic kidney disease type I. Cell 87, 979–987.

Qian, F., Germino, F.J., Cai, Y., Zhang, X., Somlo, S., and Germino, G.G. (1997). PKD1 interacts with PKD2 through a probable coiled-coil domain. Nat. Genet. 16, 179–183.

Qian, F., Boletta, A., Bhunia, A.K., Xu, H., Liu, L., Ahrabi, A.K., Watnick, T.J., Zhou, F., and Germino, G.G. (2002). Cleavage of polycystin-1 requires the receptor for egg jelly domain and is disrupted by human autosomal-dominant polycystic kidney disease 1-associated mutations. Proc. Natl. Acad. Sci. U. S. A. 99, 16981–16986.

Qiao, F., and Bowie, J.U. (2005). The Many Faces of SAM. Sci STKE 2005, re7-re7.

Quamme, G.A., and Freeman, H.J. (1987). Evidence for a high-affinity sodium- dependent D-glucose transport system in the kidney. Am. J. Physiol.-Ren. Physiol. 253, F151–F157.

Quinn, P.G., and Yeagley, D. (2005). Insulin regulation of PEPCK gene expression: a model for rapid and reversible modulation. Curr. Drug Targets Immune Endocr. Metab. Disord. 5, 423–437.

Ramachandran, H., Engel, C., Müller, B., Dengjel, J., Walz, G., and Yakulov, T.A. (2015). Anks3 alters the sub-cellular localization of the Nek7 kinase. Biochem. Biophys. Res. Commun. 464, 901–907.

Ran, F.A., Hsu, P.D., Wright, J., Agarwala, V., Scott, D.A., and Zhang, F. (2013). Genome engineering using the CRISPR-Cas9 system. Nat. Protoc. 8, 2281–2308.

Ran, J., Yang, Y., Li, D., Liu, M., and Zhou, J. (2015). Deacetylation of ?-tubulin and cortactin is required for HDAC6 to trigger ciliary disassembly. Sci. Rep. 5.

Ravichandran, K., and Edelstein, C.L. (2014). Polycystic Kidney Disease: A Case of Suppressed Autophagy? Semin. Nephrol. 34, 27–33.

Regelmann, J., Schüle, T., Josupeit, F.S., Horak, J., Rose, M., Entian, K.-D., Thumm, M., and Wolf, D.H. (2003). Catabolite degradation of fructose-1, 6-bisphosphatase in the yeast Saccharomyces cerevisiae: a genome-wide screen identifies eight novel

211 GID genes and indicates the existence of two degradation pathways. Mol. Biol. Cell 14, 1652–1663.

Rhee, E.P. (2015). Metabolomics and Renal Disease. Curr. Opin. Nephrol. Hypertens. 24, 371–379.

Riwanto, M., Kapoor, S., Rodriguez, D., Edenhofer, I., Segerer, S., and Wüthrich, R.P. (2016). Inhibition of Aerobic Glycolysis Attenuates Disease Progression in Polycystic Kidney Disease. PloS One 11, e0146654.

Romero, C.A., Remor, A., Latini, A., De Paul, A.L., Torres, A.I., and Mukdsi, J.H. (2017). Uric acid activates NRLP3 inflammasome in an in-vivo model of epithelial to mesenchymal transition in the kidney. J. Mol. Histol. 48, 209–218.

Rosenbaum, J.L., and Witman, G.B. (2002). Intraflagellar transport. Nat. Rev. Mol. Cell Biol. 3, 813–825.

Rossetti, S., Kubly, V.J., Consugar, M.B., Hopp, K., Roy, S., Horsley, S.W., Chauveau, D., Rees, L., Barratt, T.M., Hoff, W.G. van’t, et al. (2009). Incompletely penetrant PKD1 alleles suggest a role for gene dosage in cyst initiation in polycystic kidney disease. Kidney Int. 75, 848.

Rowe, I., and Boletta, A. (2014). Defective metabolism in polycystic kidney disease: potential for therapy and open questions. Nephrol. Dial. Transplant. 29, 1480–1486.

Rowe, I., Chiaravalli, M., Mannella, V., Ulisse, V., Quilici, G., Pema, M., Song, X.W., Xu, H., Mari, S., Qian, F., et al. (2013). Defective Glucose Metabolism in Polycystic Kidney Disease Identifies A Novel Therapeutic Paradigm. Nat. Med. 19, 488–493.

Roy, A., Al-bataineh, M.M., and Pastor-Soler, N.M. (2015). Collecting Duct Intercalated Cell Function and Regulation. Clin. J. Am. Soc. Nephrol. CJASN 10, 305–324.

Rysz, J., Gluba-Brzózka, A., Franczyk, B., Jabłonowski, Z., and Ciałkowska-Rysz, A. (2017). Novel Biomarkers in the Diagnosis of Chronic Kidney Disease and the Prediction of Its Outcome. Int. J. Mol. Sci. 18, 1702.

Saburi, S., Hester, I., Fischer, E., Pontoglio, M., Eremina, V., Gessler, M., Quaggin, S.E., Harrison, R., Mount, R., and McNeill, H. (2008). Loss of Fat4 disrupts PCP

212 signaling and oriented cell division and leads to cystic kidney disease. Nat. Genet. 40, 1010–1015.

Saffman, E.E., Styhler, S., Rother, K., Li, W., Richard, S., and Lasko, P. (1998). Premature translation of oskar in oocytes lacking the RNA-binding protein bicaudal- C. Mol. Cell. Biol. 18, 4855–4862.

Salemi, L.M., Loureiro, S.O., and Schild-Poulter, C. (2015). Characterization of RanBPM Molecular Determinants that Control Its Subcellular Localization. PLOS ONE 10, e0117655.

Salemi, L.M., Maitland, M.E.R., McTavish, C.J., and Schild-Poulter, C. (2017). Cell signalling pathway regulation by RanBPM: molecular insights and disease implications. Open Biol. 7.

Salomon, R., Saunier, S., and Niaudet, P. (2009). Nephronophthisis. Pediatr. Nephrol. Berl. Ger. 24, 2333–2344.

Sánchez, I., and Dynlacht, B.D. (2016). Cilium assembly and disassembly. Nat. Cell Biol. 18, 711–717.

Sang, L., Miller, J.J., Corbit, K.C., Giles, R.H., Brauer, M.J., Otto, E.A., Baye, L.M., Wen, X., Scales, S.J., Kwong, M., et al. (2011). Mapping the NPHP-JBTS-MKS Protein Network Reveals Ciliopathy Disease Genes and Pathways. Cell 145, 513– 528.

Santt, O., Pfirrmann, T., Braun, B., Juretschke, J., Kimmig, P., Scheel, H., Hofmann, K., Thumm, M., and Wolf, D.H. (2008). The Yeast GID Complex, a Novel Ubiquitin Ligase (E3) Involved in the Regulation of Carbohydrate Metabolism. Mol. Biol. Cell 19, 3323–3333.

Satir, P., Pedersen, L.B., and Christensen, S.T. (2010). The primary cilium at a glance. J. Cell Sci. 123, 499–503.

Schwartz, E.A., Leonard, M.L., Bizios, R., and Bowser, S.S. (1997). Analysis and modeling of the primary cilium bending response to fluid shear. Am. J. Physiol. - Ren. Physiol. 272, F132–F138.

213 Scott, J., Duongh, M., Young, A., Subbarao, P., Gauvreau, G., and Grasemann, H. (2014). Asymmetric Dimethylarginine in Chronic Obstructive Pulmonary Disease (ADMA in COPD). Int. J. Mol. Sci. 15, 6062–6071.

Sdelci, S., Bertran, M.T., and Roig, J. (2011). Nek9, Nek6, Nek7 and the separation of centrosomes. Cell Cycle 10, 3816–3817.

Semakova, J., Hyroššová, P., Méndez-Lucas, A., Cutz, E., Bermudez, J., Burgess, S., Alcántara, S., and Perales, J.C. (2017). PEPCK-C reexpression in the liver counters neonatal hypoglycemia in Pck1del/del mice, unmasking role in non- gluconeogenic tissues. J. Physiol. Biochem. 73, 89–98.

Semmo, M., Köttgen, M., and Hofherr, A. (2014). The TRPP Subfamily and Polycystin-1 Proteins. In Mammalian Transient Receptor Potential (TRP) Cation Channels, (Springer, Berlin, Heidelberg), pp. 675–711.

Seok, S., Fu, T., Choi, S.-E., Li, Y., Zhu, R., Kumar, S., Sun, X., Yoon, G., Kang, Y., Zhong, W., et al. (2014). Transcriptional regulation of autophagy by an FXR-CREB axis. Nature 516, 108–111.

Shamseldin, H.E., Yakulov, T.A., Hashem, A., Walz, G., and Alkuraya, F.S. (2016). ANKS3 is mutated in a family with autosomal recessive laterality defect. Hum. Genet. 135, 1233–1239.

She, P., Shiota, M., Shelton, K.D., Chalkley, R., Postic, C., and Magnuson, M.A. (2000). Phosphoenolpyruvate carboxykinase is necessary for the integration of hepatic energy metabolism. Mol. Cell. Biol. 20, 6508–6517.

Shi, H., Wang, Y., Li, X., Zhan, X., Tang, M., Fina, M., Su, L., Pratt, D., Bu, C.H., Hildebrand, S., et al. (2015). NLRP3 activation and mitosis are mutually exclusive events coordinated by NEK7, a new inflammasome component. Nat. Immunol. 17, 250–258.

Shiba, D., Manning, D.K., Koga, H., Beier, D.R., and Yokoyama, T. (2010). Inv acts as a molecular anchor for Nphp3 and Nek8 in the proximal segment of primary cilia. Cytoskeleton 67, 112–119.

Shibazaki, S., Yu, Z., Nishio, S., Tian, X., Thomson, R.B., Mitobe, M., Louvi, A., Velazquez, H., Ishibe, S., Cantley, L.G., et al. (2008). Cyst formation and activation

214 of the extracellular regulated kinase pathway after kidney specific inactivation of Pkd1. Hum. Mol. Genet. 17, 1505–1516.

Shillingford, J.M., Murcia, N.S., Larson, C.H., Low, S.H., Hedgepeth, R., Brown, N., Flask, C.A., Novick, A.C., Goldfarb, D.A., Kramer-Zucker, A., 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, 5466– 5471.

Simons, M., Gloy, J., Ganner, A., Bullerkotte, A., Bashkurov, M., Krönig, C., Schermer, B., Benzing, T., Cabello, O.A., Jenny, A., et al. (2005). Inversin, the gene product mutated in nephronophthisis type II, functions as a molecular switch between Wnt signaling pathways. Nat. Genet. 37, 537–543.

Sohara, E., Luo, Y., Zhang, J., Manning, D.K., Beier, D.R., and Zhou, J. (2008). Nek8 regulates the expression and localization of polycystin-1 and polycystin-2. J. Am. Soc. Nephrol. JASN 19, 469–476.

Sorokin, S. (1962). Centrioles and the Formation of Rudimentary Cilia by Fibroblasts and Smooth Muscle Cells. J. Cell Biol. 15, 363–377.

Sparks, M.A., Crowley, S.D., Gurley, S.B., Mirotsou, M., and Coffman, T.M. (2014). Classical Renin-Angiotensin System in Kidney Physiology. In Comprehensive Physiology, R. Terjung, ed. (Hoboken, NJ, USA: John Wiley & Sons, Inc.), pp. 1201– 1228.

Stagner, E.E., Bouvrette, D.J., Cheng, J., and Bryda, E.C. (2009). The polycystic kidney disease-related proteins Bicc1 and SamCystin interact. Biochem. Biophys. Res. Commun. 383, 16–21.

Stephanis, L. de, Bonon, A., Varani, K., Lanza, G., Gafà, R., Pinton, P., Pema, M., Somlo, S., Boletta, A., and Aguiari, G. (2017). Double inhibition of cAMP and mTOR signalling may potentiate the reduction of cell growth in ADPKD cells. Clin. Exp. Nephrol. 21, 203–211.

Stumvoll, M., Meyer, C., Perriello, G., Kreider, M., Welle, S., and Gerich, J. (1998). Human kidney and liver gluconeogenesis: evidence for organ substrate selectivity. Am. J. Physiol.-Endocrinol. Metab. 274, E817–E826.

215 Suzuki, T., Ueda, A., Kobayashi, N., Yang, J., Tomaru, K., Yamamoto, M., Takeno, M., and Ishigatsubo, Y. (2008). Proteasome-dependent degradation of alpha-catenin is regulated by interaction with ARMc8alpha. Biochem. J. 411, 581–591.

Sweeney, W.E., Kusner, L., Carlin, C.R., Chang, S., Futey, L., Cotton, C.U., Dell, K.M., and Avner, E.D. (2001). Phenotypic analysis of conditionally immortalized cells isolated from the BPK model of ARPKD. Am. J. Physiol. - Cell Physiol. 281, C1695– C1705.

Swenson-Fields, K.I., Vivian, C.J., Salah, S.M., Peda, J.D., Davis, B.M., van Rooijen, N., Wallace, D.P., and Fields, T.A. (2013). Macrophages promote polycystic kidney disease progression. Kidney Int. 83, 855–864.

Szymanska, K., and Johnson, C.A. (2012). The transition zone: an essential functional compartment of cilia. Cilia 1, 10.

Tain, Y., and Hsu, C. (2017). Toxic Dimethylarginines: Asymmetric Dimethylarginine (ADMA) and Symmetric Dimethylarginine (SDMA). Toxins 9, 92.

Takakura, A., Contrino, L., Zhou, X., Bonventre, J.V., Sun, Y., Humphreys, B.D., and Zhou, J. (2009). Renal injury is a third hit promoting rapid development of adult polycystic kidney disease. Hum. Mol. Genet. 18, 2523–2531.

Takiar, V., Nishio, S., Seo-Mayer, P., King, J.D., Li, H., Zhang, L., Karihaloo, A., Hallows, K.R., Somlo, S., and Caplan, M.J. (2011). Activating AMP-activated protein kinase (AMPK) slows renal cystogenesis. Proc. Natl. Acad. Sci. 108, 2462–2467.

Tautenhahn, R., Cho, K., Uritboonthai, W., Zhu, Z., Patti, G.J., and Siuzdak, G. (2012). An accelerated workflow for untargeted metabolomics using the METLIN database. Nat. Biotechnol. 30, 826–828.

Taylor, S.L., Ganti, S., Bukanov, N.O., Chapman, A., Fiehn, O., Osier, M., Kim, K., and Weiss, R.H. (2010). A metabolomics approach using juvenile cystic mice to identify urinary biomarkers and altered pathways in polycystic kidney disease. AJP Ren. Physiol. 298, F909–F922.

Tengholm, A., and Gylfe, E. (2017). cAMP signalling in insulin and glucagon secretion. Diabetes Obes. Metab.

216 The Diabetes Control and Complications Trial/Epidemiology of Diabetes Interventions (2005). Intensive Diabetes Treatment and Cardiovascular Disease in Patients with Type 1 Diabetes. N. Engl. J. Med. 353, 2643–2653.

Torres, V.E., and Harris, P.C. (2014). Strategies Targeting cAMP Signaling in the Treatment of Polycystic Kidney Disease. J. Am. Soc. Nephrol. 25, 18–32.

Torres, V.E., Mujwid, D.K., Wilson, D.M., and Holley, K.H. (1994a). Renal cystic disease and ammoniagenesis in Han:SPRD rats. J. Am. Soc. Nephrol. 5, 1193–1200.

Torres, V.E., Keith, D.S., Offord, K.P., Kon, S.P., and Wilson, D.M. (1994b). Renal ammonia in autosomal dominant polycystic kidney disease. Kidney Int. 45, 1745– 1753.

Tran, U., Mary Pickney, L., Duygu Özpolat, B., and Wessely, O. (2007). Xenopus Bicaudal-C Is Required for the Differentiation of the Amphibian Pronephros. Dev. Biol. 307, 152–164.

Tran, U., Zakin, L., Schweickert, A., Agrawal, R., Doger, R., Blum, M., De Robertis, E.M., and Wessely, O. (2010). The RNA-binding protein bicaudal C regulates polycystin 2 in the kidney by antagonizing miR-17 activity. Development 137, 1107– 1116.

Upadhya, P., Birkenmeier, E.H., Birkenmeier, C.S., and Barker, J.E. (2000). Mutations in a NIMA-related kinase gene, Nek1, cause pleiotropic effects including a progressive polycystic kidney disease in mice. Proc. Natl. Acad. Sci. U. S. A. 97, 217–221.

Vallance, P., and Leiper, J. (2004). Cardiovascular Biology of the Asymmetric Dimethylarginine:Dimethylarginine Dimethylaminohydrolase Pathway. Arterioscler. Thromb. Vasc. Biol. 24, 1023–1030.

Vander Heiden, M.G., Cantley, L.C., and Thompson, C.B. (2009). Understanding the Warburg effect: the metabolic requirements of cell proliferation. Science 324, 1029– 1033.

Vivante, A., Mark-Danieli, M., Davidovits, M., Harari-Steinberg, O., Omer, D., Gnatek, Y., Cleper, R., Landau, D., Kovalski, Y., Weissman, I., et al. (2013). Renal

217 Hypodysplasia Associates with a Wnt4 Variant that Causes Aberrant Canonical Wnt Signaling. J. Am. Soc. Nephrol. 24, 550–558.

Wallace, D.P., Rome, L.A., Sullivan, L.P., and Grantham, J.J. (2001). cAMP- dependent fluid secretion in rat inner medullary collecting ducts. Am. J. Physiol. - Ren. Physiol. 280, F1019–F1029.

Wallingford, J.B., and Mitchell, B. (2011). Strange as it may seem: the many links between Wnt signaling, planar cell polarity, and cilia. Genes Dev. 25, 201–213.

Walz, G., Budde, K., Mannaa, M., Nürnberger, J., Wanner, C., Sommerer, C., Kunzendorf, U., Banas, B., Hörl, W.H., Obermüller, N., et al. (2010). Everolimus in Patients with Autosomal Dominant Polycystic Kidney Disease. N. Engl. J. Med. 363, 830–840.

Wang, L., Eckmann, C.R., Kadyk, L.C., Wickens, M., and Kimble, J. (2002). A regulatory cytoplasmic poly (A) polymerase in Caenorhabditis elegans. Nature 419, 312–316.

Wang, X., Ward, C.J., Harris, P.C., and Torres, V.E. (2010). Cyclic nucleotide signaling in polycystic kidney disease. Kidney Int. 77, 129–140.

Ward, C.J., Hogan, M.C., Rossetti, S., Walker, D., Sneddon, T., Wang, X., Kubly, V., Cunningham, J.M., Bacallao, R., Ishibashi, M., et al. (2002). The gene mutated in autosomal recessive polycystic kidney disease encodes a large, receptor-like protein. Nat. Genet. 30, 259–269.

Ward, H.H., Brown-Glaberman, U., Wang, J., Morita, Y., Alper, S.L., Bedrick, E.J., Gattone, V.H., Deretic, D., and Wandinger-Ness, A. (2011). A conserved signal and GTPase complex are required for the ciliary transport of polycystin-1. Mol. Biol. Cell 22, 3289–3305.

Warner, G., Hein, K.Z., Nin, V., Edwards, M., Chini, C.C.S., Hopp, K., Harris, P.C., Torres, V.E., and Chini, E.N. (2016). Food Restriction Ameliorates the Development of Polycystic Kidney Disease. J. Am. Soc. Nephrol. 27, 1437–1447.

Watanabe, M., Suliman, M.E., Qureshi, A.R., Garcia-Lopez, E., Bárány, P., Heimbürger, O., Stenvinkel, P., and Lindholm, B. (2008). Consequences of low

218 plasma histidine in chronic kidney disease patients: associations with inflammation, oxidative stress, and mortality. Am. J. Clin. Nutr. 87, 1860–1866.

Watnick, T.J., Torres, V.E., Gandolph, M.A., Qian, F., Onuchic, L.F., Klinger, K.W., Landes, G., and Germino, G.G. (1998). Somatic Mutation in Individual Liver Cysts Supports a Two-Hit Model of Cystogenesis in Autosomal Dominant Polycystic Kidney Disease. Mol. Cell 2, 247–251.

Wei, W., Hackmann, K., Xu, H., Germino, G., and Qian, F. (2007). Characterization of cis-Autoproteolysis of Polycystin-1, the Product of Human Polycystic Kidney Disease 1 Gene. J. Biol. Chem. 282, 21729–21737.

Wessely, O., and De Robertis, E.M. (2000). The Xenopus homologue of Bicaudal-C is a localized maternal mRNA that can induce endoderm formation. Development 127, 2053–2062.

Wessely, O., Tran, U., Zakin, L., and De Robertis, E.M. (2001). Identification and expression of the mammalian homologue of Bicaudal-C. Mech. Dev. 101, 267–270.

Wiczer, B.M., Kalender, A., and Thomas, G. (2010). Bending the path to TOR. Nat. Cell Biol. 12, 1031–1033.

Williams, C.L., Li, C., Kida, K., Inglis, P.N., Mohan, S., Semenec, L., Bialas, N.J., Stupay, R.M., Chen, N., Blacque, O.E., et al. (2011). MKS and NPHP modules cooperate to establish basal body/transition zone membrane associations and ciliary gate function during ciliogenesis. J. Cell Biol. 192, 1023–1041.

Winer, L.S.P., and Wu, M. (2014). Rapid Analysis of Glycolytic and Oxidative Substrate Flux of Cancer Cells in a Microplate. PLOS ONE 9, e109916.

Wishart, D.S., Jewison, T., Guo, A.C., Wilson, M., Knox, C., Liu, Y., Djoumbou, Y., Mandal, R., Aziat, F., Dong, E., et al. (2013). HMDB 3.0—The Human Metabolome Database in 2013. Nucleic Acids Res. 41, D801–D807.

Wong, G.K., and Crawford, M.W. (2011). Carnitine deficiency increases susceptibility to bupivacaine-induced cardiotoxicity in rats. Anesthesiology 114, 1417–1424.

Woo, D. (1995). Apoptosis and Loss of Renal Tissue in Polycystic Kidney Diseases. N. Engl. J. Med. 333, 18–25.

219 Wu, G., Bazer, F.W., Datta, S., Johnson, G.A., Li, P., Satterfield, M.C., and Spencer, T.E. (2008). Proline metabolism in the conceptus: implications for fetal growth and development. Amino Acids 35, 691–702.

Wu, M., Neilson, A., Swift, A.L., Moran, R., Tamagnine, J., Parslow, D., Armistead, S., Lemire, K., Orrell, J., Teich, J., et al. (2007). Multiparameter metabolic analysis reveals a close link between attenuated mitochondrial bioenergetic function and enhanced glycolysis dependency in human tumor cells. Am. J. Physiol. - Cell Physiol. 292, C125–C136.

Wu, Y., Dai, X.-Q., Li, Q., Chen, C.X., Mai, W., Hussain, Z., Long, W., Montalbetti, N., Li, G., Glynne, R., et al. (2006). Kinesin-2 mediates physical and functional interactions between polycystin-2 and fibrocystin. Hum. Mol. Genet. 15, 3280–3292.

Xia, J., and Wishart, D.S. (2016). Using MetaboAnalyst 3.0 for Comprehensive Metabolomics Data Analysis: Using MetaboAnalyst 3.0 for Comprehensive Metabolomics Data Analysis. In Current Protocols in Bioinformatics, A. Bateman, W.R. Pearson, L.D. Stein, G.D. Stormo, and J.R. Yates, eds. (Hoboken, NJ, USA: John Wiley & Sons, Inc.), p. 14.10.1-14.10.91.

Yaguchi, S., Yaguchi, J., Wei, Z., Shiba, K., Angerer, L.M., and Inaba, K. (2010). ankAT-1 is a novel gene mediating the apical tuft formation in the sea urchin embryo. Dev. Biol. 348, 67–75.

Yaguchi, S., Yaguchi, J., and Inaba, K. (2015). Bicaudal-C is required for the formation of anterior neurogenic ectoderm in the sea urchin embryo. Sci. Rep. 4.

Yakulov, T.A., Yasunaga, T., Ramachandran, H., Engel, C., Müller, B., Hoff, S., Dengjel, J., Lienkamp, S.S., and Walz, G. (2015). Anks3 interacts with nephronophthisis proteins and is required for normal renal development. Kidney Int. 87, 1191–1200.

Yamaguchi, T., Wallace, D.P., Magenheimer, B.S., Hempson, S.J., Grantham, J.J., and Calvet, J.P. (2004). Calcium restriction allows cAMP activation of the B-Raf/ERK pathway, switching cells to a cAMP-dependent growth-stimulated phenotype. J. Biol. Chem. 279, 40419–40430.

220 Yoder, B.K., Hou, X., and Guay-Woodford, L.M. (2002). The Polycystic Kidney Disease Proteins, Polycystin-1, Polycystin-2, Polaris, and Cystin, Are Co-Localized in Renal Cilia. J. Am. Soc. Nephrol. 13, 2508–2516.

Yu, S., Hackmann, K., Gao, J., He, X., Piontek, K., García González, M.A., Menezes, L.F., Xu, H., Germino, G.G., Zuo, J., et al. (2007). Essential role of cleavage of Polycystin-1 at G protein-coupled receptor proteolytic site for kidney tubular structure. Proc. Natl. Acad. Sci. U. S. A. 104, 18688–18693.

Zakrzewicz, D., and Eickelberg, O. (2009). From arginine methylation to ADMA: A novel mechanism with therapeutic potential in chronic lung diseases. BMC Pulm. Med. 9, 5.

Zatti, A., Chauvet, V., Rajendran, V., Kimura, T., Pagel, P., and Caplan, M.J. (2005). The C-terminal tail of the polycystin-1 protein interacts with the Na,K-ATPase alpha- subunit. Mol. Biol. Cell 16, 5087–5093.

Zhang, D., Yang, H., Kong, X., Wang, K., Mao, X., Yan, X., Wang, Y., Liu, S., Zhang, X., Li, J., et al. (2011). Proteomics analysis reveals diabetic kidney as a ketogenic organ in type 2 diabetes. AJP Endocrinol. Metab. 300, E287–E295.

Zhang, J., Nuebel, E., Wisidagama, D.R.R., Setoguchi, K., Hong, J.S., Van Horn, C.M., Imam, S.S., Vergnes, L., Malone, C.S., Koehler, C.M., et al. (2012). Measuring energy metabolism in cultured cells, including human pluripotent stem cells and differentiated cells. Nat. Protoc. 7, 1068–1085.

Zhang, M.-Z., Mai, W., Li, C., Cho, S., Hao, C., Moeckel, G., Zhao, R., Kim, I., Wang, J., Xiong, H., et al. (2004). PKHD1 protein encoded by the gene for autosomal recessive polycystic kidney disease associates with basal bodies and primary cilia in renal epithelial cells. Proc. Natl. Acad. Sci. U. S. A. 101, 2311–2316.

Zhang, Y., Cooke, A., Park, S., Dewey, C.N., Wickens, M., and Sheets, M.D. (2013). Bicaudal-C spatially controls translation of vertebrate maternal mRNAs. RNA 19, 1575–1582.

Zhao, Y., Peng, S., Jia, C., Xu, F., Xu, Y., and Dai, C. (2016). Armc8 regulates the invasive ability of hepatocellular carcinoma through E-cadherin/catenin complex. Tumor Biol. 37, 11219–11224.

221 Zhou, X., Fan, L.X., Sweeney, W.E., Denu, J.M., Avner, E.D., and Li, X. (2013). Sirtuin 1 inhibition delays cyst formation in autosomal-dominant polycystic kidney disease. J. Clin. Invest. 123, 3084–3098.

Zhu, Y., Wang, Y., Xia, C., Li, D., Li, Y., Zeng, W., Yuan, W., Liu, H., Zhu, C., Wu, X., et al. (2004). WDR26: A novel G?-like protein, suppresses MAPK signaling pathway. J. Cell. Biochem. 93, 579–587.

Zoncu, R., Efeyan, A., and Sabatini, D.M. (2011). mTOR: from growth signal integration to cancer, diabetes and ageing. Nat. Rev. Mol. Cell Biol. 12, 21–35.

222 Curriculum Vitae Lucía C. Leal-Esteban

Chemin du Chêne, 16A 1020, Renens, Switzerland E-mail: [email protected] Date of birth: 25th of January 1986 Colombian

EDUCATION 2012-2017 Ph.D in Molecular Life Sciences Lausanne (CH)

Ecole Polytechnique Fédérale de Lausanne (EPFL) Thesis: Functional analysis of Bicc1 and novel interacting proteins in renal tubule cell metabolism and polycystic kidneys. D. Constam Group.

2011-2012 M.Sc in Clinical Laboratory Barcelona (SP) Universitat Pompeu Fabra (UPF) Thesis: Epigenetic and biological characterization of Derlins in colorectal cancer. M. Esteller Group

2005-2009 B.Sc in Microbiology and Clinical Laboratory Bucaramanga (CO)

WORK EXPERIENCE 2012-2017 PhD student in Molecular Biology, EPFL Lausanne (CH) UPCDA SV-ISREC Molecular biology, biochemistry, cell culture, mice dissection skills

2009-2011 Research assistant at Group of Primary Medellín (CO) Immunodeficiencies (IDP), University of Antioquia NK cells functionality tests in patients with immunodeficiencies, serological analysis in patients with pneumococcal infections. 223 LANGUAGES • Spanish: mother language • English: professional level • French: basic level

EXTRACURRICULAR ACTIVITIES • Member BioScience Network Lausanne (BSNL) since 2015 • Member Colombian association of researchers in Switzerland (ACIS) since 2015 • Member Life Science Switzerland (LS2) since 2016

PUBLICATIONS • Rothé B, Leal-Esteban L, Bernet F, Urfer S, Doerr N, Weimbs T, Iwaszkiewicz J, Constam DB. 2015. Bicc1 polymerization regulates the localization and silencing of bound mRNA. Mol Cell Biol 35:3339–3353. doi:10.1128/MCB.00341-15 • Lopez-Serra, P. et al. A DERL3-associated defect in the degradation of SLC2A1 mediates the Warburg effect. Nat. Commun. 5:3608 doi: 10.1038/ncomms4608 (2014). • Sánchez, I.P, Leal-Esteban, L.C, Orrego-Arango, J.C, Garcés-Samudio, C.G, Gómez-Arias, R.D, Franco, J.L, & Trujillo-Vargas, C.M. (2014). Variation in NK cell number and function in individuals with recurrent or severe infections. Biomédica, 34 (1), 118-131. • Sánchez, I.P., Leal-Esteban, L.C., Álvarez-Álvarez, J.A. et al. Analyses of the PRF1 gene in individuals with Hemophagocytic Lymphohystiocytosis reveal the common haplotype R54C/A91V in Colombian unrelated families associated with late onset disease. J Clin Immunol (2012) 32: 670. https://doi.org/10.1007/s10875-012-9680-5. • Leal-Esteban, L.C, Rojas, J. L, Jaimes, A. L, Montoya, J. D, Montoya, N. E, Leiva, L, & Trujillo-Vargas, C. M. (2012). An immunoenzymatic test for IgG antibody levels against 10 serotypes of Streptococcus pneumoniae. Biomédica, 32 (1), 92-102.

224