Chapter 1 7 Introduction, Aim, and Outline

UvA-DARE (Digital Academic Repository)

ATP8B1 and cellular trafficking van der Mark, V.A.

Publication date 2017 Document Version Final published version License Other Link to publication

Citation for published version (APA): van der Mark, V. A. (2017). ATP8B1 and cellular trafficking.

General rights It is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s), other than for strictly personal, individual use, unless the work is under an open content license (like Creative Commons).

Disclaimer/Complaints regulations If you believe that digital publication of certain material infringes any of your rights or (privacy) interests, please let the Library know, stating your reasons. In case of a legitimate complaint, the Library will make the material inaccessible and/or remove it from the website. Please Ask the Library: https://uba.uva.nl/en/contact, or a letter to: Library of the University of Amsterdam, Secretariat, Singel 425, 1012 WP Amsterdam, The Netherlands. You will be contacted as soon as possible.

UvA-DARE is a service provided by the library of the University of Amsterdam (https://dare.uva.nl)

Download date:27 Sep 2021

ATP8B1 and cellular trafficking

Vincent Alexander van der Mark

ATP8B1 and cellular trafficking

Cover: Vera de Groot van der Mark

Printing: Print Service Ede

ISBN: 978-94-91602-98-6

No part of this book may be reproduced, stored, or transmitted in any form without prior permission of the author.

The research conducted in this thesis was performed at the Tytgat Institute for Liver and Intestinal Research, Academic Medical Center, University of Amsterdam, The Netherlands.

The printing of this thesis was financially supported by the Dutch Society for Hepatology (NVH).

ATP8B1 and cellular trafficking

ACADEMISCH PROEFSCHRIFT

ter verkrijging van de graad van doctor

aan de Universiteit van Amsterdam

op gezag van de Rector Magnificus

prof. dr. ir. K.I.J. Maex ten overstaan van een door het College voor Promoties ingestelde commissie,

in het openbaar te verdedigen in de Agnietenkapel

op dinsdag 11 april 2017, te 10:00 uur

door Vincent Alexander van der Mark

geboren te Alkmaar

Promotiecommissie:

Promotor: prof. dr. R.P.J. Oude Elferink Universiteit van Amsterdam

Copromotor: dr. C.C. Paulusma Universiteit van Amsterdam

Overige leden: prof. dr. G.R. van den Brink Universiteit van Amsterdam

prof. dr. T.B.H. Geijtenbeek Universiteit van Amsterdam

prof. dr. R.H.J. Houwen Universiteit Utrecht

prof. dr. A.J. Verhoeven Universiteit van Amsterdam

dr. K.F.J. van de Graaf Universiteit van Amsterdam

dr. H.R. de Jonge Erasmus Universiteit Rotterdam

Faculteit der Geneeskunde

Table of contents

Chapter 1 7 Introduction, aim, and outline

Chapter 2 P4-ATPases: flippases in health and disease 11

Chapter 3 49 Differential effects of progressive familial intrahepatic cholestasis type 1 and benign recurrent intrahepatic cholestasis type 1 mutations on canalicular localization of ATP8B1

Chapter 4 71 The lipid flippase heterodimer ATP8B1-CDC50A is essential for surface expression of the apical sodium-dependent bile acid transporter (SLC10A2/ASBT) in intestinal Caco-2 cells

Chapter 5 97 The phospholipid flippase ATP8B1 mediates apical localization of the cystic fibrosis transmembrane regulator

Chapter 6 119 Phospholipid flippases attenuate LPS-induced TLR4 signaling by mediating endocytic retrieval of Toll- like receptor 4

Chapter 7 153 Summary and perspectives

Appendices Samenvatting 164 Dankwoord 168 Publications 171 Authors and affiliations 173 Portfolio 174

CHAPTER 1

Introduction, aim, and outline

Chapter 1

Progressive familial intrahepatic cholestasis type 1 (PFIC1), also called Byler’s disease, is an autosomal recessive inherited liver disease characterized by chronic cholestasis. PFIC1 presents during infancy or childhood and is lethal unless treated. Current treatment consists of external biliary drainage, which temporally improves the cholestatic phenotype and postpones or sometimes prevents liver transplantation. PFIC1 patients and patients diagnosed with the less severe form benign intrahepatic cholestasis type 1 (BRIC1) often suffer from additional extrahepatic symptoms, such as hearing loss, pancreatitis, pulmonary infection, and diarrhea. The etiology of these extrahepatic symptoms has barely been studied. PFIC1/BRIC1 is caused by mutations in the gene encoding the protein ATP8B1. ATP8B1 is a phospholipid translocating membrane protein that is involved in the maintenance of membrane asymmetry and intracellular vesicle transport.

The aim of this thesis was to study the pathogenesis of several extrahepatic phenotypes in ATP8B1 deficiency, including diarrhea and pulmonary infection.

Chapter 2 recapitulates the current scientific knowledge on the biochemical and physiological role of ATP8B1. In addition, it provides an overview of its orthologues in other organisms and the role that its family members, P4-ATPases, play in human disease.

In PFIC1 and BRIC1 patients ATP8B1 can be mutated at different positions giving rise to different degrees of malfunctioning proteins. Chapter 3 shows to what extent several of the most common mutations result in different intracellular localization, interaction with its obligatory beta subunit CDC50A, and stability of the protein, providing an explanation for the difference in presentation of PFIC1 and BRIC1 patients.

PFIC1 patients are treated with intestinal bile salt binding resins to ameliorate diarrhea. Especially in liver-transplanted PFIC1 patients diarrhea is worsened. In chapter 4 the role of ATP8B1 on the expression and function of the intestinal apical sodium-dependent bile salt transporter (ASBT) in human intestinal Caco-2 cells is analyzed. Moreover, fecal content of several PFIC1 patients, including post- transplant patients, is examined in order to determine alterations in intestinal bile

Chapter 1 salt, nutrient and electrolyte transport. From these data we conclude that diarrhea in PFIC1 patients has a secretory origin to which intestinal bile acid malabsorption can contribute.

Chapter 5 continues the research on the function of ATP8B1 in the intestine and extends its role to the lung by studying its effects on the expression and function of the cystic fibrosis transmembrane regulator (CFTR). CFTR is the primary chloride channel and indirectly drives paracellular transport of water and solutes, which is essential for proper clearance of irritants and pathogens. Here we study the effects of ATP8B1 on CFTR function and expression in human intestinal Caco-2 and T84 cells, human lung epithelial Calu-3 cells, and ex-vivo mouse intestines. From this study we conclude that ATP8B1 is essential for apical targeting of CFTR.

PFIC1 patients are more frequently affected by pulmonary infections. The potential role of P4-ATPases and their beta-subunit CDC50A in the innate immune system is studied for the first time in chapter 6 by determining the localization and activity of the Toll-like receptor 4 (TLR4) in THP-1 macrophages and primary human and murine macrophages depleted for ATP8B1, ATP11A, or CDC50A. This work presents novel insights into the molecular mechanisms of TLR4-mediated signaling and indicates an important novel role for some members of the P4-ATPase family in innate immunity.

CHAPTER 2

P4-ATPases: Flippases in Health and Disease

Vincent A. van der Mark, Ronald P. Oude Elferink, Coen C. Paulusma

Adapted from: Int J Mol Sci. 2013 Apr 11;14(4):7897-922

doi: 10.3390/ijms14047897

Chapter 2

Abstract

P4-ATPases catalyze the translocation of phospholipids from the exoplasmic to the cytosolic leaflet of biological membranes, a process termed “lipid flipping”. Accumulating evidence obtained in lower eukaryotes points to an important role for P4-ATPases in the maintenance of membrane phospholipid asymmetry and vesicular protein trafficking. The human genome encodes fourteen P4-ATPases of which the cellular and physiological functions are slowly emerging. Thus far, deficiencies of at least two P4-ATPases, ATP8B1, and ATP8A2 are the cause of human disease. However, various mouse models and in vitro studies are contributing to our understanding of the cellular and physiological functions of P4-ATPases. This review summarizes current knowledge on the basic function of these phospholipid translocating proteins, their proposed action in intracellular vesicle transport and their physiological role.

Chapter 2

1. Phospholipid Asymmetry in Biological Membranes

In 1925, the Dutch scientists Gorter and Grendel were the first to demonstrate that the plasma membrane of the erythrocyte was a bilayer of phospholipids (1). From their experiments in “chromocytes of the blood” (i.e. erythrocytes) they concluded that the surface area of lipids extracted from erythrocytes was about twice the surface area of the cells themselves. Gorter and Grendel concluded that “chromocytes are covered by a layer of fatty substances that is two molecules thick” (1). Indeed, biological membranes are an intricate mixture of many different lipid species that are organized as two back-to-back facing leaflets of phospholipids with hydrophilic head groups facing the hydrophilic environment and hydrophobic acyl tails facing the core of the bilayer (2). Eukaryotic cell membranes are composed of glycerophospholipids (~65 mol%), cholesterol (~25 mol%), and sphingolipids (~10 mol%). The main structural phospholipids are phosphatidylcholine (PC), phosphatidylserine (PS), phosphatidylethanolamine (PE), and phosphatidylinositol (PI). Most glycerophospholipids are synthesized in the endoplasmic reticulum (ER), while sphingolipids are synthesized in the Golgi. Cardiolipin, a glycerophospholipid exclusively localized to the inner membrane of mitochondria, is generated in the mitochondrion. Each leaflet has a different composition which also varies from organelle to organelle. For example, the plasma membrane contains the highest percentage of PS (~10 mol%) as opposed to late endosomes (~2 mol%) or mitochondria (~1 mol%) (3).

One of the hallmarks of eukaryotic membranes in the secretory and endocytic pathways is the asymmetric distribution of the different phospholipid species over the exoplasmic- and cytoplasmic leaflet of the bilayer. Maintaining and dissipating the non-random distribution of phospholipids is crucial for normal regulated membrane (protein) function. For instance, PS exposure at the plasma membrane is an important signal both in the recognition and phagocytosis of apoptotic cells and in the activation of blood coagulation (4). In the cytosolic leaflet of the plasma membrane, PS is important for the recruitment of protein kinase C (PKC), Src, Ras, and Rho proteins via binding to their C2 domain or other cationic domains (5, 6).

In the early 1970s, Bretscher, Gordesky, and Marinetti demonstrated that the aminophospholipids PS and PE are concentrated in the cytoplasmic leaflet of the erythrocyte bilayer (7, 8). In addition, Verkleij et al. showed that PC was equally distributed over both leaflets, whereas

Chapter 2 sphingomyelin (SM) almost exclusively localized to the exoplasmic leaflet (9). Less abundant phospholipids, like PI (and derivatives of PI) and phosphatidic acid (PA) are mostly (about 80%) confined to the cytoplasmic leaflet (10-12). Cholesterol can partition equally and rapidly (milliseconds to seconds) between the two leaflets (13). It may be enriched in the exoplasmic leaflet due to its high affinity for SM (14), although fluorescent sterols predominantly localize to the cytosolic leaflet in Chinese Hamster Ovary (CHO) cells (15). Differences in membrane lipid composition likely alter the chemical stability of cholesterol leading to differences in cholesterol distribution, as discussed in (16). Phospholipids tend to equilibrate between the two leaflets, a process called “scrambling”, which is very slow and depends on lipid head group structure and polarity, as well as hydrophobicity and saturation of the acyl chains and lipid packing in the bilayer (17). For instance, the equilibration time of phosphatidylglycerol across a membrane takes several days, whereas that of a diacylglycerol (which lacks the phosphate group) is a matter of seconds (18). However, due to extensive membrane fusion and budding events by intracellular transport vesicles with and/or from the plasma membrane and cell organelles, phospholipid scrambling is accelerated. Consequently, the non-random transbilayer distribution of phospholipids must be actively maintained, which is accomplished by phospholipid translocating proteins termed flippases and floppases. Lipid flippases and floppases are ATP-dependent proteins that confer transbilayer distribution of phospholipids by translocating phospholipids from the exoplasmic to the cytoplasmic leaflet of the bilayer and vice versa, respectively (19). Actual ATP- dependent flipping of PS and PE, but not PC, was first observed in human erythrocytes by determining the distribution of spin-labeled analogs after their incorporation in the outer leaflet (20). It is presently well-accepted that proteins from the type 4 subfamily (P4) of the P-type ATPase superfamily (P4-ATPases) are essential in the flipping of phospholipids, which is the subject of the current review. The flopping of lipids is accomplished by members of the ATP-binding cassette (ABC) transporter superfamily, which mediate the ATP-dependent transport of a wide variety of compounds across biological membranes, including (short-chain) phospholipids, sterols, and (very long-chain) fatty acids and other amphipatic compounds. Apart from ATP-dependent floppases and flippases, the membrane contains ATP-independent bidirectional scramblases that support the Ca2+- dependent translocation of phospholipids between leaflets (4, 21, 22). Scrambling is of physiological importance for processes such as apoptosis, fission and fusion of transport vesicles, sperm capacitation, blood coagulation, and intracellular signaling.

Chapter 2

Recently, a Xk-related family of proteins has been identified as primary candidates for scrambling and exposing PS in apoptotic cells (23, 24). Upon apoptotic stimuli caspase-3 cleaves the C-terminal tail of family members Xkr4, 8, and 9, and thereby promotes PS exposure to ensure phagocytosis by macrophages. In addition to Xkr proteins another PS scramblase has been identified in patients suffering from Scott syndrome. Scott syndrome is a congenital bleeding disorder associated with mutations in the transmembrane protein TMEM16F. After platelet activation Ca2+- dependent PS exposure is reduced in patients suffering from Scott syndrome and a TMEM16F-/- mouse model, preventing the formation of tenase and prothrombinase complexes required for blood coagulation (21, 25). Lipid flopping and scrambling are not the subject of this review and have been excellently highlighted elsewhere (4, 26-28).

2. The P4-ATPase Family of Lipid Flippases

The P-type ATPase superfamily is a large and evolutionary conserved family of proteins of which members are widely expressed in both prokaryotes and eukaryotes. P-type ATPases are integral membrane proteins which, in most cases, mediate the ATP-dependent transport of small cations across biological membranes, including those of intracellular organelles. Based on phylogenetic analysis, this large family is divided into 5 subfamilies (P1-P5), each unique in their class of transported substrates and in subfamily-specific sequence motifs (29) (Figure 1). For instance, the P2 subfamily includes well known proteins as the Ca2+-ATPase (30) and the Na+/K+-ATPase (31), which have been extensively characterized for their molecular activity and (patho)physiological functions; in addition, crystal structures have been obtained for these P-type ATPases (32). The P4 subfamily members are exclusively expressed in eukaryotic cells and are deviant from the other P-type ATPase subfamilies in that they mediate the transport of phospholipids instead of cations. The first evidence that P4-ATPases are involved in translocating phospholipids was presented by studies on the Saccharomyces cerevisiae P4-ATPase Drs2p. This protein was demonstrated to be involved in the transport of fluorescent, NBD-labeled analogs of PS and PE across the plasma and Golgi membranes of yeast (33, 34). Since then, P4-ATPases from other species such as Caenorhabditis elegans (35), Mus musculus (36, 37), Homo sapiens (38, 39), Leishmania donovani (40), and Arabidopsis thaliana (41, 42) have been shown to flip NBD- or spin-labeled phospholipids or to be involved in the transbilayer distribution of endogenous

Chapter 2 phospholipids (39, 43-45). Like all other P-type ATPases, P4-ATPases are multiple transmembrane-spanning proteins with a large cytoplasmic loop harboring the aspartic acid residue which is essential in the reaction cycle of the protein, and both N- and C-terminal tails protruding into the cytoplasm (Figure 2). The mechanism via which substrates are pumped is based on the autophosphorylation of the invariant aspartic acid residue, which drives the transport cycle of these proteins, hence the name P-type ATPase. At present, none of the P4-ATPases have been crystallized.

Chapter 2

Figure 1. Phylogenetic tree of the P-type ATPase superfamily and of the P4 branch. Substrates of P1–P5 branches are in between brackets. Phylogenetic analyses of the P4-ATPase protein family of mammalian, A. thaliana, S. cerevisiae, and C. elegans is shown (adapted from (46)). Database accession numbers: C. elegans: TAT-1 (NP_001022894), TAT-2 (NP_001023252), TAT-3 (NP_499363), TAT-4 (NP_495244), TAT-5 (NP_001021457), TAT-6 (NP_503858); A. thaliana: ALA1 (P98204), ALA2 (P98205), ALA3 (Q9XIE6), ALA4 (Q9LNQ4), ALA5 (Q9SGG3), ALA6 (Q9SLK6), ALA7 (Q9LVK9), ALA8 (Q9LK90), ALA9 (Q9SX33), ALA10 (Q9LI83), ALA11 (Q9SAF5), ALA12 (P57792); S. cerevisiae: Drs2p (P39524), Dnf1p (P32660), Dnf2p (Q12675), Dnf3p (Q12674), Neo1p (P40527). H. sapiens: ATP8A1 (P70704), ATP8A2 (P98200), ATP8B1 (O43520), ATP8B2 (P98198), ATP8B3 (O60423), ATP8B4 (Q8TF62), ATP9A (O75110), ATP9B (O43861), ATP10A (O60312), ATP10B (O94823), ATP10D (Q9P241), ATP11A (P98196), ATP11B (Q9Y2G3), ATP11C (Q8NB49).

Chapter 2

Figure 2. Simplified topological model of a P4-ATPase and its CDC50 subunit. P4-ATPases consist of an actuator (A), phosphorylation (P), nucleotide binding (N), and 10 predicted membrane spanning helices. CDC50 subunits consist of 2 membrane spanning domains with a large extracellular loop containing four possible N-linked glycosylation sites and two stabilizing disulfide bridges. Modified from Coleman et al. (47).

3. Beta Subunits for P4-ATPases

Similar to other P-type ATPases such as the Na+/K+- and gastric H+/K+-ATPases (48, 49), P4-ATPases require a heterodimeric interaction with a β-subunit in order to function properly. These β-subunits are glycosylated 50–60 kDa transmembrane proteins of the evolutionary conserved family of CDC50 proteins present in yeast, plants, mammals, and Leishmania (46). CDC50 proteins have two putative transmembrane domains and a large loop that protrudes into the exoplasmic space (Figure 2). Although they do not show sequence similarity to the Na+/K+-ATPase β- and γ-subunits, they do display structural and functional similarities (50). For the Na+/K+-ATPase it has been shown that assembly of the α- and β-subunits in the ER is essential for α-subunit maturation, ER exit, subcellular trafficking, and modulation of α-subunit activity (51, 52). Protein sorting signals are present in both the α- and β-subunits (53, 54). Similarly, heterodimerization of a P4-ATPase with a CDC50 protein is pivotal for release of

Chapter 2 the P4-ATPase from the ER (39, 55-58). Although the contribution of CDC50 β- subunits to subcellular trafficking of the P4-ATPase is not completely clear, Lopez- Marques et al. demonstrate for plant P4-ATPases that it is the P4-ATPase rather than the β-subunit that determines the subcellular localization (41). Similarly, these authors show that the substrate specificity is determined by the P4-ATPase rather than by the β-subunit. Indeed, in a very elegant study, Baldridge and Graham identify amino acid residues in the yeast P4-ATPases Drs2 and Dnf1, involved in the flipping of PS and PC, respectively, which determine these substrate specificities (59). Using an exhaustive number of chimeric Drs2 and Dnf1 proteins that were analyzed for their ability to translocate PC or PS, they identify specific amino acid residues in transmembrane domains (TM) 1–4 of both P4-ATPases that determine substrate preferences. They ultimately show that substituting a tyrosine for a phenylalanine at residue 618 in TM 4 of Dnf1 results in the flipping of PS, whereas substitution of the reciprocal phenylalanine for a tyrosine at amino acid 511 in TM4 abrogates PS recognition by Drs2. These data indicate that it is not the β-subunit that determines the substrate specificity, but the P4-ATPase itself.

In mammals, three CDC50 proteins are expressed that are termed CDC50A-C (60, 61); CDC50A and CDC50B are ubiquitously expressed while CDC50C is predominantly expressed in testis and brain (60). Since the human genome encodes fourteen P4-ATPases (fifteen in mice (62)) and only three CDC50 proteins, this indicates that one CDC50 protein can interact with multiple P4-ATPases. Indeed, co-immunoprecipitation studies show that CDC50A, which is the most abundantly-expressed CDC50 protein, can interact with eleven out of fourteen P4-ATPases, whereas CDC50B interacts with at least two P4-ATPases (39, 58, 63). ATP8B3, which is highly expressed in the testis, does not interact with CDC50A or CDC50B, and may heterodimerize with CDC50C (58). Interestingly there are several P4-ATPases that do not interact with any of the CDC50 proteins; ATP9A, ATP9B, and the yeast ortholog Neo1p (63, 64). It is possible that these P4-ATPases do not require a subunit for correct trafficking or activity, since the cytoplasmic N-terminus of ATP9B is implicated to contain a TGN- localizing signal sequence (63). It remains to be determined which combinations of α- and ß-subunits are of physiological relevance. For example, Coleman and Molday show that endogenous CDC50A and ATP8A2 are present as a heterodimer in photoreceptor disc membranes (44).

Chapter 2

4. The Reaction Cycle

It is presently not clear how P4-ATPases transport phospholipids across lipid bilayers (65). However, apart from a role in stabilization of the P4-ATPase and ER exit, there is experimental evidence to support a role for CDC50 proteins in the reaction cycle of the P4-ATPases Drs2p, ATP8B1, ATP8B2, and ATP8A2 (44, 55, 66, 67). Coleman and Molday demonstrate PS flipping activity upon reconstitution of CDC50A and ATP8A2 into liposomes (44). For the purified Cdc50p/Drs2p heterodimer it was found that the β-subunit prefers to bind the E2-P conformation of the P4-ATPase (explained in the next paragraph and Figure 3a) (55, 67). Based on these findings, Stone and Williamson propose that the phospholipid could be loaded on the P4-ATPase/CDC50 heterodimer in the E2-P conformation, which is exposed on the luminal side of the bilayer (68).

It has been suggested that the reaction cycle is analogous to those of other established P-type ATPases (29), i.e., the classic Post-Albers or E1E2 model (69, 70) (Figure 3a,b). In the E1 unbound conformational state cytosolic ligands can easily bind to the P-type ATPase. Ligand binding (such as Na+ for the Na+/K+- ATPase) facilitates interaction of Mg2+-ATP with the nucleotide binding domain (N); however, for P4-ATPases it is elusive whether there are intracellular ligands required to initiate the reaction cycle. Binding of ATP and the substrate to be transported results in phosphorylation of the conserved aspartate in the phosphorylation domain (P) (66). This autophosphorylation introduces a small conformational change in the protein generating the E1-P state. Release of ADP, movement of the P-domain stretching the link between TM 3 and the actuator domain (A), and rotation of the A-domain generates a large conformational change transforming the protein into the E2-P state (32). In the case of P-type ATPases, affinity for cytosolic bound ligands is decreased during this conversion granting their release at the exoplasmic side (this step is unclear for P4-type ATPases). Exoplasmic ligands are now able to bind to a high affinity region in the membrane domain (M): K+ binds to the Na+/K+-ATPase and a phospholipid from the outer membrane leaflet to the P4-ATPase. In the E2-P state the Na+/K+-ATPase is stabilized by its β-subunit (71), while the P4-type ATPase has its highest affinity for its CDC50 subunit, possibly via interactions of the transmembrane regions and/or the subunit’s large exoplasmic loop (67). The CDC50 subunit assists in binding of the phospholipid to the P4-ATPase, opens up a

Chapter 2 pathway for translocation, or occludes the bound phospholipid, but, together with the bound extracellular ligand, is likely an important component in dephosphorylation of the E2-P state to the E2 state (67). The transition from E2 to E1 occurs when the A-domain returns to its original position away from the P- domain. Affinities for the subunit and exoplasmic ligands are decreased resulting in the release of ligands towards the cytoplasmic side.

Figure 3. Proposed reaction cycles of a P4-ATPase (a) and a P2C-ATPase (Na+/K+-ATPase) (b) complexed with their subunit. P-type ATPases cycle through four main separate conformations when transporting ligands. In the E1 state the P-type ATPases have high affinity for intracellular ligands; Na+ in the case of the Na+/K+-ATPase, unknown or none for the P4-ATPase. Binding of ATP to the N-domain and subsequent phosphorylation of the P domain results in the E1-P state. While converting from E1-P to E2-P, intracellular ligands (3 Na+ for the Na+/K+ ATPase) are released into the exoplasmic milieu and the A-domain rotates. This allows binding of extracellular ligands (2 K+ for the Na+/K+-ATPase) or a phospholipid (depicted in pink) from the exoplasmic leaflet. Affinity for the subunit is highest in this state and this interaction may assist in binding of the phospholipid. Dephosphorylation changes the enzyme from the E2-P to the E2 state. Movement of the A-domain away from the P-domain reverts the ATPase back to the E1 state thereby translocating the extracellular ligands or the phospholipid to the cytoplasmic side. Adapted from Coleman et al. and Lenoir et al. (47, 67).

Other protein and lipid factors are probably involved in completing the reaction cycle of the complex. In yeast, phosphatidylinositol-4 phosphate (PI4P) synthesis by the phosphatidylinositol-4 kinase (PI4K) Pik1 and binding of PI4P to the C-terminal tail of Drs2p are required for flippase activity (66, 72, 73). Dephosphorylation of the

Chapter 2

Drs2p-Cdc50p complex in crude membranes in the presence of PS only occurs when PI4P was present as well (66). Direct regulatory phosphorylation of yeast P4- ATPases occurs via the serine/threonine protein kinases (Fpk) 1 and 2 (74). Cells lacking these kinases are deficient in (NBD-labeled) phospholipid uptake.

The E1E2 model provides an explanation for the pumping mechanism of P4- ATPases, but does not take the “giant substrate problem” into account: compared to an ion, a phospholipid with its large apolar acyl chains is a large, bulky substrate that might prove too large to fit in its proposed binding pocket in its entirety. The “credit card” or “hydrophobic gate” model provides a way to circumvent this problem; only the charged phospholipid head group is bound to the P4-ATPase while the hydrophobic tails project out of the protein and remain within the lipid bilayer during the translocation process (59). It is of interest to mention that the amino acids responsible for H+ binding and transport in H+-ATPases of the P3A subfamily (i.e., a conserved aspartic acid and arginine residue) form a water-filled space inside the transporter large enough to contain a phospholipid head group (68, 75). In analogy to this transport mechanism, Coleman et al. presented evidence to suggest a similar mode of transport for PS by the P4-ATPase ATP8A2 (45). Recently, Vestergaard et al. found that I364 and adjacent hydrophobic residues are necessary for PS translocation by ATP8A2 (76). In the transition from E2-P to E2 this “hydrophobic gate” induces a sequential formation and annihilation of water-filled spaces. PS’ hydrophilic head group uses this water gradient to slide along an outlined groove in the transmembrane segments while its hydrophobic tail is pulled through the lipid bilayer.

In an extensive screen for residues in TM 1-6 in Drs2p and Dnf1p Baldridge and Graham identified two clusters of residues involved in phospholipid selection (77). One of these clusters is located in TM 1 and 2 on the exoplasmic side of the membrane and is proposed to be the “entry gate”; important for PS recognition in Drs2p and PC recognition in Dnf1p. The other cluster, proposed as the “exit gate”, is located in TM 3 and 4 near the cytosolic side of the membrane and is suggested to be involved in selection of the phospholipid before or during the E2-P to E2 transition. This “two-gate” mechanism thus proposes alternative binding sites for the phospholipid at the membrane facing surfaces of P4-ATPases.

The “two-gate” and “hydrophobic gate” models can be reconciled by imagining an initial selection of phospholipids at the entry gate followed by transport through

Chapter 2 the path outlined by the “hydrophobic gate” model. This hypothesis is also more in line with the classical alternating-access model for membrane transporters (78). For a more detailed review of the reaction cycle, see an excellent recent paper by Montigny et al. (79). Nevertheless, the elucidation of the exact mechanism of transbilayer phospholipid translocation by P4-ATPases remains a major challenge. Crystallization of P4-ATPases together with their subunit in their natural membrane environment at different phases of the reaction cycle should provide more insights into the exact mechanism.

5. Cellular Roles of P4-ATPases

Accumulating evidence obtained in yeast, worms, and plants points to an important role for P4-ATPases in the biogenesis of intracellular transport vesicles (reviewed in (46, 80)). Vesicular trafficking is a continuous cellular activity in which cells recycle a membrane area equivalent to their cell surface one to five times per hour by endocytic activity alone (81). P4-ATPases are implicated in the initiation of vesicle biogenesis. P4-ATPase-mediated flipping of a phospholipid can result in a local increase in the concentration of phospholipids in the cytoplasmic leaflet of the lipid bilayer. In turn, this will generate a cytoplasmic facing membrane curvature, which can be explained by the “bilayer couple hypothesis” (82): a sufficiently large local expansion of the number of phospholipids in one leaflet of a lipid bilayer relative to the opposite leaflet will create an increased area and force the coupled leaflet to minimize its energy state and maintain its hydrophobic interactions between the leaflets and thereby induces local bending of the bilayer. Farge et al. (83) show that exogenous addition of the aminophospholipids PS and PE, which are flipped to the cytosolic leaflet (resulting in an increase of the cytosolic surface area) of human erythroleukemia K562 cells, causes an increase in cellular endocytosis. In contrast, addition of the non-flippable PS analog lyso-alpha-phosphatidylserine does not cause an increase in endocytosis; an observation that supports the “bilayer couple hypothesis”. Thus, the induction of membrane curvature, possibly mediated by P4- ATPases, is the initiating event in the generation of transport vesicles, e.g. at the trans-Golgi network (TGN) or at the plasma membrane. Also, a local increase in PS and/or PE concentration in the cytosolic leaflet of the membrane can provide a docking platform for curvature-stabilizing and vesicle-forming proteins (84, 85). Alternatively, membrane curvature can be induced by insertion of hydrophobic proteins, through tension generating protein scaffolds such as BAR-domain proteins

Chapter 2 or force transmission from SNARE complexes (86); whether P4-ATPases or BAR- domain proteins act alone or whether it is a collaborative effort of these type of proteins to induce initial curvature is presently not known. Either way, membrane curvature is essential for the biogenesis of transport vesicles as it creates a binding scaffold for various coat proteins, e.g. clathrin, COPI, or COPII, which induces the mobilization of the vesicle-generating protein machinery (87-91).

S. cerevisiae expresses five P4-ATPases which are all involved in the biogenesis of intracellular transport vesicles in the biosynthetic and endocytic pathways (92). For the most extensively studied P4-ATPase Drs2p it has been demonstrated that it is involved in the initiation of clathrin-coated vesicle biogenesis (92-94). Drs2p, which physically interacts with adaptor protein-1 (AP-1) (94), shuttles between the TGN and early endosomes in an AP-1/clathrin-dependent pathway. Mutant drs2Δ cells are impaired in clathrin-coated vesicle formation; however, Drs2p is not essential for clathrin recruitment since in drs2Δ mutant cells clathrin and AP-1 are normally localized in the TGN (64, 93, 95). Importantly, Drs2p activity is essential for the formation of clathrin-coated vesicles, possibly by concentrating phospholipids in the cytosolic leaflet of the TGN. Thus, Drs2p- mediated lipid flipping may induce membrane curvature and drive the formation of a clathrin-coated vesicle at the TGN or endosome. In addition to AP-1, Drs2p interacts with many more proteins of the vesicle trafficking machinery. For instance, Drs2p interacts with the F-box protein Rcy1, which is involved in recycling between the endosomes and TGN (56), but Drs2p also forms a ternary complex with the Arf guanine nucleotide-exchange factor Gea2p and the Arf-like protein Arl1 in order to regulate its flippase activity in the Golgi (72, 96, 97). Recently, evidence has been presented to support a role for P4-ATPases in the early stages of membrane curvature, the initiating event in the biogenesis of a transport vesicle. Xu et al. show that stimulation of Drs2p flippase activity by Arl1p creates a local concentration of PS, which induces the formation of a negatively charged, curved membrane structure that recruits the ADP-ribolysation factor GTPase activating protein (ARF-GAP) Gcs1p, a major regulator of the vesicle-generating machinery through a curvature sensing motif (98). Gcs1p then proceeds to inactivate Arl1p via GTP hydrolysis thereby completing a regulated feedback loop of Drs2p flippase activity (99). Genetic associations with drs2 have been described for ADP-ribosylation factors (recruitment of coat protein complexes) and Rab proteins (56, 97, 100, 101). Using tandem-affinity purification and mass spectrometric analyses, Puts et al. have

Chapter 2 identified additional Drs2p-interacting proteins with a role in vesicular trafficking (100). Amongst others, three proteins involved in phosphoinositide metabolism were identified, including a PI4P phosphatase. Phosphoinositides are critical regulators of membrane and protein trafficking, mainly via binding and activation of downstream target proteins (102). PI4P is shown to be essential for the recruitment of the clathrin coat protein machinery to the TGN, and thus for the initiation of clathrin-coated vesicle generation (103, 104). Natarajan et al. show that interaction between Drs2p and PI4P is essential to catalyze NBD-PS flipping across isolated TGN membranes (72). Interestingly, removal of the cytosolic C-terminus of Drs2p causes an increased basal activity that can no longer be stimulated by PI4P, indicating an auto-inhibitory domain in the C-terminal tail that can be displaced by PI4P (105). All these observations suggest that Drs2p-catalyzed phospholipid flipping is the initiating event in the biogenesis of clathrin-coated vesicles and the mobilization of the vesicle-trafficking protein machinery in the TGN. In addition, the different subcellular distributions of the newly identified Drs2p-interacting proteins (i.e., Golgi apparatus (GA), ER, plasma membrane, and vacuole) suggest that Drs2p resides in different protein complexes within distinct trafficking pathways (100).

In yeast, the plasma membrane-associated P4-ATPases Dnf1p and Dnf2p are important in the biogenesis of endocytic transport vesicles. Dnf1Δdnf2Δ double mutant cells have a defect in the internalization step of fluid-phase endocytosis, drs2Δdnf1Δ double mutant cells are impaired in the TGN-directed transport of alkaline phosphatase while dnf1Δdnf2Δdrs2Δ triple mutants are deficient in receptor-mediated endocytosis (64, 106, 107). Importantly, the phenotypes of the dnf1Δdnf2Δ and dnf1Δdnf2Δdrs2Δ mutants coincide with plasma membrane exposition of small amounts of natural PE and PS (107). These observations justify the speculation that Dnf1p and Dnf2p flip glycerophospholipids to initiate membrane vesiculation. Alder-Baerens, et al. show that Dnf3p in isolated post-Golgi secretory vesicles is involved in the flipping of NBD-PS, PE, and PC (108). Finally, Neo1p, the only essential P4-ATPase in yeast, is implicated in the retrograde, COPI- dependent transport between Golgi apparatus and ER (109, 110).

In the plant Arabidopsis thaliana, the P4-ATPase ALA3 is involved in phospholipid flipping and vesicle biogenesis (42). Ala3 mutant plants display impaired growth and have a defect in the secretion of TGN-derived vesicles, which contain enzymes required for cell wall breakdown and release of the peripheral cell layer. This process is required for growth of roots and shoots. In a recent study by

Chapter 2 the same group, evidence is presented indicating that ALA10 participates in the uptake of exogenous phospholipids, i.e. NBD-labeled PS, PE, PC and lyso-PC (111). Since ALA10 is highly expressed in epidermal cells of the root tip the authors find that this P4-ATPase has a role in the uptake of lyso-PC, a signaling lipid in root development. Also in C. elegans P4-ATPases are involved in vesicle transport. The P4-ATPase TAT-1 has been implicated in the early steps of endocytosis in intestinal epithelial cells and oocytes and in the biogenesis of lysosomes (112). In another study both the C. elegans Cdc50 β-subunit CHAT-1 and TAT-1 itself are shown to be important in maintaining normal endocytic recycling by promoting membrane tubulation of the early endosome (35). Recently, a role for the P4-ATPase TAT-5 was suggested in the regulation of ectosome shedding, a process that is relevant for the cellular excretion of proteins, RNA, and miRNA, and is critical for the modulation of cellular processes (113).

In Drosophila melanogaster proper expression and cilial localization of Or67d olfactory receptors is necessary for pheromone sensing. In Or67d olfactory neurons Ha et al. find that flippase activity of the P4-ATPase dATP8B (the ortholog of vertebrate ATP8B1) is required for proper olfactory function (114). Expression of Or67d olfactory receptors and dATP8B is normally concentrated in the cilia, while dATP8B mutants have a severely reduced expression of Or67d receptor subunits in the cilia, which can be rescued by overexpression of bovine ATP8A2.

Also in mammalian cells accumulating evidence points towards an important role for P4-ATPases in vesicular transport. ATP11C ensures correct targeting of basolateral bile acid transporters in murine central hepatocytes (115-117). ATP11B co-localizes with cisplatin to intracellular vesicles that originated from the TGN and are bound for the plasma membrane, in that way conferring resistance to cisplatin (40). ATP10A has been implicated in the regulation of insulin-stimulated plasma membrane mobilization of the glucose uptake transporter GLUT4 from an intracellular vesicle pool (41). Verhulst et al. show impaired surface expression of apical resident proteins, including alkaline phosphatase, sucrose isomaltase, and aminopeptidase N in ATP8B1-depleted intestinal Caco-2 cells (118). In addition, ATP8B1 mediates the targeting of the apical sodium-dependent bile-acid transporter (ASBT, chapter 3 (119)) and the cystic fibrosis transmembrane receptor (CFTR, chapter 4 (120)) in human intestinal and pulmonary cells. A role for ATP8A1 and ATP8A2 in vesicle transport is proposed based on in vitro data obtained in monkey kidney fibroblast-like COS-1 cells (121). The authors show that ATP8A1, which

Chapter 2 predominantly localized to recycling endosomes, is required for the localization of EHD1, a protein that regulates traffic from recycling endosome membranes by aiding the fission of protruding tubulating structures. Furthermore, studies in human and rat pancreatic beta cells show that glucose-stimulated insulin secretion is inhibited upon knockdown of either ATP8B1, ATP8B2, ATP9A, and CDC50A (122). ATP8B1 and CDC50A are highly concentrated in PS-rich insulin-secreting granules, while ATP9A is present in the TGN and at the plasma membrane, suggesting a role for these P4-ATPases in mediating the binding and activity of positively charged SNARE proteins during exocytic vesicle fusion.

All these studies indicate that P4-ATPases are important mediators of vesicle trafficking in multiple parts of intracellular trafficking pathways, while their specific roles are probably dependent on cell type and differentiation status (Figure 4). It seems likely that, in many cases, P4-ATPase-mediated flipping causes an initial local membrane curvature that provides a scaffold for binding of proteins of the vesicle generation machinery.

Chapter 2

Figure 4. Proposed roles of P4-ATPases at the plasma membrane and in intracellular vesicle trafficking routes in S. cerevisiae, A. thaliana, C. elegans, D. melanogaster, and mammalian cells. Although many P4-ATPases are linked to intracellular trafficking defects, only a few have been specifically linked to certain organelles. S. cerevisiae Neo1p has been implicated in retrograde, COPI-dependent trafficking from the Golgi to the ER (109, 110). Drs2p in yeast is involved in the formation of AP-1/clathrin coated vesicles back and forth between the TGN and early endosomes (43, 56, 64, 93, 94). Yeast Dnf1p and Dnf2p play a role in the formation of endocytic vesicles (64, 106, 107). A. thaliana ALA3 is necessary for the synthesis of secretory vesicles from the TGN (42) and ALA10 participates in the uptake of exogenous phospholipids (111). C. elegans TAT-1 is important in maintaining normal endocytic recycling and biogenesis of lysosomes (35, 112) whereas TAT-5 is suggested to be involved in the regulation of ectosome shedding (113). D. melanogaster dATP8B is implicated in K-Ras clustering at the plasma membrane and proper cilial localization of olfactory receptors (114, 123). In CHO cells ATP8A1 plays a role in cell migration by assisting in the formation of plasma membrane ruffles, while in COS-1 cells ATP8A1 and ATP8A2 are indirectly responsible for directing traffic from recycling endosomes (121, 124). ATP8A2 promotes vesicular trafficking to retinal photoreceptor outer segments in M. musculus (125). Murine ATP8B1 is necessary for apical hepatocyte membrane integrity (126-128). ATP8B1, ATP8B2, and ATP9A are responsible for glucose-stimulated insulin secretion in human and rat pancreatic beta cells (129). ATP11B directs exocytic vesicular traffic of cisplatin- containing vesicles from the TGN to the plasma membrane in human cancer cells (130). Human and murine ATP11C are critical for the maintenance of PS asymmetry as an apoptotic signal and for the expression of basolateral bile acid transporters (115-117, 131, 132). Possible functional locations of P4-ATPases are represented by colored circles; red for S. cerevisiae, yellow for A. thaliana, green for C. elegans, pink for D. melanogaster, and blue for mammalian P4-ATPases. Insulin-secreting granule; ISG. See text for further details.

In addition, cytosolic membrane PS is essential for the clustering of signal transducers like the small GTPases K-RAS and Cdc42. Ras proteins are membrane- bound signaling proteins involved in cell proliferation and differentiation that assemble in spatially distinct nanoclusters on the plasma membrane and activate mitogen-activated protein kinase (MAPK) signaling (133). In this tightly controlled process, membrane voltage and local PS concentration seem to play crucial roles (123). Plasma membrane depolarization of baby hamster kidney cells or mouse neuroblastoma cells selectively enhances the nanoclustering and association of K- Ras with PS, as is shown by biochemical and electron microscopical analyses, resulting in downstream MAPK activation (123). Hyperpolarization, on the other hand, had the opposite effect i.e. reduced K-Ras / PS clustering and MAPK signaling. Subsequent in vivo studies in D. melanogaster deficient for the P4-ATPase dATP8B show that dATP8B is essential to induce K-Ras clustering and MAPK signaling. Possibly, depolarization of the plasma membrane leads to scrambling of PS, which is flipped back by dATP8B to induce a local concentration of PS enabling K-Ras clustering and MAPK signaling, providing a mechanism for the induction of cell signaling cascades by electrical signals.

Similarly, P4-ATPase-mediated PS concentration and concomitant Cdc42 clustering has been implicated in the establishment of enterocyte polarization (129).

Chapter 2

The authors report structural abnormalities in apical enterocyte membranes of both ATP8B1-deficient mouse intestinal organoids and intestinal biopsies of human patients. In polarized Ls174T:W4 cells, ATP8B1 depletion results in an enlarged apical surface area while in ~9% of the cells more than 1 apical domain is observed. Cdc42 knockout cells and cells expressing Cdc42(D185), which is unable to interact with PS, phenocopied ATP8B1-depleted cells. ATP8B1 depletion disturbs the interaction of Cdc42 with the apical pole leading to increased apical diffusion of Cdc42 and a consequent enlarged apical surface area.

In contrast, S. cerevisiae Dnf1p- or Dnf2p-Lem3p heterodimer-mediated flipping of PE from the exoplasmic to the cytosolic leaflet is required for the dissociation of Cdc42 from the polar cortex (134). The Cdc42 GTPase is a key regulator of cell polarity in yeast via its role in the regulation of actin polymerization. The release of Cdc42 from the cortex, which leads to a blockage of actin polymerization, triggers the growth switch of the growing bud tip of the daughter cell. Cdc42 release is initiated by locally increasing the concentration of uncharged PE versus charged PS, which results in the disruption of the charge interaction of Cdc42 with the cytosolic leaflet of the plasma membrane. Alternatively, one may speculate that Cdc42p is released from the membrane via endocytosis, initiated by a P4-ATPase-mediated increase in local concentration of PE in the cytosolic leaflet of the bilayer.

In CHO cells the ATP8A1-CDC50A heterodimer is important for cell migration (124). Cell motility occurs through reorganization of cortical actin filaments at the leading edge, which subsequently move the plasma membrane forward. In contrast to the P4-ATPase-mediated PE internalization and subsequent release of Cdc42 from the growing bud tip of the yeast cell (134), ATP8A1 activity may provide a docking platform for actin polymerization proteins such as Rac1, which localizes to membrane ruffles, where its localization is diminished in ATP8A1-depleted cells. The authors show that absence of Rac1 mobilization coincides with impaired plasma membrane PE internalization and suggests that ATP8A1 is involved in the generation of membrane ruffles resulting in cell motility.

6. The (Patho) Physiological Function of Mammalian P4-ATPases

In contrast to lower eukaryotes, much less is known about the cellular and physiological functions of mammalian P4-ATPases. The mammalian P4-ATPase subfamily consists of fourteen proteins (fifteen in mice) of which presently two are

Chapter 2 implicated in protein trafficking. Thus far, two P4-ATPases have been associated with human disease (ATP8A2 and ATP8B1); however, various in vitro studies and studies in mouse models are contributing to our understanding of the cellular and (patho) physiological functions of mammalian P4-ATPases, and will be highlighted below. The pathophysiological characteristics of P4-ATPase deficiencies in mice and humans are summarized in Table 1.

Class P4-ATPase Pathophysiology in mice Pathophysiology in humans References 1A ATP8A1 impaired learning, increased (135) physical activity ATP8A2 neurodegenerative disease, mental retardation, hypotonia, (125, 136, axonal degeneration, growth CAMRQ 137) retardation, reduced visual and auditory function 1B ATP8B1 intrahepatic cholestasis, PFIC1, BRIC1 (126, 127, hearing loss 138-140) ATP8B2 ATP8B3 sperm capacitation anomalies (141, 142) ATP8B4 Alzheimer’s disease, systemic (143, 144) sclerosis ATP8B5 not present in humans 2 ATP9A ATP9B 5 ATP10A insulin resistance, diet-induced type 2 diabetes, insulin (145-149) obesity, hyperlipidemia, resistance in African hyperinsulinemia Americans, diet-induced obesity ATP10B ATP10D diet-induced obesity, myocardial infarction, (150-152) hyperinsulinemia, artherosclerosis hyperglycemia 6 ATP11A metastasis in colorectal cancer, (153, 154) pulmonary fibrosis in non- Hispanic whites ATP11B cisplatin resistance in ovarian (130) cancer ATP11C arrested B cell development, hemolytic anemia (37, 131, dystocia, anemia, 155, 156) stomatocytosis, hepatocellular carcinoma, conjugated hyperbilirubinemia, unconjugated hypercholanemia

Table 1. Overview of mammalian P4-ATPase deficiencies and their pathophysiological characteristics in mice and humans.

Chapter 2

ATP8A1

Atp8a1 knockout mice are characterized by impaired hippocampus-dependent learning (assessed by water-avoidance experiments) and increased activity (assessed by open field testing). These behavioural problems coincide with increased PS externalization in hippocampal neurons (135). Interestingly, Atp8a1−/− erythrocytes display no PS exposure on the exoplasmic leaflet, which, according to the authors, is most likely due to compensatory expression of ATP8A2 in these cells. ATP8A1 is present in erythrocyte precursors and present on the membrane of mature erythrocytes and has a PS flippase activity (157). Taken together with the compensatory expression of ATP8A2 in Atp8a1−/− mice, both ATP8A1 and ATP8A2 are suggested to be the prime candidates for the ATP-dependent aminophospholipid translocase activity that was first discovered in erythrocytes (20). Recently, however, ATP11C has also been implicated in the maintenance of PS distribution in erythrocyte membranes (see below).

ATP8A2

Wabbler-lethal mutant mice display neurodegenerative disease and axonal degeneration in the central and peripheral nervous system. Zhu et al. identify Atp8a2 as the causative gene in these mice, which are growth retarded and do not survive beyond 4 months of age (137). Mutant animals display central chromatolysis in neuronal cells, a characteristic of axon dystrophy without cell loss. Axonal transport of phosphorylated neurofilaments is disturbed in the lumbar motor neurons, indicating a role for ATP8A2 in the vesicular, axonal transport of this protein. Additionally, Atp8a2 knockout mice have a progressive shortening and degeneration of the retinal photoreceptor outer segments, despite having a normal morphology (125). Interestingly, while Atp8a2 reconstituted in UPS-1 cells flips PS (137), in vivo internalization of exposed PS and subsequent phagocytosis of photoreceptor outer membrane segment is unaltered (125). Since ATP8A2 is located in the TGN and endosomes in the inner segments of these photoreceptors, the authors propose that the shortening is caused by diminished vesicle trafficking from the inner segment. These in vivo data suggest a role for ATP8A2 in the maintenance of axon polarity and photoreceptor integrity. In vitro data obtained in neuronal PC12 cells and rat hippocampal neurons also indicate a role for ATP8A2 (in conjunction with its β- subunit CDC50A) in promoting neurite outgrowth (158). Indeed, a mutated ATP8A2 gene has been identified in a patient with severe mental retardation and decreased

Chapter 2 muscle tone (hypotonia) (159). However, screening of an additional 37 patients with a similar phenotype did not result in the identification of ATP8A2 mutations. Additionally, a recessive missense mutation in ATP8A2 has been detected in three members of a consanguineous family affected with cerebellar ataxia, mental retardation and disequilibrium syndrome (CAMRQ) (136). Although only a small number of patients have been identified with mutations in ATP8A2 it seems likely that it is a risk factor for neurodegenerative diseases.

ATP9A

ATP9A has been implicated in the release of insulin from pancreatic beta cells. Depletion of this protein from human pancreatic beta cells (and rat pancreatic islands) results in reduced glucose-dependent insulin release by these cells (122). Otherwise, no data on this protein have been reported yet.

ATP9B

No published data are available.

ATP8B1

ATP8B1 is discussed below in detail.

ATP8B2

Like ATP9A, ATP8B2 has been implicated in the release of insulin from pancreatic beta cells. Depletion of this protein from human pancreatic beta cells (and rat pancreatic islands) results in reduced glucose-dependent insulin release by these cells (122). Otherwise, no data on this protein have been reported yet.

ATP8B3

In mice, ATP8B3 is implicated in sperm cell acrosome formation and capacitation (141, 142). The acrosome is a Golgi-derived organelle in the head of the sperm cell that contains digestive enzymes. Upon fertilization, the acrosome fuses with the sperm cell membrane to release, amongst others, hydrolases that digest the zona pellucida of the oocyte, after which the sperm cell can fuse with the oocyte. Capacitation is a cascade of membrane remodelling events in sperm cells to prepare them for penetration of the zona pellucida. Capacitation is associated with PS

Chapter 2 exposure in the sperm cell head. In contrast to control cells, Atp8b3−/− cells exposed PS in the outer membrane leaflet even before capacitation. Although the in vitro fertilization capacity of Atp8b3−/− sperm cells is reduced (due to reduced rate of zona pellucida penetration), litter sizes are not significantly reduced (142). Besides ATP8B3, mice express ATP8B5 (also termed FetA) in the acrosomal membrane of mature and developing sperm cells (62). Knockdown of Atp8b5 in a mouse mastocytoma cell line shows profound effects on Golgi structure and protein secretion. Although presently unclear, ATP8B5 may (partially) compensate for the loss of ATP8B3 expression in the acrosomal membrane.

ATP8B4

A significant association between Alzheimer’s disease and the ATP8B4 locus on chromosome 15 has been reported (143). One of the single-nucleotide polymorphisms (SNPs) described localizes close to the ATP8B4 gene. Although a follow-up on this association never occurred, this observation might suggest that mutations in ATP8B4 may predispose to Alzheimer’s disease. Recently, whole- exome sequencing of 493 patients with systemic sclerosis, a rare multisystem autoimmune disease, indicates that mutations in ATP8B4 are associated with a significant increased risk of developing this disorder (144).

ATP10A

Mice heterozygous for the P4-ATPase Atp10a (also named Atp10c) develop insulin resistance, hyperlipidemia, and are hyperinsulinemic, and provide a model for type- 2 diabetes mellitus and diet-induced obesity (145, 146). ATP10A has been implicated in the regulation of insulin-stimulated glucose uptake by plasma membrane mobilization of GLUT4- containing vesicles (145, 146, 160). Moreover, mRNA expression levels of a canine ATP10A ortholog in visceral adipose tissue have been found to be five times higher in obese dogs in comparison with lean dogs, suggesting an involvement of ATP10A in response to diet-induced obesity (161). It remains to be demonstrated if ATP10A is directly involved in insulin-stimulated mobilization of GLUT4-containing vesicles to the plasma membrane, or if ATP10A exerts its effects on glucose metabolism via regulation of insulin receptor-induced signalling. Although ATP10A has not been identified as a risk gene for type 2 diabetes in Caucasian Europeans (147), the genomic region encompassing ATP10A has been identified as a risk locus in a

Chapter 2 genome-wide association study (GWAS) of insulin resistance in an African American cohort (148). In addition, the CpG methylation state of the ATP10A gene has been found to be a good biomarker for prediction of weight-loss of obese or overweight men in response to a hypocaloric diet (149). A methylation microarray followed by mass spectrometry profiling of specific CpG methylation sites of high and low responders to a hypocaloric diet shows that a higher methylation state of ATP10A predicts a decreased response to the diet. This could indicate that an altered epigenomic regulation of ATP10A may determine the degree of insulin-resistance in humans.

ATP10B

No published data are available.

ATP10D

ATP10D has been implicated in obesity and hyperinsulinemia in mice. Atp10d is mutated (i.e., a premature stop codon in exon 12) in C57BL/6J mice (150), which have a predisposition to develop obesity, hyperglycemia, and hyperinsulinemia when placed on a high-fat diet (162). ATP10D is expressed in kidney and placenta, possibly in macrophages. At present, the relation between these phenotypes and ATP10D deficiency is not known. However, the rs2351791 ATP10D polymorphism has been associated with increased plasma HDL levels and increased artherosclerotic disease in a post-mortem study of 1397 Japanese subjects (152). An earlier GWAS shows that ATP10D polymorphisms rs10938494 and rs2351791 are associated with an increased risk of myocardial infarction and increased circulating levels of ceramide and glucosylceramide species (151). Because sphingolipids like ceramide and glucosylceramide have been linked to the pathogenesis of artherosclerosis one can speculate that ATP10D functions as a ceramide importer.

ATP11A

Recently, the ATP11A gene has been identified as a predictive marker for metastasis in colorectal cancer (CRC) (153). The authors report that ATP11A mRNA levels are significantly elevated in CRC tissue compared to control tissue. Patients expressing high ATP11A levels show reduced, disease-free survival rates, an observation on which the authors conclude that ATP11A is a good predictive marker for metastasis in CRC. How ATP11A activity correlates with metastasis in CRC remains elusive. ATP11A

Chapter 2 has also been identified as a susceptibility locus for pulmonary fibrosis in a GWAS of 1616 non-Hispanic white subjects with fibrotic idiopathic interstitial pneumonias (154).

ATP11B

In human ovarian cancer cells, increased ATP11B expression has been correlated with higher tumour grade and cisplatin resistance (130), while silencing of ATP11B in vitro or in vivo restores the sensitivity of these tumour cells to cisplatin. More detailed colocalization and cisplatin transport studies indicate a strong involvement of ATP11B in the transport of secretory cisplatin-containing vesicles from the Golgi apparatus to the plasma membrane.

ATP11C

The X-linked Atp11c mutant mouse has been characterized by an arrest in B cell development, conjugated hyperbilirubinemia, unconjugated hypercholanemia, hepatocellular carcinoma, anemia, dystocia, and stomatocytosis (37, 155, 156, 163). The relation between these diverse phenotypes and ATP11C deficiency is presently unclear. However, it is speculated that the B cell lymphopenia is caused by a defect in the transition from the pro- to the pre-B cell stage (37). This transition requires the clathrin-mediated endocytosis of ligand-bound interleukin-7 receptor (IL-7R). ATP11C deficient pro-B cells display higher surface expression of IL-7R, which may be caused by impaired internalization of the IL-7 bound IL-7R and subsequent impaired signalling. ATP11C may be important in the clathrin-mediated endocytosis of the activated IL-7R. In humans, ATP11C has been mapped to Xq27, a genomic region associated with X-linked inherited disorders, including hypoparathyroidism, albinism-deafness, and thoracoabdominal syndrome (164). If and how ATP11C deficiency is associated with these phenotypes remains to be demonstrated.

Very recently a patient bearing the ATP11C C1253A mutation has been identified with symptoms of mild congenital hemolytic anemia (131). Flippase activity in erythrocytes of this patient is reduced 10-fold compared to control, with residual activity possibly caused by ATP8A1, ATP8A2, and ATP11A. The proportion of PS-exposing erythrocytes is limited, but sufficient to cause a persistent and mild hemolysis accompanied by mild jaundice. PS translocation in mature

Chapter 2 murine Atp11c−/− erythrocytes is also reduced, resulting in anemia and an abnormal erythrocyte stomatocyte shape (163).

The etiology of the conjugated hyperbilirubinemia and unconjugated hypercholanemia in ATP11C-deficient mice has recently been elucidated (115-117). ATP11C localizes to the basolateral membranes of central hepatocytes in control mice. However, in ATP11C-deficient mice, basolateral expression of all bile salt uptake transporters, including NTCP (SLC10A1) and all OATP family members is absent only from central hepatocytes, while apical bile salt transport is unaffected.

Activation of Ca2+-dependent scramblases and inhibition of inward PS transport cause PS exposure during apoptosis, which is an important signal for phagocytosis by macrophages (165). Until recently, the responsible flippase had not been identified. However, in a study by Segawa et al., mouse W3 ganglion-cells, which undergo apoptosis upon treatment with Fas ligand, show cleavage of ATP11C by caspases, followed by exoplasmic PS exposure and macrophage engulfment (132). In addition, caspase-resistant ATP11C rescues PS transport in ATP11C-depleted cells and prevents PS exposure after treatment with Fas ligand, indicating that activation of scramblases and inhibition of flippases are equally important for apoptotic PS exposure.

7. ATP8B1

ATP8B1 is the most extensively studied human P4-ATPase. Mutations in ATP8B1 cause progressive familial intrahepatic cholestasis type 1 (PFIC1) and benign recurrent intrahepatic cholestasis type 1 (BRIC1) (166, 167), two liver disorders which were first described by Clayton et al. (168) and Summerskill and Walshe (169) (reviewed in (138)). PFIC1 and the less severe BRIC1 have an onset during the early stages of life and are considered as two ends of a continuum of liver disease, characterized by impaired bile flow (i.e., cholestasis). The exact prevalence of PFIC1 or BRIC1 is unknown, but for PFIC1 is estimated to be 1 in 100,000−900,000 (170, 171). PFIC1 is characterized by fierce pruritus, fat malabsorption, failure to thrive, and progressive liver disease leading to fibrosis and cirrhosis. BRIC1 patients suffer from intermittent bouts of cholestasis and pruritus and sustain less pronounced liver damage (172). Liver transplantation or bile-diverting procedures during childhood are necessary for most PFIC1 and BRIC1 patients who develop a chronic cholestasis (173, 174). PFIC1 and BRIC1 patients can also develop extrahepatic disease, such

Chapter 2 as diarrhea, hearing loss, pancreatitis, rickets, and pneumonia (140, 175-177). There is evidence for clinical heterogeneity in the group of PFIC1 patients, but apart from differences in severity related to the type of mutation, the background of this heterogeneity is not known (167, 178). After liver transplantation or biliary diversion, PFIC1 patients usually suffer from an exacerbated form of diarrhea, which can be ameliorated by cholestyramine, a bile salt binding resin. This suggests that the intestine of PFIC1 patients has increased sensitivity for a restoration of bile salt secretion. Interestingly, PFIC1 patients who received a liver transplant develop liver steatosis the etiology of which is unknown (173, 175, 179).

ATP8B1 is expressed in many tissues, including the liver, pancreas, small intestine, bladder, stomach, and prostate, and localizes to the apical membrane of many epithelial cells, including the canalicular membrane of hepatocytes (39, 180, 181). Our laboratory has extensively studied the etiology of ATP8B1-associated cholestasis using knock-in mice for a prototypic PFIC1 mutation, a glycine-to-valine substitution at amino acid 308 (G308V) that results in near complete absence of the protein (127, 139, 182). We and others have shown that the apical membrane of ATP8B1-deficient hepatocytes is sensitive to bile salt-induced membrane damage evidenced by extraction of membrane components, including cholesterol and PS (126, 127). From our observations we have hypothesized that ATP8B1 is important for reducing the outer leaflet content of PS, so as to increase the relative sphingomyelin content, which together with cholesterol forms a rigid, liquid-ordered membrane that is resistant towards detergents such as bile salts. ATP8B1-deficiency thus leads to loss of normal phospholipid asymmetry of the canalicular membrane. As a result the canalicular membrane becomes more sensitive to extraction of cholesterol by hydrophobic bile salts, which impairs the activity of the major bile salt transporter ABCB11 and, as a consequence, causes cholestasis (128) (reviewed in (138)). Enhanced biliary cholesterol output in Atp8b1-deficient mice turns out to be independent of the cholesterol transporter ABCG5/G8, since Atp8b1- deficient mice with inactivated ABCG5/G8 also display enhanced cholesterol output (183). In the same mouse model a progressive degeneration of the cochlear hair cells is present, explaining the hearing-loss in PFIC1 and BRIC1 patients (140).

Recently, a role for ATP8B1 in pulmonary cardiolipin uptake has been suggested (36). Using ATP8B1 deficient mice, the authors show that pulmonary infection in these mice is associated with elevated cardiolipin levels. They conclude that enhanced pulmonary cardiolipin in ATP8B1-deficiency impairs lung and lung

Chapter 2 surfactant function, and may be the underlying cause of the pulmonary problems observed in some PFIC1 patients. The authors suggest that ATP8B1 is a flippase for cardiolipin, however, the mechanism of pulmonary cardiolipin accumulation and the role of ATP8B1 therein is not illuminated (184). Recently, we have shown that CFTR surface expression is strongly impaired in ATP8B1-depleted lung pulmonary cells (chapter 4 (120)), which provides an alternative explanation for the pulmonary (and other extrahepatic) problems observed in PFIC1 patients.

An alternative hypothesis for the etiology of ATP8B1-associated cholestasis is based on findings that ATP8B1 deficiency impairs farnesoid X Receptor (FXR)- dependent signaling. FXR is a bile salt sensing receptor that regulates bile salt homeostasis in hepatocytes, by inactivating bile salt synthesis and by upregulating ABCB11. Using CHO cells, Caco-2 cells, and human hepatocytes, Chen et al. show that ATP8B1 depletion results in diminished FXR signaling caused by impaired PKCζ-mediated phosphorylation of cytosolic FXR and subsequent absence of nuclear translocation (185-187). Inhibition of PKCζ activity coincides with reduced ABCB11 promoter activation, whereas PKCζ overexpression results in activation of ABCB11 promoter activity. Based on these findings, the authors hypothesize that ATP8B1 activity is essential for PKCζ activation, which leads to FXR phosphorylation, nuclear translocation, and transcriptional regulation of target genes. In addition, transient overexpression of mutant ATP8B1 in HepG2 cells causes a decrease in FXR luciferase reporter activity when compared to wild type ATP8B1-induced activity (188). Thus far, however, no experimental evidence is published to show that ATP8B1-mediated PS flipping is essential for PKCζ activation. Furthermore and in contrast to Chen et al., Cai et al., using Caco-2 cells and human and rat hepatocytes, show that ATP8B1 depletion actually leaves FXR activity unaffected (126).

Other than a crucial role in hepatocanalicular membrane asymmetry, ATP8B1 may also have a role in membrane trafficking and vesicular transport. Verhulst et al. observe that knockdown of ATP8B1 in Caco-2 cells leads to a loss in microvilli, an unorganized apical actin cytoskeleton and a posttranscriptional defect in apical protein expression (118). It seems unlikely that a similar dramatic impairment of apical membrane assembly of intestinal epithelial cells also occurs in patients with PFIC1. There is, however, clearly room for a more subtle defect in this process as there is an intestinal phenotype in these patients. In the study by Verhulst et al. ATP8B1 depletion by RNAi does not cause a reduction in PS transport in the apical

Chapter 2 membrane (118). This discrepancy could be explained by the fact that PS flipping by ATP8B1 only represents a small fraction of total flippase activity in the plasma membrane, that ATP8B1 plays a more important role in intracellular vesicular trafficking, and/or that ATP8B1 flips PS to a lesser extent than previously assumed. ATP8B1 substrate specificity has recently been challenged by Takatsu et al. (189). In this study human cervical cancer HeLa and CHO-K1 cells overexpressing HA- tagged P4-ATPases are evaluated for transporting NBD-labeled PC, PE, or PS at the plasma membrane. While ATP11A and ATP11C import both PE and PS, ATP8B1, ATP8B2, and ATP10A import only PC (189, 190). Overexpression of ATP8B1 in CHO-K1 cells without CDC50A results in plasma membrane surface expression of ATP8B1 (189). This is contrary to earlier work by Paulusma et al. where only upon overexpression of both ATP8B1 and CDC50A in CHO-K1 and UPS-1 cells (mutagenized CHO-K1 cells deficient for ATP11C (191)) ATP8B1 localizes to the plasma membrane (39). In these overexpressing cells NBD-PS internalization is increased while endogenous PS flipping is increased as measured by annexin V labeling.

Currently, a number of novel treatments correcting the function and expression of ATP8B1 have been analyzed. One promising method is the correction of aberrant splicing of mutant ATP8B1 pre-mRNA. By expressing modified U1 small nuclear RNAs complementary to mutated splice donor sites, correct splicing of several of these aberrantly spliced ATP8B1 pre-mRNAs can be restored in human bone osteosarcoma epithelial U2OS cells (192). Similar to PFIC1, many of the mutations causing cystic fibrosis lead to misfolding of CFTR, and a subsequent disturbed apical plasma membrane localization. In a study by Van der Woerd et al. apical membrane targeting of I661T-ATP8B1 was partly restored in vitro by the use of several typical CFTR correctors such as 4-phenyl butyric acid (4PB) and suberoylanilide hydroxamic acid (193).

Recently, 4BP has been successfully used to treat several cholestatic PFIC2 patients, a disease characterized by mutations in the bile salt exporter ABCB11, by partially restoring canalicular localization of ABCB11 (194). 4BP is now also sparingly used to treat PFIC1: gradually increasing doses of 4PB in three Japanese PFIC1 patients resulted in a significant reduction of intractable itch, despite a lack of improvement in liver biochemical or histological parameters (195). Although membrane expression of ABCB11 is increased in these patients, expression of ATP8B1 was not, indicating a lack of correction in folding and/or trafficking of these

Chapter 2 mutated proteins. In addition, care should be taken by administrating 4BP in cholestatic PFIC1 patients undergoing standard pruritus-relieving treatment with rifampin. In a case study reported by Shneider et al., withdrawal of rifampin during 4BP treatment of a 4-year old PFIC1 patient resulted in severe acute liver injury (196).

8. Concluding Remarks

P4-ATPases serve important functions in cellular physiology. Accumulating evidence obtained in lower eukaryotes point to an important role for P4-ATPases in initiating membrane vesicle biogenesis, which has important implications for P4- ATPases in polarized membrane and protein transport. As of yet, only one study in S. cerevisiae has shown that P4-ATPases can induce membrane curvature through binding of a curvature sensing polypeptide motif. In addition, P4-ATPases are important in maintaining an optimal physical state of the membrane, which is essential for proper membrane barrier and membrane protein function. Obviously, the physiological and cellular functions of most mammalian P4-ATPases are only slowly emerging. Studies in cell lines suggest important functions for P4-ATPases in transport and signalling processes. In organelles where membrane asymmetry has been established (i.e., the late secretory and endocytic pathways), studies in S. cerevisiae, A. thaliana and C. elegans, and multiple mammalian cell types have shown the importance of P4-ATPases in intracellular trafficking. Specifically, most of the P4-ATPases studied seem to be involved in transport between the TGN and early endosomes, exocytosis, endocytosis, or endocytic recycling. Live-cell imaging studies and tracking of P4-ATPases inside the cell could perhaps reveal an even clearer picture regarding the intracellular pathways they are involved in. Multiple mouse models are presently available and although their phenotype is not always reconciled with results in human cells, many of them do show that P4-ATPases fulfil multiple important physiological functions. Deficiencies result in a wide variety of neurological phenotypes, liver disease, immunological problems, and type 2 diabetes and diet-induced obesity. How all these distinct phenotypes relate to a defective flippase activity remains to be elucidated.

The existence of a large family of functionally similar flippases suggests there may be a role for redundancy amongst the P4-ATPases. Nevertheless, human ATP8B1- and ATP8A2-related disease, and mouse models indicate that, if there is redundancy, it is limited as there is room for severe phenotypes. For example; despite

Chapter 2 the compensatory upregulation of ATP8A2 in ATP8A1 deficient erythrocytes, ATP8A1 deficient mice still present neurological problems. An increasing number of GWAS reports have mentioned increased susceptibilities to several P4-ATPases: ATP8B4 (Alzheimer’s disease and systemic sclerosis), ATP10A (insulin resistance in African Americans), ATP10D (artherosclerosis), and ATP11A (pulmonary fibrosis). In addition, in recent years several interesting case reports have been published on patients with congenital defects in ATP8A2 (CAMRQ) and ATP11C (haemolytic anemia). Although future GWAS analyses and case reports may reveal even more P4-ATPases as risk genes, it may be that humans P4-ATPases encompass critical activities of which deficiencies are embryonic lethal, or that diseases associated with P4-ATPases are just rare (as is the case for PFIC1/BRIC1 disease). Moreover, P4-ATPases may play an important part in the treatment of cancer given their role in intracellular trafficking and the recently identified resistance to the anticancer drug cisplatin due to ATP11B induction. In contrast, P4-ATPases can also mediate uptake of certain drugs as CDC50A has been found to be responsible for the uptake of the anti-cancer lysophospholipid perifosine in human epidermal carcinoma KB cells (197). Therefore, in vitro studies and studies in mouse models will be crucial to identify the cellular and (patho) physiological functions of P4-ATPases and will reveal yet undiscovered P4-ATPase-related diseases.

References

1. Gorter E, Grendel F. On bimolecular layers of lipoids on the chromocytes of the blood. JExpMed. 1925;41(4):439-43. 2. van Meer G, Voelker DR, Feigenson GW. Membrane lipids: where they are and how they behave. NatRevMolCell Biol. 2008;9(2):112-24. 3. van Meer G, de Kroon AI. Lipid map of the mammalian cell. JCell Sci. 2011;124(Pt 1):5-8. 4. Bevers EM, Williamson PL. Phospholipid scramblase: an update. FEBS Lett. 2010;584(13):2724-30. 5. Leonard TA, Hurley JH. Regulation of protein kinases by lipids. CurrOpinStructBiol. 2011;21(6):785-91. 6. Leventis PA, Grinstein S. The distribution and function of phosphatidylserine in cellular membranes. AnnuRevBiophys. 2010;39:407-27. 7. Bretscher MS. Asymmetrical lipid bilayer structure for biological membranes. NatNew Biol. 1972;236(61):11-2. 8. Gordesky SE, Marinetti GV. The asymetric arrangement of phospholipids in the human erythrocyte membrane. BiochemBiophysResCommun. 1973;50(4):1027-31. 9. Verkleij AJ, Zwaal RF, Roelofsen B, Comfurius P, Kastelijn D, van Deenen LL. The asymmetric distribution of phospholipids in the human red cell membrane. A combined study using phospholipases and freeze-etch electron microscopy. BiochimBiophysActa. 1973;323(2):178- 93. 10. Butikofer P, Lin ZW, Chiu DT, Lubin B, Kuypers FA. Transbilayer distribution and mobility of phosphatidylinositol in human red blood cells. JBiolChem. 1990;265(27):16035-8. 11. Gascard P, Tran D, Sauvage M, Sulpice JC, Fukami K, Takenawa T, et al. Asymmetric distribution of phosphoinositides and phosphatidic acid in the human erythrocyte membrane. BiochimBiophysActa. 1991;1069(1):27-36. 12. Op den Kamp JA. Lipid asymmetry in membranes. AnnuRevBiochem. 1979;48:47-71. 13. Maxfield FR, van MG. Cholesterol, the central lipid of mammalian cells. CurrOpinCell Biol. 2010;22(4):422-9. 14. de Almeida RF, Fedorov A, Prieto M. Sphingomyelin/phosphatidylcholine/cholesterol phase diagram: boundaries and composition of lipid rafts. BiophysJ. 2003;85(4):2406-16. 15. Mondal M, Mesmin B, Mukherjee S, Maxfield FR. Sterols are mainly in the cytoplasmic leaflet of the plasma membrane and the endocytic recycling compartment in CHO cells. MolBiolCell. 2009;20(2):581-8. 16. Mesmin B, Maxfield FR. Intracellular sterol dynamics. BiochimBiophysActa. 2009;1791(7):636-45. 17. Sprong H, van der Sluijs P, van MG. How proteins move lipids and lipids move proteins. NatRevMolCell Biol. 2001;2(7):504-13. 18. Ganong BR, Bell RM. Transmembrane movement of phosphatidylglycerol and diacylglycerol sulfhydryl analogues. Biochemistry. 1984;23(21):4977-83. 19. Holthuis JC, Levine TP. Lipid traffic: floppy drives and a superhighway. NatRevMolCell Biol. 2005;6(3):209-20. 20. Seigneuret M, Devaux PF. ATP-dependent asymmetric distribution of spin-labeled phospholipids in the erythrocyte membrane: relation to shape changes. ProcNatlAcadSciUSA. 1984;81(12):3751-5. 21. Suzuki J, Umeda M, Sims PJ, Nagata S. Calcium-dependent phospholipid scrambling by TMEM16F. Nature. 2010;468(7325):834-8.

Chapter 2

22. Yang H, Kim A, David T, Palmer D, Jin T, Tien J, et al. TMEM16F forms a Ca2+-activated cation channel required for lipid scrambling in platelets during blood coagulation. Cell. 2012;151(1):111-22. 23. Suzuki J, Denning DP, Imanishi E, Horvitz HR, Nagata S. Xk-related protein 8 and CED-8 promote phosphatidylserine exposure in apoptotic cells. Science. 2013;341(6144):403-6. 24. Suzuki J, Imanishi E, Nagata S. Exposure of phosphatidylserine by Xk-related protein family members during apoptosis. J Biol Chem. 2014;289(44):30257-67. 25. Fujii T, Sakata A, Nishimura S, Eto K, Nagata S. TMEM16F is required for phosphatidylserine exposure and microparticle release in activated mouse platelets. Proc Natl Acad Sci U S A. 2015;112(41):12800-5. 26. Paulusma CC, Oude Elferink RP. Diseases of intramembranous lipid transport. FEBS Lett. 2006;580(23):5500-9. 27. Pohl A, Devaux PF, Herrmann A. Function of prokaryotic and eukaryotic ABC proteins in lipid transport. BiochimBiophysActa. 2005;1733(1):29-52. 28. Pomorski T, Holthuis JC, Herrmann A, van MG. Tracking down lipid flippases and their biological functions. JCell Sci. 2004;117(Pt 6):805- 13. 29. Bublitz M, Morth JP, Nissen P. P-type ATPases at a glance. JCell Sci. 2011;124(Pt 15):2515-9. 30. Schatzmann HJ. ATP-dependent Ca++-extrusion from human red cells. Experientia. 1966;22(6):364-5. 31. Skou JC. The influence of some cations on an adenosine triphosphatase from peripheral nerves. BiochimBiophysActa. 1957;23(2):394-401. 32. Palmgren MG, Nissen P. P-type ATPases. AnnuRevBiophys. 2011;40:243-66. 33. Natarajan P, Wang J, Hua Z, Graham TR. Drs2p-coupled aminophospholipid translocase activity in yeast Golgi membranes and relationship to in vivo function. ProcNatlAcadSciUSA. 2004;101(29):10614-9. 34. Tang X, Halleck MS, Schlegel RA, Williamson P. A subfamily of P-type ATPases with aminophospholipid transporting activity. Science. 1996;272(5267):1495-7. 35. Chen B, Jiang Y, Zeng S, Yan J, Li X, Zhang Y, et al. Endocytic sorting and recycling require membrane phosphatidylserine asymmetry maintained by TAT-1/CHAT-1. PLoSGenet. 2010;6(12):e1001235. 36. Ray NB, Durairaj L, Chen BB, McVerry BJ, Ryan AJ, Donahoe M, et al. Dynamic regulation of cardiolipin by the lipid pump Atp8b1 determines the severity of lung injury in experimental pneumonia. NatMed. 2010;16(10):1120-7. 37. Yabas M, Teh CE, Frankenreiter S, Lal D, Roots CM, Whittle B, et al. ATP11C is critical for the internalization of phosphatidylserine and differentiation of B lymphocytes. NatImmunol. 2011;12(5):441-9. 38. Coleman JA, Kwok MC, Molday RS. Localization, purification, and functional reconstitution of the P4-ATPase Atp8a2, a phosphatidylserine flippase in photoreceptor disc membranes. JBiolChem. 2009;284(47):32670-9. 39. Paulusma CC, Folmer DE, Ho-Mok KS, de Waart DR, Hilarius PM, Verhoeven AJ, et al. ATP8B1 requires an accessory protein for endoplasmic reticulum exit and plasma membrane lipid flippase activity. Hepatology. 2008;47(1):268-78. 40. Weingartner A, Drobot B, Herrmann A, Sanchez-Canete MP, Gamarro F, Castanys S, et al. Disruption of the lipid-transporting LdMT- LdRos3 complex in Leishmania donovani affects membrane lipid asymmetry but not host cell invasion. PLoSOne. 2010;5(8):e12443. 41. Lopez-Marques RL, Poulsen LR, Hanisch S, Meffert K, Buch-Pedersen MJ, Jakobsen MK, et al. Intracellular targeting signals and lipid specificity determinants of the ALA/ALIS P4-ATPase complex reside in the catalytic ALA alpha-subunit. MolBiolCell. 2010;21(5):791- 801. 42. Poulsen LR, Lopez-Marques RL, McDowell SC, Okkeri J, Licht D, Schulz A, et al. The Arabidopsis P4-ATPase ALA3 localizes to the golgi and requires a beta-subunit to function in lipid translocation and secretory vesicle formation. Plant Cell. 2008;20(3):658-76. 43. Chen S, Wang J, Muthusamy BP, Liu K, Zare S, Andersen RJ, et al. Roles for the Drs2p-Cdc50p complex in protein transport and phosphatidylserine asymmetry of the yeast plasma membrane. Traffic. 2006;7(11):1503-17. 44. Coleman JA, Molday RS. Critical role of the beta-subunit CDC50A in the stable expression, assembly, subcellular localization, and lipid transport activity of the P4-ATPase ATP8A2. JBiolChem. 2011;286(19):17205-16. 45. Coleman JA, Vestergaard AL, Molday RS, Vilsen B, Peter AJ. Critical role of a transmembrane lysine in aminophospholipid transport by mammalian photoreceptor P4-ATPase ATP8A2. ProcNatlAcadSciUSA. 2012;109(5):1449-54. 46. Paulusma CC, Elferink RP. P4 ATPases--the physiological relevance of lipid flipping transporters. FEBS Lett. 2010;584(13):2708-16. 47. Coleman JA, Quazi F, Molday RS. Mammalian P(4)-ATPases and ABC transporters and their role in phospholipid transport. BiochimBiophysActa. 2013;1831(3):555-74. 48. Puts CF, Holthuis JC. Mechanism and significance of P4 ATPase-catalyzed lipid transport: lessons from a Na+/K+-pump. BiochimBiophysActa. 2009;1791(7):603-11. 49. Shin JM, Munson K, Vagin O, Sachs G. The gastric HK-ATPase: structure, function, and inhibition. Pflugers Arch. 2009;457(3):609-22. 50. Poulsen LR, Lopez-Marques RL, Palmgren MG. Flippases: still more questions than answers. Cell MolLife Sci. 2008;65(20):3119-25. 51. Blanco G. Na,K-ATPase subunit heterogeneity as a mechanism for tissue-specific ion regulation. SeminNephrol. 2005;25(5):292-303. 52. Geering K. Functional roles of Na,K-ATPase subunits. CurrOpinNephrolHypertens. 2008;17(5):526-32. 53. Muth TR, Gottardi CJ, Roush DL, Caplan MJ. A basolateral sorting signal is encoded in the alpha-subunit of Na-K-ATPase. AmJPhysiol. 1998;274(3 Pt 1):C688-C96. 54. Vagin O, Turdikulova S, Sachs G. The H,K-ATPase beta subunit as a model to study the role of N-glycosylation in membrane trafficking and apical sorting. JBiolChem. 2004;279(37):39026-34. 55. Bryde S, Hennrich H, Verhulst PM, Devaux PF, Lenoir G, Holthuis JC. CDC50 proteins are critical components of the human class-1 P4- ATPase transport machinery. JBiolChem. 2010;285(52):40562-72. 56. Furuta N, Fujimura-Kamada K, Saito K, Yamamoto T, Tanaka K. Endocytic recycling in yeast is regulated by putative phospholipid translocases and the Ypt31p/32p-Rcy1p pathway. MolBiolCell. 2007;18(1):295-312. 57. Saito K, Fujimura-Kamada K, Furuta N, Kato U, Umeda M, Tanaka K. Cdc50p, a protein required for polarized growth, associates with the Drs2p P-type ATPase implicated in phospholipid translocation in Saccharomyces cerevisiae. MolBiolCell. 2004;15(7):3418-32. 58. van der Velden LM, Wichers CG, van Breevoort AE, Coleman JA, Molday RS, Berger R, et al. Heteromeric interactions required for abundance and subcellular localization of human CDC50 proteins and class 1 P4-ATPases. JBiolChem. 2010;285(51):40088-96. 59. Baldridge RD, Graham TR. Identification of residues defining phospholipid flippase substrate specificity of type IV P-type ATPases. ProcNatlAcadSciUSA. 2012;109(6):E290-E8. 60. Folmer DE, Mok KS, de Wee SW, Duijst S, Hiralall JK, Seppen J, et al. Cellular localization and biochemical analysis of mammalian CDC50A, a glycosylated beta-subunit for P4 ATPases. JHistochemCytochem. 2012;60(3):205-18. 61. Katoh Y, Katoh M. Identification and characterization of CDC50A, CDC50B and CDC50C genes in silico. OncolRep. 2004;12(4):939-43. 62. Xu P, Okkeri J, Hanisch S, Hu RY, Xu Q, Pomorski TG, et al. Identification of a novel mouse P4-ATPase family member highly expressed during spermatogenesis. JCell Sci. 2009;122(Pt 16):2866-76. 63. Takatsu H, Baba K, Shima T, Umino H, Kato U, Umeda M, et al. ATP9B, a P4-ATPase (a putative aminophospholipid translocase), localizes to the trans-Golgi network in a CDC50 protein-independent manner. JBiolChem. 2011;286(44):38159-67. 64. Hua Z, Fatheddin P, Graham TR. An essential subfamily of Drs2p-related P-type ATPases is required for protein trafficking between Golgi complex and endosomal/vacuolar system. MolBiolCell. 2002;13(9):3162-77.

Chapter 2

65. Stone A, Chau C, Eaton C, Foran E, Kapur M, Prevatt E, et al. Biochemical characterization of P4-ATPase mutations identified in patients with progressive familial intrahepatic cholestasis. JBiolChem. 2012. 66. Jacquot A, Montigny C, Hennrich H, Barry R, M. lM, Jaxel C, et al. Phosphatidylserine stimulation of Drs2p.Cdc50p lipid translocase dephosphorylation is controlled by phosphatidylinositol-4-phosphate. JBiolChem. 2012;287(16):13249-61. 67. Lenoir G, Williamson P, Puts CF, Holthuis JC. Cdc50p plays a vital role in the ATPase reaction cycle of the putative aminophospholipid transporter Drs2p. JBiolChem. 2009;284(27):17956-67. 68. Stone A, Williamson P. Outside of the box: recent news about phospholipid translocation by P4 ATPases. JChemBiol. 2012;5(4):131-6. 69. Albers RW. Biochemical aspects of active transport. AnnuRevBiochem. 1967;36:727-56. 70. Post RL, Hegyvary C, Kume S. Activation by adenosine triphosphate in the phosphorylation kinetics of sodium and potassium ion transport adenosine triphosphatase. JBiolChem. 1972;247(20):6530-40. 71. Durr KL, Tavraz NN, Dempski RE, Bamberg E, Friedrich T. Functional significance of E2 state stabilization by specific alpha/beta-subunit interactions of Na,K- and H,K-ATPase. JBiolChem. 2009;284(6):3842-54. 72. Natarajan P, Liu K, Patil DV, Sciorra VA, Jackson CL, Graham TR. Regulation of a Golgi flippase by phosphoinositides and an ArfGEF. NatCell Biol. 2009;11(12):1421-6. 73. Azouaoui H, Montigny C, Ash MR, Fijalkowski F, Jacquot A, Gronberg C, et al. A high-yield co-expression system for the purification of an intact Drs2p-Cdc50p lipid flippase complex, critically dependent on and stabilized by phosphatidylinositol-4-phosphate. PLoS One. 2014;9(11):e112176. 74. Nakano K, Yamamoto T, Kishimoto T, Noji T, Tanaka K. Protein kinases Fpk1p and Fpk2p are novel regulators of phospholipid asymmetry. MolBiolCell. 2008;19(4):1783-97. 75. Pedersen BP, Buch-Pedersen MJ, Morth JP, Palmgren MG, Nissen P. Crystal structure of the plasma membrane proton pump. Nature. 2007;450(7172):1111-4. 76. Vestergaard AL, Coleman JA, Lemmin T, Mikkelsen SA, Molday LL, Vilsen B, et al. Critical roles of isoleucine-364 and adjacent residues in a hydrophobic gate control of phospholipid transport by the mammalian P4-ATPase ATP8A2. Proc Natl Acad Sci U S A. 2014;111(14):E1334-43. 77. Baldridge RD, Graham TR. Two-gate mechanism for phospholipid selection and transport by type IV P-type ATPases. ProcNatlAcadSciUSA. 2013;110(5):E358-E67. 78. Jardetzky O. Simple allosteric model for membrane pumps. Nature. 1966;211(5052):969-70. 79. Montigny C, Lyons J, Champeil P, Nissen P, Lenoir G. On the molecular mechanism of flippase- and scramblase-mediated phospholipid transport. Biochim Biophys Acta. 2015. 80. Sebastian TT, Baldridge RD, Xu P, Graham TR. Phospholipid flippases: building asymmetric membranes and transport vesicles. BiochimBiophysActa. 2012;1821(8):1068-77. 81. Steinman RM, Mellman IS, Muller WA, Cohn ZA. Endocytosis and the recycling of plasma membrane. JCell Biol. 1983;96(1):1-27. 82. Sheetz MP, Singer SJ. Biological membranes as bilayer couples. A molecular mechanism of drug-erythrocyte interactions. ProcNatlAcadSciUSA. 1974;71(11):4457-61. 83. Farge E, Ojcius DM, Subtil A, Dautry-Varsat A. Enhancement of endocytosis due to aminophospholipid transport across the plasma membrane of living cells. AmJPhysiol. 1999;276(3 Pt 1):C725-C33. 84. Graham TR. Flippases and vesicle-mediated protein transport. Trends Cell Biol. 2004;14(12):670-7. 85. Yeung T, Gilbert GE, Shi J, Silvius J, Kapus A, Grinstein S. Membrane phosphatidylserine regulates surface charge and protein localization. Science. 2008;319(5860):210-3. 86. Kozlov MM, McMahon HT, Chernomordik LV. Protein-driven membrane stresses in fusion and fission. Trends BiochemSci. 2010;35(12):699-706. 87. Gillon AD, Latham CF, Miller EA. Vesicle-mediated ER export of proteins and lipids. BiochimBiophysActa. 2012;1821(8):1040-9. 88. Horvath CA, Vanden Broeck D, Boulet GA, Bogers J, De Wolf MJ. Epsin: inducing membrane curvature. IntJBiochemCell Biol. 2007;39(10):1765-70. 89. McMahon HT, Boucrot E. Molecular mechanism and physiological functions of clathrin-mediated endocytosis. NatRevMolCell Biol. 2011;12(8):517-33. 90. Pinot M, Goud B, Manneville JB. Physical aspects of COPI vesicle formation. MolMembrBiol. 2010;27(8):428-42. 91. Robinson MS. Adaptable adaptors for coated vesicles. Trends Cell Biol. 2004;14(4):167-74. 92. Muthusamy BP, Natarajan P, Zhou X, Graham TR. Linking phospholipid flippases to vesicle-mediated protein transport. BiochimBiophysActa. 2009;1791(7):612-9. 93. Gall WE, Geething NC, Hua Z, Ingram MF, Liu K, Chen SI, et al. Drs2p-dependent formation of exocytic clathrin-coated vesicles in vivo. CurrBiol. 2002;12(18):1623-7. 94. Liu K, Surendhran K, Nothwehr SF, Graham TR. P4-ATPase requirement for AP-1/clathrin function in protein transport from the trans- Golgi network and early endosomes. MolBiolCell. 2008;19(8):3526-35. 95. Chen CY, Ingram MF, Rosal PH, Graham TR. Role for Drs2p, a P-type ATPase and potential aminophospholipid translocase, in yeast late Golgi function. JCell Biol. 1999;147(6):1223-36. 96. Chantalat S, Park SK, Hua Z, Liu K, Gobin R, Peyroche A, et al. The Arf activator Gea2p and the P-type ATPase Drs2p interact at the Golgi in Saccharomyces cerevisiae. JCell Sci. 2004;117(Pt 5):711-22. 97. Tsai PC, Hsu JW, Liu YW, Chen KY, Lee FJ. Arl1p regulates spatial membrane organization at the trans-Golgi network through interaction with Arf-GEF Gea2p and flippase Drs2p. ProcNatlAcadSciUSA. 2013;110(8):E668-E77. 98. Xu P, Baldridge RD, Chi RJ, Burd CG, Graham TR. Phosphatidylserine flipping enhances membrane curvature and negative charge required for vesicular transport. JCell Biol. 2013;202(6):875-86. 99. Hsu JW, Chen ZJ, Liu YW, Lee FJ. Mechanism of action of the flippase Drs2p in modulating GTP hydrolysis of Arl1p. J Cell Sci. 2014;127(Pt 12):2615-20. 100. Puts CF, Lenoir G, Krijgsveld J, Williamson P, Holthuis JC. A P4-ATPase protein interaction network reveals a link between aminophospholipid transport and phosphoinositide metabolism. JProteomeRes. 2010;9(2):833-42. 101. Sakane H, Yamamoto T, Tanaka K. The functional relationship between the Cdc50p-Drs2p putative aminophospholipid translocase and the Arf GAP Gcs1p in vesicle formation in the retrieval pathway from yeast early endosomes to the TGN. Cell StructFunct. 2006;31(2):87- 108. 102. Mayinger P. Phosphoinositides and vesicular membrane traffic. BiochimBiophysActa. 2012;1821(8):1104-13. 103. Graham TR, Burd CG. Coordination of Golgi functions by phosphatidylinositol 4-kinases. Trends Cell Biol. 2011;21(2):113-21. 104. Wang YJ, Wang J, Sun HQ, Martinez M, Sun YX, Macia E, et al. Phosphatidylinositol 4 phosphate regulates targeting of clathrin adaptor AP-1 complexes to the Golgi. Cell. 2003;114(3):299-310. 105. Zhou X, Sebastian TT, Graham TR. Auto-inhibition of Drs2p, a yeast phospholipid flippase, by its carboxyl-terminal tail. J Biol Chem. 2013;288(44):31807-15.

Chapter 2

106. Liu K, Hua Z, Nepute JA, Graham TR. Yeast P4-ATPases Drs2p and Dnf1p are essential cargos of the NPFXD/Sla1p endocytic pathway. MolBiolCell. 2007;18(2):487-500. 107. Pomorski T, Lombardi R, Riezman H, Devaux PF, G. vM, Holthuis JC. Drs2p-related P-type ATPases Dnf1p and Dnf2p are required for phospholipid translocation across the yeast plasma membrane and serve a role in endocytosis. MolBiolCell. 2003;14(3):1240-54. 108. Alder-Baerens N, Lisman Q, Luong L, Pomorski T, Holthuis JC. Loss of P4 ATPases Drs2p and Dnf3p disrupts aminophospholipid transport and asymmetry in yeast post-Golgi secretory vesicles. MolBiolCell. 2006;17(4):1632-42. 109. Hua Z, Graham TR. Requirement for neo1p in retrograde transport from the Golgi complex to the endoplasmic reticulum. MolBiolCell. 2003;14(12):4971-83. 110. Wicky S, Schwarz H, Singer-Kruger B. Molecular interactions of yeast Neo1p, an essential member of the Drs2 family of aminophospholipid translocases, and its role in membrane trafficking within the endomembrane system. MolCell Biol. 2004;24(17):7402- 18. 111. Poulsen LR, Lopez-Marques RL, Pedas PR, McDowell SC, Brown E, Kunze R, et al. A phospholipid uptake system in the model plant Arabidopsis thaliana. Nat Commun. 2015;6:7649. 112. Ruaud AF, Nilsson L, Richard F, Larsen MK, Bessereau JL, Tuck S. The C. elegans P4-ATPase TAT-1 regulates lysosome biogenesis and endocytosis. Traffic. 2009;10(1):88-100. 113. Wehman AM, Poggioli C, Schweinsberg P, Grant BD, Nance J. The P4-ATPase TAT-5 inhibits the budding of extracellular vesicles in C. elegans embryos. CurrBiol. 2011;21(23):1951-9. 114. Ha TS, Xia R, Zhang H, Jin X, Smith DP. Lipid flippase modulates olfactory receptor expression and odorant sensitivity in Drosophila. Proc Natl Acad Sci U S A. 2014;111(21):7831-6. 115. de Waart DR, Naik J, Utsunomiya KS, Duijst S, Ho-Mok K, Bolier AR, et al. ATP11C targets basolateral bile salt transporter proteins in mouse central hepatocytes. Hepatology. 2016. 116. Matsuzaka Y, Hayashi H, Kusuhara H. Impaired Hepatic Uptake by Organic Anion-Transporting Polypeptides Is Associated with Hyperbilirubinemia and Hypercholanemia in Atp11c Mutant Mice. Mol Pharmacol. 2015;88(6):1085-92. 117. Naik J, de Waart DR, Utsunomiya K, Duijst S, Mok KH, Oude Elferink RP, et al. ATP8B1 and ATP11C: Two Lipid Flippases Important for Hepatocyte Function. Dig Dis. 2015;33(3):314-8. 118. Verhulst PM, van der Velden LM, Oorschot V, van Faassen EE, Klumperman J, Houwen RH, et al. A flippase-independent function of ATP8B1, the protein affected in familial intrahepatic cholestasis type 1, is required for apical protein expression and microvillus formation in polarized epithelial cells. Hepatology. 2010;51(6):2049-60. 119. van der Mark VA, de Waart DR, Ho-Mok KS, Tabbers MM, Voogt HW, Oude Elferink RP, et al. The lipid flippase heterodimer ATP8B1- CDC50A is essential for surface expression of the apical sodium-dependent bile acid transporter (SLC10A2/ASBT) in intestinal Caco-2 cells. Biochim Biophys Acta. 2014;1842(12 Pt A):2378-86. 120. van der Mark VA, de Jonge HR, Chang JC, Ho-Mok KS, Duijst S, Vidovic D, et al. The phospholipid flippase ATP8B1 mediates apical localization of the cystic fibrosis transmembrane regulator. Biochim Biophys Acta. 2016;1863(9):2280-8. 121. Lee S, Uchida Y, Wang J, Matsudaira T, Nakagawa T, Kishimoto T, et al. Transport through recycling endosomes requires EHD1 recruitment by a phosphatidylserine translocase. EMBO J. 2015;34(5):669-88. 122. Ansari IU, Longacre MJ, Paulusma CC, Stoker SW, Kendrick MA, MacDonald MJ. Characterization of P4 ATPase Phospholipid Translocases (Flippases) in Human and Rat Pancreatic Beta Cells: THEIR GENE SILENCING INHIBITS INSULIN SECRETION. J Biol Chem. 2015;290(38):23110-23. 123. Zhou Y, Wong CO, Cho KJ, van der Hoeven D, Liang H, Thakur DP, et al. SIGNAL TRANSDUCTION. Membrane potential modulates plasma membrane phospholipid dynamics and K-Ras signaling. Science. 2015;349(6250):873-6. 124. Kato U, Inadome H, Yamamoto M, Emoto K, Kobayashi T, Umeda M. Role for phospholipid flippase complex of ATP8A1 and CDC50A in cell migration. JBiolChem. 2012. 125. Coleman JA, Zhu X, Djajadi HR, Molday LL, Smith RS, Libby RT, et al. Phospholipid flippase ATP8A2 is required for normal visual and auditory function and photoreceptor and spiral ganglion cell survival. J Cell Sci. 2014;127(Pt 5):1138-49. 126. Cai SY, Gautam S, Nguyen T, Soroka CJ, Rahner C, Boyer JL. ATP8B1 deficiency disrupts the bile canalicular membrane bilayer structure in hepatocytes, but FXR expression and activity are maintained. Gastroenterology. 2009;136(3):1060-9. 127. Paulusma CC, Groen A, Kunne C, Ho-Mok KS, Spijkerboer AL, de Waart DR, et al. Atp8b1 deficiency in mice reduces resistance of the canalicular membrane to hydrophobic bile salts and impairs bile salt transport. Hepatology. 2006;44(1):195-204. 128. Paulusma CC, de Waart DR, Kunne C, Mok KS, Elferink RP. Activity of the bile salt export pump (ABCB11) is critically dependent on canalicular membrane cholesterol content. JBiolChem. 2009;284(15):9947-54. 129. Bruurs LJ, Donker L, Zwakenberg S, Zwartkruis FJ, Begthel H, Knisely AS, et al. ATP8B1-mediated spatial organization of Cdc42 signaling maintains singularity during enterocyte polarization. J Cell Biol. 2015;210(7):1055-63. 130. Moreno-Smith M, Halder JB, Meltzer PS, Gonda TA, Mangala LS, Rupaimoole R, et al. ATP11B mediates platinum resistance in ovarian cancer. JClinInvest. 2013;123(5):2119-30. 131. Arashiki N, Takakuwa Y, Mohandas N, Hale J, Yoshida K, Ogura H, et al. ATP11C is a major flippase in human erythrocytes and its defect causes congenital hemolytic anemia. Haematologica. 2016;101(5):559-65. 132. Segawa K, Kurata S, Yanagihashi Y, Brummelkamp TR, Matsuda F, Nagata S. Caspase-mediated cleavage of phospholipid flippase for apoptotic phosphatidylserine exposure. Science. 2014;344(6188):1164-8. 133. Tian T, Harding A, Inder K, Plowman S, Parton RG, Hancock JF. Plasma membrane nanoswitches generate high-fidelity Ras signal transduction. Nat Cell Biol. 2007;9(8):905-14. 134. Das A, Slaughter BD, Unruh JR, Bradford WD, Alexander R, Rubinstein B, et al. Flippase-mediated phospholipid asymmetry promotes fast Cdc42 recycling in dynamic maintenance of cell polarity. NatCell Biol. 2012;14(3):304-10. 135. Levano K, Punia V, Raghunath M, Debata PR, Curcio GM, Mogha A, et al. Atp8a1 deficiency is associated with phosphatidylserine externalization in hippocampus and delayed hippocampus-dependent learning. JNeurochem. 2012;120(2):302-13. 136. O. EO, Gulsuner S, Bilguvar K, Nazli BA, Topaloglu H, Tan M, et al. Missense mutation in the ATPase, aminophospholipid transporter protein ATP8A2 is associated with cerebellar atrophy and quadrupedal locomotion. EurJHumGenet. 2012. 137. Zhu X, Libby RT, de Vries WN, Smith RS, Wright DL, Bronson RT, et al. Mutations in a P-type ATPase gene cause axonal degeneration. PLoSGenet. 2012;8(8):e1002853. 138. Paulusma CC, Elferink RP, Jansen PL. Progressive familial intrahepatic cholestasis type 1. SeminLiver Dis. 2010;30(2):117-24. 139. Pawlikowska L, Groen A, Eppens EF, Kunne C, Ottenhoff R, Looije N, et al. A mouse genetic model for familial cholestasis caused by ATP8B1 mutations reveals perturbed bile salt homeostasis but no impairment in bile secretion. HumMolGenet. 2004;13(8):881-92. 140. Stapelbroek JM, Peters TA, van Beurden DH, Curfs JH, Joosten A, Beynon AJ, et al. ATP8B1 is essential for maintaining normal hearing. ProcNatlAcadSciUSA. 2009;106(24):9709-14. 141. Gong EY, Park E, Lee HJ, Lee K. Expression of Atp8b3 in murine testis and its characterization as a testis specific P-type ATPase. Reproduction. 2009;137(2):345-51.

Chapter 2

142. Wang L, Beserra C, Garbers DL. A novel aminophospholipid transporter exclusively expressed in spermatozoa is required for membrane lipid asymmetry and normal fertilization. DevBiol. 2004;267(1):203-15. 143. Li H, Wetten S, Li L, St Jean PL, Upmanyu R, Surh L, et al. Candidate single-nucleotide polymorphisms from a genomewide association study of Alzheimer disease. ArchNeurol. 2008;65(1):45-53. 144. Gao L, Emond MJ, Louie T, Cheadle C, Berger AE, Rafaels N, et al. Identification of Rare Variants in ATP8B4 as a Risk Factor for Systemic Sclerosis by Whole-Exome Sequencing. Arthritis Rheumatol. 2016;68(1):191-200. 145. Dhar MS, Sommardahl CS, Kirkland T, Nelson S, Donnell R, Johnson DK, et al. Mice heterozygous for Atp10c, a putative amphipath, represent a novel model of obesity and type 2 diabetes. JNutr. 2004;134(4):799-805. 146. Dhar MS, Yuan JS, Elliott SB, Sommardahl C. A type IV P-type ATPase affects insulin-mediated glucose uptake in adipose tissue and skeletal muscle in mice. JNutrBiochem. 2006;17(12):811-20. 147. Dupuis J, Langenberg C, Prokopenko I, Saxena R, Soranzo N, Jackson AU, et al. New genetic loci implicated in fasting glucose homeostasis and their impact on type 2 diabetes risk. NatGenet. 2010;42(2):105-16. 148. Irvin MR, Wineinger NE, Rice TK, Pajewski NM, Kabagambe EK, Gu CC, et al. Genome-wide detection of allele specific copy number variation associated with insulin resistance in African Americans from the HyperGEN study. PLoSOne. 2011;6(8):e24052. 149. Milagro FI, Campion J, Cordero P, Goyenechea E, Gomez-Uriz AM, Abete I, et al. A dual epigenomic approach for the search of obesity biomarkers: DNA methylation in relation to diet-induced weight loss. FASEB J. 2011;25(4):1378-89. 150. Flamant S, Pescher P, Lemercier B, Clement-Ziza M, Kepes F, Fellous M, et al. Characterization of a putative type IV aminophospholipid transporter P-type ATPase. MammGenome. 2003;14(1):21-30. 151. Hicks AA, Pramstaller PP, Johansson A, Vitart V, Rudan I, Ugocsai P, et al. Genetic determinants of circulating sphingolipid concentrations in European populations. PLoS Genet. 2009;5(10):e1000672. 152. Kengia JT, Ko KC, Ikeda S, Hiraishi A, Mieno-Naka M, Arai T, et al. A gene variant in the Atp10d gene associates with atherosclerotic indices in Japanese elderly population. Atherosclerosis. 2013;231(1):158-62. 153. Miyoshi N, Ishii H, Mimori K, Tanaka F, Nagai K, Uemura M, et al. ATP11A is a novel predictive marker for metachronous metastasis of colorectal cancer. OncolRep. 2010;23(2):505-10. 154. Fingerlin TE, Murphy E, Zhang W, Peljto AL, Brown KK, Steele MP, et al. Genome-wide association study identifies multiple susceptibility loci for pulmonary fibrosis. Nat Genet. 2013;45(6):613-20. 155. Siggs OM, Schnabl B, Webb B, Beutler B. X-linked cholestasis in mouse due to mutations of the P4-ATPase ATP11C. ProcNatlAcadSciUSA. 2011;108(19):7890-5. 156. Siggs OM, Arnold CN, Huber C, Pirie E, Xia Y, Lin P, et al. The P4-type ATPase ATP11C is essential for B lymphopoiesis in adult bone marrow. NatImmunol. 2011;12(5):434-40. 157. Soupene E, Kemaladewi DU, Kuypers FA. ATP8A1 activity and phosphatidylserine transbilayer movement. JReceptor LigandChannelRes. 2008;1:1-10. 158. Xu Q, Yang GY, Liu N, Xu P, Chen YL, Zhou Z, et al. P4-ATPase ATP8A2 acts in synergy with CDC50A to enhance neurite outgrowth. FEBS Lett. 2012;586(13):1803-12. 159. Cacciagli P, Haddad MR, Mignon-Ravix C, El-Waly B, Moncla A, Missirian C, et al. Disruption of the ATP8A2 gene in a patient with a t(10;13) de novo balanced translocation and a severe neurological phenotype. EurJHumGenet. 2010;18(12):1360-3. 160. Hurst SE, Minkin SC, Biggerstaff J, Dhar MS. Transient Silencing of a Type IV P-Type ATPase, Atp10c, Results in Decreased Glucose Uptake in C2C12 Myotubes. JNutrMetab. 2012;2012:152902. 161. Roshwalb S, Gorman S, Hurst S, Bartges J, Agarwal S, Sommardahl C, et al. mRNA expression of canine ATP10C, a P4-type ATPase, is positively associated with body condition score. VetJ. 2011;190(1):173-5. 162. Surwit RS, Feinglos MN, Rodin J, Sutherland A, Petro AE, Opara EC, et al. Differential effects of fat and sucrose on the development of obesity and diabetes in C57BL/6J and A/J mice. Metabolism. 1995;44(5):645-51. 163. Yabas M, Coupland LA, Cromer D, Winterberg M, Teoh NC, D'Rozario J, et al. Mice deficient in the putative phospholipid flippase ATP11C exhibit altered erythrocyte shape, anemia, and reduced erythrocyte life span. J Biol Chem. 2014;289(28):19531-7. 164. Andrew Nesbit M, Bowl MR, Harding B, Schlessinger D, Whyte MP, Thakker RV. X-linked hypoparathyroidism region on Xq27 is evolutionarily conserved with regions on 3q26 and 13q34 and contains a novel P-type ATPase. Genomics. 2004;84(6):1060-70. 165. Fadok VA, Voelker DR, Campbell PA, Cohen JJ, Bratton DL, Henson PM. Exposure of phosphatidylserine on the surface of apoptotic lymphocytes triggers specific recognition and removal by macrophages. J Immunol. 1992;148(7):2207-16. 166. Bull LN, van Eijk MJ, Pawlikowska L, DeYoung JA, Juijn JA, Liao M, et al. A gene encoding a P-type ATPase mutated in two forms of hereditary cholestasis. NatGenet. 1998;18(3):219-24. 167. Klomp LW, Vargas JC, van Mil SW, Pawlikowska L, Strautnieks SS, van Eijk MJ, et al. Characterization of mutations in ATP8B1 associated with hereditary cholestasis. Hepatology. 2004;40(1):27-38. 168. Clayton RJ, Iber FL, Ruebner BH, McKusick VA. Byler disease. Fatal familial intrahepatic cholestasis in an Amish kindred. AmJDisChild. 1969;117(1):112-24. 169. Summerskill WH, WALSHE JM. Benign recurrent intrahepatic "obstructive" jaundice. Lancet. 1959;2(7105):686-90. 170. Orphanet PFIC1 2011 [updated 5/1/2011]. Available from: http://www.orpha.net/consor/cgi-bin/OC_Exp.php?lng=en&Expert=79306. 171. Davit-Spraul A, Gonzales E, Baussan C, Jacquemin E. Progressive familial intrahepatic cholestasis. OrphanetJRareDis. 2009;4:1. 172. Summerskill WH. The syndrome of benign recurrent cholestasis. AmJMed. 1965;38:298-305. 173. Bustorff-Silva J, Sbraggia NL, Olimpio H, de Alcantara RV, Matsushima E, De Tommaso AM, et al. Partial internal biliary diversion through a cholecystojejunocolonic anastomosis--a novel surgical approach for patients with progressive familial intrahepatic cholestasis: a preliminary report. JPediatrSurg. 2007;42(8):1337-40. 174. Emond JC, Whitington PF. Selective surgical management of progressive familial intrahepatic cholestasis (Byler's disease). JPediatrSurg. 1995;30(12):1635-41. 175. Lykavieris P, van MS, Cresteil D, Fabre M, Hadchouel M, Klomp L, et al. Progressive familial intrahepatic cholestasis type 1 and extrahepatic features: no catch-up of stature growth, exacerbation of diarrhea, and appearance of liver steatosis after liver transplantation. JHepatol. 2003;39(3):447-52. 176. Morotti RA, Suchy FJ, Magid MS. Progressive familial intrahepatic cholestasis (PFIC) type 1, 2, and 3: a review of the liver pathology findings. SeminLiver Dis. 2011;31(1):3-10. 177. Pawlikowska L, Strautnieks S, Jankowska I, Czubkowski P, Emerick K, Antoniou A, et al. Differences in presentation and progression between severe FIC1 and BSEP deficiencies. JHepatol. 2010;53(1):170-8. 178. Folmer DE, van der Mark VA, Ho-Mok KS, Oude Elferink RP, Paulusma CC. Differential effects of progressive familial intrahepatic cholestasis type 1 and benign recurrent intrahepatic cholestasis type 1 mutations on canalicular localization of ATP8B1. Hepatology. 2009;50(5):1597-605. 179. van Mil SW, Klomp LW, Bull LN, Houwen RH. FIC1 disease: a spectrum of intrahepatic cholestatic disorders. SeminLiver Dis. 2001;21(4):535-44.

Chapter 2

180. Eppens EF, van Mil SW, de Vree JM, Mok KS, Juijn JA, Oude Elferink RP, et al. FIC1, the protein affected in two forms of hereditary cholestasis, is localized in the cholangiocyte and the canalicular membrane of the hepatocyte. JHepatol. 2001;35(4):436-43. 181. van Mil SW, van Oort MM, van den Berg IE, Berger R, Houwen RH, Klomp LW. Fic1 is expressed at apical membranes of different epithelial cells in the digestive tract and is induced in the small intestine during postnatal development of mice. PediatrRes. 2004;56(6):981- 7. 182. Groen A, Kunne C, Paulusma CC, Kramer W, Agellon LB, Bull LN, et al. Intestinal bile salt absorption in Atp8b1 deficient mice. JHepatol. 2007;47(1):114-22. 183. Groen A, Kunne C, Jongsma G, van den Oever K, Mok KS, Petruzzelli M, et al. Abcg5/8 independent biliary cholesterol excretion in Atp8b1-deficient mice. Gastroenterology. 2008;134(7):2091-100. 184. Paulusma CC, Houwen RH, Williamson PL. The flip side of cardiolipin import. NatMed. 2011;17(4):413-4. 185. Frankenberg T, Miloh T, Chen FY, Ananthanarayanan M, Sun AQ, Balasubramaniyan N, et al. The membrane protein ATPase class I type 8B member 1 signals through protein kinase C zeta to activate the farnesoid X receptor. Hepatology. 2008;48(6):1896-905. 186. Chen F, Ellis E, Strom SC, Shneider BL. ATPase Class I Type 8B Member 1 and protein kinase C zeta induce the expression of the canalicular bile salt export pump in human hepatocytes. PediatrRes. 2010;67(2):183-7. 187. Chen F, Ananthanarayanan M, Emre S, Neimark E, Bull LN, Knisely AS, et al. Progressive familial intrahepatic cholestasis, type 1, is associated with decreased farnesoid X receptor activity. Gastroenterology. 2004;126(3):756-64. 188. Koh S, Takada T, Kukuu I, Suzuki H. FIC1-mediated stimulation of FXR activity is decreased with PFIC1 mutations in HepG2 cells. JGastroenterol. 2009;44(6):592-600. 189. Takatsu H, Tanaka G, Segawa K, Suzuki J, Nagata S, Nakayama K, et al. Phospholipid flippase activities and substrate specificities of human type IV P-type ATPases localized to the plasma membrane. J Biol Chem. 2014;289(48):33543-56. 190. Naito T, Takatsu H, Miyano R, Takada N, Nakayama K, Shin HW. Phospholipid Flippase ATP10A Translocates Phosphatidylcholine and Is Involved in Plasma Membrane Dynamics. J Biol Chem. 2015;290(24):15004-17. 191. Takada N, Takatsu H, Miyano R, Nakayama K, Shin HW. ATP11C mutation is responsible for the defect in phosphatidylserine uptake in UPS-1 cells. J Lipid Res. 2015;56(11):2151-7. 192. van der Woerd WL, Mulder J, Pagani F, Beuers U, Houwen RH, van de Graaf SF. Analysis of aberrant pre-messenger RNA splicing resulting from mutations in ATP8B1 and efficient in vitro rescue by adapted U1 small nuclear RNA. Hepatology. 2015;61(4):1382-91. 193. van der Woerd WL, Wichers CG, Vestergaard AL, Andersen JP, Paulusma CC, Houwen RH, et al. Rescue of defective ATP8B1 trafficking by CFTR correctors as a therapeutic strategy for familial intrahepatic cholestasis. J Hepatol. 2016;64(6):1339-47. 194. Gonzales E, Grosse B, Schuller B, Davit-Spraul A, Conti F, Guettier C, et al. Targeted pharmacotherapy in progressive familial intrahepatic cholestasis type 2: Evidence for improvement of cholestasis with 4-phenylbutyrate. Hepatology. 2015;62(2):558-66. 195. Hasegawa Y, Hayashi H, Naoi S, Kondou H, Bessho K, Igarashi K, et al. Intractable itch relieved by 4-phenylbutyrate therapy in patients with progressive familial intrahepatic cholestasis type 1. Orphanet J Rare Dis. 2014;9:89. 196. Shneider BL, Morris A, Vockley J. Possible Phenylacetate Hepatotoxicity During 4-Phenylbutyrate Therapy of Byler Disease. J Pediatr Gastroenterol Nutr. 2016;62(3):424-8. 197. Munoz-Martinez F, Torres C, Castanys S, Gamarro F. CDC50A plays a key role in the uptake of the anticancer drug perifosine in human carcinoma cells. BiochemPharmacol. 2010;80(6):793-800.

CHAPTER 3

Differential effects of progressive familial intrahepatic cholestasis type 1 and benign recurrent intrahepatic cholestasis type 1 mutations on canalicular localization of ATP8B1

Dineke E. Folmer, Vincent van der Mark, Kam S. Ho-Mok, Ronald P. Oude Elferink, Coen C. Paulusma

Hepatology. 2009 Nov;50(5):1597-605

doi: 10.1002/hep.23158

Chapter 3

Abstract

Mutations in ATP8B1 cause progressive familial intrahepatic cholestasis type 1 (PFIC1) and benign recurrent intrahepatic cholestasis type 1 (BRIC1), forming a spectrum of cholestatic disease. Whereas PFIC1 is a progressive, end-stage liver disease, BRIC1 patients suffer from episodic periods of cholestasis that resolve spontaneously. At present it is not clear how the type and location of the mutations relate to the clinical manifestations of PFIC1 and BRIC1. ATP8B1 localizes to the canalicular membrane of hepatocytes where it mediates the inward translocation of phosphatidylserine. ATP8B1 interacts with CDC50A, which is required for endoplasmic reticulum exit and plasma membrane localization. In this study we have analyzed a panel of missense mutations causing PFIC1 (G308V, D554N, G1040R) or BRIC1 (D70N, I661T). In addition, we included two mutations that have been associated with ICP (D70N, R867C). We examined the effect of these mutations on protein stability and interaction with CDC50A in Chinese hamster ovary cells, and studied the subcellular localization in WIF-B9 cells. Protein stability was reduced for three out of six mutations studied. Two out of three PFIC1 mutant proteins did not interact with CDC50A, whereas BRIC1/ICP mutants displayed reduced interaction. Importantly, none of the PFIC1 mutants were detectable in the canalicular membrane of WIF-B9 cells, while all BRIC1/ICP mutants displayed the same cellular staining pattern as wild-type ATP8B1. Our data indicate that PFIC1 mutations lead to complete absence of canalicular expression, whereas in BRIC1/ICP residual protein is expressed in the canalicular membrane. In conclusion, these data provide an explanation for the difference in severity between the phenotypes of PFIC1 and BRIC1.

Chapter 3

1. Introduction

Mutations in ATP8B1 cause two forms of cholestatic disease, Progressive Familial Intrahepatic Cholestasis type 1 (PFIC1) and Benign Recurrent Intrahepatic Cholestasis type 1 (BRIC1) (1). In both disorders, patients present with pruritis, jaundice, and fat malabsorption. Some PFIC1 patients develop extrahepatic symptoms, including pancreatitis, diarrhea, and hearing loss (2). PFIC1 patients develop progressive, end-stage liver disease and need to undergo liver transplantation before adolescence. Although this relieves the cholestasis, extrahepatic symptoms may remain (3). BRIC1 patients suffer from episodes of cholestasis that resolve spontaneously without leaving any detectable liver damage, and are symptom-free between the episodes of disease (2). Recently, mutations in ATP8B1 have been associated with Intrahepatic Cholestasis of Pregnancy (ICP) (4, 5). ICP affects pregnant women in the third trimester of pregnancy, but resolves spontaneously within a few weeks after delivery. ICP is associated with increased risk of preterm birth, fetal distress, and fetal loss (6). ATP8B1 is a member of the type 4 subfamily of P-type ATPases (P4-ATPases) and is expressed in the apical membrane of epithelial cells, including the canalicular membrane of hepatocytes (7-9). ATP8B1 mediates the inward translocation of phosphatidylserine (PS) from the exoplasmic to the cytoplasmic leaflet of the plasma membrane (9-12). Furthermore, ATP8B1 physically interacts with CDC50A, which is required for endoplasmic reticulum (ER) exit and plasma membrane localization (11). The etiology of ATP8B1-associated cholestasis is not completely understood. We have previously shown that in Atp8b1-deficient mice the canalicular membrane is more sensitive to the detergent action of hydrophobic bile salts (10, 13). Importantly, cholic acid challenge induced cholestasis in Atp8b1-deficient mice, which coincided with decreased cholesterol to phospholipid ratios of the canalicular membrane and a subsequent reduction in Abcb11-mediated transport (14). Cai et al. (12) have recently demonstrated that Abcb11 activity is significantly reduced in isolated, Atp8b1-deficient rat hepatocytes. Furthermore, in these rat hepatocytes the authors showed disrupted canalicular membrane areas after exposure to chenodeoxycholic acid. From all these observations Paulusma et al. (10, 14) and Cai et al. (12) have proposed that ATP8B1 deficiency results in reduced lipid packing of the canalicular membrane outer leaflet. Subsequent enhanced extraction of cholesterol from the canalicular membrane impairs ABCB11 activity and causes the cholestasis. An

Chapter 3 alternative hypothesis proposes that ATP8B1 deficiency leads to impaired FXR signaling resulting in reduced ABCB11 and enhanced ASBT activation. This leads to bile salt accumulation and could explain the cholestasis in PFIC1/BRIC1 patients (15-17). However, it remains to be elucidated why particular mutations result in severe, end-stage liver disease, whereas other mutations give rise to episodes of cholestasis. In this study we examined a panel of ATP8B1 mutant proteins that cause different forms of ATP8B1-associated cholestasis. We examined the effect of these mutations on protein stability and interaction with CDC50A in Chinese hamster ovary cells, and studied the subcellular localization in WIF-B9 cells. Two out of three PFIC1 mutations result in reduced protein stability and impaired interaction with CDC50A, while all mutants displayed impaired protein trafficking to the canalicular membrane. On the other hand, two out of three BRIC1/ICP mutations do not affect protein stability, while all mutants interacted with CDC50A and displayed proper canalicular membrane localization.

2. Materials and methods

2.1 Constructs and mutagenesis

Lentiviral vectors containing C-terminal enhanced green fluorescent protein-tagged ATP8B1 (ATP8B1-eGFP) cDNA and N-terminal hemagglutinin antigen-tagged CDC50A (HA-CDC50A) have been described previously (11). Missense mutations were introduced using the QuikChange® Multi Site-Directed Mutagenesis Kit (Stratagene, Amsterdam, The Netherlands), according to the manufacturer’s instructions. PFIC1 mutations: 923G>T (G308V), 1660G>A (D554N), 3118G>A (G1040R); BRIC1 mutation: 1982T>C (I661T); BRIC1/ICP-associated mutation 208G>A (D70N); ICP-associated mutation: 2599C>T (R867C). Recombinant lentivirus was produced as described (18).

2.2 Cell culture

UPS-1 and WIF-B9 cells were cultured in Dulbecco's modified Eagle's medium (DMEM) (Lonza, Verviers, Belgium) supplemented with 10% fetal calf serum, 1% glutamine (vol/vol), and 1% (vol/vol) penicillin/streptomycin at 37°C in a 10% CO2 atmosphere. For lentiviral transduction, UPS-1 and WIF-B9 cells were grown to 30-

Chapter 3

50% confluence, incubated with virus-containing supernatants/DMEM (1:1) supplemented with 10 μg/ml diethylaminoethyl-dextran for 4 hours. Cells were transduced with equal multiplicity of infection. The cell lines generated stably express ATP8B1 proteins and were examined from passage number 3 post- transduction. Proteasomal breakdown was inhibited by incubating the cells for 6 hours with 10 µM MG-132 (Calbiochem, La Jolla, CA, USA). UPS-1 cells were used to study ATP8B1 mutant protein-CDC50A interaction and cell surface expression of ATP8B1 mutant proteins. UPS-1 cells do not express endogenous ATP8B1 and CDC50A, which render these cells very suitable to study their interaction. Furthermore, we have previously demonstrated that co-expression of both proteins results in plasma membrane localization of ATP8B1, which makes this cell line very suitable for studying cell surface expression of ATP8B1 mutant proteins (11).

2.3 Co-immunoprecipitation

Co-immunoprecipitation was performed as described (11). Briefly, cells were lysed in RIPA buffer (50mM Tris pH7.4, 0.1% SDS, 1% NP-40, 150mM NaCl). Cleared supernatant was incubated with monoclonal anti-HA agarose conjugate clone HA-7 (Sigma-Aldrich, St. Louis, MO, USA) for 2 hours at 20°C. Immunoprecipitated proteins were analyzed by Western blotting as described (11). Protein was detected with rabbit polyclonal anti-ATP8B1 (2K), rat monoclonal anti-HA (clone 3F10) and rabbit polyclonal anti- Atp1a1 (C356-M09) (19), and peroxidase-conjugated IgGs. Immune complexes were visualized with Lumi-Light Western blotting substrate (Roche). Chemiluminescence was detected and quantified with a Lumi-Imager F1 and LumiAnalyst 3.1 software (Roche).

2.4 Cell surface biotinylation

The cell surface protein isolation kit (Pierce, Rockford, IL, USA) was used according to the manufacturer’s instructions. Briefly, cells were incubated with 0.25mg/ml biotin in PBS for 30 minutes at 4°C. Cells were scraped and lysed in ice-cold RIPA buffer supplemented with Complete Protease Inhibitor Cocktail (Roche Diagnostics, Mannheim, Germany). Cleared supernatant was incubated with neutravidin beads for 1 hour at room temperature. Precipitated proteins were eluted and analyzed as described above.

Chapter 3

2.5 Indirect immunofluorescence and confocal laser scanning microscopy

2x105 WIF-B9 cells were seeded on glass coverslips, grown for 5-7 days until approximately 80% confluence was reached. By that time, 10-35% of the cells had formed a canalicular vacuole. Cells were fixed in 2% paraformaldehyde in phosphate-buffered saline (PBS) for 20 minutes at room temperature, permeabilized in 0.1% Triton-X100 in PBS, and incubated with antibodies to ABCC2 (M2III6) (20). Immunoreactivity of ABCC2 was visualized using goat-anti-mouse alexa594 (Molecular Probes, Leiden, The Netherlands). ATP8B1-eGFP was detected by studying GFP fluorescence. Cells were mounted in Vectashield mounting medium with 4’6-diamidino-2-phenylindole (Vector Laboratories, Burlingame, CA, USA) and were studied with a Leica TCS-SP2 confocal microscope.

3. Results

3.1 Description of the mutations

We analyzed three PFIC1 mutations (G308V, D554N, G1040R), two BRIC1 mutations (I661T, D70N), and two mutations associated with ICP (D70N, R867C) (Figure 1). All mutations were introduced in ATP8B1-eGFP. In a previous study we have demonstrated that the eGFP-tag does not interfere with ATP8B1 localization nor activity (11).

Chapter 3

Figure 1. Schematic representation of ATP8B1 protein in which the PFIC1, BRIC1, and ICP mutations presented in this study are indicated. PFIC1 mutations in red, BRIC1 mutations in blue, and ICP mutations in green (created with the TOPO2 program; http://www.sacs.ucsf.edu/TOPO-run/wtopo.pl). Amino acid sequence motifs that are conserved in P-type ATPases or P4 ATPases (25, 30) are depicted in yellow. The catalytically important aspartic acid residue is indicated in orange.

The G308V mutation was identified in the Amish population (1, 21). ATP8B1(G308V) protein is not detectable in liver from patients homozygous for this mutation (8), nor in liver from knock-in mice bearing this mutation (22). Glycine-308 is a highly conserved residue present in all P-type ATPases, which suggests a critical function for this residue in protein function (23). The D554N mutation is in the signature motif of P4-ATPases (24, 25), and was described in Greenland Familial Cholestasis (GFC) (26, 27). GFC is at the most severe end of the spectrum of ATP8B1-disease (26, 28). The G1040R mutation was identified in two Saudi families and substitutes a highly conserved glycine into an arginine in predicted transmembrane region 7 (21). The I661T mutation is the most common BRIC1 mutation in Western Europe (1, 21, 29). The I661T affects an amino acid at

Chapter 3 a position where most P4-ATPases have a leucine or isoleucine residue (1). The D70N mutation was identified in a German BRIC1 patient as a compound heterozygous mutation in combination with the R600Q substitution, but was also present on 0.5% of the chromosomes of healthy control subjects (21). The D70N substitution has also been identified in three ICP cases, and was not detected in healthy pregnant controls that were included in the same study (4). The R867C mutation has been reported in one ICP patient and was absent in parous controls (4). Arginine-867 is in one of the ‘signature sequences’ and is highly conserved in all P4 ATPases (4, 30). In addition to the mutations reported in patients, we included the D454A substitution, which affects the aspartic acid residue that is essential in the catalytic cycle of all P-type ATPases and leads to an inactive protein (31).

3.2 Protein expression and proteasomal breakdown

To examine whether the mutations affected protein stability and/or localization, ATP8B1-eGFP mutants were co-expressed together with HA-CDC50A in CHO-K1- derived UPS-1 cells. Previously we have shown that interaction of ATP8B1 with CDC50A is required for ER exit and plasma membrane localization of ATP8B1 (11). Cells were transduced with recombinant lentivirus expressing the different mutant proteins with the same multiplicity of infection. Although all mRNA levels were similar to the wild-type control (data not shown), protein expression levels were dramatically decreased for the PFIC1 mutants G308V and D554N (14% and 10% of wild-type, respectively), the BRIC1 mutant I661T (18% of wild-type), and the ATPase-dead D454A mutant (9% of wild-type) (Figure 2A,B). Protein levels for the D70N (BRIC1/ICP), R867C (ICP) and for the G1040R (PFIC1) mutants were not significantly reduced. Incubation with the proteasomal inhibitor MG-132 resulted in recovery of the mutant expression levels to wild-type (Figure 2C, D), suggesting that the reduced protein levels of the mutant proteins were in all cases due to enhanced proteasomal degradation. The I661T mutant had a lower apparent molecular mass, which may be explained by affected posttranslational modification. These data suggest that mutations in ATP8B1-eGFP result in decreased stability of the protein.

Chapter 3

Figure 2. (A) ATP8B1-eGFP mutant proteins and HA-CDC50A were coexpressed in UPS-1 cells. Ten µg of total protein was analyzed by immunoblotting for ATP8B1-eGFP expression. ATP1A1 was used as a loading control. Cells transduced with HA-CDC50A were used as a negative control. (B) Quantification of ATP8B1-eGFP expression of four independent experiments. ATP8B1-eGFP expression levels were quantified by densitometry, normalized to ATP1A1 levels, and expressed as a percentage of wild-type ATP8B1-eGFP. Statistical analysis was performed with SPSS (ANOVA, Bonferroni). Asterisks indicate values that are significantly decreased compared to wild-type (p<0.05). (C) Representative Western blot showing ATP8B1-eGFP expression levels in UPS-1 cells (10 µg of total protein was separated by SDS-PAGE) after a 5h incubation period with the proteasome inhibitor MG- 132 (10 μM). ATP1A1 was used as a loading control. (D) Quantification of ATP8B1-eGFP expression levels after incubation with MG-132. Data are of three independent experiments and were quantified by densitometry, normalized to ATP1A1 levels, and expressed as a percentage of wild-type ATP8B1-eGFP. For the R867C mutant the mean of two experiments is shown.

3.3 Binding to CDC50A

Subsequently, we investigated the ability of the ATP8B1 mutants to interact with CDC50A. Co-immunoprecipitation studies were performed on UPS-1 cells co- expressing ATP8B1-eGFP and HA-CDC50A. Since several mutants displayed very low stability (Figure 2), cells were pre-incubated with the proteasomal inhibitor MG- 132 to enhance protein levels. We could not detect any ATP8B1-eGFP in immunoprecipitates of UPS-1 cells expressing only ATP8B1-eGFP or HA- CDC50A. In the immunoprecipitates we could not detect a signal for Atp1a1, indicating that the solubilization of the membrane preparations was efficient, and that there was no aspecific binding of membrane proteins to the beads (not shown).

Chapter 3

Apart from wild-type protein, the I661T (BRIC1), D70N (BRIC1/ICP), and the R867C mutants (ICP) also co-immunoprecipitated with HA-CDC50A (Figure 3A, B). In contrast, the PFIC1 mutants G308V and D554N, as well as the ATPase-dead mutant D454A, were not detectable in the HA-CDC50 immunoprecipitates. Interestingly, the PFIC1 mutant G1040R was also detectable after co- immunoprecipitation, indicating that this mutation did not fully disrupt the interaction with HA-CDC50A. These data indicate that all mutations interfere with HA-CDC50A interaction.

3.4 Cell surface biotinylation

Next we quantified plasma membrane expression of ATP8B1-eGFP mutants. The PFIC1 mutants G308V and D554N were not detectable in the plasma membrane, while the G1040R (PFIC1) was (albeit strongly reduced) (Figure 4A); Also MG-132 treatment did not result in detectable protein in the plasma membrane (not shown). The BRIC1/ICP mutants D70N, I661T, and R867C, on the other hand, were clearly detectable in the plasma membrane. Despite the reduced expression levels of these mutant proteins, a similar proportion of the protein reached the plasma membrane when compared to wild-type ATP8B1-eGFP (Figure 4B). The ATPase-dead protein D454A was also not detectable in the plasma membrane. These data indicate that BRIC1 and ICP mutations do not affect protein trafficking to the plasma membrane while all PFIC1 mutations do interfere with trafficking.

Chapter 3

Figure 3. (A) Co-immunoprecipitation of ATP8B1-eGFP mutant proteins with HA-CDC50A. ATP8B1-eGFP mutants and HA-CDC50A were co-expressed in UPS-1 cells. Cells were incubated with the proteasomal inhibitor MG-132 (10µM) to obtain maximal protein levels (see figure 2D). Total cell lysates were incubated with anti-HA- coupled agarose beads and the co-immunoprecipitates were analyzed by immunoblotting for ATP8B1 expression. The upper panel displays ATP8B1-eGFP and HA-CDC50A expression in the total lysates (20 µg of total protein was separated by SDS-PAGE). The lower panel displays the co-immunoprecipitated ATP8B1-eGFP mutants (eluate). (B) Quantification of the interaction capacity of ATP8B1-eGFP mutant proteins with HA-CDC50A. The amount of HA-CDC50A-associated ATP8B1-eGFP (present in the eluate, figure 3A lower panel) is expressed as percentage of total ATP8B1-eGFP present in the total lysates (figure 3A, upper panel). Results shown are from a representative experiment that was repeated at least three times.

Chapter 3

Figure 4. (A) Cell surface biotinylation of UPS-1 co-expressing ATP8B1-eGFP mutant proteins and HA-CDC50A. Total ATP8B1-eGFP expression (upper panel) was analyzed by immunoblotting with anti-ATP8B1 antibodies (20 µg of total protein was separated by SDS-PAGE). Biotin was allowed to bind to the cells, total cell lysates were prepared and were incubated with Neutravidin beads. Cell surface expressed ATP8B1-eGFP was analyzed in the eluates (lower panel). Please note that the wild-type ATP8B1-eGFP expression level in the total lysate is low compared to four of the mutant proteins, however, the fraction of wild-type protein that reaches the cell surface is comparable to that of the three BRIC1/ICP mutants. (B) Quantification of the cell surface expression levels of ATP8B1-eGFP mutant proteins when co-expressed with HA-CDC50A. The ratio of ATP8B1-eGFP at the cell surface vs. ATP8B1-eGFP in the total cell lysate was calculated and values were expressed as a percentage of wild- type ATP8B1-eGFP. Results shown are from a representative experiment that was repeated at least three times.

Chapter 3

3.5 Subcellular localization in polarized WIF-B9 cells

Finally, we studied the effect of the mutations on polarized ATP8B1-eGFP trafficking in WIF-B9 cells, a hepatocyte model cell line (32). Wild-type and mutant ATP8B1-eGFP proteins were expressed in WIF-B9 cells and the canalicular vacuole was visualized by immunofluorescent staining of the canalicular membrane protein ABCC2. Wild-type ATP8B1-eGFP localized to the canalicular membrane, intracellular vesicles, and (to less extent) the basolateral membrane (Figure 5A), which is consistent with previous observations (11). The ATPase-dead mutant D454A was not detectable in the canalicular membrane (Figure 5B). The PFIC1 mutants G308V, D554N, and G1040R were not detectable in the canalicular membrane but displayed intracellular staining, most likely representing the ER (Figure 6; supplementary figure 1). The I661T (BRIC1), D70N (BRIC1/ICP), and R867C (ICP) mutants all displayed a staining pattern similar to wild-type (Figure 7).

Figure 5. Confocal microscopical optical sections showing the localization of ATP8B1-eGFP wild-type and ATPase-dead protein in WIF-B9 cells (left panels). Cells were co-stained for the canalicular marker protein ABCC2 (middle panels). The right panels show the merged images of ATP8B1-eGFP and ABCC2 (ATP8B1-eGFP in green; ABCC2 in red; nuclear DAPI staining in blue; colocalization of ATP8B1-eGFP and ABCC2 in yellow). Arrows indicate the canalicular vacuoles. A, wild-type ATP8B1-eGFP; B, D454A. The scale bar equals 9.38 µm.

Chapter 3

Figure 6. Confocal microscopical optical sections showing the localization of three PFIC1 ATP8B1-eGFP mutant proteins in WIF-B9 cells. A, G308V; B, D554N; C, G1040R. See figure 5 for details.

Chapter 3

Figure 7. Confocal microscopical optical sections showing the localization of three BRIC1/ICP ATP8B1-eGFP mutant proteins in WIF-B9 cells. A, I661T; B, D70N; C, R867C. See figure 5 for details.

Chapter 3

4. Discussion

Mutations in ATP8B1 give rise to a spectrum of cholestatic disease, but it is currently unclear how the different mutations determine their specific clinical outcome. In this study we analyzed the consequence of several PFIC1, BRIC1 and ICP mutations on ATP8B1 protein stability, interaction with CDC50A, and trafficking to the plasma membrane. Our data provide an explanation for the phenotypic differences observed between patients with PFIC1 and BRIC1. Firstly, two out of three PFIC1 mutants and one of the two BRIC1 mutant proteins studied are less stable resulting in reduced protein expression levels (Figure 2). However, reduced ATP8B1 protein expression levels cannot solely explain the differences between PFIC1 and BRIC1. Although we found strongly decreased ATP8B1 protein levels for the PFIC1 mutants G308V and D554N, the expression level of the G1040R mutant was not significantly different from wild-type. Furthermore, the expression level of the BRIC1 I661T mutant was also strongly reduced and comparable to the G308V and D554N mutants. Apparently, some mutations affect protein folding more dramatically, cannot pass the ER quality control system, and are targeted for proteasomal breakdown. Indeed, protein levels of the mutants were restored after incubation with a proteasomal inhibitor. The strongly reduced protein expression of the G308V mutant is consistent with the near- absence of Atp8b1 protein in liver and intestine of the Atp8b1 deficient mouse, which is a knock-in mouse for the G308V mutation (22). Thus, the phenotypic differences between PFIC1 and BRIC1 disease are not solely determined by the protein expression levels per se. Secondly, a physical CDC50A-ATP8B1 interaction may be an important determinant of the differences between PFIC1 and BRIC1. While the interaction of the PFIC1 mutants G308V and D554N with CDC50A is below the level of detection, the BRIC1/ICP mutants D70N, I661T, and R867C do show residual interaction with CDC50A (Figure 3). In contrast to the G308V and D554N mutation, the PFIC1 G1040R mutation did show residual, though strongly reduced, interaction with CDC50A (see below). We have previously demonstrated that this interaction is a prerequisite for protein stability, ER exit, and plasma membrane localization and activity (11). Indeed, the PFIC1 mutants G308V and D554N are not detected in the plasma membrane of UPS-1 cells, whereas we did detect the PFIC1 G1040R (albeit strongly reduced) and the BRIC1/ICP mutants (Figure 4). Still, an interaction of ATP8B1 with CDC50A is not a guarantee for stable ATP8B1 expression as

Chapter 3 evidenced by the strongly reduced I661T BRIC1 mutant expression levels (Figure 2); this mutant is, however, still capable of interacting with CDC50A and subsequent trafficking to the plasma membrane. Thirdly, when expressed in WIF-B9 cells, the PFIC1 mutants were not detectable in the canalicular membrane whereas the BRIC1/ICP mutants were, evidenced by co-staining with the canalicular maker ABCC2. Previously, we have shown in WIF-B9 cells that, upon co-expression of ATP8B1-eGFP and HA- CDC50A, both proteins physically interact and localize to the canalicular membrane, to lysosomal/late endosomal compartments, and to the basolateral membrane (11). Since WIF-B9 cells express endogenous human CDC50A (unpublished results, DF) we did not co-express the mutant ATP8B1-eGFP proteins together with HA- CDC50A in this study. Also in this case, ATP8B1-eGFP stained the same membrane domains and intracellular structures (Figure 5A) indicating that endogenous CDC50A sufficiently interacts with ATP8B1-eGFP to ensure canalicular targeting of ATP8B1-eGFP. Although we do detect small amounts of G1040R mutant protein in the plasma membrane of UPS-1 cells, we could not detect this mutant in the canalicular membrane of WIF-B9 cells (Figure 6C and supplementary figure 1). There are several possible explanations for this discrepancy. First, in the UPS-1 cells both ATP8B1 and CDC50A are overexpressed, which (in combination with the sparsely affected stability of the G1040R mutant) may explain the residual interaction of G1040R with CDC50A and faint plasma membrane localization. Secondly, in WIF-B9 cells only ATP8B1 was exogenously expressed. The endogenous CDC50A expression may be limited and restricts exit of G1040R protein from the ER resulting in non-detectable canalicular localization of this protein. Thirdly, in WIF-B9 cells additional (yet unknown) proteins important for ER-to-Golgi or trans-Golgi-network-to-canalicular membrane trafficking may have impaired interaction with the G1040R, which can explain absence of this mutant from the canalicular membrane. Interestingly, the ATPase-dead D454A mutant protein was very instable, did not interact with CDC50A, and was not detectable in the canalicular membrane of WIF-B9 cells. One explanation is that this mutation leads to an improperly folded protein which is rapidly degraded by the proteasome. Alternatively, the ATPase activity of ATP8B1 may be essential for the interaction of the protein with CDC50A, which stabilizes the complex and results in ER exit and proper trafficking, but this needs to be addressed in future research. Here we demonstrate that ATP8B1 mutations in PFIC1 lead to absence of canalicular expression due to premature breakdown and/or impaired interaction with

Chapter 3

CDC50A. Our data do not provide insight into the cause of recurrent episodes of cholestasis in BRIC1. BRIC1 mutant proteins although less stable, do reach the canalicular membrane and are likely to have residual protein activity. It was not possible, however, to measure PS translocation activity in polarized WIF-B9 cells. The factors that cause the onset and resolution of the cholestasis remain elusive. Based on our previous work, we have proposed that ATP8B1 deficiency in the canalicular membrane results in loss of the rigid, liquid-ordered state of this membrane; subsequent bile salt-induced extraction of membrane cholesterol impairs ABCB11 activity (10, 13, 14). Having this in mind, it can be speculated that the canalicular expression levels of ATP8B1 mutants in a non-cholestatic period in BRIC1 are sufficient to maintain a rigid canalicular membrane, in which ABCB11 activity is not impaired. The onset of a BRIC1 episode possibly coincides with a ‘second hit’ in which ABCB11 expression may be impaired. For that matter, it may be speculated that infections associated with a proinflammatory cytokine response are important for initiating the cholestatic attacks in BRIC1. For instance, endotoxin/lipopolysaccharide (LPS) induced cholestasis is accompanied by a posttranscriptional downregulation (i.e. due to canalicular endocytosis of the protein) of human and rat ABCB11 (33, 34). In mice, downregulation of Abcb11 was observed at the transcriptional level upon exposure to LPS, interleukin-6, tumor necrosis factor-alpha or interleukin-1 beta (35, 36). In conclusion, we show that two out of three PFIC1 mutations result in decreased protein stability and impaired interaction with CDC50A. Importantly, none of the PFIC1 mutants is detectable in the canalicular membrane of WIF-B9 cells. BRIC1/ICP mutations, on the other hand, do not result in decreased protein stability, except for the BRIC1 I661T mutation. None of the BRIC1/ICP mutants display impaired CDC50A interaction and all maintain a proper canalicular membrane localization allowing residual activity at its final destination. Our data provide an explanation for the difference in severity of the phenotype of PFIC1 and BRIC1 patients.

Acknowledgments

We thank Drs R.E. Pagano and D. Cassio for the cell lines, and Drs L.W.J. Klomp and J.B. Koenderink for the antibodies to ATP8B1 and ATP1A1.

Chapter 3

References

1. Bull LN, van Eijk MJ, Pawlikowska L, DeYoung JA, Juijn JA, Liao M, et al. A gene encoding a P-type ATPase mutated in two forms of hereditary cholestasis. Nat Genet 1998 Mar;18(3):219-224. 2. van Mil SW, Klomp LW, Bull LN, Houwen RH. FIC1 disease: a spectrum of intrahepatic cholestatic disorders. Semin Liver Dis 2001 Nov;21(4):535-544. 3. Lykavieris P, van MS, Cresteil D, Fabre M, Hadchouel M, Klomp L, et al. Progressive familial intrahepatic cholestasis type 1 and extrahepatic features: no catch-up of stature growth, exacerbation of diarrhea, and appearance of liver steatosis after liver transplantation. J Hepatol 2003 Sep;39(3):447-452. 4. Mullenbach R, Bennett A, Tetlow N, Patel N, Hamilton G, Cheng F, et al. ATP8B1 mutations in British cases with intrahepatic cholestasis of pregnancy. Gut 2005 Jun;54(6):829-834. 5. Painter JN, Savander M, Ropponen A, Nupponen N, Riikonen S, Ylikorkala O, et al. Sequence variation in the ATP8B1 gene and intrahepatic cholestasis of pregnancy. Eur J Hum Genet 2005 Apr;13(4):435-439. 6. Pusl T, Beuers U. Intrahepatic cholestasis of pregnancy. Orphanet J Rare Dis 2007 May;2:26. 7. van Mil SW, van Oort MM, van dB, I, Berger R, Houwen RH, Klomp LW. Fic1 is expressed at apical membranes of different epithelial cells in the digestive tract and is induced in the small intestine during postnatal development of mice. Pediatr Res 2004 Dec;56(6):981-987. 8. Eppens EF, van Mil SW, de Vree JM, Mok KS, Juijn JA, Oude Elferink RP, et al. FIC1, the protein affected in two forms of hereditary cholestasis, is localized in the cholangiocyte and the canalicular membrane of the hepatocyte. J Hepatol 2001 Oct;35(4):436-443. 9. Ujhazy P, Ortiz D, Misra S, Li S, Moseley J, Jones H, et al. Familial intrahepatic cholestasis 1: studies of localization and function. Hepatology 2001 Oct;34(4 Pt 1):768-775. 10. Paulusma CC, Groen A, Kunne C, Ho-Mok KS, Spijkerboer AL, Rudi de WD, et al. Atp8b1 deficiency in mice reduces resistance of the canalicular membrane to hydrophobic bile salts and impairs bile salt transport. Hepatology 2006 Jul;44(1):195-204. 11. Paulusma CC, Folmer DE, Ho-Mok KS, de Waart DR, Hilarius PM, Verhoeven AJ, et al. ATP8B1 requires an accessory protein for endoplasmic reticulum exit and plasma membrane lipid flippase activity. Hepatology 2008 Jan;47(1):268-278. 12. Cai SY, Gautam S, Nguyen T, Soroka CJ, Rahner C, Boyer JL. ATP8B1 Deficiency Disrupts the Bile Canalicular Membrane Bilayer Structure in Hepatocytes, but FXR Expression and Activity Are Maintained. Gastroenterology 2009 Mar;136(3):1060-1069. 13. Groen A, Kunne C, Jongsma G, van den OK, Mok KS, Petruzzelli M, et al. Abcg5/8 independent biliary cholesterol excretion in Atp8b1- deficient mice. Gastroenterology 2008 Jun;134(7):2091-2100. 14. Paulusma CC, de Waart DR, Kunne C, Mok KS, Elferink RP. Activity of the bile salt export pump (ABCB11) is critically dependent on canalicular membrane cholesterol content. J Biol Chem 2009 Apr 10;284(15):9947-9954. 15. Alvarez L, Jara P, Sanchez-Sabate E, Hierro L, Larrauri J, Diaz MC, et al. Reduced hepatic expression of farnesoid X receptor in hereditary cholestasis associated to mutation in ATP8B1. Hum Mol Genet 2004 Oct 15;13(20):2451-2460. 16. Chen F, Ananthanarayanan M, Emre S, Neimark E, Bull LN, Knisely AS, et al. Progressive familial intrahepatic cholestasis, type 1, is associated with decreased farnesoid X receptor activity. Gastroenterology 2004 Mar;126(3):756-764. 17. Frankenberg T, Miloh T, Chen FY, Ananthanarayanan M, Sun AQ, Balasubramaniyan N, et al. The membrane protein ATPase class I type 8B member 1 signals through protein kinase C zeta to activate the farnesoid X receptor. Hepatology 2008 May 27;48(6):1896-1905. 18. Seppen J, Rijnberg M, Cooreman MP, Oude Elferink RP. Lentiviral vectors for efficient transduction of isolated primary quiescent hepatocytes. J Hepatol 2002 Apr;36(4):459-465. 19. Koenderink JB, Geibel S, Grabsch E, De Pont JJ, Bamberg E, Friedrich T. Electrophysiological analysis of the mutated Na,K-ATPase cation binding pocket. J Biol Chem 2003 Dec 19;278(51):51213-51222. 20. Paulusma CC, Bosma PJ, Zaman GJ, Bakker CT, Otter M, Scheffer GL, et al. Congenital jaundice in rats with a mutation in a multidrug resistance-associated protein gene. Science 1996 Feb 23;271(5252):1126-1128. 21. Klomp LW, Vargas JC, van Mil SW, Pawlikowska L, Strautnieks SS, van Eijk MJ, et al. Characterization of mutations in ATP8B1 associated with hereditary cholestasis. Hepatology 2004 Jul;40(1):27-38. 22. Pawlikowska L, Groen A, Eppens EF, Kunne C, Ottenhoff R, Looije N, et al. A mouse genetic model for familial cholestasis caused by ATP8B1 mutations reveals perturbed bile salt homeostasis but no impairment in bile secretion. Hum Mol Genet 2004 Apr 15;13(8):881-892. 23. Moller JV, Juul B, le MM. Structural organization, ion transport, and energy transduction of P-type ATPases. Biochim Biophys Acta 1996 May 6;1286(1):1-51. 24. Klomp LW, Bull LN, Knisely AS, van Der Doelen MA, Juijn JA, Berger R, et al. A missense mutation in FIC1 is associated with greenland familial cholestasis. Hepatology 2000 Dec;32(6):1337-1341. 25. Catty P, de Kerchove dA, Goffeau A. The complete inventory of the yeast Saccharomyces cerevisiae P-type transport ATPases. FEBS Lett 1997 Jun 16;409(3):325-332. 26. Eiberg H, Norgaard-Pedersen B, Nielsen IM. Cholestasis Familiaris Groenlandica/Byler-like disease in Greenland-a population study. Int J Circumpolar Health 2004;63 Suppl 2:189-191. 27. Andersen S, Okkels H, Krarup H, Laurberg P. Geographical clustering and maintained health in individuals harbouring the mutation for Greenland familial cholestasis: A population-based study. Scand J Gastroenterol 2006 Apr;41(4):445-450. 28. Nielsen IM, Eiberg H. Cholestasis Familiaris Groenlandica: an epidemiological, clinical and genetic study. Int J Circumpolar Health 2004;63 Suppl 2:192-194. 29. Tygstrup N, Steig BA, Juijn JA, Bull LN, Houwen RH. Recurrent familial intrahepatic cholestasis in the Faeroe Islands. Phenotypic heterogeneity but genetic homogeneity. Hepatology 1999 Feb;29(2):506-508. 30. Axelsen KB, Palmgren MG. Evolution of substrate specificities in the P-type ATPase superfamily. J Mol Evol 1998 Jan;46(1):84-101. 31. Lutsenko S, Kaplan JH. Organization of P-type ATPases: significance of structural diversity. Biochemistry 1995 Dec 5;34(48):15607-15613.

Chapter 3

32. Decaens C, Rodriguez P, Bouchaud C, Cassio D. Establishment of hepatic cell polarity in the rat hepatoma-human fibroblast hybrid WIF-B9. A biphasic phenomenon going from a simple epithelial polarized phenotype to an hepatic polarized one. J Cell Sci 1996 Jun;109 ( Pt 6):1623- 1635. 33. Elferink MG, Olinga P, Draaisma AL, Merema MT, Faber KN, Slooff MJ, et al. LPS-induced downregulation of MRP2 and BSEP in human liver is due to a posttranscriptional process. Am J Physiol Gastrointest Liver Physiol 2004 Nov;287(5):G1008-G1016. 34. Muhlfeld A, Kubitz R, Dransfeld O, Haussinger D, Wettstein M. Taurine supplementation induces multidrug resistance protein 2 and bile salt export pump expression in rats and prevents endotoxin-induced cholestasis. Arch Biochem Biophys 2003 May 1;413(1):32-40. 35. Geier A, Dietrich CG, Voigt S, Ananthanarayanan M, Lammert F, Schmitz A, et al. Cytokine-dependent regulation of hepatic organic anion transporter gene transactivators in mouse liver. Am J Physiol Gastrointest Liver Physiol 2005 Nov;289(5):G831-G841. 36. Siewert E, Dietrich CG, Lammert F, Heinrich PC, Matern S, Gartung C, et al. Interleukin-6 regulates hepatic transporters during acute-phase response. Biochem Biophys Res Commun 2004 Sep 10;322(1):232-238.

Supplemental data

Supplementary figure 1. Confocal microscopical optical sections showing the localization of the G1040R mutant in WIF-B9 cells (left panel). Cells were co-stained for the canalicular marker protein ABCC2 (middle panel). The right panel show the merged image of ATP8B1G1040R-eGFP and ABCC2 (ATP8B1G1040R-eGFP in green; ABCC2 in red; nuclear DAPI staining in blue). The G1040R protein is clearly absent from the canalicular membrane and resides in the endoplasmic reticulum (white arrows). Bar equals 18.75 µm.

CHAPTER 4

The lipid flippase heterodimer ATP8B1-CDC50A is essential for surface expression of the apical sodium-dependent bile acid transporter (SLC10A2/ASBT) in intestinal Caco-2 cells

Vincent A. van der Mark, D. Rudi de Waart, Kam S. Ho-Mok, Merit M. Tabbers, Heleen W. Voogt, Ronald P. Oude Elferink, Alex S. Knisely, Coen C. Paulusma

Biochim Biophys Acta. 2014 Dec;1842(12 Pt A):2378-86

doi: 10.1016/j.bbadis.2014.09.003

Chapter 4

Abstract

Deficiency of the phospholipid flippase ATP8B1 causes progressive familial intrahepatic cholestasis type 1 (PFIC1) and benign recurrent intrahepatic cholestasis type 1 (BRIC1). Apart from cholestasis, many patients also suffer from diarrhea of yet unknown etiology. Here we have studied the hypothesis that intestinal ATP8B1 deficiency results in bile salt malabsorption as a possible cause of PFIC1/BRIC1 diarrhea. Bile salt transport was studied in ATP8B1-depleted intestinal Caco-2 cells. Apical membrane localization was studied by a biotinylation approach. Fecal bile salt and electrolyte contents were analysed in stool samples of PFIC1 patients, of whom some had undergone biliary diversion or liver transplantation. Bile salt uptake by the apical sodium-dependent bile salt transporter SLC10A2 was strongly impaired in ATP8B1-depleted Caco-2 cells. The reduced SLC10A2 activity coincided with strongly reduced apical membrane localization, which was caused by impaired apical membrane insertion of SLC10A2. Moreover, we show that endogenous ATP8B1 exists in a functional heterodimer with CDC50A in Caco-2 cells. Analyses of stool samples of post-transplant PFIC1 patients demonstrated that bile salt content was not changed, whereas sodium and chloride concentrations were elevated and potassium levels were decreased. The ATP8B1-CDC50A heterodimer is essential for the apical localization of SLC10A2 in Caco-2 cells. Diarrhea in PFIC1/BRIC1 patients has a secretory origin to which SLC10A2 deficiency may contribute. This results in elevated luminal bile salt concentrations and consequent enhanced electrolyte secretion and/or reduced electrolyte resorption.

Chapter 4

1. Introduction

Progressive familial intrahepatic cholestasis type 1 (PFIC1) and benign recurrent intrahepatic cholestasis type 1 (BRIC1) are rare autosomal recessive liver diseases caused by mutations in the gene encoding the P-type ATPase ATP8B1 (1, 2). ATP8B1 is a P4-ATPase and is a phospholipid flippase that translocates phosphatidylserine from the exoplasmic to the cytosolic leaflet of the plasma membrane (3, 4). ATP8B1 is expressed in the apical membrane of many epithelial cell types, including hepatocytes, cholangiocytes and enterocytes (5, 6). In vitro studies have indicated that ATP8B1 exists as a heterodimer with its β-subunit CDC50A; heterodimerization is pivotal for endoplasmic reticulum (ER) exit and for activity of the protein (3, 7-9). PFIC1 and BRIC1 patients present with jaundice and pruritus caused by impaired bile flow (10). While BRIC1 patients suffer from intermittent bouts of cholestasis and pruritus without liver scarring, PFIC1 patients have non-remitting and progressive liver disease leading to fibrosis and cirrhosis. Biliary diversion (BD) procedures relieve symptoms for BRIC1 and PFIC1 patients and generally suffice as therapy for hepatobiliary disease; however, for many PFIC1 patients liver transplantation (LTx) is the only curative therapy (11, 12). Extrahepatic symptoms such as diarrhea and hearing loss are common in PFIC1 and BRIC1 patients (10, 13, 14). After LTx, PFIC1 patients have a high risk (~85%) of developing an exacerbated form of diarrhea, possibly as a consequence of restored bile flow (15-19). Diarrhea is ameliorated upon treatment with bile salt-absorptive resins such as cholesevelam and cholestyramine (20). Although the etiology of diarrhea in ATP8B1-deficient patients is not known, the observations in liver-transplanted patients allow the hypothesis that intestinal bile salt malabsorption is the cause of diarrhea in these patients. In the intestine, approximately 95% of conjugated bile salts are reabsorbed via the apical sodium-dependent bile salt transporter (ASBT/SLC10A2), which is located in the apical membrane of enterocytes in the terminal ileum (21). Congenital ileal bile salt transport defects or ileal resection are usually associated with diarrhea. Indeed, mutations in SLC10A2 are the cause of primary bile acid malabsorption, an intestinal disorder associated with congenital diarrhea (22). Here we have studied the hypothesis that diarrhea in ATP8B1 deficiency is caused by intestinal bile salt malabsorption. Bile salt transport was studied in ATP8B1-depleted Caco-2 cells and fecal bile salt output was analysed in PFIC1 patients, of whom some had undergone BD or LTx.

Chapter 4

2. Materials and methods

2.1. Antibodies and reagents

Primary antibodies: rabbit polyclonals to SLC10A2 (kindly provided by Dr. P.A. Dawson) (23), CDC50A (24), ATP8B1 (25), ATP1A1 (26), CD13 (H-300, Santa Cruz Biotechnology), CD26 (A28340, Abcam), SLC2A2 (07-1402, Millipore), and mouse monoclonals to calnexin (H-70, Santa Cruz Biotechnology), P-glycoprotein

(C219, Abcam), ABCC2 (M2III6) (27), and villin (MAB1671, Millipore); SLC10A2-specific inhibitor S910960 (Aventis) (28). All other, non-specified chemicals and reagents were obtained from Sigma-Aldrich.

2.2. Cell culture and lentiviral transduction

The human colonic adenocarcinoma Caco-2 cell line (ATCC, HTB-37) was cultured in Dubecco’s Modified Eagle Medium (DMEM; Lonza) supplemented with 10% bovine fetal calf serum, 2mM L-glutamine, 100U/ml penicillin, and 100U/ml streptomycin at 37ºC in a 10% CO2 humidified atmosphere. Culture medium was refreshed twice per week. Stable knockdown cell lines were generated by lentiviral transduction with shRNA vectors from the Mission shRNA library (Sigma-Aldrich) as described (29). Vector TRCN0000050127 was used for depletion of ATP8B1, vector TRCN0000159317 was used for depletion of CDC50A. All experiments were performed on cells grown on microporous polycarbonate membrane filters (12 mm diameter Transwell, 0.4μm pore size, Corning) or on polystyrene tissue culture plates (Corning) at 3 weeks post-confluency (at least 4 weeks after seeding). Cells were used not more than 10 passages after lentiviral transduction at which point they were found to have a similar knockdown to those of earlier passages.

2.3. Quantitative RT-PCR

Total RNA was isolated using Trizol (Invitrogen). cDNA was synthesized with an oligo-dT primer and Superscript III RT (Invitrogen). RT-PCR was performed in a Lightcycler 480 (Roche) with SYBR-Green master mix (Roche). Expression levels were calculated with the LinregPCR program (30) using acidic ribosomal

Chapter 4 phosphoprotein P0 (36B4) as a reference gene. Primer sequences are available on request.

2.4. Cell surface biotinylation

Cells were washed with ice-cold phosphate buffered saline containing 1mM

MgCl2/0.1mM CaCl2 (PBS-CM) and incubated 2 x 30 minutes with 1 mg/ml sulfo- NH-ss biotin (Thermo-Scientific) in ice-cold borate buffer (10mM borate, 137mM

NaCl, 3.8mM KCl, 0.9mM CaCl2, 0.52mM MgCl2, 0.16mM MgSO4 pH9.0). To quench non-bound biotin, cells were washed twice with PBS-CM containing 100 mM glycine for 5 minutes. Cells were lysed in RIPA buffer (50mM Tris-HCl, pH8.0, 150mM NaCl, 1% NP-40, 0.5% Na-deoxycholate, 0.1% SDS) containing protease inhibitor cocktail (Roche). Biotinylated protein was precipitated with neutravidin agarose resin (Thermo-Scientific) overnight at 4°C. Beads were extensively washed with RIPA buffer. Biotinylated protein was eluted with SDS-containing sample buffer. Samples were analyzed by SDS-PAGE and Western blotting.

2.5. Biotinylation pulse-chase assay

Cells were washed four times with ice-cold PBS-CM and biotin binding sites were blocked on ice with 1.5 mg/ml NHS-acetate (Thermo-Scientific) in PBS-CM for two times 45 minutes. Cells were washed twice with ice-cold PBS-CM containing 100 mM glycine for 10 minutes and were incubated at 37°C in Hanks’ balanced salt solution (HBSS; Lonza) for the indicated periods. Biotinylated proteins were isolated and detected as described above.

2.6. SDS-PAGE and Western blotting

Proteins were separated on a poly-acrylamide gel and transferred to a polyvinylidene difluoride (Immobilon-P) membrane. Membranes were blocked for 1 hour in PBS/0.05% Tween-20 containing 4% low-fat milk (Nutricia Profitar-plus) (block buffer), and incubated for 1 hour with primary antibodies in block buffer, and 1 hour with HRP-conjugated secondary antibodies (Bio-Rad) in PBS/0.05% Tween-20. After antibody incubations, blots were extensively washed in PBS/0.05% Tween-20. Immune complexes were visualized with Lumi-Light Western Blotting Substrate

Chapter 4

(Roche). Chemiluminescence was detected and quantified with a Lumi-Imager F1 and LumiAnalyst 3.1 software (Roche).

2.7. Co-immunoprecipitation

Cells were washed with ice-cold PBS, scraped into suspension, centrifuged and resuspended in hypotonic lysis buffer (1mM NaHCO3 supplemented with protease inhibitors) on ice for 30 minutes. Cells were pestle-homogenized (30x, tight pestle) and spun down. The supernatant was centrifuged for 1 hour at 45,000 rpm (Ti70 rotor) in an Optima L-90K ultracentrifuge (Beckman/Coulter). The membrane pellet was resuspended in RIPA buffer. The membrane lysates was precleared with Protein G sepharose 4 fast flow beads (GE Healthcare) and goat-serum IgG for 1 hour at 4°C. Cleared supernatants were incubated overnight at 4°C with affinity-purified anti-CDC50A antibody or control IgG. CDC50A-associated protein was immunoprecipitated by Protein G sepharose incubation for 2 hours at 4°C. Beads were extensively washed with RIPA buffer and immunoprecipitates were eluted with 1% SDS-containing sample buffer. Samples were analyzed by SDS-PAGE and Western blotting.

2.8. Taurocholate uptake assays

Caco-2 cells were incubated at 37°C with 1 μCi/ml [3H]-[G]-taurocholate (Perkin Elmer) in incubation buffer (143 mM NaCl or choline chloride (in case of sodium- free buffer), 5.4 mM KCl, 1.3 mM CaCl2, 0.4 mM MgSO4, 0.5 mM MgCl2, 0.44 mM

KH2PO4, 10 mM Tris-HEPES, pH7.4, 5.4mM D-glucose) supplemented with 20μM taurocholate (Calbiochem). Cells were washed with PBS and lysed in RIPA buffer and radioactivity was counted in a Packard Tri-carb 2900TR Liquid Scintillation Analyzer (GMI).

2.9. 2,4-dinitrophenyl-S-glutathione (DNP-GS) excretion assay

DNP-GS excretion, as a measure of ABCC2 activity, was determined as described (31). Briefly, Caco-2 cells were incubated at 37°C from the apical and basolateral compartment with the hydrophobic compound 0.5 μCi/ml [14C]-labeled 1-chloro- 2,4-dinitrobenzene (CDNB; GE/Healthcare) in HBSS supplemented with 20 mM Tris-buffered HEPES, pH 7.4. Samples were taken at the indicated time points and

Chapter 4 were extracted with ethylacetate to separate the hydrophobic CDNB from its hydrophilic metabolite DNP-GS. Radioactivity was measured by liquid-scintillation counting.

2.10. Patients

Twenty-four hour stool collections were obtained from 15 Amish PFIC1 patients with documented ATP8B1 mutation (p.G308V homozygotes) among whom six had undergone LTx, six had undergone external BD and three had received no surgical treatment. Twenty-four hour stool collections from 20 healthy, age-matched controls were obtained in The Netherlands. The study was approved by the AMC medical ethical committee. Informed consent was obtained from all parents of affected children and from all adult participants.

2.11. Fecal bile salt and electrolyte measurements

Bile salts were extracted from feces according to the method of Bligh and Dyer (32). Fecal bile salts were determined in the extracted supernatant by reverse-phase high- performance liquid chromatography (HPLC) (33). Electrolytes and glucose concentrations were measured in fecal fluid by routine clinical chemistry procedures. Stool osmotic gap (expressed in mOsmol/kg) calculations were performed according + + to Binder (34) using the following equation: 290-2*([Na ]stool +[K ]stool), in which + 290 is the fecal fluid osmolality reference value (mOsmol/kg) (35) and [Na ]stool and + [K ]stool the measured fecal sodium and potassium concentrations (in mM).

2.12. Statistics

Data are expressed as mean ± SD and were generated with Prism 5 (Graphpad). Statistical significance was determined by performing Student’s t-test or one-way ANOVA with Bonferroni’s correction for multiple testing as indicated in the legend of the figures.

Chapter 4

3. Results

3.1. SLC10A2-mediated taurocholate uptake is impaired in ATP8B1-depleted Caco- 2 cells

To study the consequences of ATP8B1 deficiency for bile salt transport, we constructed ATP8B1-depleted Caco-2 cells by lentiviral transduction and compared them with cells transduced with a scrambled shRNA control (figures 1A, B). ATP8B1-depleted cells displayed normal growth and differentiation, the latter evidenced by the presence of apical microvilli (figure 1C). Quantitative RT-PCR analyses revealed no significant changes between control and ATP8B1-depleted cells for SLC10A2 (figure 1D), nor for the bile salt receptor farnesoid-X receptor (FXR) and short heterodimeric partner (SHP) (supplementary figure 1).

Figure 1. ATP8B1-depleted Caco-2 cells display strongly reduced ATP8B1 expression without loss of apical microvilli. (A) Relative ATP8B1 mRNA levels in shRNA control and ATP8B1-depleted Caco-2 cells (n=4). Data are normalized for 36B4. Statistical significance determined by Student’s t-test. ** p<0.01 (B) Total cell lysates of control or ATP8B1-depleted cells were analyzed by immunoblotting for ATP8B1 and ATP1A1 (loading control). Six different isolations are shown. (C) Transmission electron microscopical images of Transwell-grown shRNA control

Chapter 4

and ATP8B1-depleted Caco-2 cells. N; nucleus, mp; microporous membrane. Bar represents 5 µm. (D) Relative SLC10A2 mRNA levels in shRNA control and ATP8B1-depleted Caco-2 cells (n=4). Data are normalized for 36B4.

Analysis of time-dependent taurocholate (TC) uptake revealed that intracellular TC accumulation was strongly reduced in ATP8B1-depleted cells (figure 2A). ATP8B1- depleted cells grown on Transwell inserts displayed strongly impaired apical TC uptake, evidenced by reduced intracellular accumulation and basolateral appearance of TC (figures 2B, C). The reduced TC uptake involved a sodium-dependent system, since replacement of sodium by choline led to similar low levels of TC uptake in control and knockdown cells (figure 2D), suggesting a defect in SLC10A2 activity. Indeed, TC uptake was inhibited upon co-incubation with the SLC10A2-specific inhibitor S910960 (figure 2E). These results demonstrate that SLC10A2 activity is strongly impaired in ATP8B1-depleted Caco-2 cells.

Chapter 4

Figure 2. SLC10A2-dependent uptake of taurocholate is reduced in ATP8B1-depleted Caco-2 cells. (A) shRNA control or ATP8B1-depleted cells grown on plastic were incubated with 20 µM taurocholate (TC) and a trace amount of [3H]TC ([3H]TC/TC) for the indicated time periods. Intracellular TC (mol/mg protein) is plotted in time. (B) Intracellular TC accumulation in Transwell-grown Caco-2 cells 15 and 30 minutes after apical administration of [3H]TC/TC (n=4). (C) TC in the basolateral medium in the experiment shown in Fig. 2B. (D) Cells grown on plastic were incubated in the absence (cholineCl) and presence of sodium (NaCl) with [3H]TC/TC for 30 minutes. Intracellular accumulation of TC is plotted (n=5). (E) Competitive inhibition of bile salt uptake by the SLC10A2- specific inhibitor S910960 (100 µM) in cells grown on plastic (n=3). Cells were incubated for 30 minutes with [3H]TC/TC. Intracellular accumulation of TC is plotted. Statistical significance determined by Student’s t-test. *p<0.05, ***p<0.001.

Chapter 4

3.2. Apical membrane expression and insertion of SLC10A2 is strongly reduced in ATP8B1-depleted Caco-2 cells

Next, SLC10A2 protein expression levels were analyzed. Compared to human ileum, SLC10A2 expression is very low in Caco-2 cells, appearing as two bands on a Western blot (figure 3A). These likely represent the core- and terminal- glycosylated forms of the protein (36). Membrane-associated SLC10A2 was quantified by cell surface biotinylation. Whereas total SLC10A2 expression was not affected, membrane-associated SLC10A2 was hardly detectable in ATP8B1- depleted cells (figure 3B). The expression level of the ER resident calnexin was assessed as a control for proper surface biotinylation. Apparently, biotin could pass the tight junctions since the basolateral membrane resident ATP1A1 was also detected in the biotinylated fraction.

Figure 3. Apical membrane expression and insertion of SLC10A2 is strongly reduced in ATP8B1-depleted Caco- 2 cells. (A) Cell lysates of Chinese Hamster Ovary (CHO) cells, Caco-2 cells and human ileum were immunoblotted for SLC10A2. (B) Total cell lysates and biotinylated surface fractions of shRNA control and ATP8B1-depleted Caco- 2 cells were analyzed by immunoblotting for SLC10A2. ATP1A1 and calnexin were included as loading- and biotinylation control, respectively. Quantification of SLC10A2 protein expression is included next to its respective blot. (C) Biotinylation sites on membrane proteins were blocked with NHS-acetate. Cells were subsequently incubated at 37°C for the indicated time periods (t = 0, 1, 2 hour) and then incubated with biotin for 1 hour on ice,

Chapter 4

after which cells were harvested and analyzed by immunoblotting for SLC10A2 and ATP1A1. (D) Quantification of reappearance of SLC10A2 (n=3). Statistical significance was determined by a Student’s t-test. ***p<0.001

Next, we quantified the re-appearance of SLC10A2 on the apical membrane by a biotinylation approach. First, biotin binding sites on surface proteins were blocked with NHS-acetate. Subsequently, cells were chased for the indicated times to allow insertion of apical proteins from the intracellular pool, after which the surface was labeled with biotin. While in the control cells surface SLC10A2 expression was completely normalized after a one hour chase (figures 3C, D), we could not detect any increase in SLC10A2 membrane expression in ATP8B1- depleted cells. Apparently, NHS-acetate could not pass the tight junctions since the ATP1A1 signal was not reduced after NHS-acetate incubation (figure 3C, t=0h). These results strongly suggest that ATP8B1 plays an essential role in the apical delivery of SLC10A2.

3.3. Apical organic anion excretion is unaffected in ATP8B1-depleted Caco-2 cells

To determine if ATP8B1 depletion caused a more general defect in apical membrane protein function, we examined the activity of multidrug resistance-associated protein 2 (ABCC2/MRP2), an apical protein involved in the secretion of anionic compounds, in cells grown on Transwell inserts. Measurement of apically secreted DNP-GS showed that ABCC2 activity was not affected in ATP8B1-depleted Caco-2 cells (figure 4).

Chapter 4

Figure 4. Apical ABCC2-mediated organic anion excretion is unaffected in ATP8B1-depleted Caco-2 cells. Transwell-grown shRNA control and ATP8B1-depleted cells were incubated with CDNB. At indicated time points samples were taken and extracted with ethylacetate to separate [14C]CDNB from its metabolite [14C]DNP-GS. Apical secretion of DNP-GS is plotted in time.

3.4. ATP8B1 and CDC50A form a functional heterodimer in Caco-2 cells

Using an over-expression approach, we and others have previously shown that CDC50A and ATP8B1 heterodimerize (3, 8, 9). We analyzed whether endogenous CDC50A and ATP8B1 were present as a heterodimer in Caco-2 cells. CDC50A was immunoprecipitated from crude membranes using affinity-purified CDC50A antibody and the immunoprecipitates were analyzed for the presence of ATP8B1 by Western blotting (figure 5A). ATP8B1 did not co-immunoprecipitate with a rabbit IgG control antibody. However, ATP8B1 was co-immunoprecipitated in CDC50A immunoprecipitates, demonstrating that in Caco-2 cells endogenous CDC50A and ATP8B1 are present as a heterodimer.

Chapter 4

Figure 5. ATP8B1 and CDC50A exist as a functional heterodimer in Caco-2 cells. (A) ATP8B1 was co- immunoprecipitated with affinity-purified antibody to CDC50A from Caco-2 crude membrane fractions. ip; immunoprecipitates, ft; flow through. Two different concentrations of anti-CDC50A serum were used. Rabbit IgGs were included as a negative control. (B) shRNA control and CDC50A-depleted cells grown on plastic were incubated with [3H]TC/TC for the indicated time periods. Intracellular TC is plotted in time. Statistical significance was determined by a Student’s t-test. ***p<0.001

To analyze the functional relevance of this interaction, we studied bile salt uptake in CDC50A-depleted cells. Indeed, and analogous to ATP8B1-depleted cells, TC accumulation was significantly reduced in CDC50A-depleted cells (figure 5B). In line with this, plasma membrane expression of both ATP8B1 and SLC10A2 was strongly reduced in CDC50A-depleted cells (figure 6). Intriguingly, surface expression of other apical residents, including CD13, CD26, ABCC2 and P- glycoprotein, was not affected in ATP8B1- and CDC50A-depleted cells. These data indicate that the ATP8B1-CDC50A heterodimer is essential in the apical delivery of SLC10A2 protein.

Chapter 4

Figure 6. Expression of surface proteins in ATP8B1- and CDC50A-depleted Caco-2 cells. ATP8B1-, CDC50A- depleted and shRNA control cells were biotinylated and analyzed for expression of several apical and basolateral resident proteins. Total and surface protein fractions are indicated. ATP1A1 and calnexin were included as loading and biotinylation control, respectively.

3.5. PFIC1 patients have a secretory diarrhea

Our in vitro findings supported us in hypothesizing that intestinal bile salt malabsorption is a major determinant of diarrhea in PFIC1/BRIC1 patients. Based on this hypothesis, we expect increased fecal bile salt losses, especially after LTx, when diarrhea usually is aggravated. We determined bile salt content in 24-hour stool samples of 15 PFIC1 patients of whom six had undergone LTx, six external BD and three had received no surgical treatment. All LTx patients and two BD patients had (semi-)liquid stools. Four BD patients, all untreated patients and all age-matched,

Chapter 4 healthy controls had solid stools. Two LTx patients and one BD patient received anti-diarrhetic treatment with clonidine (an α2 adrenergic agonist that prolonges intestinal transit time to facilitate greater absorption (37, 38)), however, their stools were still (semi-)liquid in appearance. We analyzed bile salt, glucose and electrolyte content and plotted this to the different treatments (figure 7). Although fecal bile salt output tended to be higher in the LTx group, this only significantly differed from the BD group (figure 7A). Fecal glucose output was also unaffected (figure 7B). However, fecal sodium (figure 7C) and chloride (figure 7D) concentrations were significantly increased whereas potassium (figure 7E) was significantly decreased in the LTx group compared to the other groups. Sodium (but not chloride) was also mildly elevated in the BD and untreated PFIC1 patients compared to healthy controls (figure 7C). To analyze whether the severity of diarrhea is related to fecal concentrations of bile salts and/or electrolytes, we graded the stools according to the Bristol stool chart (39). These concentrations were plotted against stool form types (supplemental figure 2). Indeed, higher fecal bile salt and electrolyte output corresponds with a higher Bristol stool type. From the fecal fluid data of the LTx patients with diarrhea, we calculated the stool osmotic gap, an indicative value for the cause of diarrhea. An osmotic gap below 50 mOsm/kg indicates a secretory diarrhea while a larger osmotic gap indicates an osmotic origin (34). The mean osmotic gap was -4±47 mOsm/kg for LTx. These data suggest that diarrhea in PFIC1 patients is caused by intestinal hypersecretion and/or malabsorption of electrolytes and water.

Chapter 4

Figure 7. Fecal bile salt and electrolyte analyses are indicative for a secretory diarrhea in PFIC1 patients. 24- Hour stool samples of PFIC1 patients who had undergone liver transplantation (LTx, six patients), biliary diversion (BD, six patients) or were not treated (NO, three patients) and age-matched, healthy controls (20 subjects) were extracted and analyzed for bile salts. Sodium, potassium, chloride and glucose concentrations were determined in fecal fluids. Patients treated with the anti-diarrhetic compound clonidine are indicated in red. (A) Fecal bile salt excretion. (B) Glucose concentration. (C) Sodium concentrations are significantly increased in PFIC1 LTx patients and also moderately increased in BD and NO PFIC1 patients when compared to controls. (D) Chloride

Chapter 4

concentrations are significantly increased in LTx PFIC1 patients compared to controls and to BD and NO PFIC1 patients. (E) Potassium concentrations are significantly reduced in LTx PFIC1 patients when compared to controls and to BD and NO PFIC1 patients. Statistical significance was determined by one-way ANOVA with Bonferroni’s post-hoc correction. *p<0.05, **p<0.01, ***p<0.001

4. Discussion

Here we show that ATP8B1 is essential for apical membrane localization of the apical sodium-dependent bile acid transporter SLC10A2 in Caco-2 cells. Little is known about trafficking itineraries and regulation of SLC10A2 in epithelial cells. Sarwar and colleagues have presented data that suggest a regulatory role of PKC in SLC10A2 recycling; PKCζ activation in Caco-2 cells resulted in reduced apical membrane localization of SLC10A2 (40). Using primary rat cholangiocytes, Alpini and coworkers showed that SLC10A2 was recruited to the apical membrane from intracellular stores upon stimulation with secretin (41). In our ATP8B1-depleted cells, strongly reduced SLC10A2 localization coincided with unchanged total SLC10A2 protein and mRNA content, which suggests a post-translational defect. ATP8B1 depletion thus may impair the transport of SLC10A2 to the apical membrane. This may concern newly synthesized SLC10A2 on its way from the TGN or recycling of SLC10A2 between the membrane and a subapical vesicle pool. Alternatively, ATP8B1 deficiency may interfere with the biophysical properties of the apical membrane, leading to reduced SLC10A2 membrane expression. We and others have previously proposed such a role for ATP8B1 in the canalicular membrane of hepatocytes (42, 43). Chen and colleagues showed induced SLC10A2 transcript levels in ATP8B1-depleted Caco-2 cells (44). The authors hypothesized that ATP8B1 deficiency leads to impaired FXR signaling, which leads to reduced SHP expression and consequent induction of SLC10A2. In our ATP8B1-depleted cells, however, none of these transcript levels were different from control cells. One explanation for this discrepancy can be differences in the generation of ATP8B1-depleted cell lines; Chen and colleagues transiently depleted ATP8B1 using a siRNA approach, whereas our study was performed in stable cell lines. Another explanation can be the use of different Caco-2 cell clones and/or culture conditions, which can affect morphology and function of Caco-2 cells (45). An essential difference, however, between the studies is that we have analyzed the consequences of ATP8B1 depletion on endogenous SLC10A2 activity, expression and localization whereas Chen and colleagues studied ectopic SLC10A2 promoter regulation. Chen and colleagues also

Chapter 4 analyzed intestinal biopsy specimens from three PFIC1 patients and found reduced FXR and SHP levels accompanied by induced SLC10A2 (44). Since biopsies from two cholestatic non-PFIC1 patients did not show elevated SLC10A2 expression, induced expression in PFIC1 samples cannot be explained by reduced luminal bile salt concentrations as this is also expected in the non-PFIC1 patient samples. However, an explanation for SLC10A2 induction in PFIC1 samples can be that, in line with our present findings, SLC10A2 localization is impaired, which results in low intracellular bile salt levels, reduced FXR signaling and SHP expression with consequent induction of SLC10A2. A surprising finding of our study is that, of all proteins for which apical membrane expression was analyzed, only that of SLC10A2 apical membrane expression was impaired. Verhulst and colleagues previously showed that (membrane) expression of several apical proteins, including CD13 and alkaline phosphatase (AP), was strongly reduced in ATP8B1-depleted Caco-2 cells (46). Future analyses of the apical membrane proteome should elucidate additional proteins that are differentially expressed in ATP8B1-depleted Caco-2 cells. Along with deficient synthesis of apical proteins, Verhulst and colleagues also described loss of microvilli, a phenotype not observed in our ATP8B1-depleted Caco-2 cells. These discrepant findings may be attributed to artifacts introduced by clonal selection of cells, to use of different Caco-2 cell clones, to variation in culture conditions or to combinations of these factors. One must note that enterocytes of ATP8B1-deficient mice do not lose microvilli (Paulusma et al., unpublished observation). Our observation of reduced SLC10A2 activity in ATP8B1-depleted Caco-2 cells sheds new light on the diarrhea that is observed in PFIC1 patients. The etiology of diarrhea in these patients is poorly understood. During cholestatic periods, the intestines of PFIC1 patients encounter reduced concentrations of bile salts. However, after restoration of bile flow, i.e. after LTx, high concentrations of bile salts enter the lumen of the intestine, which in combination with bile salt malabsorption may result in spillover of bile salts to the colon and in exacerbation of diarrhea that can be ameliorated by application of bile salt-binding resins (15-19). This favors the hypothesis that intestinal bile salt malabsorption may be the underlying cause of diarrhea. Indeed, Egawa and colleagues showed that fecal bile salt loss was increased in three of six transplanted PFIC1 patients with diarrhea (16). Treatment with bile salt-binding resins improved their condition (i.e., led to normal stool and stooling patterns), but also improved the condition of patients without increased stool bile salt

Chapter 4 concentrations. In addition, Bijleveld and colleagues demonstrated significantly augmented fecal bile salt losses in seven of 10 BRIC1 patients (47). In our present study, all six liver-transplanted patients had diarrhea. In this small patient group only one patient displayed elevated fecal bile salt loss. Still, our observations, and those previously published strongly suggest that intestinal bile salt malabsorption is an important determinant of PFIC1/BRIC1 diarrhea. Without access to intestinal biopsy specimens from PFIC1/BRIC1 patients, however, it remains to be established whether apical SLC10A2 localization is reduced in enterocytes of these patients. In all six LTx patients, fecal sodium and chloride levels were significantly increased, while potassium levels were decreased. In case of SLC10A2 dysfunction, spillover of bile salts to the colon will lead to high luminal concentrations of dihydroxy bile salts, which can induce secretion of water and electrolytes. Mekhjian and colleagues have shown that perfusion of human large intestine with dihydroxy bile salts such as deoxycholic acid and chenodeoxycholic acid (CDCA) actively induces water, sodium and chloride secretion (48). In addition, CDCA has been shown to inhibit sodium absorption in the human colon (49). Thus, in PFIC1 patients, elevated colonic bile salt levels may interfere with the activity of electrolyte transporters, resulting in elevated fecal sodium and chloride concentrations. Strikingly, the effect of ATP8B1 deficiency on SLC10A2 function appears to differ importantly between man and mouse. We have previously demonstrated that, in contrast to ATP8B1-depleted Caco-2 cells, ATP8B1-deficient mice have no defect in SLC10A2 function (50). Apparently, SLC10A2 localization in mouse and human enterocytes is regulated by distinct mechanisms. The mouse and human SLC10A2 sequences deviate in the C-terminus, which was shown to be essential for apical localization of rat SLC10A2 (51). Alternatively, SLC10A2 localization in mouse and man may be regulated by different P4-ATPases, or in mice redundant activity of other P4-ATPases may compensate for the loss of ATP8B1. In conclusion, our results indicate that a functional ATP8B1-CDC50A heterodimer is essential for apical membrane localization of SLC10A2 in Caco-2 cells. Furthermore, PFIC1/BRIC1 diarrhea has a secretory origin to which intestinal bile salt malabsorption can contribute. Whether ileal SLC10A2 expression and activity are impaired in these patients remains to be determined in intestinal biopsy specimens.

Chapter 4

Acknowledgements

We thank Drs. P.A. Dawson, J.B. Koenderink and H.R. de Jonge respectively for kindly providing us with antibodies against SLC10A2 and ATP1A1 and with an inhibitor to SLC10A2. We thank Ms. L. Rose and Dr. B. Dahms for facilitating stool collection from Amish PFIC1 patients.

References

1. Bull LN, van Eijk MJ, Pawlikowska L, DeYoung JA, Juijn JA, Liao M, et al. A gene encoding a P-type ATPase mutated in two forms of hereditary cholestasis. NatGenet. 1998;18(3):219-24. 2. Klomp LW, Vargas JC, van Mil SW, Pawlikowska L, Strautnieks SS, van Eijk MJ, et al. Characterization of mutations in ATP8B1 associated with hereditary cholestasis. Hepatology. 2004;40(1):27-38. 3. Paulusma CC, Folmer DE, Ho-Mok KS, de Waart DR, Hilarius PM, Verhoeven AJ, et al. ATP8B1 requires an accessory protein for endoplasmic reticulum exit and plasma membrane lipid flippase activity. Hepatology. 2008;47(1):268-78. 4. van der Mark VA, Elferink RP, Paulusma CC. P4 ATPases: Flippases in Health and Disease. IntJMolSci. 2013;14(4):7897-922. 5. Eppens EF, van Mil SW, de Vree JM, Mok KS, Juijn JA, Oude Elferink RP, et al. FIC1, the protein affected in two forms of hereditary cholestasis, is localized in the cholangiocyte and the canalicular membrane of the hepatocyte. JHepatol. 2001;35(4):436-43. 6. van Mil SW, van Oort MM, van den Berg IE, Berger R, Houwen RH, Klomp LW. Fic1 is expressed at apical membranes of different epithelial cells in the digestive tract and is induced in the small intestine during postnatal development of mice. PediatrRes. 2004;56(6):981-7. 7. Bryde S, Hennrich H, Verhulst PM, Devaux PF, Lenoir G, Holthuis JC. CDC50 proteins are critical components of the human class-1 P4- ATPase transport machinery. JBiolChem. 2010;285(52):40562-72. 8. Takatsu H, Baba K, Shima T, Umino H, Kato U, Umeda M, et al. ATP9B, a P4-ATPase (a putative aminophospholipid translocase), localizes to the trans-Golgi network in a CDC50 protein-independent manner. JBiolChem. 2011;286(44):38159-67. 9. van der Velden LM, Wichers CG, van Breevoort AE, Coleman JA, Molday RS, Berger R, et al. Heteromeric interactions required for abundance and subcellular localization of human CDC50 proteins and class 1 P4-ATPases. JBiolChem. 2010;285(51):40088-96. 10. Paulusma CC, Elferink RP, Jansen PL. Progressive familial intrahepatic cholestasis type 1. SeminLiver Dis. 2010;30(2):117-24. 11. Bustorff-Silva J, Sbraggia NL, Olimpio H, de Alcantara RV, Matsushima E, De Tommaso AM, et al. Partial internal biliary diversion through a cholecystojejunocolonic anastomosis--a novel surgical approach for patients with progressive familial intrahepatic cholestasis: a preliminary report. JPediatrSurg. 2007;42(8):1337-40. 12. Emond JC, Whitington PF. Selective surgical management of progressive familial intrahepatic cholestasis (Byler's disease). JPediatrSurg. 1995;30(12):1635-41. 13. Pawlikowska L, Strautnieks S, Jankowska I, Czubkowski P, Emerick K, Antoniou A, et al. Differences in presentation and progression between severe FIC1 and BSEP deficiencies. JHepatol. 2010;53(1):170-8. 14. Stapelbroek JM, Peters TA, van Beurden DH, Curfs JH, Joosten A, Beynon AJ, et al. ATP8B1 is essential for maintaining normal hearing. ProcNatlAcadSciUSA. 2009;106(24):9709-14. 15. Davit-Spraul A, Fabre M, Branchereau S, Baussan C, Gonzales E, Stieger B, et al. ATP8B1 and ABCB11 analysis in 62 children with normal gamma-glutamyl transferase progressive familial intrahepatic cholestasis (PFIC): phenotypic differences between PFIC1 and PFIC2 and natural history. Hepatology. 2010;51(5):1645-55. 16. Egawa H, Yorifuji T, Sumazaki R, Kimura A, Hasegawa M, Tanaka K. Intractable diarrhea after liver transplantation for Byler's disease: successful treatment with bile adsorptive resin. Liver Transpl. 2002;8(8):714-6. 17. Hori T, Egawa H, Takada Y, Ueda M, Oike F, Ogura Y, et al. Progressive familial intrahepatic cholestasis: a single-center experience of living-donor liver transplantation during two decades in Japan. ClinTransplant. 2011;25(5):776-85. 18. Lykavieris P, van MS, Cresteil D, Fabre M, Hadchouel M, Klomp L, et al. Progressive familial intrahepatic cholestasis type 1 and extrahepatic features: no catch-up of stature growth, exacerbation of diarrhea, and appearance of liver steatosis after liver transplantation. JHepatol. 2003;39(3):447-52. 19. Miyagawa-Hayashino A, Egawa H, Yorifuji T, Hasegawa M, Haga H, Tsuruyama T, et al. Allograft steatohepatitis in progressive familial intrahepatic cholestasis type 1 after living donor liver transplantation. Liver Transpl. 2009;15(6):610-8. 20. Walters JR, Pattni SS. Managing bile acid diarrhoea. TherapAdvGastroenterol. 2010;3(6):349-57. 21. Dawson PA, Oelkers P. Bile acid transporters. CurrOpinLipidol. 1995;6(2):109-14. 22. Oelkers P, Kirby LC, Heubi JE, Dawson PA. Primary bile acid malabsorption caused by mutations in the ileal sodium-dependent bile acid transporter gene (SLC10A2). JClinInvest. 1997;99(8):1880-7. 23. Jung D, Fantin AC, Scheurer U, Fried M, Kullak-Ublick GA. Human ileal bile acid transporter gene ASBT (SLC10A2) is transactivated by the glucocorticoid receptor. Gut. 2004;53(1):78-84. 24. Folmer DE, Mok KS, de Wee SW, Duijst S, Hiralall JK, Seppen J, et al. Cellular localization and biochemical analysis of mammalian CDC50A, a glycosylated beta-subunit for P4 ATPases. JHistochemCytochem. 2012;60(3):205-18. 25. Pawlikowska L, Groen A, Eppens EF, Kunne C, Ottenhoff R, Looije N, et al. A mouse genetic model for familial cholestasis caused by ATP8B1 mutations reveals perturbed bile salt homeostasis but no impairment in bile secretion. HumMolGenet. 2004;13(8):881-92. 26. Koenderink JB, Geibel S, Grabsch E, De Pont JJ, Bamberg E, Friedrich T. Electrophysiological analysis of the mutated Na,K-ATPase cation binding pocket. JBiolChem. 2003;278(51):51213-22. 27. Paulusma CC, Bosma PJ, Zaman GJ, Bakker CT, Otter M, Scheffer GL, et al. Congenital jaundice in rats with a mutation in a multidrug resistance-associated protein gene. Science. 1996;271(5252):1126-8. 28. Bijvelds MJ, Jorna H, Verkade HJ, Bot AG, Hofmann F, Agellon LB, et al. Activation of CFTR by ASBT-mediated bile salt absorption. AmJPhysiol GastrointestLiver Physiol. 2005;289(5):G870-G9.

Chapter 4

29. Seppen J, Rijnberg M, Cooreman MP, Oude Elferink RP. Lentiviral vectors for efficient transduction of isolated primary quiescent hepatocytes. JHepatol. 2002;36(4):459-65. 30. Ruijter JM, Ramakers C, Hoogaars WM, Karlen Y, Bakker O, van den Hoff MJ, et al. Amplification efficiency: linking baseline and bias in the analysis of quantitative PCR data. Nucleic Acids Res. 2009;37(6):e45. 31. Paulusma CC, van Geer MA, Evers R, Heijn M, Ottenhoff R, Borst P, et al. Canalicular multispecific organic anion transporter/multidrug resistance protein 2 mediates low-affinity transport of reduced glutathione. BiochemJ. 1999;338 ( Pt 2):393-401. 32. BLIGH EG, DYER WJ. A rapid method of total lipid extraction and purification. CanJBiochemPhysiol. 1959;37(8):911-7. 33. Kunne C, Acco A, Hohenester S, Duijst S, de Waart DR, Zamanbin A, et al. Defective bile salt biosynthesis and hydroxylation in mice with reduced cytochrome P450 activity. Hepatology. 2013;57(4):1509-17. 34. Binder HJ. The gastroenterologist's osmotic gap: fact or fiction? Gastroenterology. 1992;103(2):702-4. 35. Topazian M, Binder HJ. Brief report: factitious diarrhea detected by measurement of stool osmolality. NEnglJMed. 1994;330(20):1418-9. 36. Zhang EY, Phelps MA, Banerjee A, Khantwal CM, Chang C, Helsper F, et al. Topology scanning and putative three-dimensional structure of the extracellular binding domains of the apical sodium-dependent bile acid transporter (SLC10A2). Biochemistry. 2004;43(36):11380-92. 37. Rubinoff MJ, Piccione PR, Holt PR. Clonidine prolongs human small intestine transit time: use of the lactulose-breath hydrogen test. AmJGastroenterol. 1989;84(4):372-4. 38. Schiller LR, Santa Ana CA, Morawski SG, Fordtran JS. Studies of the antidiarrheal action of clonidine. Effects on motility and intestinal absorption. Gastroenterology. 1985;89(5):982-8. 39. Lewis SJ, Heaton KW. Stool form scale as a useful guide to intestinal transit time. ScandJGastroenterol. 1997;32(9):920-4. 40. Sarwar Z, Annaba F, Dwivedi A, Saksena S, Gill RK, Alrefai WA. Modulation of ileal apical Na+-dependent bile acid transporter ASBT by protein kinase C. AmJPhysiol GastrointestLiver Physiol. 2009;297(3):G532-G8. 41. Alpini G, Glaser S, Baiocchi L, Francis H, Xia X, Lesage G. Secretin activation of the apical Na+-dependent bile acid transporter is associated with cholehepatic shunting in rats. Hepatology. 2005;41(5):1037-45. 42. Cai SY, Gautam S, Nguyen T, Soroka CJ, Rahner C, Boyer JL. ATP8B1 deficiency disrupts the bile canalicular membrane bilayer structure in hepatocytes, but FXR expression and activity are maintained. Gastroenterology. 2009;136(3):1060-9. 43. Paulusma CC, de Waart DR, Kunne C, Mok KS, Elferink RP. Activity of the bile salt export pump (ABCB11) is critically dependent on canalicular membrane cholesterol content. JBiolChem. 2009;284(15):9947-54. 44. Chen F, Ananthanarayanan M, Emre S, Neimark E, Bull LN, Knisely AS, et al. Progressive familial intrahepatic cholestasis, type 1, is associated with decreased farnesoid X receptor activity. Gastroenterology. 2004;126(3):756-64. 45. Sambuy Y, de A, I, Ranaldi G, Scarino ML, Stammati A, Zucco F. The Caco-2 cell line as a model of the intestinal barrier: influence of cell and culture-related factors on Caco-2 cell functional characteristics. Cell BiolToxicol. 2005;21(1):1-26. 46. Verhulst PM, van der Velden LM, Oorschot V, van Faassen EE, Klumperman J, Houwen RH, et al. A flippase-independent function of ATP8B1, the protein affected in familial intrahepatic cholestasis type 1, is required for apical protein expression and microvillus formation in polarized epithelial cells. Hepatology. 2010;51(6):2049-60. 47. Bijleveld CM, Vonk RJ, Kuipers F, Havinga R, Boverhof R, Koopman BJ, et al. Benign recurrent intrahepatic cholestasis: altered bile acid metabolism. Gastroenterology. 1989;97(2):427-32. 48. Mekjian HS, Phillips SF, Hofmann AF. Colonic secretion of water and electrolytes induced by bile acids: perfusion studies in man. JClinInvest. 1971;50(8):1569-77. 49. Roediger WE, Rigol G, Rae D. Sodium absorption with bacterial fatty acids and bile salts in the proximal and distal colon as a guide to colonic resection. DisColon Rectum. 1984;27(1):1-5. 50. Groen A, Kunne C, Paulusma CC, Kramer W, Agellon LB, Bull LN, et al. Intestinal bile salt absorption in Atp8b1 deficient mice. JHepatol. 2007;47(1):114-22. 51. Sun AQ, Salkar R, Sachchidanand, Xu S, Zeng L, Zhou MM, et al. A 14-amino acid sequence with a beta-turn structure is required for apical membrane sorting of the rat ileal bile acid transporter. JBiolChem. 2003;278(6):4000-9.

Supplemental data

Supplemental figure 1. Relative mRNA expression levels of FXR and SHP in shRNA control and ATP8B1-depleted Caco-2 cells.

Chapter 4

Supplemental figure 2. Fecal bile salt and electrolyte concentrations in PFIC1 patients graded according to stool form. 24-Hour stool samples of PFIC1 patients and age-matched, healthy controls were extracted and analyzed for bile salts, sodium, potassium, chloride and glucose concentrations. Patients treated with the anti-diarrhetic compound clonidine are indicated in red. Stools were graded according to the Bristol stool form scale indicated in supplemental table 1 (A) Fecal bile salt excretion. (B) Glucose concentration. (C) Sodium concentrations are significantly increased in PFIC1 patients with stools Type 5-6 and Type 7 when compared to controls and PFIC1

Chapter 4

patients with normal Type 3-4 stools. (D) Chloride concentration. (E) Potassium concentration. Statistical significance was determined by one-way ANOVA with Bonferroni’s post-hoc correction. *p<0.05, ***p<0.001

Type 1 Separate hard lumps, like nuts Type 2 Sausage-shaped but lumpy Type 3 Like a sausage or snake but with cracks on its surface Type 4 Like a sausage or snake, smooth and soft Type 5 Soft blobs with clear-cut edges Type 6 Fluffy pieces with ragged edges, a mushy stool Type 7 Watery, no solid pieces

Supplemental table 1. Bristol stool form scale, taken from (39).

CHAPTER 5

The phospholipid flippase ATP8B1 mediates apical localization of the cystic fibrosis transmembrane regulator

Vincent A. van der Mark, Hugo R. de Jonge, Jung-Chin Chang, Kam S. Ho-Mok, Suzanne Duijst, Dragana Vidović, Marianne S. Carlon, Ronald P. Oude Elferink, Coen C. Paulusma

Biochim Biophys Acta. 2016 Sep;1863(9):2280-8

doi: 10.1016/j.bbamcr.2016.06.005

Chapter 5

Abstract

Progressive familial intrahepatic cholestasis type 1 (PFIC1) is caused by mutations in the gene encoding the phospholipid flippase ATP8B1. Apart from severe cholestatic liver disease, many PFIC1 patients develop extrahepatic symptoms characteristic of cystic fibrosis (CF), such as pulmonary infection, sweat gland dysfunction, and failure to thrive. CF is caused by mutations in the cystic fibrosis transmembrane conductance regulator (CFTR), a chloride channel essential for epithelial fluid transport. Previously it was shown that CFTR transcript levels were strongly reduced in livers of PFIC1 patients. Here we have investigated the hypothesis that ATP8B1 is important for proper CFTR expression and function.

We analyzed CFTR expression in ATP8B1-depleted intestinal and pulmonary epithelial cell lines and assessed CFTR function by measuring short-circuit currents across Transwell-grown ATP8B1-depleted intestinal T84 cells and by a genetically- encoded fluorescent chloride sensor. In addition, we studied CFTR surface expression upon induction of CFTR transcription.

We show that CFTR protein levels are strongly reduced in the apical membrane of human ATP8B1-depleted intestinal and pulmonary epithelial cell lines, a phenotype that coincided with reduced CFTR activity. Apical membrane insertion upon induction of ectopically-expressed CFTR was strongly impaired in ATP8B1- depleted cells.

We conclude that ATP8B1 is essential for correct apical localization of CFTR in human intestinal and pulmonary epithelial cells, and that impaired CFTR localization underlies some of the extrahepatic phenotypes observed in ATP8B1 deficiency.

Chapter 5

1. Introduction

ATP8B1 is a class 4 P-type ATPase (P4-ATPase) and a member of a conserved family of proteins that catalyze the translocation of (amino)phospholipids from the exoplasmic to the cytoplasmic leaflet of biological membranes (1). P4-ATPases are essential for maintaining the asymmetric distribution of phospholipids within biological membranes but also for the biogenesis of transport vesicles in the biosynthetic and endocytic pathways (reviewed in (2-4)).

Mutations in the gene encoding ATP8B1 cause progressive familial intrahepatic cholestasis type 1 (PFIC1), a rare autosomal recessively inherited liver disease (5). PFIC1 patients usually present at a young age with typical cholestatic symptoms, including elevated serum bile salt, bilirubin and transaminase levels, low gamma-glutamyltranspeptidase levels, and intractable pruritus (6). Previous in vivo and in vitro studies indicate an important function for ATP8B1 in the maintenance of apical membrane structure of hepatocytes (7, 8). Besides cholestatic liver disease, many PFIC1 patients develop extrahepatic symptoms such as diarrhea, pulmonary infection, defects in sweat gland function, and failure to thrive (9-12). Apart from diarrhea, these symptoms are commonly associated with cystic fibrosis (CF) (13). CF is the most prevalent genetically inherited disease in Caucasians and is mainly manifested by airway disease (14). CF is caused by mutations in the gene encoding the cystic fibrosis transmembrane conductance regulator (CFTR), a chloride and bicarbonate channel that is expressed in the apical membrane of many epithelial cells, where it plays a principal role in transepithelial water and salt transport (15, 16). CFTR activity is tightly regulated by a multitude of signaling mechanisms and by the balance between endocytic retrieval and insertion from an intracellular vesicular pool (reviewed in (17, 18)). Interestingly, ATP8B1 deficiency has been associated with reduced expression of CFTR, both in liver of two PFIC1 patients and in an ATP8B1-depleted human biliary epithelial cell line (19).

The observation that PFIC1 patients share similar symptoms of CF patients and that CFTR expression is reduced in PFIC1 patients led us to investigate the hypothesis that ATP8B1 deficiency leads to reduced CFTR activity. In this study, we have analyzed the effects of depletion of ATP8B1 in intestinal and pulmonary epithelial cells on the expression and activity of CFTR. Our results underline the importance of ATP8B1 in proper apical surface expression of CFTR in epithelial cells.

Chapter 5

2. Materials and methods

2.1 Cell culture and lentiviral transduction

The human colonic carcinoma cell line T84 was cultured in 1:1 DMEM:HAM’s F12 (Lonza) supplemented with 16 mM HEPES, pH 7.5 (Sigma-Aldrich), 14 mM

NaHCO3 (Sigma-Aldrich), 10% bovine calf serum, 2 mM L-glutamine, 100 U/ml penicillin, 100 U/ml streptomycin at 37 ºC in a 10% CO2 humidified atmosphere. The human colorectal adenocarcinoma cell line Caco-2 and the lung adenocarcinoma cell line Calu-3 were cultured in respectively DMEM (Lonza) and Alpha MEM (Lonza) supplemented with 10% bovine calf serum, 2 mM L-glutamine, 100 U/ml penicillin, 100 U/ml streptomycin at 37 ºC in a 5% CO2 humidified atmosphere. Stable ATP8B1 knockdown (‘ATP8B1.7’) and control (‘control’) cell lines were generated by lentiviral transduction with validated MISSION shRNA (Sigma- Aldrich) vectors as described previously (20). TRCN0000050127 was used to target ATP8B1 and a scrambled, non-targeting shRNA SHC002 was used to generate control cells. Cells were incubated with virus-containing supernatants/DMEM (1:1) supplemented with 10 µg/ml diethylaminoethyl-dextran for 4 hours after which the medium was refreshed. Two days post-transduction, cells were selected on 10 µg/ml (Caco-2, T84) or 2 µg/ml (Calu-3) puromycin. Knockdown of ATP8B1 was checked regularly in subsequent passages via immunoblotting and quantitative RT-PCR. Cell lines overexpressing CFTR (driven by a cytomegalovirus (CMV) promoter) (21) or a ratiometric chloride sensor (driven by a CMV promoter (22)) were generated by lentiviral transduction. All experiments were performed on cells grown on microporous polycarbonate membrane filters (12 mm diameter transwell, 0.4 μm pore size, Corning) or on cells grown as a monolayer at 3 weeks post-confluency.

2.2 Animals

Male, age-matched (3-6 months) Atp8b1G308V/G308V mutant and wild type mice in a C57BL/6 background (23) were kept in a pathogen-free environment on a controlled 12-hour light/dark regimen in the animal facility of the Academic Medical Center (AMC). All animal experiments were approved by the institutional animal care and use committee of the AMC.

100

Chapter 5

2.3 Ussing chamber experiments

Mouse intestinal explants were isolated as follows: mice were anesthetized. The intestines from jejunum to colon were removed and flushed with ice cold PBS. The mucosal layer was turned inside out and the muscular layer was stripped with a cotton swab. Intestinal segments and cut-open gallbladders were mounted in an Ussing chamber setup (Physiological Instruments Inc). For transwell-grown T84 monolayers, filters were excised from their holders and mounted in an Ussing chamber setup. Ussing chamber experiments were performed as follows: samples were incubated in modified Meyler buffer (128 mM NaCl, 4.7 mM KCl, 1.3 mM

CaCl2, 1 mM MgCl2, 20.2 mM NaHCO3, 0.4 mM NaH2PO4, 0.33 mM Na2HPO4, 10 mM HEPES). Serosal/basolateral buffer for the assessment of apical CFTR activity contained equimolar amounts of Na+-gluconate and K+-gluconate instead of NaCl and KCl. Buffers and samples in the Ussing chamber were incubated at 37 °C and continuously gassed with carbogen. Voltage over the explant or transwell was clamped at 0 during experiments. Electrical currents were measured on a DVC-1000 (World Precision Instruments), digitalized with Powerlab 4/26 (AD Instruments) and analyzed with Labchart 6 (AD Instruments).

2.4 Ratiometric intracellular chloride measurements

T84 cells stably expressing a ratiometric chloride sensor (22) were grown on glass coverslips. After 2 weeks of confluence the glass slides were mounted in a perfusion chamber at 37 °C. Cells were perfused with equilibration buffer (144 mM NaCl, 1 mM KH2PO4, 2 mM MgSO4, 1 mM Ca gluconate, 5 mM HEPES pH 7.5, 5.6 mM D-glucose) for 10 min before the start of an experiment. NaI buffer contained equimolar amounts of NaI instead of NaCl. Buffers were changed after a stable fluorescence emission was reached (typically after 10-15 min). Fluorescence was detected on a NOVOSTAR microplate reader (BMG Labtech) with excitation filters of 440 ± 10 nm and 485 nm and an emission filter of 540 ± 10nm. Fluorescence was acquired every 5 s. After background subtraction, fluorescence emission of the YFP excited signal at 485 nm was divided by fluorescence emission of the CFP excited signal at 440 nm. Ratio changes over time were calculated by subtracting the averages of 3 sequential measurement points at stabilized fluorescence emissions before and after buffer change or addition of forskolin.

101

Chapter 5

2.5 Quantitative RT-PCR

Total RNA was isolated using Trizol reagent (Invitrogen). cDNA was synthesized from 2 μg of total RNA with oligo dT 12-18 and randomhexamer primers and Superscript III RT (Invitrogen). RT-PCR measurements were performed on a Lightcycler 480 (Roche) with SYBR-Green master mix (Roche). Expression levels of ATP8B1 or CFTR were calculated with the LinregPCR program (24) using human acidic ribosomal phosphoprotein P0 (RPLP0) or mouse villin (Vil1) as a reference gene. Primer sequences used: ATP8B1: forward 5’- GTCTTGGACAGAGTCACTTC-3’, reverse 5’- CGTCTTATCAGAGAAGATATAAT-3’; human CFTR: forward 5’- ACCTGGGAAAGGGCTGTTAT-3’, reverse 5’- CCTGGAGCTTCTGTGGAAAG-3’; human RPLP0: forward 5’- TCGACAATGGCAGCATCTAC-3’, reverse 5’- ATCCGTCTCCACAGACAAGG-3’; mouse Cftr: forward 5’- GCACAGCAGCTCAAACAACTGGAA, reverse 5’- TTCTCATTTGGAACCAGCGCAA-3’; mouse Vil1: forward 5’- GCTCCAACCAGACCGGACGC-3’, reverse 5’-GGGGCTCGTGTCCCTGCTTC- 3’

2.6 Cell surface biotinylation

Cells were washed 3x with ice-cold phosphate buffered saline (pH 8.2) containing 1 mM MgCl2 / 0.1 mM CaCl2 (PBS-CM) and were incubated with 1 mg/ml sulfo-NH- ss biotin (Thermo Scientific) in ice-cold PBS-CM for 60 min while gently shaken. To quench the non-bound biotin, cells were washed twice for 5 min with ice-cold PBS-CM containing 100 mM glycine and once more with ice-cold PBS-CM. Cells were lysed for 30 min on ice in RIPA buffer (50 mM Tris-HCl pH 8.0, 150 mM NaCl, 1% NP-40, 0.5% Na-deoxycholate, 0.1% SDS) containing protease inhibitor cocktail (Roche). Biotinylated protein was precipitated with high-capacity neutravidin agarose resin (Thermo Scientific) for 2 h at 4°C. Beads were washed four times with RIPA buffer and in between washes spun down for 1 min at 450 x g. Biotinylated protein was eluted with DTT- and SDS-containing sample buffer at room temperature. Total lysates and eluates were analyzed by SDS-PAGE and western blotting.

102

Chapter 5

2.7 SDS PAGE and western blotting

Cell lysates were obtained after incubation of the cells at 4 ⁰C in RIPA buffer containing protease inhibitor cocktail. Proteins were separated on a poly-acrylamide gel and transferred to a Immobilon-P PVDF membrane (Millipore). Membranes were blocked for 1 hour at RT in block buffer (4% low-fat milk (Nutricia Profitar- plus) in 0.1% PBS-Tween 20 (PBS-T)), washed 3 x in wash buffer (PBS-T), incubated for 1 hour at RT with rabbit polyclonal antibodies to ATP8B1 (kindly provided by dr. Leo Klomp) (25) or ATP1A1( C356-M09, kindly provided by Dr. J. Koenderink) (26), or mouse monoclonal antibodies to calnexin (H-70, Santa Cruz Biotechnology), GAPDH (MAB374, Millipore), or CFTR (MAB596, kindly provided by Dr. John R. Riordan, University of North Carolina at Chapel Hill, NC 27599, USA). Membranes were washed 3 x in PBS-T, incubated for 1 hour at RT with peroxidase-conjugated goat-anti-rabbit or mouse IgGs (Bio-Rad), developed with homemade enhanced chemiluminescence reagents (100 mM Tris-HCl pH 8.5,

1.25 mM luminol, 0.2 mM p-coumarin and freshly added 3 mM H2O2), and detected using an ImageQuant™ LAS 4000 (GE Healthcare). Densitometric analyses of band intensities were done using ImageJ 1.47v.

2.8 Statistics

Data are expressed as mean ± standard deviation (SD) and were calculated with Prism 5 (Graphpad). Statistical significance was determined by performing a Student’s T-test or one-way ANOVA with Bonferroni’s post-hoc correction for multiple testing and is indicated in the legend of the figures.

3. Results

3.1 ATP8B1 is required for expression of CFTR in T84 and Caco-2 cells

In order to study the role of ATP8B1 on the expression and activity of CFTR, we constructed a stable ATP8B1-depleted T84 cell line (hereafter referred to as ‘ATP8B1.7’ cells). T84 is a human colon carcinoma cell line with a crypt-like phenotype and high CFTR expression, and is generally used to study CFTR activity (27). As a control we used T84 cells transduced with a scrambled short hairpin sequence (‘control’ cells). Compared to control cells, ATP8B1 mRNA expression and protein levels in ATP8B1.7 cells were reduced by 83% (figure 1A) and by 70% (figure 1B, quantified in figure 1C), respectively. Similar to what was reported

103

Chapter 5 previously for human biliary epithelial cells (19), CFTR mRNA expression was reduced by 34% in ATP8B1.7 cells (figure 1D).

Figure 1. Depletion of ATP8B1 in intestinal T84 and Caco-2 cells leads to reduced transcript and protein levels of CFTR. (A) Mean relative ATP8B1 mRNA expression levels ± standard deviation in control and ATP8B1.7 T84 cells (n=4). (B) ATP8B1 protein levels in total lysates of control and ATP8B1.7 T84 cells (n=3). ATP1A1 is included as a loading control. (C) Quantification of ATP8B1 protein levels by densitometric analysis normalized to ATP1A1. (D) Mean relative CFTR mRNA expression levels ± standard deviation in control and ATP8B1.7 T84 cells (n=4). (E) CFTR protein levels in total lysates control and ATP8B1.7 T84 cells (n=3). ATP1A1 is included as a loading control. ‘C’ = CFTR complex-glycosylated, ‘B’ = CFTR core-glycosylated. (F) Quantification of CFTR protein levels by densitometric analysis normalized to ATP1A1. (G) Mean relative ATP8B1 and CFTR mRNA expression levels ± standard deviation in Caco2.control and Caco-2.7 cells (n=3-4). (H) ATP8B1 and CFTR protein levels in total lysates of Caco2.control and Caco2.7 cells (n=4). ATP1A1 is included as a loading control. (I) Quantification of ATP8B1 and CFTR protein levels by densitometric analysis normalized to ATP1A1.

Despite this mild reduction in CFTR expression, total CFTR protein levels were reduced by 53% compared to control cells (figure 1E, quantified in figure 1F). The

104

Chapter 5 expression levels of the core-glycosylated CFTR (B band) and the terminally- glycosylated CFTR (C band) were reduced by respectively 31% and 56% in ATP8B1.7 cells. A similar down-regulation of CFTR at the mRNA (figure 1G) and protein level (figure 1H and I) was observed in ATP8B1-depleted Caco-2 cells (Caco2.7), a colorectal adenocarcinoma cell line with characteristics of small- intestinal enterocytes.

3.2 ATP8B1 is required for normal physiological activity of CFTR in T84 cells

Although it is well established that as little as ~20% residual CFTR expression can guarantee a normal CFTR-mediated chloride current (28-31), we analyzed if the reduced expression of CFTR in ATP8B1.7 cells had functional consequences. Chloride transport over the apical membrane, which is predominantly mediated by CFTR in intestinal cells (32), was assessed in transwell-grown T84 cells mounted in Ussing chambers via measurements of short circuit current (Isc) (see supplemental figure 1 for a description of the experimental setup).

Addition of forskolin to intact cells resulted in the activation of CFTR- mediated activity as witnessed by an increased Isc (figure 2A). However, under these conditions we did not detect a significant difference in CFTR activity between control and ATP8B1.7 cells (figure 2B, pre-nystatin). To sensitize our assay, we exposed the cells at the basolateral side to the ionophore nystatin, which selectively permeabilizes the basolateral plasma membrane for monovalent ions. This procedure eliminates the membrane resistance and allows the specific measurement of chloride transport across the apical membrane (32). Removal of chloride from the basolateral compartment reversed the subsequent forskolin-induced CFTR chloride current (figure 2A). After addition of nystatin we observed a significant, though moderate, decrease in CFTR activity of 19% in ATP8B1.7 cells when compared to control cells (figure 2B, post-nystatin).

105

Chapter 5

Figure 2. Short-circuit currents (Isc) in ATP8B1-depleted T84 cells show reduced CFTR activity. (A) Example tracing of Isc across transwell-grown T84 cell monolayers mounted in Ussing chambers. Consecutively, forskolin (10 µM) is added to activate CFTR, forskolin is washed out, chloride in the serosal buffer is replaced by gluconate, nystatin (1 mg/ml) is added to the serosal buffer to selectively permeabilize the basolateral membrane, followed by a final re-stimulation with forskolin (10 µM). (B) Quantification of forskolin-induced Isc (per cm2) in ATP8B1.7 and control T84 cells before and after semi-permeabilization of the basolateral membrane by nystatin (n=8).

To confirm these observations in a different assay, we generated control and ATP8B1-depleted T84 cells expressing a genetically-encoded chloride sensor, a fusion of yellow fluorescent protein (YFP) with high sensitivity to chloride and cyan fluorescent protein (CFP) with no sensitivity to chloride (22). This CFP-YFP-based sensor allows ratiometric fluorescent analysis of intracellular chloride concentrations. The sensitivity of the sensor varies in the presence of different halides (F->I->Br->Cl- (33)) that quench the florescence emitted by the YFP moiety. Since CFTR is also able to transport iodide, we exploited the variable sensitivity of the chloride sensor to assess the transmembrane flux of chloride and iodide ions (see supplemental figure 2 for a description of the experimental setup).

After replacement of extracellular chloride by iodide, addition of forskolin resulted in a faster quenching of the sensor in control cells compared to ATP8B1.7 cells, with a reduction of 59% in ATP8B1.7 cells compared to control cells (figure 3). Collectively, these results indicate that CFTR function is reduced in ATP8B1- depleted T84 cells.

106

Chapter 5

Figure 3. ATP8B1-depleted T84 cells display reduced CFTR channel activity as shown with a genetically-encoded fluorescent chloride sensor. (A) Example tracing of YFP/CFP (F485/F450) fluorescence emission ratios in T84 cells. Consecutively, equilibration buffer is added, extracellular chloride is replaced by iodide, forskolin (10 µM) is added to activate CFTR, extracellular iodide is replaced by chloride. (B) Quantification of the change in YFP/CFP (F485/F450) fluorescence emission ratios as a measure of CFTR activity in ATP8B1.7 and control cells after addition of forskolin (n=8). Ratio changes over time were calculated by subtracting the averages of 3 sequential measurement points at stabilized fluorescence emissions before and after buffer change or addition of forskolin.

3.3 ATP8B1 is required for apical membrane localization of CFTR in T84 cells

Recently we and others have shown that apical membrane localization of several proteins was strongly impaired in ATP8B1-depleted Caco-2 cells (20, 34). To assess whether ATP8B1 was also involved in the apical localization of CFTR, we generated CFTR-overexpressing T84 cells in which ATP8B1 was depleted (‘ATP8B1.7CFTR cells’) and analyzed CFTR surface expression by surface biotinylation. In these cells CFTR expression is driven by a cytomegalovirus (CMV) promoter which enabled us to study CFTR surface expression upon its induction by sodium butyrate. In ATP8B1.7CFTR cells, ATP8B1 mRNA levels were reduced by 88% compared to shControlCFTR cells (figure 4A). In contrast to ATP8B1.7 cells in which CFTR expression was reduced (figure 1D), CFTR transcript levels were similar in shControlCFTR and ATP8B1.7CFTR cells (figure 4A). To quantify CFTR surface expression, shControlCFTR and ATP8B1.7CFTR cells were pretreated with 5 mM sodium butyrate for 24 hours to induce CFTR protein expression. In mock-treated shControlCFTR and ATP8B1.7CFTR cells total CFTR levels were similar whereas butyrate treatment resulted in a ~50% increase in total CFTR protein in both cell lines (figure 4B). Importantly, surface expression of CFTR in mock-treated

107

Chapter 5

ATP8B1.7CFTR cells was only 29% of that in shControlCFTR cells (figure 4C). Butyrate treatment increased CFTR surface expression in shControlCFTR cells to 172% of mock-treated cells whereas this was only 58% in ATP8B1.7CFTR cells (figure 4C). These results show that in T84 cells ATP8B1 is essential for proper surface localization of CFTR.

Figure 4. Surface expression of CFTR is strongly reduced in CFTR-overexpressing ATP8B1-depleted T84 cells. shControlCFTR and ATP8B1.7CFTR T84 cells (overexpressing CFTR) were incubated with or without 5 mM butyrate for 24 hours. Subsequently, the cells were biotinylated after which the biotinylated membrane proteins were precipitated. Proteins in the eluate as well as in the total lysate were immunoblotted for CFTR. Densitometric quantification of CFTR protein is shown next to its respective blot. (A) Mean relative ATP8B1 and CFTR mRNA expression levels ± standard deviation in shControlCFTR and ATP8B1.7CFTR T84 cells (overexpressing CFTR) (n=3). (B) CFTR protein levels in total lysates of shControlCFTR and ATP8B1.7CFTR T84 cells. Calnexin is included as a loading and a biotinylation control. CFTR was quantified by densitometric analysis and corrected for calnexin. (C) CFTR protein levels at the apical surface (eluate) of mock- and butyrate-treated shControlCFTR and ATP8B1.7CFTR T84 cells. Calnexin is used as a biotinylation control. CFTR was quantified by densitometric analysis and corrected for calnexin levels from the total lysates. The relative CFTR level in untreated shControlCFTR cells is set as 1 (n=3).

108

Chapter 5

3.4 ATP8B1 is required for apical membrane localization of CFTR in Calu-3 cells

Since CFTR plays a crucial role in the lung and PFIC1 patients (like CF patients) are susceptible to pulmonary infections, we determined if ATP8B1 was also important for apical localization of CFTR in the lung epithelial cell line Calu-3. We generated ATP8B1-depleted Calu-3 (Calu-3.7) cells and short hairpin control Calu-3 (Calu- 3.control) cells. Interestingly, and similar to ATP8B1-depleted T84 and Caco-2 cells (figure 1), a 77% reduction of ATP8B1 expression in Calu-3.7 cells coincided with a significant, 49% reduction in endogenous CFTR expression (figure 5A). In line with our findings in T84 cells, total and surface CFTR protein levels in Calu-3.7 cells were reduced by 36% (figure 5B) and 60% (figure 5C), respectively, compared to Calu-3.control cells. These results show that ATP8B1is important for apical localization of CFTR not only in intestinal- but also in pulmonary epithelial cells.

3.5 CFTR-mediated chloride secretion is unaltered in intestinal explants of mutant Atp8b1G308V/G308V mice

The in vitro findings in T84 cells lead us to study intestinal CFTR activity in the mouse model for PFIC1, the Atp8b1G308V/G308V mutant mouse, a knock-in for a prototypic PFIC1 mutation that leads to a glycine-to-valine substitution at amino acid 308 that results in absence of the protein (25). In contrast to ATP8B1-depleted human cells, Cftr mRNA expression in intestinal segments (duodenum, jejunum, ileum, and colon) and gallbladders of Atp8b1G308V/G308V mice was unaltered when compared to wild types (figure 6A). Also, and in contrast to ATP8B1-depleted human cells, CFTR total protein levels were unaffected as shown for gallbladder (figure 6B). Intestinal segments stripped from the mucosal muscle layer and gallbladders were analyzed for their ability to generate short circuit currents in Ussing chambers after stimulation with forskolin. Although forskolin induced a robust Isc response in all tissues examined (gallbladder, jejunal, ileal or colonic segments) we could not detect any significant difference in forskolin-induced, CFTR-mediated chloride secretion between Atp8b1G308V/G308V mutant and wild type mice (figure 6C).

109

Chapter 5

Figure 5. CFTR mRNA and surface protein expression are reduced in ATP8B1-depleted lung epithelial Calu-3 cells. (A) Mean relative ATP8B1 and CFTR mRNA expression levels ± standard deviation in Calu-3.control and Calu-3.7 cells (n=4). (B) ATP8B1 and CFTR protein levels in total lysates of Calu-3.control and Calu-3.7 cells. GAPDH is included as a loading and biotinylation control. Quantification of protein levels is shown next to the respective blot. ATP8B1 and CFTR levels were quantified by densitometric analysis and corrected for GAPDH. (C) ATP8B1 and CFTR protein levels at the apical surface (eluate) of Calu-3.control and Calu-3.7 cells. ATP8B1 and CFTR were quantified by densitometric analysis and corrected for GAPDH levels from the total lysates. The relative ATP8B1 and CFTR levels in Calu-3.control cells are set as 1 (n=3).

110

Chapter 5

Figure 6. Atp8b1G308V/G308V mutant mice show no difference in CFTR expression or activity in intestinal explants. (A) Mean relative Atp8b1 mRNA expression levels ± standard deviation in intestinal segments and gallbladders of wild type and Atp8b1G308V/G308V mutant mice (n=4). (B) ATP8B1, CFTR and ATP1A1 protein expression in gallbladder homogenates from wild type and Atp8b1G308V/G308V mutant mice (n=5). (C) Forskolin-induced (10 µM) Isc measurements in Ussing chamber-mounted intestinal explants and gall bladders of wild type and Atp8b1G308V/G308V mutant mice (jejunum n=8-10 mice, 2-6 explants each; ileum n=4 mice, 4 explants each; colon n=4 mice, 4 explants each, gallbladders n=4 mice, 1 explant each).

111

Chapter 5

4. Discussion

In this paper we show that the phospholipid flippase ATP8B1 is crucial for apical membrane localization of CFTR in human intestinal and pulmonary epithelial cells.

Mammalian P4-ATPases have been shown to play an important role in the regulation of membrane protein localization by catalyzing the biogenesis of transport vesicles at the plasma membrane and at the TGN. For example, we have recently shown that ATP8B1 mediates the targeting of the apical sodium-dependent bile acid transporter (ASBT) in intestinal Caco-2 cells (20). Similarly, Van der Velden et al. previously showed impaired surface expression of apical resident proteins, including alkaline phosphatase (AP), sucrose isomaltase (SI) and aminopeptidase N (APN), in differentiated Caco-2 cells (34). In addition, the mammalian P4-ATPases ATP11B and ATP11C have been implicated in the generation of transport vesicles at the trans-Golgi network (TGN) that are bound for the plasma membrane (35-37), while ATP8B1, ATP8B2 and ATP9A have been implicated in facilitating glucose- stimulated insulin release from human pancreatic beta cells (38). Based on our present data, we hypothesize that ATP8B1 mediates the apical targeting of CFTR either directly from the trans-Golgi network (TGN), on route from the TGN to the plasma membrane or during endocytosis and recirculation to the apical membrane. Firstly, induction of ectopically-expressed CFTR by butyrate resulted in equal amounts of total CFTR protein levels in control and ATP8B1-depleted cells, however, this only resulted in significantly increased CFTR surface expression in control cells. Secondly, the glycosylation status of CFTR was not affected in ATP8B1-depleted cells, which indicates that in these cells CFTR normally traverses the glycosylation machinery of the Golgi apparatus.

The endogenous CFTR protein levels in ATP8B1-depleted T84, Caco-2 and Calu-3 cells were moderately reduced which may be explained by enhanced proteasomal breakdown due to impaired apical membrane localization of CFTR. Alternatively, reduced CFTR protein levels are the consequence of reduced CFTR mRNA levels that were observed in these cells. Reduced CFTR mRNA expression was previously reported by Demeilliers et al. in ATP8B1-depleted Mz-ChA-2 cells, a human biliary epithelial cell line, and in livers of two PFIC1 patients (19). The reason for the reduced CFTR expression remains elusive as the mechanisms underlying the transcriptional regulation of CFTR are complex and poorly

112

Chapter 5 understood, however, we believe that this downregulation is not a primary consequence of ATP8B1 depletion. Firstly, the previously reported reduced apical surface expression of ASBT, AP and APN was not accompanied by reduced mRNA expression (20, 34). Secondly, in cells ectopically over-expressing CFTR, ATP8B1 depletion did not reduce expression levels of CFTR nor did it lead to elevated surface expression. These observations indicate that the effect of ATP8B1-depletion is at the post-transcriptional level and underscores a role for ATP8B1 in the vesicular targeting of CFTR to the apical plasma membrane. Possibly, disturbed trafficking of other (yet) unknown apical membrane proteins (or of CFTR itself) may induce transcriptional downregulation of CFTR.

Our data show that in ATP8B1-depleted T84 cells the activity of CFTR is moderately impaired due to reduced apical localization of the protein. However, it remains to be determined if patients with ATP8B1 deficiency have a defect in the apical localization of CFTR in epithelial cells. Obviously, there is not a complete defect in apical plasma membrane localization of CFTR as the CF-like symptoms are mild. Hence, there may be redundancy in the trafficking itineraries of CFTR that partially compensate for the loss of ATP8B1. Still, our present findings may explain the pathogenesis of several extrahepatic phenotypes observed in patients with ATP8B1 deficiency, possibly caused by partial inactivation of CFTR function. For instance, many ATP8B1 deficient patients present with elevated sweat sodium and chloride concentrations (9-11), a phenotype commonly observed in CF patients (13). Furthermore, many ATP8B1 deficient patients develop pulmonary infection that presents with wheezing and sometimes pneumonia (11, 12). Pulmonary infection in ATP8B1 deficient patients was proposed to be caused by accumulation of cardiolipin in lung surfactant that impaired lung and lung surfactant function (39). The authors suggested that ATP8B1 was a flippase for cardiolipin, however, the mechanism of pulmonary cardiolipin accumulation and the role of ATP8B1 therein was not illuminated (40). Our observation of reduced surface expression of CFTR in ATP8B1-depleted lung epithelial cells indicates that this may contribute to the pulmonary infections in ATP8B1 deficient patients. Impaired CFTR activity in airway epithelial cells severely reduces the clearance of mucus and pathogens resulting in increased susceptibility to pulmonary infections (41). In addition, CFTR has been implicated in the acidification of phagosomes in alveolar macrophages; impaired acidification leads to a failure in phagocytic killing of pulmonary bacteria and sensitizes to pulmonary infection (42, 43).

113

Chapter 5

In contrast, diarrhea is a fairly common extrahepatic symptom in ATP8B1- deficient patients (11), as opposed to CF patients that suffer from intestinal obstruction and constipation (44). If intestinal CFTR expression is reduced in ATP8B1 deficient patients, it may be that partial CFTR activity is sufficient for proper intestinal function. On the other hand, the impaired apical localization of ASBT, AP, SI and APN (and possibly other, yet unknown apical proteins) can contribute to the diarrhea in ATP8B1 deficiency. For example, reduced apical localization of ASBT and SI in ATP8B1 deficiency lead to impaired intestinal reabsorption of bile salts and impaired intestinal degradation of sucrose/maltose, respectively, both of which contribute to diarrhea (45, 46).

Our findings in three human cell lines could not be replicated in ATP8B1- deficient mice, as these showed no impairment of forskolin-activated, CFTR- mediated chloride secretion under Ussing chamber conditions. Previous Isc measurements in intestinal epithelium of CF mice and patients have indicated that basolateral transporters rather than apical CFTR are rate-limiting for transepithelial chloride secretion and that >80% depletion of CFTR is needed to reach the dynamic window for a CFTR assay (47). Unfortunately, rendering the Isc response to forskolin more CFTR-dependent through nystatin permeabilization of the basolateral membrane, as done successfully in the T84 cell monolayers, is technically challenging and poorly reproducible in a multi-cell layer system like intestinal explants. Therefore our Ussing chamber experiments on native epithelia clearly demonstrate that transepithelial chloride secretion is not affected in the intestine and gallbladder of ATP8B1 deficient mice but do not rule out a partial depletion of CFTR in the apical membrane, similar to our findings in the human intestinal cell lines. However, and in contrast to human ATP8B1-depleted cells, Cftr mRNA levels were not affected suggesting that ATP8B1 deficiency in mice does not impair CFTR targeting. A strikingly similar mouse-human difference was reported by us for ASBT, whose apical localization and activity were strongly reduced in human ATP8B1-depleted Caco-2 cells but remained unaffected in ATP8B1 deficient mice (20, 48). It is possible therefore, that intestinal CFTR trafficking and/or localization is differently regulated in mice and humans, potentially with other redundant mechanisms in mice.

In conclusion, we show a crucial role for the phospholipid flippase ATP8B1 in mediating the apical localization of CFTR in human intestinal and pulmonary epithelial cells. Our data provide novel insights into the intracellular trafficking of

114

Chapter 5

CFTR and shed light on the extrahepatic phenotypes observed in patients with ATP8B1 deficiency.

Acknowledgements

We thank Dr P. Bregestovski, INMED, INSERM, Marseilles, France for providing a vector expressing the chloride sensor and Drs. L.W. Klomp and J.B. Koenderink for polyclonal antibodies to ATP8B1 and ATP1A1, respectively.

References

1. Holthuis JC, Menon AK. Lipid landscapes and pipelines in membrane homeostasis. Nature. 2014;510(7503):48-57. 2. Sebastian TT, Baldridge RD, Xu P, Graham TR. Phospholipid flippases: building asymmetric membranes and transport vesicles. Biochim Biophys Acta. 2012;1821(8):1068-77. 3. van der Mark VA, Elferink RP, Paulusma CC. P4 ATPases: Flippases in Health and Disease. Int J Mol Sci. 2013;14(4):7897-922. 4. Lopez-Marques RL, Theorin L, Palmgren MG, Pomorski TG. P4-ATPases: lipid flippases in cell membranes. Pflugers Arch. 2014;466(7):1227-40. 5. Bull LN, van Eijk MJ, Pawlikowska L, DeYoung JA, Juijn JA, Liao M, et al. A gene encoding a P-type ATPase mutated in two forms of hereditary cholestasis. Nat Genet. 1998;18(3):219-24. 6. Paulusma CC, Elferink RP, Jansen PL. Progressive familial intrahepatic cholestasis type 1. Semin Liver Dis. 2010;30(2):117-24. 7. Paulusma CC, Groen A, Kunne C, Ho-Mok KS, Spijkerboer AL, Rudi de Waart D, et al. Atp8b1 deficiency in mice reduces resistance of the canalicular membrane to hydrophobic bile salts and impairs bile salt transport. Hepatology. 2006;44(1):195-204. 8. Cai SY, Gautam S, Nguyen T, Soroka CJ, Rahner C, Boyer JL. ATP8B1 deficiency disrupts the bile canalicular membrane bilayer structure in hepatocytes, but FXR expression and activity are maintained. Gastroenterology. 2009;136(3):1060-9. 9. Bourke B, Goggin N, Walsh D, Kennedy S, Setchell KD, Drumm B. Byler-like familial cholestasis in an extended kindred. Arch Dis Child. 1996;75(3):223-7. 10. Knisely AS, Agostini RM, Zitelli BJ, Kocoshis SA, Boyle JT. Byler's syndrome. Arch Dis Child. 1997;77(3):276-7. 11. Pawlikowska L, Strautnieks S, Jankowska I, Czubkowski P, Emerick K, Antoniou A, et al. Differences in presentation and progression between severe FIC1 and BSEP deficiencies. J Hepatol. 2010;53(1):170-8. 12. Whitington PF, Freese DK, Alonso EM, Schwarzenberg SJ, Sharp HL. Clinical and biochemical findings in progressive familial intrahepatic cholestasis. J Pediatr Gastroenterol Nutr. 1994;18(2):134-41. 13. Farrell PM, Rosenstein BJ, White TB, Accurso FJ, Castellani C, Cutting GR, et al. Guidelines for diagnosis of cystic fibrosis in newborns through older adults: Cystic Fibrosis Foundation consensus report. J Pediatr. 2008;153(2):S4-S14. 14. Davies JC, Alton EW, Bush A. Cystic fibrosis. BMJ. 2007;335(7632):1255-9. 15. Riordan JR. CFTR function and prospects for therapy. Annu Rev Biochem. 2008;77:701-26. 16. Rowe SM, Verkman AS. Cystic fibrosis transmembrane regulator correctors and potentiators. Cold Spring Harb Perspect Med. 2013;3(7). 17. Ameen N, Silvis M, Bradbury NA. Endocytic trafficking of CFTR in health and disease. J Cyst Fibros. 2007;6(1):1-14. 18. Guggino WB, Stanton BA. New insights into cystic fibrosis: molecular switches that regulate CFTR. Nat Rev Mol Cell Biol. 2006;7(6):426- 36. 19. Demeilliers C, Jacquemin E, Barbu V, Mergey M, Paye F, Fouassier L, et al. Altered hepatobiliary gene expressions in PFIC1: ATP8B1 gene defect is associated with CFTR downregulation. Hepatology. 2006;43(5):1125-34. 20. van der Mark VA, de Waart DR, Ho-Mok KS, Tabbers MM, Voogt HW, Oude Elferink RP, et al. The lipid flippase heterodimer ATP8B1- CDC50A is essential for surface expression of the apical sodium-dependent bile acid transporter (SLC10A2/ASBT) in intestinal Caco-2 cells. Biochim Biophys Acta. 2014;1842(12 Pt A):2378-86. 21. Vidovic D, Carlon MS, M FdC, Dekkers JF, Hollenhorst MI, Bijvelds MJ, et al. rAAV-CFTRDeltaR Rescues the Cystic Fibrosis Phenotype in Human Intestinal Organoids and CF Mice. Am J Respir Crit Care Med. 2015. 22. Markova O, Mukhtarov M, Real E, Jacob Y, Bregestovski P. Genetically encoded chloride indicator with improved sensitivity. JNeurosciMethods. 2008;170(1):67-76. 23. Groen A, Kunne C, Jongsma G, van den Oever K, Mok KS, Petruzzelli M, et al. Abcg5/8 independent biliary cholesterol excretion in Atp8b1-deficient mice. Gastroenterology. 2008;134(7):2091-100. 24. Ruijter JM, Ramakers C, Hoogaars WM, Karlen Y, Bakker O, van den Hoff MJ, et al. Amplification efficiency: linking baseline and bias in the analysis of quantitative PCR data. Nucleic Acids Res. 2009;37(6):e45. 25. Pawlikowska L, Groen A, Eppens EF, Kunne C, Ottenhoff R, Looije N, et al. A mouse genetic model for familial cholestasis caused by ATP8B1 mutations reveals perturbed bile salt homeostasis but no impairment in bile secretion. Hum Mol Genet. 2004;13(8):881-92. 26. Koenderink JB, Geibel S, Grabsch E, De Pont JJ, Bamberg E, Friedrich T. Electrophysiological analysis of the mutated Na,K-ATPase cation binding pocket. J Biol Chem. 2003;278(51):51213-22. 27. Silvis MR, Bertrand CA, Ameen N, Golin-Bisello F, Butterworth MB, Frizzell RA, et al. Rab11b regulates the apical recycling of the cystic fibrosis transmembrane conductance regulator in polarized intestinal epithelial cells. Mol Biol Cell. 2009;20(8):2337-50. 28. de Nooijer RA, Nobel JM, Arets HG, Bot AG, van Berkhout FT, de Rijke YB, et al. Assessment of CFTR function in homozygous R117H- 7T subjects. JCystFibros. 2011;10(5):326-32. 29. Kerem E. Pharmacological induction of CFTR function in patients with cystic fibrosis: mutation-specific therapy. Pediatr Pulmonol. 2005;40(3):183-96. 30. Dorin JR, Farley R, Webb S, Smith SN, Farini E, Delaney SJ, et al. A demonstration using mouse models that successful gene therapy for cystic fibrosis requires only partial gene correction. Gene Ther. 1996;3(9):797-801.

115

Chapter 5

31. Chu CS, Trapnell BC, Curristin SM, Cutting GR, Crystal RG. Extensive posttranscriptional deletion of the coding sequences for part of nucleotide-binding fold 1 in respiratory epithelial mRNA transcripts of the cystic fibrosis transmembrane conductance regulator gene is not associated with the clinical manifestations of cystic fibrosis. J Clin Invest. 1992;90(3):785-90. 32. Anderson MP, Welsh MJ. Calcium and cAMP activate different chloride channels in the apical membrane of normal and cystic fibrosis epithelia. Proc Natl Acad Sci U S A. 1991;88(14):6003-7. 33. Kuner T, Augustine GJ. A genetically encoded ratiometric indicator for chloride: capturing chloride transients in cultured hippocampal neurons. Neuron. 2000;27(3):447-59. 34. Verhulst PM, van der Velden LM, Oorschot V, van Faassen EE, Klumperman J, Houwen RH, et al. A flippase-independent function of ATP8B1, the protein affected in familial intrahepatic cholestasis type 1, is required for apical protein expression and microvillus formation in polarized epithelial cells. Hepatology. 2010;51(6):2049-60. 35. Moreno-Smith M, Halder JB, Meltzer PS, Gonda TA, Mangala LS, Rupaimoole R, et al. ATP11B mediates platinum resistance in ovarian cancer. J Clin Invest. 2013;123(5):2119-30. 36. Matsuzaka Y, Hayashi H, Kusuhara H. Impaired Hepatic Uptake by Organic Anion-Transporting Polypeptides Is Associated with Hyperbilirubinemia and Hypercholanemia in Atp11c Mutant Mice. Mol Pharmacol. 2015;88(6):1085-92. 37. de Waart DR, Naik J, Utsunomiya KS, Duijst S, Ho-Mok K, Bolier AR, et al. ATP11C targets basolateral bile salt transporter proteins in mouse central hepatocytes. Hepatology. 2016. 38. Ansari IH, Longacre MJ, Paulusma CC, Stoker SW, Kendrick MA, MacDonald MJ. Characterization of P4 ATPase Phospholipid Translocases (Flippases) in Human and Rat Pancreatic Beta Cells: Their Gene Silencing Inhibits Insulin Secretion. J Biol Chem. 2015. 39. Ray NB, Durairaj L, Chen BB, McVerry BJ, Ryan AJ, Donahoe M, et al. Dynamic regulation of cardiolipin by the lipid pump Atp8b1 determines the severity of lung injury in experimental pneumonia. Nat Med. 2010;16(10):1120-7. 40. Paulusma CC, Houwen RH, Williamson PL. The flip side of cardiolipin import. Nat Med. 2011;17(4):413; author reply -4. 41. Boucher RC. Airway surface dehydration in cystic fibrosis: pathogenesis and therapy. Annu Rev Med. 2007;58:157-70. 42. Di A, Brown ME, Deriy LV, Li C, Szeto FL, Chen Y, et al. CFTR regulates phagosome acidification in macrophages and alters bactericidal activity. Nat Cell Biol. 2006;8(9):933-44. 43. Painter RG, Valentine VG, Lanson NA, Jr., Leidal K, Zhang Q, Lombard G, et al. CFTR Expression in human neutrophils and the phagolysosomal chlorination defect in cystic fibrosis. Biochemistry. 2006;45(34):10260-9. 44. van der Doef HP, Kokke FT, van der Ent CK, Houwen RH. Intestinal obstruction syndromes in cystic fibrosis: meconium ileus, distal intestinal obstruction syndrome, and constipation. Curr Gastroenterol Rep. 2011;13(3):265-70. 45. Ament ME, Perera DR, Esther LJ. Sucrase-isomaltase deficiency-a frequently misdiagnosed disease. J Pediatr. 1973;83(5):721-7. 46. Oelkers P, Kirby LC, Heubi JE, Dawson PA. Primary bile acid malabsorption caused by mutations in the ileal sodium-dependent bile acid transporter gene (SLC10A2). J Clin Invest. 1997;99(8):1880-7. 47. De Boeck K, Derichs N, Fajac I, de Jonge HR, Bronsveld I, Sermet I, et al. New clinical diagnostic procedures for cystic fibrosis in Europe. J Cyst Fibros. 2011;10 Suppl 2:S53-66. 48. Groen A, Kunne C, Paulusma CC, Kramer W, Agellon LB, Bull LN, et al. Intestinal bile salt absorption in Atp8b1 deficient mice. J Hepatol. 2007;47(1):114-22.

Supplemental data

Supplemental figure 1. Ussing chamber experimental setup. (A) T84 cells grown on Transwell permeable supports or muscle-stripped intestinal explants of wild type or Atp8b1G308V/G308V mutant mice are mounted into Ussing chambers and voltage clamped to 0 mV. (B) CFTR is activated by adding forskolin to the serosal compartment inducing a chloride current across the cell layer. (C) Serosal chloride is replaced by gluconate, after which the basolateral membrane of T84 cells is semi-permeabilized by nystatin. Subsequent addition of forskolin induces an absorptive chloride current across the cell layer.

116

Chapter 5

Supplemental figure 2. Chloride sensor experimental setup. (A) T84 cells are stably transduced with a lentivirus encoding a CFP-YFP fusion protein sensitive to halides. The capacity of iodide to quench YFP fluorescence is much larger than that of chloride. (B) The cells are constantly perfused and equilibrated with a physiological buffer after which the chloride in the buffer is replaced by iodide. (C) After further equilibration, CFTR is stimulated by adding forskolin. After buffer change or forskolin addition, CFP fluorescence remains unchanged, while YFP fluorescence is altered.

117

CHAPTER 6

Phospholipid flippases attenuate LPS-induced TLR4 signaling by mediating endocytic retrieval of Toll-like receptor 4

Vincent A. van der Mark*, Mohammed Ghiboub*, Casper Marsman, Jing Zhao, Remco van Dijk, Johan K. Hiralall, Kam S. Ho-Mok, Zoë Castricum, Wouter J. de Jonge, Ronald P. Oude Elferink, Coen C. Paulusma

* both authors contributed equally

Cell Mol Life Sci. 2017 Feb; 74(4):715-730

doi: 10.1007/s00018-016-2360-5

Chapter 6

Abstract

P4-ATPases are lipid flippases that catalyze the transport of phospholipids to create membrane phospholipid asymmetry and to initiate the biogenesis of transport vesicles. Here we show, for the first time, that lipid flippases are essential to dampen the inflammatory response and to mediate the endotoxin-induced endocytic retrieval of Toll-like receptor 4 (TLR4) in human macrophages. Depletion of CDC50A, the ß-subunit that is crucial for the activity of multiple P4-ATPases, resulted in endotoxin-induced hypersecretion of proinflammatory cytokines, enhanced MAP kinase signaling and constitutive NF-қB activation. In addition, CDC50A-depleted THP-1 macrophages displayed reduced tolerance to endotoxin. Moreover, endotoxin-induced internalization of TLR4 was strongly reduced and coincided with impaired endosomal MyD88-independent signaling. The phenotype of CDC50A- depleted cells was also induced by separate knockdown of two P4-ATPases, namely ATP8B1 and ATP11A. We conclude that lipid flippases are novel elements of the innate immune response that are essential to attenuate the inflammatory response, possibly by mediating endotoxin-induced internalization of TLR4.

120

Chapter 6

1. Introduction

Toll-like receptor 4 (TLR4) is expressed on myeloid-derived cells, including macrophages and dendritic cells, and on some epithelial and endothelial cells, and is an essential component of the innate immune response (1). Activation of TLR4 involves multiple co-receptors, including lipopolysaccharide-binding protein, myeloid differentiation protein 2 and CD14. These co-receptors mediate ligation and transfer of lipopolysaccharides (LPS), an inflammatory mediator of the cell wall of Gram-negative bacteria, to TLR4 (2). This ligation induces dimerization of TLR4 and activates a MyD88-dependent early-response signaling pathway that results in the production of pro-inflammatory cytokines such as TNFα, IL-1β and IL-6 (3). Consequently, TLR4 is endocytosed, which induces MyD88-independent signaling from the TLR4-containing endosomal recycling compartment (4-6), inflicting a type I interferon response that is critical to counter an ongoing infection (7). Importantly, TLR4 internalization reduces signaling from the plasma membrane and dampens the inflammatory reaction. Failure to curtail the TLR4-mediated immune response can lead to pervasive tissue injury and may give rise to immunopathology such as sepsis, autoimmune diseases, metabolic diseases, neurodegeneration and chronic inflammation (1, 8). Although the molecular mechanisms underlying the endocytic retrieval of TLR4 are poorly understood, it is a dynamin-driven process that strongly depends on CD14 (5, 9).

P4-ATPases are a family of phospholipid flippases, i.e. integral membrane proteins that translocate phospholipids from the exoplasmic leaflet to the cytoplasmic leaflet of biological membranes (10-12). Emerging evidence in S. cerevisiae, C. elegans, A. thaliana and mammalian cells indicate important functions for members of the P4-ATPase protein family in the biogenesis of intracellular transport vesicles in the biosynthetic and endocytic pathways (10, 13). For instance, yeast cells deficient for the P4-ATPase Drs2p have a defect in the biogenesis of clathrin-coated vesicles at the trans-Golgi network (TGN), while combined deficiency of different P4-ATPases leads to defects in fluid-phase endocytosis and intracellular protein transport (14-16). Similarly, P4-ATPases in C. elegans have been implicated in receptor-mediated and fluid-phase endocytosis and in endocytic sorting and recycling (17, 18), while in the plant A. thaliana, P4-ATPases are involved in generation of a specific class of TGN-derived secretory vesicles important for root development (19). In mammalian cells P4-ATPases probably also fulfill important roles in vesicle biogenesis, but evidence for this is much more

121

Chapter 6 scarce. For instance, ATP11B is thought to confer resistance to cis-platinum in ovarian cancer via the biogenesis of cisplatin-containing vesicles at the TGN (20). ATP8B1 mediates the apical targeting of the apical sodium-dependent bile acid transporter SLC10A2/ASBT, either directly from the TGN or via recycling from a subapical vesicle pool in intestinal Caco-2 cells (21). Recently, ATP8A1 and ATP8A2 were shown to be essential for membrane fission of recycling endosomes (22).

Thus far, deficiency of two P4-ATPases were shown to cause severe hereditary disease. Mutations in ATP8B1 cause progressive familial intrahepatic cholestasis type 1, a severe liver disease characterized by impaired bile formation (23). Mutations in ATP8A2 are associated with a severe neurological disorder described as cerebellar axatia, mental retardation and dysequilibrium syndrome (CAMRQ) (24).

Most P4-ATPases function as a heterodimer with a member of the CDC50 protein family, in which the P4 ATPase is the α-subunit and CDC50 the ß-subunit (11, 25). The human genome encodes fourteen P4-ATPases and three CDC50 proteins, eleven of which form a heterodimer with CDC50A. CDC50A is a ~50-kDa complex-glycosylated transmembrane protein (26), and its interaction with individual P4-ATPases is essential for endoplasmic reticulum exit and activity of the heterodimer (27-30).

Here we have investigated the hypothesis that lipid flippases of the P4- ATPase family are important mediators of TLR4-mediated signaling. We have analyzed the inflammatory reaction of CDC50A-depleted THP-1 and primary human monocyte-derived macrophages upon LPS challenge. Our data point to important functions for multiple P4-ATPases in attenuating TLR4-mediated signaling.

2. Materials and Methods

2.1 Cell culture and lentiviral transduction

The human monocytic leukemia cell line THP-1 was cultured in RPMI 1640 medium (Gibco) supplemented with 10% fetal bovine serum (FBS) (Lonza), 2 mM L- glutamine (Lonza), 100 U/ml penicillin (Lonza), and 100 U/ml streptomycin (Lonza) at 37 ºC in a 5% CO2 humidified atmosphere. Knockdown cell lines for CDC50A

122

Chapter 6 and P4 ATPases were generated by lentiviral transduction of undifferentiated THP- 1 cells (31). Briefly, 0.5 x 106 cells were incubated with virus-containing supernatants / RPMI 1640 (1:1) supplemented with 10 µg/ml diethylaminoethyldextran for 4 hours. Two days post-transduction, cells were selected with 2 µg/ml puromycin. Validated short-hairpin RNA (shRNA) vectors to CDC50A (TRCN0000159317 (4) and TRCN0000160267 (1)), ATP8B1 (TRCN0000050127) and ATP11A (TRCN0000051887) were obtained through the MISSION shRNA library (Sigma-Aldrich). SHC002 was included as a control.

Lentiviral constructs to haemagglutinin antigen (HA)-tagged CDC50A (HA- CDC50A) and enhanced green fluorescent protein (eGFP)-tagged ATP8B1 (ATP8B1-eGFP) were described previously (27). A Rab11-eGFP plasmid (32) was AgeI/XbaI-digested and subcloned into a second generation lentiviral transfer vector (31). ATP11A cDNA was obtained from the PlasmID Repository / DF / HCC DNA Resource Core (http://plasmid.med.harvard.edu). A FLAG-tag was introduced on the 3’end by PCR using forward oligo 5’- cattagctacgaccggtatggactgcagcctcgtgcggacg-3’and reverse oligo 5’- ggctggtctagactaCTTGTCATCGTCGTCCTTGTAGTCgaaactcaggctgctggaag-3’, in which the FLAG sequence is capitalized. All experiments were performed on THP- 1 cells differentiated to macrophages with 100 nM phorbol-12-myristate-13-acetate (PMA) (Sigma-Aldrich) for 3 days, after which they were rested for 2-3 days in PMA-free culture medium (33). Differentiated THP-1 cells were challenged with o LPS from Escherichia coli 0111:B4 (Sigma-Aldrich) at 37 C, 5% CO2 for indicated time-points and at indicated concentrations.

2.2 Isolation, maturation, polarization and siRNA transfection of monocyte-derived macrophages

Peripheral blood mononuclear cells (PBMCs) were obtained from whole blood of healthy donors by density gradient centrifugation using Ficoll (Invitrogen). Briefly, 13 ml Ficoll was added below 30 ml PBS-diluted blood and centrifuged at 2000 rpm for 20 minutes at room temperature, with acceleration 3 and no break. The interphase was recovered using a Pasteur pipet into a new 50 ml tube containing 10 ml PBS, followed by twice washing with PBS. 5 x 106 of PBMCs (containing ~ 0.5 x 106 monocytes) were incubated in 12-well plates in 1 ml Isocove’s Modified Dulbecco’s Medium (Lonza) supplemented with 10% fetal bovine serum (FBS) (Lonza), 2 mM L-glutamine (Lonza), 100 U/ml penicillin (Lonza) and 100 U/ml streptomycin

123

Chapter 6

(Lonza) for 90 minutes at 37°C, 5% CO2. After 2 hours, medium was aspirated and the cells were washed several times with sterilized warm PBS till removing all floating cells. Monocytes were matured using 72 hours treatment with 20 ng/ml of macrophage colony-stimulating factor (M-CSF). The cells were then washed with PBS and stimulated with 100 ng/ml IFNγ or 40 ng/ml IL-4 for 3 days to generate M1 or M2 macrophages respectively. Medium was added to maturated monocytes (M0 macrophages) for 3 days to keep it as M0 subset. M1 macrophages were transfected using DharmaFECT™ transfection reagents (Dharmacon) according to the manufacturer’s instructions with siGENOME human TMEM30A smartpool siRNAs to deplete CDC50A and non-targeting siRNAs as a control. Cells were analyzed at 72 hours post-transfection.

2.3 Enzyme Linked Immuno Sorbent Assay

ELISA (R&D Systems) for human TNFα, IL1β and IL6 were performed according to the manufacturer’s instructions.

2.4 Quantitative RT-PCR

Total RNA was isolated from differentiated cells using TriPURE reagent (Invitrogen). cDNA was synthesized from 2 μg of total RNA with random hexamers and oligo dT 12-18 primer and Superscript III RT (Invitrogen). Real-time PCR measurements were performed on a Lightcycler 480 (Roche) with Fast Start DNA MasterPlus SYBR Green I kit (Roche). Expression levels in THP-1 cells and human monocyte-derived M1 macrophages were calculated with the LinregPCR software (34) and were normalized to the geometric means of the three most stable reference genes (RPLP0, ACTB, GAPDH) as determined by Genorm analysis (35). Expression levels in primary mouse cells were calculated similarly and were normalized to Rplp0. Primer sequences are depicted in supplemental table 1.

2.5 Isolation of nuclear extracts

Cells were washed with ice-cold PBS, scraped into solution and pelleted by centrifugation (400 x g, 10 minutes, 4 °C). The cell pellet was washed in 5 packed cell volumes (PCV) resuspension buffer (10 mM HEPES pH 7.9, 1.5 mM MgCl2, 10 mM KCl, 0.5 mM DTT) supplemented with 0.5 mM phenylmethanesulfonylfluoride, phosSTOP phosphatase and protease inhibitor

124

Chapter 6 cocktails (Roche) and centrifuged (400 x g, 10 minutes, 4 °C). The cell pellet was incubated for 10 minutes on ice in 2 PCV resuspension buffer and homogenized with a Dounce homogenizer (tight pestle, 60 strokes). Cell lysis was confirmed via microscopic analysis and non-broken cells were pelleted by centrifugation (200 x g, 2 minutes, 4 °C). The resulting supernatant was pelleted again by centrifugation (425 x g, 10 minutes, 4 °C). The supernatant containing the cytosolic fraction was stored on ice and the nuclei-containing pellet was resuspendend in ultracentrifuge resuspension buffer (20 mM HEPES pH 7.9, 1.5 mM MgCl2, 0.42 M NaCl, 0.5 mM DTT, 0.2 mM EDTA, 25% v/v glycerol, 0.5 mM phenylmethanesulfonylfluoride, phosSTOP phosphatase and protease inhibitor cocktails) and incubated while rotating for 30 minutes at 4 °C. The nuclear and cytosolic fractions were each separately subjected to ultracentrifugation in an Optima L-90K ultracentrifuge (Beckton Dickenson, Ti70 rotor, 32,000 x g, 30 minutes, 4 °C). Cleared supernatants were stored at -80 °C until use.

2.6 SDS-PAGE and western blotting

Cells were lysed in RIPA buffer (50 mM Tris-HCl, pH 8.0, 150 mM NaCl, 1% NP- 40, 0.5% Na-deoxycholate, 0.1% SDS) containing protease inhibitor cocktail (Roche) and /or PhosSTOP phosphatase inhibitor cocktail (Roche). Proteins were separated on a polyacrylamide gel (6-12% depending on the size of the protein of interest) and were transferred to Immobilon-P PVDF membranes (Millipore) by semi-dry blotting using 10 mM CAPS, pH 10.5 / 15% methanol buffer. Membranes were blocked for 1 hour at RT in block buffer (PBS / 5% low-fat milk (Nutricia Profitar-plus)) and incubated for 1 hour at RT in block buffer with rabbit polyclonal antibodies to TLR4 (H-80, Santa Cruz), CDC50A (26), ATP8B1 (36), histone H3 ((di methyl K79) antibody - ChIP Grade (Abcam)), ATP1A1 (37), phospho- SAPK/JNK (Thr183/Tyr185), SAPK/JNK, phospho-p44/42 MAPK (Erk1/2) (Thr202/Tyr204), p44/42 MAPK (Erk1/2), mouse monoclonal to phospho-NF-kB p65 (Ser536) (7F1) (all from Cell Signaling), NF-κB p65 (F-6) (Santa Cruz), GAPDH (MAB374, Millipore), and β-actin (clone AC-15, Sigma Aldrich). Immune complexes were visualized with peroxidase-conjugated goat-anti-rabbit or mouse IgGs (Bio-Rad), developed with homemade enhanced chemiluminescence reagents (100 mM Tris-HCl pH 8.5, 1.25 mM luminol, 0.2 mM p-coumarin and freshly added

3 mM H2O2) and detected using an ImageQuant™ LAS 4000 (GE Healthcare). Densitometric analyses of band intensities were performed using ImageJ 1.49.

125

Chapter 6

2.7 Indirect immunofluorescence

Differentiated THP-1 cells (knockdowns, HA-CDC50A/Rab11eGFP-, ATP8B1eGFP- and ATP11A-Flag over-expressing) were grown on glass coverslips and fixed in 2% paraformaldehyde (PFA) for 20 min at RT. PFA fixed cells were permeabilized on PBS/0.1% triton X-100 (PBS-Tx) and incubated with rat monoclonal anti-HA (clone 3F10; Roche) to detect CDC50A, mouse monoclonal anti-Flag (M2, Sigma) to detect ATP11A or rabbit polyclonal anti-TLR4 (H-80, Santa Cruz). Cells were extensively washed in between antibody incubations in PBS- Tx. Immunoreactivity was visualized with goat-anti-rat Texas Red, goat-anti-mouse Alexa 488 or goat-ant-rabbit Alexa 594 (Molecular Probes). Sections were mounted in Vectashield/DAPI (Vector Laboratories) and were imaged on a SP8-X-SMD confocal microscope (Leica) or on a DMi8 microscope (Leica).

2.8 Measuring TLR4 / CD14 / CD11B surface expression by flow cytometry

PMA-differentiated THP-1 cells (2 x 106/well, 6-well plate) were washed with ice- cold PBS and were incubated in ice-cold PBS containing 5 mM ethylenediaminetetraacetic acid (PBS/EDTA) for 10 minutes on ice. Cells were removed using a cell scraper (Corning) and were transferred to eppendorf tubes. Cells were centrifuged for 5 minutes at 240 x g and supernatant was removed after which the cell pellet was incubated with mouse IgG2a phycoerythrin (PE)- conjugated anti-human CD284 (TLR4) antibody (clone HTA125, BioLegend, San Diego, USA), mouse IgG2b PE-conjugated anti-human CD14 (clone MϕP9, BD Biosciences) or mouse IgG1 Alexa Fluor 488-conjugated anti-human CD11B (clone ICRF44, BD Biosciences) in PBS for 45 minutes at 4 oC in the dark with constant agitation. Subsequently cells were washed with ice-cold PBS, were resuspended in PBS and immediately analyzed by flow cytometry using a BD LSRFortessa™ cell analyzer (BD Biosciences). To analyze LPS-induced TLR4 / CD14 internalization, differentiated THP-1 cells and human monocyte-derived M1 macrophages were o challenged with LPS (concentration indicated in figure legend) at 37 C, 5% CO2 for indicated time-points. Cells were analyzed as described above.

2.9 Statistics

Data are displayed as mean ± standard deviation (sd) or standard error of the mean (sem). Statistical significance was determined by performing Student’s t-test or one-

126

Chapter 6 way ANOVA with Bonferroni’s correction for multiple testing as indicated in the legend of the figures.

3. Results

3.1 CDC50A is required to dampen the inflammatory reaction triggered by LPS in human macrophages.

To investigate the role of CDC50A-interacting P4-ATPases in the innate immune response in macrophages, we analyzed the inflammatory response to LPS challenge in human THP-1 and primary human monocyte-derived macrophages (MDMs) that were depleted from CDC50A. As we could not localize endogenous CDC50A, we assessed the localization of ectopically-expressed HA-tagged CDC50A in differentiated THP-1 macrophages (figure 1A). HA-CDC50A localized to intracellular vesicles and almost completely costained with Rab11, a marker for recycling endosomes and the endocytic recycling compartment; in addition, weak plasma membrane staining was detected. Since CDC50A is only released from the endoplasmic reticulum in association with a P4-ATPase (28), this localization most likely represents a functional CDC50A-P4-ATPase association. Stable CDC50A- depleted THP-1 cells (further referred to as CDC50A.4 cells) were generated by constitutive expression of short hairpin RNA sequences targeting CDC50A. CDC50A mRNA expression was reduced by over 80% in CDC50A.4 cells compared to shControl cells (figure 1B), which coincided with strongly reduced CDC50A protein levels (figure 1C). CDC50A depletion did not interfere with cell growth, morphology, or differentiation, the latter evidenced by normal CD11B surface expression (supplementary figure 1A). Differentiated THP-1 cells were incubated for four hours with LPS, and TNFα, IL-1β, and IL6 secretion was quantified by ELISA. Compared to shControl cells, CDC50A.4 macrophages showed a strongly enhanced output (> 6-fold) of TNFα, IL-1β, and IL6 upon LPS challenge (figure 1D). This phenotype was reproduced in CDC50A-depleted THP-1 cells generated by a different shRNA sequence (supplementary figure 1B). The enhanced secretion of these cytokines was mirrored by their mRNA expression (figure 1E).

127

Chapter 6

Figure 1. CDC50A depletion leads to endotoxin-induced hypersecretion of inflammatory cytokines in THP-1 macrophages. (A) Confocal microscopical detection of HA-tagged CDC50A and GFP-tagged Rab11 in THP-1 macrophages. CDC50-HA co-stained with Rab11eGFP-positive (recycling) endosomes and the endocytic recycling compartment (indicated by an asterisk in the CDC50-HA panel). Nuclear DAPI staining in blue. (B) CDC50A mRNA levels in shControl and CDC50A-depleted THP-1 macrophages. Data shown are means ± standard deviation of triplicate wells. mRNA expression in shControl cells was set as 1. Statistical significance was tested by a Student’s t-test, *p<0.002. (C) Immunoblot analysis of CDC50A protein levels in membrane isolates of shControl and CDC50A-depleted THP-1 macrophages. Arrows indicate the different glycosylation states of CDC50A. ATP1A1 is included as a loading control. (D) TNFα, IL-1ß and IL6 secretion in shControl and CDC50A-depleted THP-1 macrophages 4 hours after 100 ng/ml LPS administration. Data shown are means ± standard deviation of quadruple wells. Statistical significance was tested by one-way ANOVA with Bonferroni’s correction for multiple testing, *p<0.005. (E) TNFα, IL-1β and IL6 mRNA expression levels in shControl and CDC50A-depleted THP-1 macrophages 3 hours post-LPS administration. Data shown are means ± standard deviation of triplicate wells. mRNA expression in shControl cells without LPS was set as 1. Statistical significance was tested by a Student’s t- test, *p<0.05, ** p<0.001, ***p<0.0001. Data shown in figure 1 are representative of three to six independent experiments with similar results.

128

Chapter 6

To verify the physiological relevance of the hyper-inflammatory phenotype observed in CDC50A-depleted THP-1 cells, we performed the same analyses in primary human MDMs in which CDC50A was transiently depleted by CDC50A- directed siRNA transfection. Analyses were performed in M1 polarized MDMs, M1 polarization evidenced by high expression of M1-specific marker CD80 and absence of M2-specific marker CD200R (supplemental figure 1C) (38). 72 Hour post- transfection MDMs were challenged for 4 hours with LPS and the inflammatory response was assessed (figure 2). CDC50A mRNA expression and protein levels were strongly reduced 72 hour post-transfection (figure 2A, B). In contrast to CDC50A-depleted THP-1 macrophages, CDC50A-depleted M1 macrophages displayed no elevated excretion of TNFα; However, IL6 and IL10 excretion was ~6- fold increased by these cells, while IL-1ß tended to be increased (figure 2C). Also at the mRNA level, expression of most of the mentioned cytokines was slightly elevated (figure 2D). Collectively these data suggest that CDC50A-associated P4- ATPases negatively regulate TLR4-mediated signaling in human macrophages.

129

Chapter 6

Figure 2. CDC50A depletion leads to endotoxin-induced hypersecretion of inflammatory cytokines in primary human monocyte-derived macrophages. (A) CDC50A mRNA levels in primary human monocyte-derived macrophages (MDM) transfected with scrambled siRNAs (siControl) and siRNAs directed to CDC50A (siCDC50A). RNA was isolated 72-hour post-transfection. Data shown are means ± standard deviation; n=8 of 2 different isolations. Statistical significance was tested by a Student’s t-test, **p < 0.0005. (B) Immunoblot analysis of CDC50A protein levels in total lysates of two independent isolations of MDMs. ß-Actin is included as a loading control. (C) TNFα, IL6, IL-1ß and IL10 secretion in siControl and siCDC50A MDMs (72h post-transfection) that were stimulated for 4 hours with 100 ng/ml LPS. Data shown are means ± sem of n=10 of 4 different isolations. Statistical significance was tested by a Student’s t.test,*p<0.05 (D) TNFα, IL6, IL-1β and IL10 mRNA expression levels in siControl and siCDC50A MDMs (72h post-transfection) 3 hours post-LPS (100ng/ml) administration. Data shown are means ± standard deviation (n=8, 2 different isolations). mRNA expression in siControl cells without LPS was set as 1. Statistical significance was tested one-way ANOVA with Bonferroni’s correction for multiple testing, *p<0.01; **p<0.005.

3.2 CDC50A attenuates TLR4-mediated signaling in human macrophages.

TLR4 activation by LPS leads to induction of the interferon signaling pathway from TLR4-containing endosomes (5, 6). This late response is characterized by the

130

Chapter 6 induction of interferon ß (IFNß) and RANTES/CCL5, which we analyzed in CDC50A-depleted THP-1 and primary MDM-derived macrophages (figure 3). In both control and CDC50A-depleted macrophages LPS challenge induced IFNß and RANTES, however, the induction was significantly less in CDC50A-depleted THP- 1 (figure 3A) and MDMs (figure 3B). Similarly, suppressor of cytokine signaling 1 (SOCS1), which is a negative regulator of TLR-mediated signaling to dampen the inflammatory reaction (39), was induced in LPS-challenged macrophages, however, this induction was significantly reduced in CDC50A-depleted THP-1 and MDMs, suggesting impaired negative feedback regulation of TLR4-mediated signaling (figure 3A,B). These data suggest that CDC50A plays an important role in the induction of the interferon response and in dampening the TLR4-mediated inflammatory reaction in macrophages. Since in both CDC50A-depleted THP-1 and primary human macrophages downstream signaling of TLR4 was affected, we further assessed the role of CDC50A in TLR4-mediated signaling in THP-1 macrophages. LPS-induced TLR4 signaling leads to nuclear translocation of the transcription factors AP-1 (via MAPK pathways) and NF-κB to promote transcription of pro-inflammatory cytokines. We thus analyzed activation (phosphorylation) of c-Jun terminal kinases (JNK) 1/2, extracellular signal-regulated kinases (ERK) 1/2 and NF-κB in LPS challenged cells (figure 3C). In CDC50A.4 macrophages JNK phosphorylation was strongly induced already 15 minutes after LPS challenge and was more sustained compared to shControl cells. Importantly, the levels of nuclear phosphorylated NF-κB were strongly increased in both LPS- stimulated and unstimulated CDC50A.4 macrophages (figure 3D). These observation provide additional evidence for a role of CDC50A-interacting P4- ATPases in the attenuation of TLR4-mediated signaling.

131

Chapter 6

Figure 3. CDC50A depletion in human macrophages leads to an impaired interferon response and sustained TLR4-mediated signaling. (A) IFNß, RANTES and SOCS1 mRNA levels in shControl and CDC50A-depleted THP- 1 cells 3 hours post-LPS administration. Data shown are means ± standard deviation of triplicate wells. mRNA expression in shControl cells without LPS was set as 1. Statistical significance was tested by a Student’s t-test, *p<0.0005. Data shown are representative of three to six independent experiments with similar results. (B) IFNß, RANTES and SOCS1 mRNA levels in siControl and siCDC50A MDMs (72h post-transfection) 3 hours post-LPS (100ng/ml) administration. Data shown are means ± standard deviation (n=7, 2 different isolations). mRNA expression in siControl cells without LPS was set as 1. Statistical significance was tested one-way ANOVA with Bonferroni’s correction for multiple testing, *p<0.05; **p<0.005. (C) Immunoblot analysis of activated JNK1/2 and Erk1/2 in shControl and CDC50A-depleted THP-1 macrophages. Cells were incubated with 100 ng/ml LPS for

132

Chapter 6

the time points indicated. ß-actin is included as a loading control. (D) Immunoblot analysis of cytosolic and nuclear total and activated NF-κB p65 in shControl and CDC50A-depleted THP-1 macrophages. Cells were incubated with 10 ng/ml LPS for 2 hours. GAPDH and histone H3 are included as cytosolic and nuclear loading controls, respectively.

3.3 CDC50A is required for LPS-induced endocytosis of TLR4 in human macrophages.

The increased and sustained activation of TLR4-dependent signaling in CDC50A- depleted macrophages led us to investigate the surface expression of TLR4 by flow cytometry. In unstimulated CDC50A-depleted THP-1 and human primary macrophages, we observed a significant 3- and 1.2-fold increase, respectively, in surface expression of TLR4 compared to controls (figure 4A, B), whereas TLR4 mRNA levels were ~2-fold decreased only in THP-1 macrophages (figure 4C and not shown). The elevated plasma membrane localization of TLR4 was visualized by immunofluorescent staining of TLR4, which outlined the apex of CDC50A.4 cells but not that of shControl cells (figure 4D). Previous work has shown that TLR4 is endocytosed upon LPS stimulation, a process that is dependent on CD14 expression (4, 5, 9). To investigate whether CDC50A was involved in TLR4 endocytosis, we stimulated CDC50A.4 macrophages with LPS and analyzed TLR4 surface expression after 30 and 60 minutes. While in shControl cells TLR4 surface expression was reduced by 16% after 30 minutes and 26% after 60 minutes, TLR4 surface expression was not reduced but increased by 38% after 60 minutes in CDC50A.4 macrophages (figure 4E). Like TLR4, surface expression of CD14 was 2-fold increased in CDC50A.4 macrophages, which was mirrored by an increase at the mRNA level, however, no loss of CD14 surface expression was observed in either shControl or CDC50A.4 macrophages after LPS challenge (supplementary figure 2). These data indicate that CDC50A is essential for LPS-induced endocytosis of TLR4 in macrophages. To verify that the hyperinflammatory reaction of CDC50A-depleted cells was due to impaired TLR4 endocytosis, we chemically blocked TLR4 internalization using the GTPase dynamin inhibitor dynasore (4, 7). We measured IL6 excretion in CDC50A-depleted human MDMs that were challenged for 4 hours with LPS in the absence and presence of 80µM dynasore. CDC50A-depleted cells displayed strongly elevated IL6 excretion when challenged with LPS alone, however and surprisingly, dynasore treatment almost completely abolished this effect (figure 4F). These observations were mirrored by IL6 mRNA expression levels (figure 4G). Collectively these data suggest that, apart from a role

133

Chapter 6 in LPS-induced internalization of TLR4, CDC50A-associated P4-ATPases may play a role in the signaling cascade downstream of TLR4.

Figure 4. CDC50A-depleted THP-1 macrophages are impaired in LPS-induced endocytosis of TLR4. (A) TLR4 surface expression in shControl and CDC50A-depleted THP-1 macrophages was determined by flow cytometry. Data are expressed as MFI ± standard deviation of triplicate wells. Statistical significance was tested by a Student’s t-test, *p<0.02. (B) TLR4 surface expression in siControl and siCDC50A MDMs was determined by flow cytometry. Data are expressed as MFI ± standard deviation of triplicate wells. Statistical significance was tested by a Student’s t-test, *p<0.05. (C) TLR4 mRNA levels in shControl and CDC50A-depleted THP-1 cells 3 hours post-LPS administration. Statistical significance was tested by a Student’s t-test, *p<0.00005. (D) Immunofluorescent

134

Chapter 6

detection of TLR4 in shControl and CDC50A.4 cells. Arrows indicate TLR4 staining at the apex of the cells. (E) TLR4 surface expression in shControl and CDC50A-depleted THP-1 macrophages after stimulation with 100 ng/ml LPS. Cells were analyzed by flow cytometry. Data shown are MFIs ± standard deviation at different time points of triplicate wells. Statistical significance was determined between shControl and CDC50A.4 cells and tested by a Student’s t-test, *p<0.05, **p<0.002. (F) IL6 secretion in siControl and siCDC50A MDMs (72h post-transfection) that were stimulated for 4 hours with 100 ng/ml LPS with/without 80µM dynasore. Dynasore-treated cells were pre- incubated for 30 minutes with 80µM dynasore. Data shown are means ± sem of n=8 of 2 different isolations. Statistical significance was tested by one-way ANOVA with Bonferroni’s correction for multiple testing, *p<0,005; **p<0,00005. (G) IL6 mRNA levels in siControl and siCDC50A MDMs (72h post-transfection) that were stimulated for 4 hours with 100 ng/ml LPS with/without 80µM dynasore. Data shown are means ± standard deviation (n=8, 2 different isolations). mRNA expression in siControl cells without LPS was set as 1. Statistical significance was tested one-way ANOVA with Bonferroni’s correction for multiple testing, *p<0,005; **p<0,0005. Data shown in figure 4 are representative of three to four independent experiments with similar results.

3.4 ATP8B1 and ATP11A-depleted THP-1 macrophages show LPS-induced hypersecretion of pro-inflammatory cytokines.

We analyzed which member(s) of the P4-ATPase family serve(s) as the active α- subunit for obligate heterodimerization with CDC50A. To investigate this, we depleted THP-1 cells from all individual P4-ATPases. THP-1 cells expressed 11 P4- ATPases (figure 5A), nine of which have been shown to heterodimerize with CDC50A (28-30). Several P4-ATPase-depleted THP-1 cells displayed poor growth or no survival, still, seven different knockdown lines could be analyzed for the inflammatory reaction to LPS. LPS-induced TNFα excretion was ~2- and ~3-fold elevated only in ATP8B1-depleted cells (ATP8B1.7) and ATP11A-depleted cells (ATP11A.21), respectively, when compared to LPS-induced shControl macrophages (figure 5B). Already without LPS stimulation, TNFα output was elevated in both P4- ATPase depleted cells. The enhanced TNFα output was associated with a ~70% reduction in ATP8B1 and ATP11A mRNA expression (figure 5C). Similar to CDC50A.4 macrophages, ATP8B1.7 and ATP11A.21 macrophages showed an increased and more sustained JNK1/2 phosphorylation already 15 minutes after LPS challenge (figure 5D). Immunolocalization of ectopically-expressed ATP8B1eGFP and ATP11A-Flag showed predominant plasma membrane localization of both proteins (figure 5E). Finally, we assessed TLR4 internalization 30 and 60 minutes after LPS administration. While after 60 minutes TLR4 surface expression was reduced by ~50% in shControl cells, this was ~30% in ATP8B1.7 cells, whereas in ATP11A.21 cells no reduction in TLR4 surface expression was observed (figure 5F). Both ATP8B1 and ATP11A were expressed in different subsets of MDMs (figure 5G). These data suggest that ATP8B1 and/or ATP11A are involved in the LPS- induced internalization of TLR4 upon an LPS insult in human macrophages (Of note:

135

Chapter 6

We could not reproduce the hyper-responsiveness to LPS observed in ATP8B1- depleted THP-1macrophages in ATP8B1 deficient bone marrow-derived macrophages (BMDM) nor alveolar macrophages (AM); see appendix).

Figure 5. ATP8B1 and ATP11A display elevated LPS-induced TLR4-mediated signaling and reduced TLR4 internalization. (A) P4-ATPases, CDC50A and CDC50B mRNA levels in THP-1 macrophages. mRNA data shown are means ± standard deviation of triplicate wells. (B) TNFα secretion in shControl, CDC50A-, ATP8B1- and

136

Chapter 6

ATP11A-depleted THP-1 macrophages 4 hours after 100 ng/ml LPS administration. Data shown are means ± standard deviation of triplicate wells. Statistical significance was tested by one-way ANOVA with Bonferroni’s correction for multiple testing, *p<0.00001, **p=0.05. (C) ATP8B1 and ATP11A mRNA levels in shControl and ATP8B1- and-ATP11A depleted THP-1 macrophages. Data shown are means ± standard deviation of triplicate wells. Statistical significance was tested by a Student’s t-test, *p<0.001. (D) Immunoblot analysis of activated JNK1/2 in shControl, ATP8B1- and ATP11A-depleted THP-1 macrophages. Cells were incubated with 100 ng/ml LPS for the time points indicated. ß-actin is included as a loading control. (E) Confocal microscopical detection of GFP-tagged ATP8B1 (ATP8B1eGFP) and flag-tagged ATP11A (ATP11A-flag) in THP-1 macrophages. Both ATP8B1eGFP and ATP11A-flag are in green and predominantly localize to the plasma membrane. Nuclear DAPI staining in blue. (F) TLR4 surface expression in shControl, ATP8B1- and ATP11A-depleted THP-1 macrophages after stimulation with 100 ng/ml LPS. Cells were analyzed and data expressed as described in figure 4A. No statistical differences by one-way ANOVA. (G) CDC50A, ATP8B1 and ATP11A mRNA levels in primary human monocytes and M1- and M2-primed macrophages. Data shown are means ± standard deviation of quadruple wells. ATP8B1 expression in the three subsets was set as 1. Data shown in figure 5 are representative of two to three independent experiments with similar results.

3.5 CDC50A is involved in endotoxin tolerance in THP-1 macrophages.

Because CDC50A-depleted macrophages displayed a hyper-inflammatory response to LPS, we investigated whether CDC50A could play a role in endotoxin tolerance. CDC50A-depleted THP-1 macrophages were stimulated for 18 hours with 100 ng/ml LPS, washed and re-stimulated with 100 ng/ml LPS for 6 hours and TNFα output and TLR4 expression were quantified. TNFα levels in the culture medium of LPS- stimulated, tolerized CDC50A.4 macrophages were ~8 times higher compared to those in LPS-stimulated tolerized shControl cells, indicating a problem with tolerization in CDC50A.4 cells (figure 6A). Endotoxin tolerance has been reported to be associated with reduced LPS-induced TLR4-mediated MAPK signaling in mouse macrophages (40), but also with prolonged down-regulation of TLR4 in murine and THP-1 macrophages (41, 42). We therefore assessed the activation of JNK1/2 after restimulation of LPS-tolerized THP-1 cells. ShControl and CDC50A.4 cells were pretreated for 24 hours with medium or 100 ng/ml LPS after which the cells were washed and restimulated with 100 ng/ml LPS for the indicated time periods. In LPS-pretreated shControl cells, we could not detect any increase in the phosphorylation status of JNK1/2 up to 120 minutes of LPS restimulation, which indicates that these cells were tolerant to LPS (figure 6B). In LPS pretreated CDC50A.4 cells, however, phosphorylated JNK1/2 was observed already 30 minutes after LPS restimulation indicating loss of endotoxin tolerization in these cells. Total protein levels of TLR4 were greatly reduced in tolerized shControl cells, however, in CDC50A.4 cells TLR4 expression was unaltered (figure 6C, quantified in figure 6D). These data suggest that CDC50A.4 macrophages display impaired

137

Chapter 6 endotoxin tolerance, which is possibly caused by sustained TLR4-mediated signaling.

Figure 6. CDC50A-depleted THP-1 macrophages display a partial loss of endotoxin tolerance. (A) TNFα secretion in LPS-tolerized and non-tolerized shControl and CDC50A-depleted THP-1 macrophages. Cells were incubated with or without 100 ng/ml LPS for 18 hours, washed and re-incubated with or without 100 ng/ml LPS for 6 hours after which TNFα secretion was measured. Data shown are means ± standard deviation of triplicate wells. Statistical significance was tested by one-way ANOVA with Bonferroni’s correction for multiple testing, *p<0.00005, med = medium. (B) LPS induced phosphorylation of JNK1/2 in tolerized and non-tolerized shControl and CDC50A- depleted THP-1 macrophages. Cells were incubated with 100 ng/ml LPS for the time points indicated and cell lysates were analysed for p-JNK1/2 expression by immunoblotting. ß-actin is included as a loading control. (C) TLR4 expression in LPS-tolerized and non-tolerized shControl and CDC50A-depleted THP-1 macrophages. Cell lysates were harvested from cells treated as described in figure 3 (A) and were analyzed for TLR4 expression by immunoblotting. ß-actin is included as a loading control. (D) Quantification of total TLR4 expression in tolerized and non-tolerized shControl and CDC50A-depleted THP-1 macrophages as represented in figure 3 (C). Protein levels were quantified by densitometric analysis, normalized to ß-actin, and expressed as mean percentage ± standard deviation of protein present in the med/med group (set to 100%). Statistical significance was tested by a Student’s t-test, *p<0.005. Data shown in figure 3 are representative of two independent experiments with similar results.

138

Chapter 6

4. Discussion

Here we report in human THP-1 and human primary monocyte-derived macrophages that CDC50A-interacting P4-ATPase are essential to attenuate endotoxin-induced TLR4-mediated signaling possibly by facilitating the endocytic retrieval of TLR4. This is not only crucial to prevent severe chronic inflammatory conditions such as sepsis, but also to guarantee a sustained immune response to suppress an ongoing infection. It is well documented that activation of TLR4 sequentially induces two signaling pathways leading to an early and a late inflammatory response (1). The early response, activated by ligand binding and dimerization of TLR4, initiates signaling via NF-κB and the MAP kinase (MAPK) pathway and leads to the production of pro-inflammatory cytokines. Our data show that depletion of CDC50A, which is an essential subunit of P4-ATPase phospholipid flippase complexes that localize to (amongst others) the plasma membrane (27-30), results in a hyperactivation of the early response pathway as evidenced by elevated TNFα, IL- 1ß and IL6 release, enhanced MAPK signaling and sustained NF-κB activation. The late response is activated upon ligand-induced TLR4 endocytosis and leads to a type I interferon response (4, 5, 7, 43). We show here that in CDC50A-depleted cells the LPS-induced internalization of TLR4 is impaired and that the induction of the interferon response in these cells is reduced; The lack of complete ablation of the interferon response in our study can be explained by residual expression of CDC50A in the knockdown cells and / or the timing of LPS treatment (3 hours in our study).

Collectively, our data suggest that LPS-induced hyperactivation of Myd88- dependent signaling in CDC50A-depleted cells is caused by impaired endocytic retrieval of TLR4 leading to sustained TLR4 signaling. This is supported by previous studies in CD14-deficient mouse bone marrow-derived macrophages (BMDM), CD14 being crucial for LPS-induced endocytic retrieval of TLR4 (5, 9). In CD14- deficient BMDMs, TNFα excretion was almost completely abolished at low dose LPS (10ng/ml), however, at higher LPS doses (≥100ng/ml) TNFα excretion was increased compared to wild type cells, an observation that remained unexplained (5, 9), but that may be caused by sustained TLR4 signaling. In our study we also applied 100ng/ml LPS, which resulted in increased cytokine excretion by CDC50A-depleted cells. We can, however, not exclude that the phenotype in CDC50A-depleted macrophages is caused by two independent processes, i.e. impaired TLR4 internalization and sustained activation of the signaling cascade downstream TLR4. This is underscored by the unexpected observation that inhibition of LPS-induced

139

Chapter 6

TLR4 internalization by the dynamin inhibitor dynasore completely abrogated IL6 excretion and mRNA expression in CDC50A-depleted macrophages. This observation suggests that the hyper-responsiveness of CDC50A-depleted cells to LPS was not caused by impaired TLR4 internalization. However, the specificity of dynasore, i.e. a selective inhibitor of the dynamin GTPase activity to inhibit clathrin- mediated endocytosis, has recently being disputed (44). Multiple dynamin- independent effects of dynasore have been described, including disruption of lipid rafts i.e. detergent-resistant membrane domains that serve as signaling platforms (45, 46). Previous studies have shown that following LPS stimulation, TLR4 and CD14 are mobilized to lipid rafts, an event crucial for TLR4 dimerization and subsequent recruitment and assembly of adaptor proteins; The use of lipid raft-disrupting compounds abrogates the LPS-induced TLR4 signaling (47, 48). Thus, dynasore treatment of macrophages may not only inhibit clathrin-mediated endocytosis of TLR4 but also the lipid raft-associated TLR4 signaling cascade, explaining the lack of inflammatory reaction in dynasore-treated CDC50A-depleted cells. Still, we cannot exclude the possibility that, apart from blocking TLR4 endocytosis, CDC50A depletion may interfere with Myd88-dependent signaling downstream of TLR4.

We showed that CDC50A-depleted macrophages were impaired in the LPS- induced internalization of TLR4, however, in the TLR4 endocytosis assay TLR4 surface expression was increased 30 and 60 minutes after LPS challenge compared to unchallenged cells. Although we have no clear explanation for this observation it may be that LPS challenge mobilizes an existing pool of TLR4 to the plasma membrane that subsequently cannot be internalized leading to an increase of TLR4 surface expression post-LPS challenge. Furthermore, and despite increased TLR4 surface expression, TLR4 mRNA levels were reduced in CDC50A-depleted cells, which may be due to the sustained activation of the Myd88-dependent signaling pathway leading to a compensatory down-regulation of TLR4 transcription. Previous studies showed that LPS-induced internalization of TLR4 associates with endocytosis of CD14 (5, 49), although Rajaiah et al. (9) recently reported CD14- independent TLR4 endocytosis. We did not observe any reduction in CD14 surface expression in LPS-challenged control macrophages. Still, CD14 surface expression in CDC50A-depleted cells was strongly increased compared to control cells. An explanation for this observation may be that, due to reduced TLR4 turn-over, the cell responds by elevating CD14 transcription with consequent elevation of CD14

140

Chapter 6 surface levels. Indeed, we found that CD14 mRNA levels were increased in CDC50- depleted cells.

CDC50A-depleted cells also displayed a partial loss of endotoxin tolerance. Endotoxin tolerance is a condition of reduced responsiveness to LPS upon a secondary LPS challenge with concomitant repression of proinflammatory cytokine production (50). Tolerization, which is crucial for dampening a recurrent inflammatory insult and protects the host to fatal infection (i.e. during sceptic shock), is associated with repression of TLR4-mediated signaling and reduced TLR4 surface expression (40-42, 50). Reduced endotoxin tolerance in CDC50A-depleted cells coincided with enhanced JNK1/2 activation and TNFα excretion upon restimulation with LPS. Furthermore, whereas in control cells total TLR4 protein levels were reduced after overnight LPS challenge, TLR4 protein levels were unaffected in CDC50A-depleted cells. Apparently, sustained TLR4 surface expression accounts for the reduced tolerized state in CDC50A-depleted cells.

CDC50A-depleted cells were phenocopied by ATP8B1- and ATP11A- depleted cells with regard to enhanced LPS-induced JNK1/2 activation, TNFα output and impaired LPS-induced internalization of TLR4. This underscores the need for obligate heterodimerization of a member of the P4-ATPase family of phospholipid flippases with CDC50A to form an active phospholipid flippase complex. These phenotypes were, however, not as dramatic as in CDC50A-depleted cells, which can be best explained by residual activity of these P4-ATPases or activity of other, redundant P4-ATPases.

How are P4-ATPases involved in the endocytic retrieval of TLR4? Although CDC50A was detected in intracellular vesicles, both ATP8B1 and ATP11A predominantly localized to the plasma membrane of THP-1 cells, which renders a role in the endocytic retrieval, recycling and/or TGN-to plasma membrane delivery likely. One of the preferred canonical substrates for ATP8B1 and ATP11A is phosphatidylserine (PS) (12, 27, 51), although recently phosphatidylcholine (PC) has been proposed to be a substrate for ATP8B1 as well (51). PS has been implicated to drive endocytic processes in human and yeast cells (52-54); for instance, Farge et al. have shown that the formation of endocytic vesicles was enhanced when the cytosolic surface area was expanded upon incubation of the cells with PS and PE analogs, a process that depended on a plasma membrane flippase activity (52). P4- ATPases can catalyze a local concentration of phospholipids and the consequential

141

Chapter 6 vesiculation according to the bilayer couple hypothesis (55). This hypothesis proposes that if one leaflet of a bilayer expands through a local increase in phospholipids, the coupled leaflet follows, which leads to bending of the bilayer. The resulting membrane curvature facilitates binding of curvature-stabilizing proteins and / or proteins of the vesicle-generating machinery. For instance, F-BAR domain- containing proteins sense and stabilize shallow membrane curvatures in the early steps of membrane invagination and contain positively charged membrane binding domains that interact with PS (56-58). In addition, the negative charge of the serine head group in PS is important for recruiting positively charged proteins of the vesicle-generating machinery to initiation sites within the endocytic pathway (59). Such a mechanism has recently been confirmed by Xu et al. (60) who showed that localized PS flipping by the S. cerevisiae P4-ATPase Drs2p creates a negatively charged, curved membrane structure that recruits the ADP-ribosylation factor GTPase activating protein (ARF-GAP) Gcs1, a major regulator of the vesicle- generating machinery (61). In addition, Lui et al. previously demonstrated an essential role for the PS flippase Drs2p in the formation of clathrin-coated vesicles at the TGN (62). Very recently Bruurs et al. have shown that ATP8B1 is essential for clustering of Cdc42, which can interact with PS (63), during the establishment of the apical membrane of enterocytes (64). Based on these observations, we hypothesize that ATP8B1 and/or ATP11A mediate local clustering of PS or other phospholipids at the plasma membrane to induce a curvature which facilitates the binding of proteins of the vesicle-generating machinery, including clathrin. This is the initiating event in the biogenesis of clathrin-coated pits via which LPS-ligated TLR4 is endocytosed. Indeed, Husebye et al. previously reported that endocytosis of TLR4 relies on clathrin (4).

Here we report an essential role for CDC50A-associated phospholipid flippases in the innate immune response. Although it remains to be determined how these proteins are involved in the inflammatory reaction and what the relative contribution is to this process, we have identified two potential candidates i.e. ATP8B1 and ATP11A, both of which are expressed in human macrophages. Deficiency of ATP8B1 or ATP11A is associated with chronic inflammatory conditions in humans. For instance, patients with ATP8B1 deficiency present with severe chronic liver disease but are also susceptible to pneumonia (65, 66). Previously, it was proposed that pulmonary infection in ATP8B1 deficiency causes pulmonary cardiolipin levels to rise leading to impaired lung (surfactant) function,

142

Chapter 6 leading to pulmonary inflammation, a hypothesis that was debated by us (67, 68). Based on our present work, an alternative explanation can be that patients have problems attenuating the TLR4-mediated inflammatory reaction in pulmonary macrophages. Similarly, in a recent genome-wide association study ATP11A was associated with fibrotic idiopathic interstitial pneumonias, a group of pulmonary disorders associated with inflammation and fibrosis (69). Whether P4-ATPase genes are novel risk genes for inflammatory disease needs to be established.

Acknowledgements

We thank Drs. L.W. Klomp and J.B. Koenderink for polyclonal antibodies to ATP8B1 and ATP1A1, respectively, and Drs. B.C. Tilly and H.R. de Jonge for the Rab11-eGFP plasmid.

References

1. Kawai T, Akira S. The role of pattern-recognition receptors in innate immunity: update on Toll-like receptors. Nature immunology. 2010;11(5):373-84. 2. Poltorak A, Ricciardi-Castagnoli P, Citterio S, Beutler B. Physical contact between lipopolysaccharide and toll-like receptor 4 revealed by genetic complementation. ProcNatlAcadSciUSA. 2000;97(5):2163-7. 3. Lee CC, Avalos AM, Ploegh HL. Accessory molecules for Toll-like receptors and their function. Nat Rev Immunol. 2012;12(3):168-79. 4. Husebye H, Halaas O, Stenmark H, Tunheim G, Sandanger O, Bogen B, et al. Endocytic pathways regulate Toll-like receptor 4 signaling and link innate and adaptive immunity. EMBO J. 2006;25(4):683-92. 5. Zanoni I, Ostuni R, Marek LR, Barresi S, Barbalat R, Barton GM, et al. CD14 controls the LPS-induced endocytosis of Toll-like receptor 4. Cell. 2011;147(4):868-80. 6. Jiang Z, Georgel P, Du X, Shamel L, Sovath S, Mudd S, et al. CD14 is required for MyD88-independent LPS signaling. Nat Immunol. 2005;6(6):565-70. 7. Kagan JC, Su T, Horng T, Chow A, Akira S, Medzhitov R. TRAM couples endocytosis of Toll-like receptor 4 to the induction of interferon-beta. NatImmunol. 2008;9(4):361-8. 8. Lucas K, Maes M. Role of the Toll Like receptor (TLR) radical cycle in chronic inflammation: possible treatments targeting the TLR4 pathway. Mol Neurobiol. 2013;48(1):190-204. 9. Rajaiah R, Perkins DJ, Ireland DD, Vogel SN. CD14 dependence of TLR4 endocytosis and TRIF signaling displays ligand specificity and is dissociable in endotoxin tolerance. Proc Natl Acad Sci U S A. 2015;112(27):8391-6. 10. Sebastian TT, Baldridge RD, Xu P, Graham TR. Phospholipid flippases: building asymmetric membranes and transport vesicles. Biochim Biophys Acta. 2012;1821(8):1068-77. 11. van der Mark VA, Elferink RP, Paulusma CC. P4 ATPases: Flippases in Health and Disease. Int J Mol Sci. 2013;14(4):7897-922. 12. Lopez-Marques RL, Theorin L, Palmgren MG, Pomorski TG. P4-ATPases: lipid flippases in cell membranes. Pflugers Arch. 2014;466(7):1227-40. 13. Holthuis JC, Menon AK. Lipid landscapes and pipelines in membrane homeostasis. Nature. 2014;510(7503):48-57. 14. Pomorski T, Lombardi R, Riezman H, Devaux PF, van Meer G, Holthuis JC. Drs2p-related P-type ATPases Dnf1p and Dnf2p are required for phospholipid translocation across the yeast plasma membrane and serve a role in endocytosis. Mol Biol Cell. 2003;14(3):1240-54. 15. Gall WE, Geething NC, Hua Z, Ingram MF, Liu K, Chen SI, et al. Drs2p-dependent formation of exocytic clathrin-coated vesicles in vivo. Curr Biol. 2002;12(18):1623-7. 16. Hua Z, Fatheddin P, Graham TR. An essential subfamily of Drs2p-related P-type ATPases is required for protein trafficking between Golgi complex and endosomal/vacuolar system. Mol Biol Cell. 2002;13(9):3162-77. 17. Ruaud AF, Nilsson L, Richard F, Larsen MK, Bessereau JL, Tuck S. The C. elegans P4-ATPase TAT-1 regulates lysosome biogenesis and endocytosis. Traffic. 2009;10(1):88-100. 18. Chen B, Jiang Y, Zeng S, Yan J, Li X, Zhang Y, et al. Endocytic sorting and recycling require membrane phosphatidylserine asymmetry maintained by TAT-1/CHAT-1. PLoS Genet. 2010;6(12):e1001235. 19. Poulsen LR, Lopez-Marques RL, McDowell SC, Okkeri J, Licht D, Schulz A, et al. The Arabidopsis P4-ATPase ALA3 localizes to the golgi and requires a beta-subunit to function in lipid translocation and secretory vesicle formation. Plant Cell. 2008;20(3):658-76. 20. Moreno-Smith M, Halder JB, Meltzer PS, Gonda TA, Mangala LS, Rupaimoole R, et al. ATP11B mediates platinum resistance in ovarian cancer. J Clin Invest. 2013;123(5):2119-30. 21. van der Mark VA, de Waart DR, Ho-Mok KS, Tabbers MM, Voogt HW, Oude Elferink RP, et al. The lipid flippase heterodimer ATP8B1-CDC50A is essential for surface expression of the apical sodium-dependent bile acid transporter (SLC10A2/ASBT) in intestinal Caco-2 cells. Biochim Biophys Acta. 2014;1842(12 Pt A):2378-86. 22. Lee S, Uchida Y, Wang J, Matsudaira T, Nakagawa T, Kishimoto T, et al. Transport through recycling endosomes requires EHD1 recruitment by a phosphatidylserine translocase. EMBO J. 2015;34(5):669-88. 23. Bull LN, van Eijk MJ, Pawlikowska L, DeYoung JA, Juijn JA, Liao M, et al. A gene encoding a P-type ATPase mutated in two forms of hereditary cholestasis. Nat Genet. 1998;18(3):219-24.

143

Chapter 6

24. Onat OE, Gulsuner S, Bilguvar K, Nazli Basak A, Topaloglu H, Tan M, et al. Missense mutation in the ATPase, aminophospholipid transporter protein ATP8A2 is associated with cerebellar atrophy and quadrupedal locomotion. Eur J Hum Genet. 2013;21(3):281-5. 25. Katoh Y, Katoh M. Identification and characterization of CDC50A, CDC50B and CDC50C genes in silico. Oncol Rep. 2004;12(4):939- 43. 26. Folmer DE, Mok KS, de Wee SW, Duijst S, Hiralall JK, Seppen J, et al. Cellular localization and biochemical analysis of mammalian CDC50A, a glycosylated beta-subunit for P4 ATPases. J Histochem Cytochem. 2012;60(3):205-18. 27. Paulusma CC, Folmer DE, Ho-Mok KS, de Waart DR, Hilarius PM, Verhoeven AJ, et al. ATP8B1 requires an accessory protein for endoplasmic reticulum exit and plasma membrane lipid flippase activity. Hepatology. 2008;47(1):268-78. 28. van der Velden LM, Wichers CG, van Breevoort AE, Coleman JA, Molday RS, Berger R, et al. Heteromeric interactions required for abundance and subcellular localization of human CDC50 proteins and class 1 P4-ATPases. J Biol Chem. 2010;285(51):40088-96. 29. Bryde S, Hennrich H, Verhulst PM, Devaux PF, Lenoir G, Holthuis JC. CDC50 proteins are critical components of the human class-1 P4- ATPase transport machinery. J Biol Chem. 2010;285(52):40562-72. 30. Takatsu H, Baba K, Shima T, Umino H, Kato U, Umeda M, et al. ATP9B, a P4-ATPase (a putative aminophospholipid translocase), localizes to the trans-Golgi network in a CDC50 protein-independent manner. J Biol Chem. 2011;286(44):38159-67. 31. Seppen J, Rijnberg M, Cooreman MP, Oude Elferink RP. Lentiviral vectors for efficient transduction of isolated primary quiescent hepatocytes. J Hepatol. 2002;36(4):459-65. 32. Choudhury A, Dominguez M, Puri V, Sharma DK, Narita K, Wheatley CL, et al. Rab proteins mediate Golgi transport of caveola- internalized glycosphingolipids and correct lipid trafficking in Niemann-Pick C cells. J Clin Invest. 2002;109(12):1541-50. 33. Daigneault M, Preston JA, Marriott HM, Whyte MK, Dockrell DH. The identification of markers of macrophage differentiation in PMA- stimulated THP-1 cells and monocyte-derived macrophages. PLoS One. 2010;5(1):e8668. 34. Ruijter JM, Ramakers C, Hoogaars WM, Karlen Y, Bakker O, van den Hoff MJ, et al. Amplification efficiency: linking baseline and bias in the analysis of quantitative PCR data. Nucleic Acids Res. 2009;37(6):e45. 35. Vandesompele J, De Preter K, Pattyn F, Poppe B, Van Roy N, De Paepe A, et al. Accurate normalization of real-time quantitative RT- PCR data by geometric averaging of multiple internal control genes. Genome Biol. 2002;3(7):RESEARCH0034. 36. Pawlikowska L, Groen A, Eppens EF, Kunne C, Ottenhoff R, Looije N, et al. A mouse genetic model for familial cholestasis caused by ATP8B1 mutations reveals perturbed bile salt homeostasis but no impairment in bile secretion. Hum Mol Genet. 2004;13(8):881-92. 37. Koenderink JB, Geibel S, Grabsch E, De Pont JJ, Bamberg E, Friedrich T. Electrophysiological analysis of the mutated Na,K-ATPase cation binding pocket. J Biol Chem. 2003;278(51):51213-22. 38. Jaguin M, Houlbert N, Fardel O, Lecureur V. Polarization profiles of human M-CSF-generated macrophages and comparison of M1- markers in classically activated macrophages from GM-CSF and M-CSF origin. Cellular immunology. 2013;281(1):51-61. 39. Mansell A, Smith R, Doyle SL, Gray P, Fenner JE, Crack PJ, et al. Suppressor of cytokine signaling 1 negatively regulates Toll-like receptor signaling by mediating Mal degradation. Nat Immunol. 2006;7(2):148-55. 40. Medvedev AE, Kopydlowski KM, Vogel SN. Inhibition of lipopolysaccharide-induced signal transduction in endotoxin-tolerized mouse macrophages: dysregulation of cytokine, chemokine, and toll-like receptor 2 and 4 gene expression. J Immunol. 2000;164(11):5564-74. 41. Nomura F, Akashi S, Sakao Y, Sato S, Kawai T, Matsumoto M, et al. Cutting edge: endotoxin tolerance in mouse peritoneal macrophages correlates with down-regulation of surface toll-like receptor 4 expression. J Immunol. 2000;164(7):3476-9. 42. Martin M, Katz J, Vogel SN, Michalek SM. Differential induction of endotoxin tolerance by lipopolysaccharides derived from Porphyromonas gingivalis and Escherichia coli. J Immunol. 2001;167(9):5278-85. 43. Yamamoto M, Sato S, Hemmi H, Uematsu S, Hoshino K, Kaisho T, et al. TRAM is specifically involved in the Toll-like receptor 4- mediated MyD88-independent signaling pathway. Nat Immunol. 2003;4(11):1144-50. 44. Preta G, Cronin JG, Sheldon IM. Dynasore - not just a dynamin inhibitor. Cell Commun Signal. 2015;13:24. 45. Preta G, Lotti V, Cronin JG, Sheldon IM. Protective role of the dynamin inhibitor Dynasore against the cholesterol-dependent cytolysin of Trueperella pyogenes. FASEB J. 2015;29(4):1516-28. 46. Plociennikowska A, Hromada-Judycka A, Borzecka K, Kwiatkowska K. Co-operation of TLR4 and raft proteins in LPS-induced proinflammatory signaling. Cell Mol Life Sci. 2015;72(3):557-81. 47. Triantafilou M, Miyake K, Golenbock DT, Triantafilou K. Mediators of innate immune recognition of bacteria concentrate in lipid rafts and facilitate lipopolysaccharide-induced cell activation. Journal of cell science. 2002;115(Pt 12):2603-11. 48. Pfeiffer A, Bottcher A, Orso E, Kapinsky M, Nagy P, Bodnar A, et al. Lipopolysaccharide and ceramide docking to CD14 provokes ligand-specific receptor clustering in rafts. European journal of immunology. 2001;31(11):3153-64. 49. Klein DC, Skjesol A, Kers-Rebel ED, Sherstova T, Sporsheim B, Egeberg KW, et al. CD14, TLR4 and TRAM Show Different Trafficking Dynamics During LPS Stimulation. Traffic. 2015. 50. Biswas SK, Lopez-Collazo E. Endotoxin tolerance: new mechanisms, molecules and clinical significance. Trends Immunol. 2009;30(10):475-87. 51. Takatsu H, Tanaka G, Segawa K, Suzuki J, Nagata S, Nakayama K, et al. Phospholipid flippase activities and substrate specificities of human type IV P-type ATPases localized to the plasma membrane. J Biol Chem. 2014;289(48):33543-56. 52. Farge E, Ojcius DM, Subtil A, Dautry-Varsat A. Enhancement of endocytosis due to aminophospholipid transport across the plasma membrane of living cells. Am J Physiol. 1999;276(3 Pt 1):C725-33. 53. Sun Y, Drubin DG. The functions of anionic phospholipids during clathrin-mediated endocytosis site initiation and vesicle formation. Journal of cell science. 2012;125(Pt 24):6157-65. 54. Zha X, Genest J, Jr., McPherson R. Endocytosis is enhanced in Tangier fibroblasts: possible role of ATP-binding cassette protein A1 in endosomal vesicular transport. J Biol Chem. 2001;276(42):39476-83. 55. Sheetz MP, Singer SJ. Biological membranes as bilayer couples. A molecular mechanism of drug-erythrocyte interactions. Proc Natl Acad Sci U S A. 1974;71(11):4457-61. 56. Coutinho-Budd J, Ghukasyan V, Zylka MJ, Polleux F. The F-BAR domains from srGAP1, srGAP2 and srGAP3 regulate membrane deformation differently. Journal of cell science. 2012;125(Pt 14):3390-401. 57. Uezu A, Umeda K, Tsujita K, Suetsugu S, Takenawa T, Nakanishi H. Characterization of the EFC/F-BAR domain protein, FCHO2. Genes Cells. 2011;16(8):868-78. 58. Rao Y, Haucke V. Membrane shaping by the Bin/amphiphysin/Rvs (BAR) domain protein superfamily. Cell Mol Life Sci. 2011;68(24):3983-93. 59. Yeung T, Gilbert GE, Shi J, Silvius J, Kapus A, Grinstein S. Membrane phosphatidylserine regulates surface charge and protein localization. Science. 2008;319(5860):210-3. 60. Xu P, Baldridge RD, Chi RJ, Burd CG, Graham TR. Phosphatidylserine flipping enhances membrane curvature and negative charge required for vesicular transport. J Cell Biol. 2013;202(6):875-86. 61. Cherfils J, Zeghouf M. Regulation of small GTPases by GEFs, GAPs, and GDIs. Physiol Rev. 2013;93(1):269-309. 62. Liu K, Surendhran K, Nothwehr SF, Graham TR. P4-ATPase requirement for AP-1/clathrin function in protein transport from the trans- Golgi network and early endosomes. MolBiolCell. 2008;19(8):3526-35.

144

Chapter 6

63. Fairn GD, Hermansson M, Somerharju P, Grinstein S. Phosphatidylserine is polarized and required for proper Cdc42 localization and for development of cell polarity. Nat Cell Biol. 2011;13(12):1424-30. 64. Bruurs LJ, Donker L, Zwakenberg S, Zwartkruis FJ, Begthel H, Knisely AS, et al. ATP8B1-mediated spatial organization of Cdc42 signaling maintains singularity during enterocyte polarization. J Cell Biol. 2015;210(7):1055-63. 65. Pawlikowska L, Strautnieks S, Jankowska I, Czubkowski P, Emerick K, Antoniou A, et al. Differences in presentation and progression between severe FIC1 and BSEP deficiencies. J Hepatol. 2010;53(1):170-8. 66. Whitington PF, Freese DK, Alonso EM, Schwarzenberg SJ, Sharp HL. Clinical and biochemical findings in progressive familial intrahepatic cholestasis. J Pediatr Gastroenterol Nutr. 1994;18(2):134-41. 67. Ray NB, Durairaj L, Chen BB, McVerry BJ, Ryan AJ, Donahoe M, et al. Dynamic regulation of cardiolipin by the lipid pump Atp8b1 determines the severity of lung injury in experimental pneumonia. NatMed. 2010;16(10):1120-7. 68. Paulusma CC, Houwen RH, Williamson PL. The flip side of cardiolipin import. Nat Med. 2011;17(4):413; author reply -4. 69. Fingerlin TE, Murphy E, Zhang W, Peljto AL, Brown KK, Steele MP, et al. Genome-wide association study identifies multiple susceptibility loci for pulmonary fibrosis. Nat Genet. 2013;45(6):613-20.

145

Chapter 6

Supplemental data

146

Chapter 6

Supplementary figure 1. (A) CD11B surface expression in shControl and CDC50A-depleted THP-1 macrophages was determined by flow cytometry. Data are expressed as mean fluorescence intensity (MFI) ± standard deviation of triplicate wells. No statistical differences by a Student’s t-test. (B) TNFα excretion and CDC50A mRNA expression in CDC50A-depleted THP-1 cells 4 h post LPS (100 ng/ml). Statistical significance was tested by one- way ANOVA with Bonferroni’s correction for multiple testing; *p< 0,05; **p < 0,0005 (C) CD80 and CD200R mRNA expression in M0, M1 and M2 human monocyte-derived macrophages.

Supplementary figure 2. (A) CD14 surface expression in shControl and CDC50A-depleted THP-1 macrophages was determined by flow cytometry. Data shown are MFIs ± standard deviation of triplicate wells. Statistical significance was tested by a Student’s t-test, *p<0.00005. (B) CD14 mRNA levels in shControl and CDC50A- depleted THP-1 cells 3 hours post-LPS administration. Statistical significance was tested by a Student’s t-test, *p<0.002, **p<0.008. (C) CD14 surface expression in shControl and CDC50A-depleted THP-1 macrophages after stimulation with 100 ng/ml LPS. Cells were analyzed and data expressed as described in figure 4 A.

147

Chapter 6

Appendix

Materials and methods

Isolation of murine bone marrow-derived macrophages and alveolar macrophages

Bone marrow-derived macrophages (BMDMs) were isolated according to standard procedures from femurs and tibias of age-matched (3-6 months) C57BL/6J male Atp8b1G308V/G308V mutant and wild type mice (1). Cells were cultured on bacteriologic plastic petri dishes in RPMI 1640 medium supplemented with 10% FBS, 2 mM L- glutamine, 100 U/ml penicillin, 100 U/ml streptomycin, and 15% L-cell conditioned medium at 37 ºC in a 5% CO2 humidified atmosphere. Nine days post-isolation, macrophages were detached using 15 mM citrate, 135 mM KCl solution at 37 °C. Cells were transferred to RPMI 1640 medium supplemented with 10% FBS, 2 mM L-glutamine, 100 U/ml penicillin, 100 U/ml streptomycin and HEPES to inactivate the citrate solution, seeded (1 x 106 cells per well) in a 6-well tissue culture plate, and cultured at 37 ºC in a 5% CO2 humidified atmosphere. The next day, cells were challenged with indicated concentrations of LPS and medium. RNA and protein lysates were harvested and analyzed as described. Alveolar macrophages were isolated according to (2). Briefly, mice were anesthetized and trachea were surgically exposed. Pre-warmed (37°C) calcium- and magnesium-free Dulbecco’s PBS was infused into the trachea and withdrawn. Isolated macrophages were seeded at 5 x 104/cm2 and cultured in RPMI 1640 medium supplemented with 10% FCS, 2 mM L-glutamine, 100 U/ml penicillin, and 100 U/ml streptomycin at 37 ºC in a 5% CO2 humidified atmosphere and used for experiments within 24 hours.

Results

ATP8B1 deficient bone marrow-derived macrophages (BMDM) nor alveolar macrophages (AM) are hyper-responsive to LPS.

Since ATP8B1 deficient mice were available (3, 4), we analyzed the inflammatory response to LPS in BMDMs isolated from wild type and ATP8B1-deficient mice. Whereas in THP-1 cells (figure 5A) and human primary monocytes and macrophages (figure 5G) ATP8B1 and ATP11A were intermediately expressed, mouse BMDMs and AMs expressed extremely low levels of Atp8b1 while Atp11a was abundantly expressed (appendix figure 1A). Despite very low Atp8b1 expression, we analyzed the inflammatory reaction to LPS in these cells. In contrast

148

Chapter 6 to ATP8B1-depleted THP-1 macrophages, neither ATP8B1-deficient BMDMs (appendix figure 3B) or AMs (appendix figure 1C) displayed elevated TNFα excretion upon a 24 hour LPS challenge. LPS-induced activation of Erk1/2 and Jnk1/2 was not affected in the ATP8B1-deficient BMDMs (appendix figure 1D). Furthermore, Tnfα, Il1ß, nor Rantes mRNA expression levels were affected in ATP8B1-deficient BMDMs (appendix figure 1E). These results indicate that in mouse primary macrophages ATP8B1 does not play a role in the regulation of TLR4- mediated signaling.

Appendix figure 1. BMDMs and AMs of Atp8b1G308V/G308V mutant mice display no LPS-induced hypersecretion of inflammatory cytokines. (A) Atp8b1 and Atp11a mRNA levels in BMDMs of wild type and Atp8b1G308V/G308V (4 mice per group). Atp8b1 expression in wild type mice was set as 1. (B) Tnfα secretion in BMDMs from wild type and Atp8b1G308V/G308V mice 24 hours post LPS administration. Data shown are means ± standard deviation of triplicate wells. No statistical differences by one-way ANOVA. (C) Tnfα secretion in AMs from wild type and Atp8b1G308V/G308V mice 4 hours post LPS administration. Data shown are means ± standard deviation of triplicate wells. No statistical differences by one-way ANOVA. (D) Immunoblot analysis of activated Erk1/2 and JNK1/2 in BMDMs from wild type and Atp8b1G308V/G308V mice. ß-actin is included as a loading control. (E) Tnfα, Il-1β and Rantes mRNA levels in BMDMs from wild type and Atp8b1G308V/G308V mice 4 hours post LPS administration. Data shown are means ± standard deviation of quadruple wells. No statistical differences by a Student’s t-test. Data shown in this figure are representative of two independent experiments with similar results.

149

Chapter 6

References

1. Peiser L, Gough PJ, Kodama T, Gordon S. Macrophage class A scavenger receptor-mediated phagocytosis of Escherichia coli: role of cell heterogeneity, microbial strain, and culture conditions in vitro. Infect.Immun. 2000;68:1953-1963. 2. Zhang X, Goncalves R, Mosser DM. The isolation and characterization of murine macrophages. Curr.Protoc.Immunol. 2008;Chapter 14:Unit. 3. Pawlikowska L, Groen A, Eppens EF, Kunne C, Ottenhoff R, Looije N, Knisely AS, et al. A mouse genetic model for familial cholestasis caused by ATP8B1 mutations reveals perturbed bile salt homeostasis but no impairment in bile secretion. Hum Mol Genet 2004;13:881- 892. 4. Paulusma CC, Groen A, Kunne C, Ho-Mok KS, Spijkerboer AL, Rudi de Waart D, Hoek FJ, et al. Atp8b1 deficiency in mice reduces resistance of the canalicular membrane to hydrophobic bile salts and impairs bile salt transport. Hepatology 2006;44:195-204.

150

CHAPTER 7

SUMMARY AND PERSPECTIVES

Chapter 7

Summary

Progressive familial intrahepatic cholestasis type 1 (PFIC1), also known as Byler’s disease, was first described in 1969 in direct descendants of Jacob Byler (1). Jacob Byler was an Amish farmer who settled in the Pennsylvania area in the 18th century. Patients have recurrent cholestatic episodes that become persistent later on. The recurrent cholestatic pattern is similar to that of benign recurrent intrahepatic cholestasis type 1 (BRIC1). In 1998 PFIC1 and BRIC1 were linked to mutations in the gene encoding ATP8B1 (2). ATP8B1 is a phospholipid translocating protein that catalyzes the ‘flipping’ of aminophospholipids in membranes in order to maintain a proper phospholipid membrane asymmetry. Previous studies in our lab, aided by the mutant Atp8b1G308V/G308V knock-in mouse, have provided a working hypothesis to explain the cholestasis: a lack of ATP8B1 in the canalicular membrane of hepatocytes results in an improper membrane asymmetry, which renders the membrane prone to extraction of critical membrane components by bile salts (3).

However, the causes of several extrahepatic phenotypes associated with PFIC1 and BRIC1, such as diarrhea and susceptibility to pulmonary infection, have barely been studied and as of yet have not been elucidated. In this thesis we investigated the physiological role of ATP8B1 in human intestinal, pulmonary, and innate immune cells. In addition, we studied the extrahepatic functions of ATP8B1 ex vivo in the mutant Atp8b1G308V/G308V mouse.

Chapter 2 gives an overview on the subject of P4-ATPases in general and ATP8B1 in specific. We summarize the current knowledge on the basic biochemical function of P4-ATPases, their proposed role in intracellular vesicle transport, and their (patho)physiological function in animals and humans.

In chapter 3 we studied several common missense mutations causing PFIC1 (G308V, D554N, G1040R) or BRIC1 (D70N, I661T), and two mutations that have been associated with intrahepatic cholestasis of pregnancy (ICP) (D70N, R867C). We analyzed the effect of these mutations on the intracellular localization of the ATP8B1 protein in WIF-B9 cells via confocal microscopy, while stability of the protein and interaction with its beta subunit CDC50A was studied in Chinese hamster ovary

154

Chapter 7

(CHO) cells using immunoblotting techniques. None of the PFIC1 mutants localized to the canalicular membrane of WIF-B9 cells, but instead displayed an intracellular staining that likely represented the endoplasmic reticulum. All BRIC1/ICP mutants showed a similar intracellular staining as wild-type ATP8B1 that included normal canalicular localization. Interestingly, in CHO cells total protein levels of the G308V, D554N, and I661T mutants were reduced, but restored to normal levels after proteasomal inhibition, indicating decreased protein stability in these mutants leading to increased proteasomal degradation. The G308V and D554N mutants did not coimmunoprecipitate with CDC50A in CHO cells, whereas the other mutants displayed reduced interaction. The reduced interaction with CDC50A correlated with the level of mutant ATP8B1 protein expression at the plasma membrane surface of CHO cells: the G308V and D554N mutant were not detectable at the surface, while the BRIC1/ICP mutants were expressed at levels comparable to those of wild type ATP8B1. These data provide an explanation for the difference in severity between the phenotypes of PFIC1 and BRIC1.

Throughout chapter 4 we studied the hypothesis that intestinal deficiency of ATP8B1 results in bile acid malabsorption, which could lead to the diarrhea observed in a proportion of PFIC1/BRIC1 patients and that may exacerbate after liver transplantation. Using surface biotinylation experiments we showed that apical membrane localization of the apical sodium-dependent bile salt transporter (ASBT) was strongly reduced in human intestinal Caco-2 cells depleted for ATP8B1, a phenotype that coincided with strongly reduced bile salt uptake. We presented data that insertion of the protein in the apical membrane was impaired by a lack of ATP8B1 expression. In addition, in coimmunoprecipation experiments we demonstrated for the first time that endogenously-expressed ATP8B1 and CDC50A exist as a functional heterodimer. Similar to ATP8B1, depletion of CDC50A resulted in reduced activity and surface expression of ASBT in Caco-2 cells. The effects of depletion of ATP8B1 and CDC50A on ASBT seemed to not be a general defect in apical membrane organization, since apical expression and activity of ABCC2 was unaffected in ATP8B1-depleted Caco-2 cells. Moreover, surface expression of other apical membrane proteins, including CD13, CD26, ABCB1, and SLC2A2 was not decreased in ATP8B1- or CDC50A-depleted Caco-2 cells. Interestingly, analyses of

155

Chapter 7 stool samples of post-transplant PFIC1 patients (with restored bile flow) demonstrated that fecal bile salt content was unchanged, whereas sodium and chloride concentrations were elevated and potassium levels were decreased. These results indicate that a functional ATP8B1-CDC50A heterodimer is necessary for apical membrane localization of ASBT in Caco-2 cells. Furthermore, PFIC1/BRIC1 diarrhea has a secretory origin to which mild bile salt malabsorption could contribute.

Chapter 5 continued our work on the role of ATP8B1 in intestinal cells by analyzing its effect on the cystic fibrosis transmembrane regulator (CFTR), a chloride channel essential for epithelial fluid transport and mucal clearance. We showed that in ATP8B1-depleted human intestinal Caco-2 and T84 cells and human lung epithelial Calu-3 cells CFTR mRNA and total protein were significantly reduced. In addition, using a surface biotinylation approach, we demonstrated that apical plasma membrane insertion of CFTR is impaired in ATP8B1-depleted T84 cells upon exogenous, inducible overexpression of CFTR. Functionally, the reduced CFTR protein levels in ATP8B1-depleted T84 cells were mirrored by reduced CFTR activity when assessed on Transwell-cultured monolayers, either by transepithelial current measurements or ratiometric fluorescent chloride-sensor experiments. In contrast, in the mouse model for PFIC1, the mutant Atp8b1G308V/G308V knock-in mouse, we could neither observe a decreased expression of CFTR on mRNA or protein levels nor a reduction of transepithelial chloride currents in explanted intestinal segments or gall-bladder. These findings may explain the pathogenesis of several extrahepatic phenotypes (pulmonary infection, elevated sweat electrolytes) observed in patients (but not in mice) with ATP8B1 deficiency, possibly caused by partial inactivation of CFTR function.

In chapter 6 we analyzed the role of ATP8B1 and CDC50A in macrophages, specifically by studying Toll-like receptor 4 (TLR4)-mediated signaling after activation with lipopolysaccharide (LPS). Depletion of CDC50A in human THP-1 macrophages and primary human peripheral blood mononuclear cell (PBMC)- derived M1 macrophages resulted in LPS-induced hypersecretion of proinflammatory cytokines, including TNFα, IL-1β, and IL-6. In addition, in CDC50A-depleted THP-1 macrophages MAP kinase signaling was enhanced and nuclear NF-κB was constitutively phosphorylated (activated). This hyperactivity

156

Chapter 7 was also observed after an additional induction with LPS in CDC50A-depleted THP- 1 macrophages, which coincided with a lack of downregulation of total TLR4 protein, indicating impaired regulation of endotoxin-induced tolerance. LPS-induced internalization of TLR4 was strongly reduced in CDC50A-depleted THP-1 macrophages, which corresponded with impaired endosomal MyD88-independent signaling as indicated by reduced IFNβ, RANTES, and SOCS1 mRNA levels. IFNβ and SOCS1 mRNA expression was also decreased in CDC50A-depleted PBMC- derived macrophages. Knockdown of ATP8B1 and ATP11A in THP-1 macrophages resulted in a similar hyperactive proinflammatory phenotype and reduced LPS- induced internalization of TLR4, albeit to a lesser extent than CDC50A depletion. In contrast, primary pulmonary- or bone marrow-derived macrophages from mutant Atp8b1G308V/G308V mice did not phenocopy human cells. These data indicate that human P4-ATPases, including ATP8B1 and ATP11A, are involved in the regulation of endocytosis and intracellular trafficking of TLR4, thereby providing a rationale for a possible role in chronic (pulmonary) inflammatory conditions.

Our results show the physiological importance of ATP8B1 in human cells. P4- ATPases are important conserved flippases with an emerging role in intracellular trafficking. As shown in this thesis, ATP8B1 seems to play a crucial part in the transport of vesicles to and/or from the plasma membrane. In addition to the proteins described in this thesis, ATP8B1 deficiency affects hepatocyte canalicular membrane- and cochlear hair cell membrane integrity and glucose-stimulated insulin secretion (chapter 1). Therefore it may be hypothesized that P4-ATPases are essential indirect mediators of protein trafficking, which has not only an impact on the physiology of the cells in which they reside, but also on the organism itself.

Perspectives

Transduction of information from the outside milieu to the cell depends on the sensing of signals via a properly organized membrane protein machinery. This is important for the maintenance of cellular homeostasis and particularly relevant for the regulation of the immune response and cellular proliferation (see chapter 1 and 5). Because P4-ATPases are essential mediators for proper localization of membrane

157

Chapter 7 proteins, membrane integrity and intracellular trafficking, inflammation and cancer are two areas where P4-ATPases may play, yet to be identified, essential roles.

Inflammation, which is defined as the biological response to harmful stimuli, such as pathogens, damaged cells, or irritants, plays a role in many human pathologies (e.g. atherosclerosis, auto-immune disease and allergies, cancer, and infection). Recognition and subsequent transduction of danger signals and attenuating safety signals occurs via a multitude of membrane proteins present in cells of the innate and adaptive immune system (4, 5). Rheumatoid arthritis (RA) is a complex autoimmune disease that primarily results in a chronic inflammatory condition of the joints. Locally reducing the increased levels of pro-inflammatory cytokines IL-1, IL- 6, IL-17, or TNFα that are present during RA, or targeting their respective receptors via antibodies or antagonists has had some therapeutic success (6, 7). A more general approach of downregulating the surface expression of multiple different receptors by targeting intracellular trafficking routes that are regulated by specific P4-ATPases could have a more potent effect. Therefore, targeting P4-ATPases as a therapeutic treatment and thereby regulating the localization of essential membrane receptors could be advantageous in instances where the immune system is dysregulated.

P4-ATPases may play important roles in the efficacy of specific cancer treatments by controlling the internalization of target proteins, the delivery and membrane insertion of newly-synthesized target proteins, or directly by mediating drug excretion and uptake. Many cancer treatments are targeted to cancer-specific plasma membrane proteins whose expression at the plasma membrane is susceptible to downregulation following treatment. For example, increased internalization of the epidermal growth factor receptor can protect against cell death induced by gefitinib (8), a drug approved for first-line treatment of non-small cell lung cancer, while death receptor 4 is constitutively endocytosed in certain breast cancer cells resulting in its escape to TRAIL-induced apoptosis (9). Reversibly, upregulation of xenobiotic efflux transporters is a well-known phenomenon in cancer cells treated with small molecule drugs (10), a process in which P4-ATPases may be involved either by increasing their delivery to the plasma membrane or inhibiting their internalization. Furthermore, several studies have confirmed the role that P4-ATPases play in the

158

Chapter 7 effective excretion or uptake of anti-cancer drugs: ATP11B directs cisplatin- containing secretory vesicles to the plasma membrane (11), in this way conferring resistance to cis-platinum, while yet unknown CDC50A-associating P4-ATPases mediate the uptake of perifosine, an antitumor alkylphospholipid that may target Akt and / or supports recruitment of death receptors into lipid rafts and induces growth arrest and apoptosis (12). If the activity of specific P4-ATPases could be directed in such a way that intracellular vesicle trafficking or endocytosis of specific membrane proteins is altered this would present an attractive option to enhance the efficacy of other therapeutic treatments.

When performing fundamental studies in animal models on the exact mechanism of action and direction of vesicular traffic of ATP8B1 and other P4-ATPases the researcher has to be wary of species differences. Although many scientific and medical advances have been made using animals as a basic translational model, many others animal models have failed during the translation from animal to the clinic (13). For example, as little as 8% of cancer treatments successful in animal models subsequently passes Phase I clinical trials (14). The work in this thesis offers a striking example of the differences between mouse and human cells with regard to the function of ATP8B1 in intestines and innate immune cells. However, in recent years alternative translational approaches such as labs-on-a-chip, 3D organoid cultures, induced pluripotent stem cells (iPSC) obtained from e.g. urine or skin, humanized rodent models, and advanced in silico models have emerged that should eventually improve the predictive power of preclinical studies.

The recent advent of the organoid technology has provided researchers with more relevant physiological models of human diseases and are a tremendous advance over traditional 2D cell culture models (15). The self-organizing capabilities of stem cells can be harnessed to create organoids that possess relevant structural and functional properties of organs. Currently, small intestinal organoids and colonoids, although still displaying a fetal-like phenotype, develop a crypt-villus axis, but can be differentiated to contain all major epithelial cell types (16). Exploiting adult stem cells to create organoids enables the establishment of a bio-bank of intestinal, liver, and lung organoids from biopsies of PFIC1/BRIC1 patients. When studying a rare

159

Chapter 7 genetic disease such as PFIC1/BRIC1 these models would greatly facilitate more relevant studies on the pathophysiology and therapeutic possibilities in the different affected organs.

In conclusion, this thesis has provided important insights into the physiological role of ATP8B1with regard to the proper expression and functional activity of various (apical) membrane proteins in the intestine, lung, and innate immune system. Although, the Atp8b1G308V/G308V mutant mouse model has proven to be useful for the study of cholestasis and hearing loss, it was unfortunately not relevant for the phenotypes studied in this thesis. Therefore, it is my recommendation that future studies on the function of ATP8B1 regarding these phenotypes should be confirmed in stem cell-derived organoid models obtained from biopsies or iPSCs of PFIC1/BRIC1 patients.

1. Clayton RJ, Iber FL, Ruebner BH, McKusick VA. Byler disease. Fatal familial intrahepatic cholestasis in an Amish kindred. AmJDisChild. 1969;117(1):112-24. 2. Bull LN, van Eijk MJ, Pawlikowska L, DeYoung JA, Juijn JA, Liao M, et al. A gene encoding a P-type ATPase mutated in two forms of hereditary cholestasis. NatGenet. 1998;18(3):219-24. 3. Paulusma CC, Groen A, Kunne C, Ho-Mok KS, Spijkerboer AL, de Waart DR, et al. Atp8b1 deficiency in mice reduces resistance of the canalicular membrane to hydrophobic bile salts and impairs bile salt transport. Hepatology. 2006;44(1):195-204. 4. Chen L, Flies DB. Molecular mechanisms of T cell co-stimulation and co-inhibition. Nat Rev Immunol. 2013;13(4):227-42. 5. Kawai T, Akira S. The role of pattern-recognition receptors in innate immunity: update on Toll-like receptors. Nature immunology. 2010;11(5):373-84. 6. Bossaller L, Rothe A. Monoclonal antibody treatments for rheumatoid arthritis. Expert Opin Biol Ther. 2013;13(9):1257-72. 7. Kunwar S, Dahal K, Sharma S. Anti-IL-17 therapy in treatment of rheumatoid arthritis: a systematic literature review and meta-analysis of randomized controlled trials. Rheumatol Int. 2016;36(8):1065-75. 8. Kwak EL, Sordella R, Bell DW, Godin-Heymann N, Okimoto RA, Brannigan BW, et al. Irreversible inhibitors of the EGF receptor may circumvent acquired resistance to gefitinib. Proc Natl Acad Sci U S A. 2005;102(21):7665-70. 9. Zhang Y, Zhang B. TRAIL resistance of breast cancer cells is associated with constitutive endocytosis of death receptors 4 and 5. Mol Cancer Res. 2008;6(12):1861-71. 10. Ozben T. Mechanisms and strategies to overcome multiple drug resistance in cancer. FEBS Lett. 2006;580(12):2903-9. 11. Moreno-Smith M, Halder JB, Meltzer PS, Gonda TA, Mangala LS, Rupaimoole R, et al. ATP11B mediates platinum resistance in ovarian cancer. JClinInvest. 2013;123(5):2119-30. 12. Munoz-Martinez F, Torres C, Castanys S, Gamarro F. CDC50A plays a key role in the uptake of the anticancer drug perifosine in human carcinoma cells. BiochemPharmacol. 2010;80(6):793-800. 13. Perel P, Roberts I, Sena E, Wheble P, Briscoe C, Sandercock P, et al. Comparison of treatment effects between animal experiments and clinical trials: systematic review. BMJ. 2007;334(7586):197. 14. Mak IW, Evaniew N, Ghert M. Lost in translation: animal models and clinical trials in cancer treatment. Am J Transl Res. 2014;6(2):114- 8. 15. Clevers H. Modeling Development and Disease with Organoids. Cell. 2016;165(7):1586-97. 16. In JG, Foulke-Abel J, Estes MK, Zachos NC, Kovbasnjuk O, Donowitz M. Human mini-guts: new insights into intestinal physiology and host-pathogen interactions. Nat Rev Gastroenterol Hepatol. 2016;13(11):633-42.

160

APPENDICES

Samenvatting

Dankwoord

Publications

Authors and affiliations

Portfolio

Appendices

Samenvatting

Progressieve familiaire intrahepatische cholestase type 1 (PFIC1), ook wel Byler’s ziekte genoemd, is voor het eerst beschreven in 1969 bij directe afstammelingen van Jacob Byler, een Amish die zich in de loop van de 18e eeuw in Pennsylvania vestigde. PFIC1 patiënten hebben een terugkerende ophoping van gal in de lever (cholestase), die later verergert tot een continue cholestase. Het terugkerende patroon van cholestase komt overeen met de symptomen van benigne recurrente intrahepatische cholestase type 1 (BRIC1). In 1998 zijn PFIC1 en BRIC1 gekoppeld aan mutaties in het gen dat codeert voor ATP8B1. ATP8B1 is een eiwit dat het ‘flippen’ van vetmoleculen in celmembranen katalyseert om een correcte membraanomgeving te creëren en / of te onderhouden. Eerdere studies van ons lab in de mutante Atp8b1G308V/G308V knock-in muis (een model voor PFIC1/BRIC1), hebben tot een hypothese geleid om de cholestasis te verklaren: een gebrek aan ATP8B1 in het canaliculaire membraan van hepatocyten resulteert in een verkeerde membraanstructuur, waardoor galzouten essentiële membraancomponenten kunnen extraheren.

De oorzaak van enkele extrahepatische fenotypen die geassocieerd zijn met PFIC1 en BRIC1, zoals diarree en gevoeligheid voor longinfecties, zijn nauwelijks bestudeerd en daarom tot op heden ook nog niet ontrafeld. In dit proefschrift hebben wij de fysiologische rol van ATP8B1 in humane darm- en longcellen en cellen van het aangeboren immuunsysteem verder uitgediept. Bovendien hebben wij de extrahepatische functies van ATP8B1 bestudeerd in de Atp8b1G308V/G308V muis.

Hoofdstuk 2 geeft een overzicht over de familie van P4-ATPases in het algemeen en ATP8B1 in het bijzonder. Wij vatten de huidige kennis van de basale biochemische functie van P4-ATPases, hun voorgestelde rol in intracellulair blaasjestransport en hun (patho)fysiologische functie in dieren en mensen samen.

In hoofdstuk 3 hebben wij enkele veelvoorkomende PFIC1 (G308V, D554N, G1040R) en BRIC1 (D70N, I661T) mutaties, en mutaties geassocieerd met zwangerschapscholestase (ICP) (D70N, R867C) bestudeerd. Wij hebben de

164

Appendices

consequentie van deze mutaties op de intracellulaire lokalisatie van ATP8B1 geanalyseerd in WIF-B9 cellen via confocale microscopie. De stabiliteit van het eiwit en de interactie met zijn beta subeenheid CDC50A hebben we bestudeerd in Chinese hamster ovariumcellen (CHO) met immunoblottechnieken. Geen van de PFIC1 mutanten lokaliseerde in het canaliculaire membraan van WIF-B9 cellen, maar liet een intracellulaire kleuring zien die waarschijnlijk overeenkomt met het endoplasmatisch reticulum. Alle BRIC1/ICP mutanten lieten een normale canaliculaire lokalisatie zien die gelijk was aan wildtype ATP8B1. De meeste mutaties leiden waarschijnlijk tot een instabiel eiwit (mogelijk door onvolledige vouwing) aangezien na remming van het proteasoom (daar waar fout gevouwen eiwitten worden afgebroken) de totale eiwitniveaus weer tot normale niveaus teruggebracht werden. De G308V en D554N mutanten coimmunoprecipiteerden niet met CDC50A in CHO cellen, terwijl de andere mutanten een verminderde interactie vertoonden. De verminderde interactie met CDC50A correleerde met het niveau van eiwitexpressie van de ATP8B1 mutanten aan het oppervlak van het plasmamembraan van CHO cellen. Deze data geven een verklaring voor het verschil in de ernst tussen de fenotypen van PFIC1 en BRIC1.

In hoofdstuk 4 hebben wij de hypothese bestudeerd dat de diarree die in een deel van PFIC1/BRIC1 patiënten wordt gezien en die verergert na levertransplantatie veroorzaakt wordt door een defect in de galzoutopname in de darm. Met behulp van oppervlaktebiotinylering hebben wij laten zien dat de lokalisatie van de apicale natrium-afhankelijke galzouttransporter (ASBT) sterk verminderd was in humane intestinale Caco-2 cellen die gedepleteerd waren van ATP8B1. Wij vonden dat de insertie van het ASBT eiwit in het apicale membraan was aangedaan. Bovendien lieten wij zien dat ATP8B1 en CDC50A een functionele endogene heterodimeer vormen. Net als in cellen waarin ATP8B1 expressie was verlaagd, resulteerde depletie van CDC50A tot een verminderde activiteit en oppervlakte expressie van ASBT in Caco-2 cellen. De effecten van depletie van ATP8B1 en CDC50A op ASBT waren geen algemeen defect in apicale membraanorganisatie, omdat de expressie en activiteit van andere membraaneiwitten niet was aangedaan. Analyses van ontlastingsmonsters van post-levertransplantatie PFIC1 patiënten (met herstelde galstuwing) lieten zien dat fecale galzoutniveaus onveranderd waren, terwijl

165

Appendices natrium- en chlorideconcentraties verhoogd waren en kaliumniveaus verminderd. Deze resultaten tonen aan dat een functionele ATP8B1-CDC50A heterodimeer noodzakelijk is voor de apicale membraanlokalisatie van ASBT in Caco-2 cellen. De diarree in PFIC1/BRIC1 patiënten heeft een secretoire oorzaak waar galzoutmalabsorptie een bijdrage aan zou kunnen leveren.

In hoofdstuk 5 hebben wij de consequentie van ATP8B1 deficiëntie op de lokalisatie en activiteit van CFTR bestudeerd. Defecten in CFTR activiteit leiden tot cystische fibrose, een ernstige longziekte die ook wel taaislijmziekte wordt genoemd. CFTR is een chloridekanaal dat essentieel is voor epitheliaal vloeistoftransport en klaring van mucus. Wij lieten zien dat in ATP8B1 gedepleteerde humane intestinale Caco- 2 en T84 cellen en humane epitheliale long Calu-3 cellen CFTR mRNA en totaal eiwit significant waren verminderd. Verder toonden wij aan dat apicale membraaninsertie van CFTR is aangetast in ATP8B1 gedepleteerde T84 cellen na overexpressie van CFTR. De gereduceerde eiwitniveaus van CFTR in ATP8B1 gedepleteerde T84 cellen leidden tot verminderde CFTR activiteit. Dit was het geval na metingen van transepitheliale stromen of na ratiometrische fluorescente chloridesensor experimenten. Daarentegen vonden wij geen verminderde expressie en activiteit van CFTR in geïsoleerde darmsegmenten of galblazen van de mutante Atp8b1G308V/G308V knock-in muis. De gedeeltelijke inactivatie van CFTR in ATP8B1 gedepleteerde cellen kan een verklaring zijn voor de extrahepatische fenotypen (longinfecties, verhoogde zweetelectrolieten) in PFIC1/BRIC1 patiënten.

In hoofdstuk 6 hebben wij de rol van CDC50A en ATP8B1 in het aangeboren immuunsysteem bestudeerd in macrofagen waarin de eiwitexpressie van deze eiwitten was verminderd. Depletie van CDC50A in gekweekte en primaire humane macrofagen resulteerde in lipopolysaccharide (LPS)-geïnduceerde hypersecretie van de pro-inflammatoire cytokines TNFα, IL-1β en IL-6 en verminderde productie van mRNA van de anti-inflammatoire genen IFNβ, RANTES en SOCS1. Het plasmamembraaneiwit Toll-like receptor 4 (TLR4) leidt het signaal van de herkenning van LPS verder de cel in. Wij vonden dat de hyperactiviteit in CDC50A gedepleteerde macrofagen mogelijk veroorzaakt wordt door een aangedane internalisatie van TLR4. Dit zorgt voor een verhoogde pro-inflammatoire signalering

166

Appendices

van TLR4 aan het plasmamembraan en verminderde anti-inflammatoire signalering van geïnternaliseerd TLR4. Knockdown van ATP8B1 en ATP11A in THP-1 macrofagen leidde tot een vergelijkbaar hyperactief pro-inflammatoir fenotype en verminderde LPS geïnduceerde internalisatie van TLR4, maar in een mindere mate dan CDC50A depletie. Daarentegen konden wij het humane fenotype niet terugvinden in long of beenmergmacrofagen van de mutante Atp8b1G308V/G308V muis. Deze resultaten laten zien dat humane P4-ATPases, inclusief ATP8B1 en ATP11A, betrokken zijn bij de regulering van internalisatie en het intracellulaire verkeer van TLR4 en geven zodoende een verklaring voor hun mogelijke rol in chronische (pulmonaire) ontsteking.

Onze resultaten laten het fysiologische belang zien van ATP8B1 in humane cellen. In dit proefschrift blijkt dat ATP8B1 een cruciaal onderdeel is van het transport van celmembraanblaasjes naar en/of van het plasmamembraan. Naast de eiwitten die beschreven zijn in dit proefschrift heeft ATP8B1 ook een effect op de integriteit van canaliculaire leveren cochleaire haarcelmembranen en op glucose gestimuleerde insuline secretie (hoofdstuk 1).

Concluderend, ATP8B1 is een essentiële indirecte bemiddelaar van cellulair eiwittransport wat niet alleen een directe invloed heeft op de fysiologie van de cellen waarin ze zich bevinden, maar ook op het gehele organisme zelf.

167

Appendices

Dankwoord

Allereerst hartelijk dank aan mijn copromotor. Coen, dank je wel dat ik onder jouw begeleiding de raadselen van ATP8B1 verder mocht ontrafelen. Je hebt mij niet alleen op wetenschappelijk gebied geweldig geholpen, maar ook als persoon had ik echt geen betere begeleider kunnen wensen. De FASEB meetings waren onvergetelijk. Je praktische en onvermoeibaar optimistische instelling is een geweldig voorbeeld voor mij geweest. Ronald, dank je wel dat ik bij jou mocht promoveren in het Tytgat Instituut. Ik ben blij dat ik in een lab terecht kwam met zoveel gedreven en slimme wetenschappers. Je scherpe en kritische blik op de details en het grote plaatje zijn bewonderenswaardig; iets wat ik hoop in de rest van mijn carrière te kunnen imiteren.

Daarnaast wil ik ook de leden van mijn promotiecommissie bedanken; prof. dr. Van den Brink, prof. dr. Geijtenbeek, prof. dr. Houwen, prof. dr. Verhoeven, dr. Van de Graaf en dr. De Jonge, bedanken voor het vrijmaken van hun tijd om plaats te nemen in de commissie en voor het beoordelen van mijn proefschrift.

Serge, mijn verhuizing naar S2 was misschien wel mijn beste beslissing tijdens mijn promotie. Ontzettend bedankt voor alle steun en advies (inclusief trainingsschema’s) en je bereidheid om mijn paranimf te zijn. Luca, thank you so much for all the good times inside and outside of the lab, and your willingness to come back from Italy to support me during my defense.

Jing, thanks for being such a great roommate. I really appreciated your window into China. Paula, despite your battles with Piter, our good conversations and time spent together made my time at the Tytgat much more enjoyable, thank you. Dineke, dank je voor je voorbereidende stappen, alhoewel die soms wel erg luid waren. Jung-Chin, your enthusiasm in the lab is truly an inspiration. I am very grateful for all your help and the wide variety of topics we discussed.

Johan, ontzettend veel dank en waardering voor je (spirituele) wijsheid, inzichten en natuurlijk de oneindige hoeveelheid ELISAs. Cindy, je was een grote hulp binnen en buiten het werk. Dank voor je advies en gezelschap. Rudi, onmisbaar aanspreekpunt in het lab, je uitgebreide inzichten waren altijd zeer nuttig. Suzanne,

168

Appendices

dank je wel voor het organiseren van alle dierexperimenten en de gesprekken. Kam, wat moet Coen ooit zonder jou doen? Dank je voor je onmisbare hulp in het lab.

Mona, wat moeten alle Tytgatters zonder jou doen? Ondanks alle drukte, dank je wel voor je open deur en natuurlijk voor het ondergaan van alle frustraties voor het regelen van mijn promotie. Orlando, je opgewekte humeur elke ochtend was een prachtig begin van de dag. Marianne en Wil, dank voor het logistiek mogelijk maken van alle experimenten. Derek, partner in crime, wie had ooit gedacht dat wij zoveel interesses gemeen hadden. Dank voor alle tips.

Mohammed, thank you a lot for your enthusiasm in helping finishing my last chapter. Hugo, dank je wel voor je praktische en theoretische hulp en het openen van de CFTR wereld voor mij. Ik heb zeer genoten van al onze gesprekken. Verder heb ik het geluk gehad dat een aantal getalenteerde studenten heeft meegeholpen aan mijn project; Aïcha, Britt, Xiaokang, Casper, Zoë. Dank voor jullie harde werk en inzet. Valentina, dank je voor het voortzetten van het ATP8B1 onderzoek, veel succes!

Stan, van je eerste vragen over ATP8B1 tot je komst in Amsterdam. Ik heb altijd genoten van onze interacties, bedankt voor je hulp bij mijn PhD project en mijn postdoc project. Ingrid en Theo, dank voor jullie voortdurende interesse en waardevolle adviezen op alle verschillende vlakken gerelateerd aan het werken in de wetenschap. Jurgen, zonder jou was ik niet op deze plek beland. Dit zal ik nooit vergeten. Bedankt voor je humor en de introductie tot het wetenschappelijk onderzoek. Ulrich, dank voor je persoonlijke interesse en altijd hulpvolle advies. Wout, een schat aan kennis. Ik prijs me gelukkig dat ik naast je zat. Piter, bedankt voor je hulp tijdens mijn onderzoek, en natuurlijk voor het in toom houden van Coen.

Verder veel dank aan de onderzoekers van de oudere en nieuwere generatie van de levergroep; Job, Ebtisam, Lucas, Vassilis, Sasha, Ruth, Jacqueline, Remco, Sem, Lowiek, Sateesh, Jyoti. Jullie maakten de promotie (en postdoc) zoveel aangenamer, zowel binnen als buiten het AMC.

Esther, Caroline, Jacqueline, Olaf, Sander, Francisca, Dagmar, Liesbeth, Connie, veel dank voor jullie technische expertise en voor het leven in het lab wat prettiger te maken. Alexandra, Karina, Monique, Fernando, Eduardo, thanks for bringing some of that Brazilian warmth to our cold country. Sara, het was geweldig om je

169

Appendices beter te leren kennen. Jochem, misschien een keer 18 holes doen als we met pensioen zijn? Ronald, José, Shobhit, Oana, Sophie, Charlotte, Lea, Laurens, dank voor jullie hulp op immunologisch gebied, de lunches en de bezigheden buiten het lab. Wouter, Manon, dank voor jullie expertise die het laatste hoofdstuk heeft mogelijk gemaakt.

En natuurlijk mijn dank aan alle andere Tytgatters die mijn promotie hebben geholpen, op praktisch of geestelijk gebied!

Tot slot wil ik ook nog alle medewerkers van de afdeling Experimentele Chirurgie bedanken, die mij voornamelijk als postdoc hebben meegemaakt. Erik, Koen, Chung, Ruud, Mans, Julia, Lionel, Kasia, Pim, Jimme, Megan, Lindy, Daniël, Goos, Adri, Esther, Joost, Zuhre en Thomas, dank voor de gezelligheid en de inzicht in de chirurgische zijde van het ziekenhuis en de wetenschap. Ruurdtje en Rob, het was geweldig om deel uit te kunnen maken van de BAL groep. Dank voor jullie flexibiliteit om mij af en toe nog aan mijn promotie te laten werken. Ik heb tijdens deze periode enorm genoten en veel geleerd: hoe het is om als een onafhankelijk wetenschapper te werken en hoeveel moeite het kost om van een academisch project een product te maken. Aziza en Martien, het was een waar plezier om met jullie samen te mogen werken in de BAL groep, veel succes met het afmaken en schrijven van jullie laatste artikelen. Albert, onmisbaar op de afdeling. Dank je voor het ophouden van het moraal, je hulp en adviezen. Michal, respect. Het was een eer om met je te werken. Maurice, dank je voor je hulp en gedrevenheid, blijf doen wat je doet, veel succes in de toekomst. Rowan, Robert-Jan, Sebastiaan, Eva, Floor, dank voor de interessante gesprekken en de altijd goede sfeer in onze kamer!

170

Appendices

Publications

In this thesis van der Mark VA, Ghiboub M, Marsman C, Zhao J, van Dijk R, Hiralall JK, Ho- Mok KS, Castricum Z, de Jonge WJ, Oude Elferink RP, Paulusma CC. Phospholipid flippases attenuate LPS-induced TLR4 signaling by mediating endocytic retrieval of Toll-like receptor 4. Cell Mol Life Sci. 2017 Feb; 74(4):715-730. doi: 10.1007/s00018-016-2360-5 van der Mark VA, de Jonge HR, Chang JC, Ho-Mok KS, Duijst S, Vidović D, Carlon MS, Oude Elferink RP, Paulusma CC. The phospholipid flippase ATP8B1 mediates apical localization of the cystic fibrosis transmembrane regulator. Biochim Biophys Acta. 2016 Sep;1863(9):2280-8. doi: 10.1016/j.bbamcr.2016.06.005. van der Mark VA, de Waart DR, Ho-Mok KS, Tabbers MM, Voogt HW, Oude Elferink RP, Knisely AS, Paulusma CC. The lipid flippase heterodimer ATP8B1- CDC50A is essential for surface expression of the apical sodium-dependent bile acid transporter (SLC10A2/ASBT) in intestinal Caco-2 cells. Biochim Biophys Acta. 2014 Dec;1842(12 Pt A):2378-86. doi: 10.1016/j.bbadis.2014.09.003. van der Mark VA, Elferink RP, Paulusma CC. P4 ATPases: flippases in health and disease. Int J Mol Sci. 2013 Apr 11;14(4):7897-922. doi: 10.3390/ijms14047897.

Folmer DE, van der Mark VA, Ho-Mok KS, Oude Elferink RP, Paulusma CC. Differential effects of progressive familial intrahepatic cholestasis type 1 and benign recurrent intrahepatic cholestasis type 1 mutations on canalicular localization of ATP8B1. Hepatology. 2009 Nov;50(5):1597-605. doi: 10.1002/hep.23158.

Other van der Mark VA, de Waart DR, Shevchenko V, Oude Elferink RP, Chamuleau RA, Hoekstra R. Stable overexpression of the constitutive androstane receptor reduces the requirement for culture with dimethyl sulfoxide for high drug metabolism in HepaRG cells. Drug Metab Dispos. 2017 Jan; 45(1):56-67. doi: 10.1124/dmd.116.072603

171

Appendices

Lionarons DA, Heger M, van Golen RF, Alles LK, van der Mark VA, Kloek JJ, de Waart DR, Marsman HA, Rusch H, Verheij J, Beuers U, Paulusma CC, van Gulik TM. Simple steatosis sensitizes cholestatic rats to liver injury and dysregulates bile salt synthesis and transport. Sci Rep. 2016 Aug 18;6:31829. doi: 10.1038/srep31829. van Dijk R, Mayayo-Peralta I, Aronson SJ, Kattentidt-Mouravieva AA, van der Mark VA, de Knegt R, Oruc N, Beuers U, Bosma PJ. Disruption of HNF1α binding site causes inherited severe unconjugated hyperbilirubinemia. J Hepatol. 2015 Dec;63(6):1525-9. doi: 10.1016/j.jhep.2015.07.027.

172

Appendices

Authors and affiliations

Zoë Castricum, Jung-Chin Chang, Remco van Dijk, Suzanne Duijst, Dineke E. Folmer, Mohammed Ghiboub, Casper Marsman, Kam S. Ho-Mok, Johan K. Hiralall, Wouter J. de Jonge, Ronald P.J. Oude Elferink, Coen C. Paulusma, D. Rudi de Waart, Jing Zhao: Tytgat Institute for Liver and Intestinal Research, Academic Medical Center, Amsterdam, The Netherlands

Merit M. Tabbers, Heleen W. Voogt: Department of Paediatric Gastroenterology and Nutrition, Emma Children's Hospital, Academic Medical Center, Amsterdam, The Netherlands

Alex S. Knisely: Institute of Liver Studies, King's College Hospital, London, UK

Hugo R. de Jonge: Department of Gastroenterology & Hepatology, Erasmus University Medical Centre, Rotterdam, The Netherlands

Marianne S. Carlon, Dragana Vidović: Laboratory for Molecular Virology and Gene Therapy, Department of Pharmaceutical and Pharmacological Sciences, KU Leuven, Belgium

173

Appendices

Portfolio

PhD period: August 2008 – March 2013

Supervisors: Dr C.C. Paulusma and Prof Dr R.P.J. Oude Elferink

Year Workload (ECTS)

Courses

AMC World of Science 2009 0.7

Practical Biostatistics 2009 1.1

Bioinformatics 2009 1.1

Basic Microscopy 2009 1.6

Laboratory Animals: licensed Article 9 2009 4.3

Radiation Protection: licensed level 5B 2010 1.7

The Macroscopic, Microscopic and Pathological 2010 1.1 Anatomy of the House Mouse Seminars Weekly department seminars 2008-2013 4

Oral presentations

Annual Spring Conference, Dutch Experimental 2011 0.5 Gastroenterology and Hepatology Meeting, Veldhoven Dutch Liver Meeting, Spier 2012 0.5

Annual Spring Conference, Dutch Experimental 2012 0.5 Gastroenterology and Hepatology Meeting, Veldhoven Federation of American Societies for 2012 0.5 Experimental Biology, Summer Research Conference, Snowmass Village, USA Poster presentations

Federation of American Societies for 2010 0.5 Experimental Biology, Summer Research Conference, Snowmass Village, USA

174

Appendices

American Association for the Study of Liver 2011 0.5 Diseases, The Liver Meeting, San Francisco, USA Federation of American Societies for 2012 0.5 Experimental Biology, Summer Research Conference, Snowmass Village, USA

Conferences

Falk Symposium 165, Bile Acid Biology and 2008 0.5 Therapeutic Actions, Amsterdam Annual Spring Conference, Dutch Experimental 2010-2012 1.5 Gastroenterology and Hepatology Meeting, Veldhoven Dutch Liver Meeting, Spier 2010-2011 1

American Association for the Study of Liver 2011 1 Diseases, The Liver Meeting, San Francisco, USA Federation of American Societies for 2010 and 2012 2 Experimental Biology, Summer Research Conference, Snowmass Village, USA

Other

Journal club 2009-2012 1

Annual PhD student retreat, Lunteren 2009-2012 3

Student supervision

B.Sc. internship Aïcha Lachmouchi on “CDC50 2010 1 proteins do not play a role in the trafficking of ATP8B1 in Caco-2 cells”

M.Sc. internship Britt van der Ploeg on 2011 2 “Analyses of P4 ATPase knockdown on ASBT- mediated bile salt uptake in Caco-2 cells”

M.Sc. internship Xiaokang Lum on 2012 2 “Intracellular trafficking of P4-ATPases”

175