Role of Modifier Gene SLC6A14 in Cystic Fibrosis and the Path to Personalized Medicine

Role of modifier gene SLC6A14 in Cystic Fibrosis and the path to personalized medicine

Saumel Bashir Ahmadi

A thesis submitted in conformity with the requirements for the degree of Doctor of Philosophy Department of Physiology University of Toronto

Role of modifier gene SLC6A14 in Cystic Fibrosis and the path to personalized medicine

Saumel Bashir Ahmadi

Doctor of Philosophy

Department of Physiology University of Toronto

2018 Abstract

Cystic Fibrosis (CF) is the most common fatal genetic disorder in Canada. It is a multi-system disorder caused by mutations in the CFTR gene, expressed in epithelial tissues. Decrease in lung function over time is the most common cause of morbidity and mortality in CF. However, affliction to other organs systems like the gastro-intestinal system, also contributes to significant disease burden. Variation in disease severity among CF patients is well established, attributable to CFTR gene and modifier gene mutations. There are over 2,000 disease causing mutations in the CFTR gene. F508del is the most common CF causing mutation present on at least one allele in 90% of the CF population. However, patients bearing the same F508del mutation on both alleles also exhibit a tremendous variation in disease severity, which has been attributed to modifier gene mutations. Two FDA approved drugs that work directly on CFTR protein –

Ivacaftor and Lumacaftor, are shown to have a heterogenous response in CF patients. The heterogeneity in response has also been attributed to modifier genes. With a greater understanding of modifier genes and CFTR genetics, the variation in patient responses to currently available CF therapies could be explained. This has led to efforts for in vitro phenotypic profiling of individual patient derived tissues, in the context of CF. Towards this we developed the apical CFTR conductance assay, to measure CFTR function in vitro using cultured

ii airway epithelia from individual patients. Later we applied this technology to murine intestinal tissue to understand the mechanism of a genetic modifier of Cystic Fibrosis – SLC6A14, which was a top hit in a recent genome wide association study. Using a murine CF model and a

SLC6A14 knockout mouse, we discovered that SLC6A14 modifies the intestinal phenotype of CF by regulating the fluid secretory capacity of the CF affected epithelium, via the nitric-oxide pathway. Thus, we explored the biologic basis of SLC6A14 as a modifier of CF. Taken together, the studies described in this thesis will facilitate the path towards personalized medicine in CF.

iii

Acknowledgments

Foremost, I would like to thank my supervisors Dr. Christine Bear and Dr. Johanna Rommens for having me in their laboratories and for being excellent mentors and caring individuals. Their hard work and conduciveness to new ideas have always been very inspiring. They have helped me grow as a scientist to tackle the next generation of challenges in Biomedical research. Dr. Bear’s question of “what would be the best experiment?” helped me push the limit of innovation. The collegial and collaborative environment at The Hospital for Sick Children was a cornerstone to the success of everyone involved in the projects of this thesis.

I am thankful to my committee members Dr. Scott Heximer and Dr. Reinhart Reithmeier for their helpful discussions, insights and keeping my progress on track.

I am very thankful to my dear friend and colleague Dr. Stan Pasyk, who helped me grow professionally as well as socially. I am grateful for his insight, critical arguments and for inspiring me by sharing his passion for science.

Over the years I have been fortunate to have many students and collaborators, without whose help this thesis would not be possible. I am thankful to Timothy Chung for discussions in physical chemistry and for maintaining a mutual inspiration for long experiments, and to Rupinder Mangat for her help. I am also thankful to Wilson Wu and Randolph Kissoon for the fun times in the animal facility, productive discussions and for their hard work with the mice. All of you have bright futures ahead of you.

The people with whom I worked the most in the lab – my dear friends Sunny Xia and Michelle Di Paola, are really special to me. Their hard work, passion for science and their support was indispensable to this thesis. They are success bound with bright futures.

I am grateful to Dr. Mohabir Ramjeesingh for supporting me, believing me and nurturing me as a scientist. I am thankful to my friend and colleague Dr. Steven Molinski for helpful discussions, help with molecular biology, and for a mutually productive life-long collaboration. I am also very thankful to Dr. Canhui (Danny) Li, who normalized my crazy ideas and helped them become a reality. His vision and passion for science has always been motivational. I have had helpful discussions with my colleagues and friends Stephanie Chin and Maurita Hung and I see bright futures for them. I am thankful to Ling Jun Huan, for her help in the lab all along the way, and maintaining a collegial environment. I am grateful to all past and present members of the Bear Lab including Elyse Watkins, Ida Szarics, Dr. Paul Eckford, Dr. Leigh Wellhauser, Dr. Kai Du, Wilson Yu, Janet Jiang, Onofrio Laselva, Donghe Yang, Angela Skoutakis, Anick Auger and Roman Pekhletski for their help at various stages of my project, and for sharing interesting conversations.

I would like to give my sincere gratitude to Catherine Luk, whose help was critical in maintaining the mouse facility for this project. Her care, devotion and tireless efforts kept the project smooth and fun for everyone.

I am thankful to all members of Dr. Johanna Rommens’ laboratory including Dr. Stéphane Gagnon, Rikesh Gandhi, Dr. Marina Tourlakis, Dr. Holly Liu, Rashmi Parekh and Fan Lin, who helped me all along especially in the area of molecular genetics.

I am grateful for collaborating with Dr. Julie Forman-Kay’s laboratory. I made a life-long friend Dr. Zoltan Bozoky, whose insight into computational biology and big data handling was unmatched. I had a very productive professional and social relationship with him.

I am also very grateful to collaborate with Dr. Theo Moraes’ and Dr. Tanja Gonska’s laboratories, and all the present and past members of their labs including Andrew Lloyd-Kuzik, Wan Ip, Claire Bartlett, Hong Ouyang and Wenming Duan. I was also fortunate to have a productive collaboration with Dr. Amy Wong in Dr. Janet Rossant’s lab, and with Dr. Felix Ratjen, Julie Avolio and Dr. Jeremy Hirota.

I am also thankful to all my medical school mentors, and also my medical school friends Dr. Parth Vaishnav, Dr. Vidhi Patel and Dr. Mihir Chauhan for their support in my endeavors.

Finally, I would also like to thank my parents Dr. Bashir Ahmadi and Nafisa Ahmadi for their constant support and trust. I am also very grateful to my grandparents Dr. Abdulmuttalib Ahmadi and Zubeda Ahmadi for constantly inspiring me to work hard, and to my sister Swela Bashir for her support.

This study would not have been possible without the help and support of the Cystic Fibrosis patients and their families, and the inspiration they provided to find a cure. I hope this thesis will be a small step towards reaching a cure for everyone.

Table of Contents Contents

Acknowledgments ...... iv

Table of Contents ...... vi

List of Tables ...... xi

Abbreviations ...... xii

List of Figures ...... xvi

List of Appendices ...... xix

Chapter 1 Introduction ...... 1

1.1 Cystic Fibrosis (CF) ...... 1

1.2 Cystic Fibrosis Transmembrane conductance Regulator (CFTR) ...... 2

1.2.1 Introduction ...... 2

1.2.2 CFTR Genetics and regulation of gene expression ...... 4

1.2.3 CFTR protein and regulation ...... 6

1.2.4 CFTR anion conductance and epithelial pH regulation ...... 11

1.2.5 Cystic Fibrosis causing mutation F508del CFTR ...... 11

1.3 Mechanisms underlying CF pathogenesis ...... 11

1.4 CF animal models ...... 14

1.4.1 Murine CF intestinal phenotype ...... 14

1.4.2 Growth of CF mice ...... 15

1.4.3 Murine pulmonary CF phenotype ...... 15

1.4.4 Other CF phenotypes in mice ...... 15

1.4.5 Other CF animal models ...... 16

1.5 Modifier genes in CF ...... 16

1.6 Cystic Fibrosis Diagnostic tests ...... 24

1.7 CFTR directed therapeutics ...... 25

1.8 Thesis Rationale ...... 26

1.9 Hypothesis ...... 27

Chapter 2 Patient specific in vitro assessment of CFTR modulators: Path towards personalized medicine in CF ...... 28

2.1 Summary ...... 29

2.2 Introduction ...... 29

2.3 Materials and Methods ...... 31

2.3.1 iPSC derived airway epithelial cell generation ...... 31

2.3.2 Nasal cell culture method ...... 31

2.3.3 Western blotting ...... 31

2.3.4 Immunofluorescence ...... 32

2.3.5 qRT-PCR 32

2.3.6 Ussing chamber ...... 33

2.3.7 Membrane potential assay ...... 33

2.3.8 Chemical reagents ...... 34

2.3.9 Statistical Analysis ...... 34

2.4 Results ...... 35

2.4.1 Fluorescence-based assay of CFTR channel activity in patient derived respiratory epithelial cultures ...... 35

2.4.2 ACC assay of patient-specific responses in nasal epithelial cultures is accurate and reproducible ...... 39

2.4.3 ACC assay correlates with Ussing chamber measurements of patient-specific responses to interventions...... 41

2.4.4 ACC assay enables profiling of existing and novel interventions in patient- specific primary nasal epithelial cultures ...... 42

2.4.5 FLIPR® based ACC assay quantifies CFTR channel activity and pharmacological rescue of F508del in iPS cell-derived lung ...... 46

2.5 Discussion ...... 48 vii

Chapter 3 SLC6A14 modulates epithelial fluid secretion ...... 51

3.1 Summary ...... 52

3.2 Introduction ...... 52

3.2.1 Intestinal amino-acid transporters ...... 52

3.2.2 SLC6A14 (ATB0,+) ...... 54

3.2.2.1 Introduction ...... 54

3.2.2.2 SLC6A14 and tissue growth ...... 56

3.2.2.3 Disease associations with SLC6A14 ...... 58

3.2.3 Murine CF phenotype ...... 58

3.3 Materials and Methods ...... 59

3.3.1 Mice ...... 59

3.3.2 Slc6a14 mouse genotyping ...... 60

3.3.3 Quantitative real-time PCR ...... 60

3.3.4 Mouse colonic organoid culture ...... 61

3.3.5 Transfection of organoids ...... 61

3.3.6 Studies of CFTR and F508del-CFTR mediated swelling of colonic organoids ...... 62

3.3.7 Mass spectrometry ...... 62

3.3.8 IGF-1 level measurements ...... 63

3.3.9 Histology analysis ...... 63

3.3.10 Morphometric Measurements ...... 64

3.3.11 In vivo fluid secretory assay ...... 64

3.3.12 Ex vivo closed loop amino-acid uptake assay ...... 65

3.3.13 Statistical Analysis ...... 66

3.4 Results ...... 66

3.4.1 Slc6a14 is a major apical amino-acid transporter in the colon ...... 66

viii

3.4.2 Disruption of Slc6a14 influence growth of FVB mice modeling the major CF mutation: F508del ...... 69

3.4.3 Loss of Slc6a14 worsens defective F508del-Cftr mediated secretion in murine colonic epithelium ...... 72

3.5 Discussion ...... 77

Chapter 4 Mechanistic insight into SLC6A14 mediated modulation of CFTR anion conductance ...... 79

4.1 Summary ...... 80

4.2 Introduction ...... 80

4.2.1 Arginine nitric-oxide pathway ...... 81

4.2.2 Rationale and Hypothesis ...... 82

4.3 Materials and Methods ...... 83

4.3.1 Mice ...... 83

4.3.2 Generation of split open organoids ...... 83

4.3.3 Apical Chloride Conductance (ACC) Assay for CFTR in “split-open” organoids . 84

4.3.4 Nitric oxide measurements in “split-open” organoids ...... 85

4.3.5 Statistical analysis ...... 86

4.4 Results ...... 87

4.4.1 Direct measurement of CFTR channel function in split-open organoids ...... 87

4.4.2 Slc6a14 is important for epithelial NO production ...... 88

4.4.3 Slc6a14 enhances F508del-CFTR mediated chloride conductance ...... 90

4.4.4 Loss of Slc6a14 and arginine-mediated NO generation contributes to worsening of defective epithelial fluid secretion ...... 91

4.5 Discussion ...... 95

Chapter 5 Discussion and Future Directions ...... 98

5.1 Path towards personalized medicine in CF ...... 98

5.1.1 Summary ...... 98

5.1.2 Improvisation of technology ...... 98 ix

5.1.3 Improving clinical trial design for personalized medicine ...... 98

5.1.4 Patient tissue on chip technologies for CF ...... 99

5.1.5 Personalized medicine in CF ...... 99

5.2 Role of genetic modifier SLC6A14 in Cystic Fibrosis ...... 100

5.2.1 Summary 100

5.2.2 Effect of SLC6A14 on CFTR channel gating ...... 100

5.2.3 Role of SLC6A14, arginine and nitric-oxide in CF ...... 101

5.2.4 Validation and translation to human intestinal tissue ...... 101

5.2.5 Role of SLC6A14 in other CF affected tissues ...... 102

5.2.6 Translation to the clinic ...... 102

5.2.7 Genetics of CF disease and application to other genetic disorders ...... 103

5.2.8 Significance ...... 104

Appendix for Tables ...... 105

Supplementary Figures ...... 108

References ...... 114

List of Tables

Table 1.5.1: Modifier genes discovered by gene specific studies in patient populations.

Table 2.3.5: Primers employed for qRT-PCR.

Table 3.3.1: Genotyping primers.

Table 3.3.2: Primers used for qRT-PCR.

Table 3.4.2.1: Slc6a14 deletion in CF mice maintains Mendelian inheritance.

Abbreviations

AA Amino-acids ACC Apical Chloride/CFTR conductance ADHD Attention Deficit Hyperactivity Disorder AMPK Adenosine Monophosphate stimulated Kinase ANOVA Analysis of Variance ARRDC3 Arrestin Domain-Containing 3 ASL Airway Surface Liquid ASO Antisense Oligonucleotide ATP Adenosine Triophosphate ATP12A H+/K+ Adenosine Triphosphatase BHK Baby Hamster Kidney BMI Body Mass Index Br- Bromide ion CaCC Calcium activated Chloride Channel CAL CFTR associated Ligand cAMP cyclic Adenosine Monophosphate Cas9 CRISPR associated protein 9 CBAVD Congenital absence of Vas Deferens cGMP cyclic Guanosine Monophosphate CF Cystic Fibrosis CFBE41o- Cystic Fibrosis Bronchial Epithelial cell line clone 41 CFFT Cystic Fibrosis Foundation Therapeutics CFTR Cystic Fibrosis Transmembrane conductance Regulator CFTR-AS1 Cystic Fibrosis Transmembrane conductance Regulator-Antisense 1 CFTRinh-172 CFTR inhibitor 172 Cl- Chloride ion CNX Calnexin COP Coat Protomer CPM scintillation counts per minute CREB cAMP Responsive Element Binding Protein CRISPR Clustered Regularly Interspaced Short Palindromic Repeats Csp Cysteine String Protein DAF-FM 4-Amino-5-Methylamino-2',7'-Difluorofluorescein DAPI 4',6-Diamidino-2-Phenylindole xii

DAT Dopamine Transporter DIOS Distal Intestinal Obstruction Syndrome DMSO Dimethyl sulfoxide EAA Essential Amino-acid EBP50 ezrin-radixin-moesin-binding phosphoprotein 50 EDTA Ethylenediaminetetraacetic acid ELISA Enzyme Linked Immunosorbent Assay ENaC Epithelial Sodium (Na+) Channel ER Endoplasmic Reticulum F508del deletion of phenylalanine at position 508 (of CFTR) FDA Food and Drug Administration FeNO Forced expiratory concentration of Nitric Oxide FEV1 Forced expiratory volume in the first second FGF2 Fibroblast Growth Factor 2 FIS Forskolin Induced Swelling FLIPR Fluorimetric Imaging Plate Reader FOXJ1 Forkhead Box J1 FSK Forskolin GAPDH Glyceraldehyde 3-phosphate dehydrogenase GFP Green Fluorescent Protein GSNO S-nitrosoglutathione GUSB Glucuronidase Beta GWAS Genome Wide Association Study HBE Human Bronchial Epithelial cell line HBSS Hanks Buffered Salt Solution (With Calcium, Magnesium, HEPES; without phenol red) HEPES 4-(2-Hydroxyethyl)piperazine-1-ethanesulfonic acid Hop Hsp70/Hsp90 organizing protein HSP Heat Shock Protein HTS High-throughput Screening HNF1α Hepatocyte Nuclear Factor 1 alpha I- Iodide ion IBMX 3-isobutyl-1-methylxanthine ICL Intracellular Loop

Ieq Equivalent Current IGF-1 Insulin like Growth Factor one

xiii

IL Interleukin iPOP integrative Personal Omics Profile iPSC induced Pluripotent Stem Cells IRT Immunoreactive Trypsinogen KO Knock-out lncRNA Long non-coding RNA MI meconium ileus miRNA micro-RNA MRP4 Multidrug Resistance Protein 4 MSD Membrane Spanning Domain mTOR mammalian Target of Rapamycin NaCl Sodium Chloride NaOH Sodium Hydroxide NBD Nucleotide Binding Domain NFκB Nuclear Factor kappa B NHERF Na+/H+ exchanger regulatory factor NKCC Na+ K+ Cl- co-transporter NMDG N-methyl-D-glucamine NO Nitric Oxide - NO3 Nitrate ion NOS Nitric Oxide Synthase NSAID Non-Steroidal Anti-inflammatory Drugs PanCK Pancytokeratin PBS Phosphate Buffered Solution PCR Polymerase Chain Reaction PKA Protein Kinase A PKC Protein Kinase C PKG Protein Kinase G PYK2 Proline-rich tyrosine kinase 2 “R” Domain Regulatory Domain RFU Relative Fluorescence Units SCN- Thiocyanate ion SD Standard Deviation SDS PAGE sodium dodecyl sulfate polyacrylamide gel electrophoresis SEM Standard Error of Mean

xiv

SERT Serotonin Transporter SHH Sonic HedgeHog protein SLC Solute Carrier SNP Single Nucleotide Polymorphisms SP3 Specificity Protein 3 Src Rous Sarcoma, Proto-Oncogene Tyrosine-Protein Kinase SSRI Selective Serotonin Reuptake Inhibitor TBP TATA Binding Protein TF Transcription Factor Th T-helper cell TGFβ Transforming Growth Factor beta TIFF Tag Image File Format TMD Transmembrane Domain UNC University of North Carolina ZO-1 Zona Occludens one

List of Figures

Figure 1.1: Cystic Fibrosis phenotype.

Figure 1.2.1.1.: Fluid secretion across the apical surface.

Figure 1.2.1.2: Channels and transporters in a CF relevant epithelium.

Figure 1.2.2.1: Regulation of CFTR functional expression by miRNA and lncRNA.

Figure 1.2.2.2: CFTR mutation classification.

Figure 1.2.3.1: CFTR protein on gel electrophoresis.

Figure 1.2.3.2: CFTR protein structure.

Figure 1.2.3.3: Regulation of CFTR protein phosphorylation by kinases.

Figure 1.3.1: Key pathophysiologic processes in the CF epithelium.

Figure 1.3.2: Inflammation and infection cycle in CF.

Figure 1.5.1: Modifier gene discovery approaches and translation into clinic.

Figure 2.4.1.1: Functional measurement of CFTR chloride conductance in CFBE41o- F508del CFTR using the apical chloride conductance assay (ACC).

Figure 2.4.1.2: Application of Apical Chloride Conductance (ACC) Assay to measure F508del CFTR function in primary bronchial tissue.

Figure 2.4.2.1: ACC Assay is accurate measure of mutant CFTR function and responses to interventions in patient-specific primary nasal epithelial cultures.

Figure 2.4.3.1: ACC Assay correlates with Ussing chamber measurements of patient-specific responses to interventions in primary nasal epithelial cultures.

Figure 2.4.4.1: ACC Assay enables profiling of existing and novel interventions in patient- specific primary nasal epithelial cultures.

xvi

Figure 2.4.5.1: ACC Assay reports primary defect and pharmacological rescue of major CF mutant in lung cultures differentiated from iPSCs.

Figure 3.2.1: Amino-acid transporters in the ileum and colon.

Figure 3.2.2.1: SLC6A14 is expressed in CF affected tissues like colon and lungs.

Figure 3.2.2.2: Role of amino-acid (AA) transporters in mTOR activation.

Figure 3.3.7: Standard curve for ELISA based detection of IGF-1 levels.

Figure 3.3.8: Representative intestinal section.

Figure 3.4.1.1: Expression of SLC6A14 in various CF affected tissues.

Figure 3.4.1.2: SLC6A14 is a major arginine transporter on the apical colonic epithelium.

Figure 3.4.2.1: Disruption of Slc6a14 in F508del CF mice leads to decrease in weight gain post weaning.

Figure 3.4.2.2: Deletion of Slc6a14 in CF mice does not cause a change in serum amino-acid levels.

Figure 3.4.2.3: Deletion of Slc6a14 in CF mice does not cause a change in serum IGF-1 levels.

Figure 3.4.3.1: Loss of Slc6a14 worsens defective F508del Cftr mediated secretion in murine colonic epithelium.

Figure 3.4.3.2: Loss of Slc6a14 worsens defective F508del Cftr mediated secretion in murine colonic organoids.

Figure 3.4.3.3: Slc6a14 disruption results in an increase in smooth muscle thickness in the ileum and colon.

Figure 4.2.1: Arginine nitric oxide metabolism.

Figure 4.2.2: Hypothesis figure.

Figure 4.3.2: Split-open organoid concept. xvii

Figure 4.3.3: Standard curve for Nitric-Oxide (NO) measurement.

Figure 4.4.1: CFTR channel function in split-open murine organoids.

Figure 4.4.2.1: Slc6a14 is a major apical arginine transporter in the colonic epithelium.

Figure 4.4.2.2: Role of SLC6A14 in colonic epithelial NO production.

Figure 4.4.3: SLC6A14 activation with L-Arginine potentiates F508del CFTR function.

Figure 4.4.4.1 Murine Colonic epithelial tissue expresses inducible nitric oxide synthase (iNOS).

Figure 4.4.4.2: SLC6A14 causes NO mediated enhancement of F508del CFTR channel function.

Figure 4.4.4.3: Mutant F508del- CFTR function in double mutant murine organoids can be enhanced by enhancing PKG mediated phosphorylation.

Figure 5.2.7: Hypothetical model for genetic disorders.

xviii

List of Appendices

Appendix Table 1: Reproducibility of apical CFTR conductance (ACC) assay between two different platings of nasal cultures.

Appendix Table 3: Serum amino-acid levels measured in 56 days old mice.

Supplementary Figure 1: Nasal cultures from three CF patients have no significant differences in expression of epithelial cell differentiation markers.

Supplementary Figure 2: Representative Ussing chamber studies of bronchial epithelial cultures.

Supplementary Figure 3: Representative ACC traces of F508del-CFTR (-/+ VX-809 pretreatment) from nasal epithelial cultures grown on 96 transwell plate.

Supplementary Figure 4: Nasal cultures grown on 96 transwell plate express epithelial differentiation markers.

Supplementary Figure 5: MRP4 expression in primary nasal cultures.

Supplementary Figure 6: CFTR protein expression in primary nasal cultures.

Supplementary Figure 7: Z prime test of ACC on 96 transwell plate of nasal epithelial cultures.

xix 1

Chapter 1 Introduction

1.1 Cystic Fibrosis (CF)

CF is the most common life-limiting autosomal recessive genetic disorder in North America (CF Canada Registry 2014). The median age of survival for the Canadian CF population is the highest in the world at 51.8 years (CF Canada Registry 2014). About 59% patients with CF are diagnosed by one year of age, and this number is increasing after introduction of new born screening programs for CF (CF patient Registry 2015).

Cystic fibrosis is a multi-system disorder, affecting many epithelial tissues in the body1. This may result in recurrent lung infections, sinusitis, pancreatic insufficiency, meconium ileus and distal intestinal obstructive syndrome (DIOS), CBAVD in men (congenital bilateral absence of vas deferens), growth retardation, malnutrition and diabetes mellitus2,3. Decrease in lung function over time is the major cause of morbidity and mortality4,5. However, the gastrointestinal (GI) phenotype of CF also leads to significant morbidity6. In fact, the GI phenotype of meconium ileus (MI) is of particular interest from the genetic perspective, since it is a highly heritable feature (~80% heritability) of CF7, occurs early and has minimal external environmental influence. Meconium is the first feces passed by the newborn, and MI results from obstruction of the ileocaecal region with meconium8. On a contrast radiogram, MI can present as micro-colon with meconium pellets in the distal ileum, in addition to dilated bowel loops, intestinal perforation or calcifications9. Presence of MI usually indicates the presence of severe CFTR mutation2,8.

As shown in Figure 1.1, CF affects multiple organ systems, and classically a diagnosis of CF is made when one or more of the organ systems (shown) are affected in conjunction with an elevated sweat chloride measurement (≥60 mmol/L). The current diagnostic criteria for CF include: (1) the presence of clinical symptoms consistent with CF in at least one organ system and (2) evidence of CFTR dysfunction evident from one or more of the following – elevated sweat chloride (≥60 mmol/L) on two occasions, presence of two disease-causing mutations in CFTR , one from each parental allele or an abnormal nasal potential difference measurement10,11.

Figure 1.1: Cystic Fibrosis phenotype. CF is a multi-system disease affecting epithelial tissues across the body3,12. Phenotypic prevalence is indicated in italicized text beside the phenotype12,13.

1.2 Cystic Fibrosis Transmembrane conductance Regulator (CFTR)

1.2.1 Introduction

CF is caused by mutations in the Cystic Fibrosis Transmembrane conductance Regulator (CFTR) gene14-16. CFTR protein is functionally expressed on the apical surface of epithelial cells, where it functions as an anion channel, important for salt homeostasis1. It is mainly important for fluid secretion across the apical surface of the epithelium, thereby maintaining a fluid layer on the tissue surface17. In the pulmonary system, this fluid layer is referred to as the Airway Surface Liquid (ASL), which is important for maintaining ciliary function (Figure 1.2.1.1). Absence of apical fluid secretion results in the formation of inspissated mucus in the lungs, which impairs the ciliary function18, making the epithelial susceptible to inflammation and infection19,20.

Figure 1.2.1.1.: Fluid secretion across the apical surface. Non-CF or healthy airway epithelium expressing Wt CFTR (green) has a ASL height sufficient to allow ciliary function (left). CF affected epithelium expressing mutant CFTR (red), has a thinner ASL, thereby impairing ciliary function (right). Cilia are shown in orange.

There are several apical and basal channels and transporters that help in maintaining directional fluid transport across the apical surface of the epithelium (Figure 1.2.1.2). These can be broadly divided into secretory and absorptive molecules21. CFTR and alternative chloride channels like Calcium activated chloride channel (CaCC)22,23 on the apical surface and sodium potassium chloride channel (NKCC) on the basal surface constitute as secretory molecules. On the other hand, epithelial sodium channel (ENaC)24 is a major absorptive channel on the apical surface of the epithelium1,21.

Figure 1.2.1.2: Channels and transporters in a CF relevant epithelium. The differential functional expression of apical and basal molecules in the epithelium, help maintain the apical salt transport and fluid homeostasis (Graphic with help from Dr. Stan Pasyk).

1.2.2 CFTR Genetics and regulation of gene expression

Human CFTR gene was discovered in 1989 at The Hospital for Sick Children, Toronto14,15. It was found that CF was frequently caused by deletion of the phenylalanine residue at position 508 in the CFTR gene14,15.

CFTR gene is located on the long arm of chromosome 7, position 31.2 (7q31.2). It consists of 27 exons, and encodes a 6,129-bp transcript, which produces a 1,480 amino-acid polypeptide, which functions on the apical plasma membrane of epithelial cells, as an anion channel14,15,25,26. In mouse, Cftr gene is located on chromosome 6, and produces a 1,476 amino-acid polypeptide which has a ~78% sequence identity to the human CFTR protein25,27. In pigs (Sus scrofa), CFTR is present on chromosome 18, and produces a 1,482 amino-acid protein sharing a ~92% sequence identity to the human CFTR protein28.

At the locus level CFTR gene expression is regulated by multiple transcription factors (TF) like CREB29, HNF1α30 among others31. CFTR is also regulated by proteins involved in inflammation like NFκB and IL-1β32. All the mentioned TFs are known to enhance CFTR gene expression29-32. Recently, there has been a great interest in targeting microRNA (miRNA) and long non-coding RNA for CF therapeutics. Towards this, various miRNAs have been discovered which can modulate the functional expression of mutant CFTR33-39 (Figure 1.2.2.1).

Figure 1.2.2.1: Regulation of CFTR functional expression by miRNA and lncRNA. The figure illustrates biologically validated microRNAs and long non-coding RNA (CFTR-AS1 or Bgas)40, that can regulate functional expression of CFTR.

This increased understanding of the genetics of CF was possible because of technological advancements which led to the Human Genome project. This in turn helped discover many more disease-causing mutations in the CFTR gene. CFTR mutations are traditionally classified into six groups based on the mechanism by which they disrupt the functional expression of CFTR41,42. However, recently in light of new CF therapeutics and the paradigm shift towards personalized medicine, a newer classification of CFTR mutations has emerged as shown in Figure 1.2.2.243,44. The phenotypic changes resulting from these CFTR mutations are conglomerated in an international database (https://www.cftr2.org/)45. The CF therapeutic approaches indicated in Figure 1.2.2.2 are discussed later in section 1.7.

Figure 1.2.2.2: CFTR mutation classification. The top panel shows CFTR gene to protein expression in an epithelial cell. After the translation of CFTR protein in the ER, it gets trafficked to the Golgi apparatus, and thereafter reaches the cell surface where it functions as an anion channel. Mutations in the CFTR gene result in decreased functional expression of the protein. Mechanistic and therapeutic classification of CFTR mutations is shown. The existing and experimental CF therapies are also shown. The bottom panel gives an example of a representative CFTR mutation in each class. (Graphic with help from Dr. Stan Pasyk and adapted from Marson et al.43)

1.2.3 CFTR protein and regulation

The CFTR protein belongs to the ABC (ATP binding cassette) family of proteins46. It is the only ABC transporter which functions as an ion channel47.

After the translation of CFTR mRNA in the rough endoplasmic reticulum (rER), the CFTR polypeptide gets trafficked to the Golgi apparatus and ultimately reaches the apical plasma membrane, where it functions as an anion channel1,41,48,49. It gets core glycosylated in the ER and

complexly glycosylated in the Golgi50. As previously reported51 and as shown in Figure 1.2.3.1, CFTR protein can be seen as two distinct bands on a SDS PAGE – a lower band B which represents the core glycosylated form of the protein and an upper band C, which represents the complexly glycosylated form of the protein. The band C is usually considered to be the mature form of the protein which is functional on the plasma membrane26,50,52, however there is a suggestion that under ER stress a portion of the core glycosylated form of the protein could reach the plasma membrane via an unconventional secretion pathway53,54.

Figure 1.2.3.1: CFTR protein on gel electrophoresis. BHK cells over-expressing Wt and F508del CFTR were lysed and subjected to SDS-PAGE. CFTR protein migrates mainly as two bands (Band B and C) representing the core and complex glycosylated forms of the protein respectively. Core glycosylation occurs in the ER and the complex glycosylation occurs in the Golgi. The most common CFTR mutation F508del, causes a trafficking defect, resulting in absence of the mature Band C of the protein.

Although glycosylation is an important post-translational modification of CFTR, other modifications like S-nitrosylation can also affect functional expression of the protein55,56. Nitric oxide (NO) is an endogenous metabolite gas, which can diffuse through cell membrane57. Experimentally, the S-nitrosylating agent S-nitrosoglutathione (GSNO) has been widely used, and it is seen that increasing GSNO concentrations affect the functional maturation of CFTR 55,56. However, at a lower concentration, GSNO enhances the functional expression of CFTR by multiple pathways namely – inhibits the interaction of Hsp70/Hsp90 organizing protein (Hop) with CFTR58, increases SP3 dependent transcription of CFTR55, increases association of Csp protein with CFTR56, and enhances NO mediated PKG phosphorylation of CFTR59. At high concentrations GSNO inhibits functional expression of CFTR by three mechanisms – by inhibition of CFTR transcription55, reversible oxidative glutathionylation of cys-134460 and CFTR degradation via tyrosine nitration61. The use of NO donors in CF has been explored because in addition to improving functional expression of CFTR at a low concentration, they also improve the pathophysiology of CF disease62 by enhancing bacterial killing63 and enabling

biofilm dispersal64, by improving ciliary function65-67, enhancing epithelial fluid secretion68, facilitating smooth muscle relaxation leading to bronchodilation69,70 and improving ventilation- perfusion in the lung71.

At the molecular level, several efforts have been made to study the structure of CFTR protein72- 75. However, it was only recently that using Cryo-electron-microscopy (Cryo-EM), was the structure of Zebrafish and human CFTR solved76,77.

As shown in Figure 1.2.3.2, CFTR protein consists of two transmembrane domains (TMDs) and two nucleotide binding domains (NBDs) and a regulatory “R” domain. ATP binds to the NBDs. Each NBD has two ATP binding sites – one site is hydrolytic and the other is not78,79. The CFTR channel is ATP dependent and binding of the ATP to the NBDs and ATP hydrolysis is required for gating of CFTR78-80. The Cryo-EM structure of CFTR revealed the presence of a lasso motif, which is also present in a related ABC transporter MRP1, but its functional significance is not clear. It could be an interaction hub for protein-protein interactions81,82. Although the structure of “R” domain was not resolved by Cryo-EM, the proposed role of this domain in regulating CFTR channel activity has been extensively investigated using structure-function studies59,77,80,83.

Figure 1.2.3.2: CFTR protein structure. A) Schematic of human CFTR protein structure, showing the two Transmembrane domains (TMDs), two nucleotide binging domains (NBDs) and the regulatory “R” domain. The bolded numbers represent the Transmembrane helices (TMs) and the italicized numbers represent the intracellular loops (ICLs). In a three-dimensional model, the ICL1 and 4 would interact with NBD1, and ICL2 and 3 would interact with NBD2. B) Human CFTR structure as recently described using Cryo-EM. The most common disease- causing CF mutation F508del, resides in the NBD1 of CFTR. The F508 residue is shown in purple. (Figure made in PyMol using PDB file 5UAK from www.rcsb.org, published in Liu et al 201777)

CFTR is an anion channel which hydrolyzes ATP during its gating cycle78-80. The “R” domain of CFTR is considered to be an intrinsically disordered portion of the protein, which serves as a regulatory hub for CFTR channel gating84, and phosphorylation of this domain is the key biochemical modification which affects channel function. Phosphorylation status of the “R” domain can be modified by various kinases and phosphatases59,85-94. Among these, PKA (protein kinase A) activation by cyclic-AMP is the key physiological process which modulates channel function78. There are nine biochemically confirmed PKA phosphorylation sites in the “R” domain of CFTR59,87,93,95,96. However, other kinases have also been shown to be important in channel gating. PKC (protein kinase C) enhances CFTR surface expression and PKA mediated CFTR gating89,97. PKG (cGMP dependent protein kinase G) has 2 isoforms: PKGI and PKGII.

Of these, PKGII is expressed is expressed in the same apical sub-cellular compartment, and can activate CFTR by phosphorylation88,98-100. Adenosine monophosphate stimulated kinase (AMPK) is the only kinase which has an inhibitory effect on CFTR channel function. AMPK phosphorylation sites at serines 737 and 768 in the “R” domain of CFTR overlap with the PKA sites, and were previously referred to as PKA inhibitory sites101. Interestingly, the replacement of serine to alanine at position 737 and 768 led to an enhanced CFTR channel function – both basal and PKA stimulated, suggesting that phosphorylation of these is inhibitory101. Calcium calmodulin can enhance CFTR channel function directly by binding to the “R” domain102, and indirectly by activation of tyrosine kinase complex Src/PYK2103-105. There is also synergy among these kinase pathways, with PKA enhancing calcium release from intracellular stores106, and PKC enhancing PKA mediated CFTR function89,97. On the other hand, increased activity of phosphatases can dephosphorylate CFTR and decrease CFTR channel function85. Figure 1.2.3.3 summarizes the key molecular players which modify CFTR channel function by affecting phosphorylation status.

Figure 1.2.3.3: Regulation of CFTR protein phosphorylation by kinases. The black arrows indicate increased channel function upon CFTR phosphorylation. Red lines indicate routes to decreased CFTR function. (graphic with help of Dr. Stan Pasyk).

1.2.4 CFTR anion conductance and epithelial pH regulation

CFTR functions as an anion channel on the apical surface of the epithelium. As discussed above, the channel gating is dependent on ATP hydrolysis and phosphorylation of the “R” of CFTR. - - - - - The anion conductances through CFTR has been determined to be: SCN > NO3 > Br > Cl > I > acetate107-109. The bicarbonate permeability of CFTR is ~20% of its chloride permeability110,111. The role of CFTR as a chloride channel has been extensively studied1,26,112,113. However, the bicarbonate transport function of CFTR is also important in maintaining the extracellular pH of the apical fluid. Normally, the pH of ASL is basic, but in CF patients ASL is acidic114,115. Also, in CF patients, the pH submucosal nasal fluid is acidic115,116. This acidic pH of CF airways leads to an increase in ASL viscosity and compromises host defense117,118.

1.2.5 Cystic Fibrosis causing mutation F508del CFTR

F508del is the most common CF disease-causing mutation. It is found in at least one of the two gene copies in 87.5% of Canadian CF patients (CF Canada Registry 2014). F508del mutation occurs in and affects the NBD1 domain of CFTR; it results in mis-trafficking of the protein and is classified as a Class II mutation. The misfolded F508del CFTR protein is trapped in the ER, and undergoes proteasome degradation. In addition in vitro studies have shown that mutant F508del CFTR also has a defective propensity for phosphorylation at the PKA sites93, and exhibits altered gating, compared to WT-CFTR119.

Traditionally, the mutant F508del CFTR protein could be rescued to the cell surface with low temperature51. However, the plasma membrane stability of this low temperature (24-28 °C) rescued mutant protein is reversed upon increasing the temperature back to physiological temperature (37 °C), the mutant protein internalizes and is rapidly degraded120,121. The functional half-life of the mutant protein on the plasma membrane is <4 hours, compared to >24 hours for the Wt CFTR protein120.

1.3 Mechanisms underlying CF pathogenesis

CF is a multi-system disease affecting epithelial tissues. The mutations in CFTR gene which leads to its decreased functional expression, is the key initiating factor for CF pathogenesis1,122. This leads to a decreased pH of the apical surface epithelial fluid123, which in turn leads to inflammation and infection of the epithelium20,124 (Figure 1.3.1).

Figure 1.3.1: Key pathophysiologic processes in the CF airway epithelium. The left panel shows a non-CF healthy airway epithelium with functional expression of Wt CFTR leading to chloride and bicarbonate secretion thereby maintaining ASL height. The right panel depicts a CF airway epithelium where the functional expression of CFTR is lost, causing a chloride and bicarbonate secretory defect, leading to a decreased ASL height and decreased ASL pH. These defects lead to epithelial inflammation and infection in CF.

One of the key pathophysiological features of the CF affected epithelium is inflammation and infection20,124,125. Patients with CF exhibit a heightened inflammatory response and chronic pulmonary infections. There is controversy in the field, with regards to heightened inflammation leading to chronic infection or vice versa. The CF ferret model argues that the inflammatory pathways are dysregulated in CF, and the first bacterial exposure while passing through the birth canal, drives excessive inflammation in the CF lung126. However, the CF pig model, argues that inflammatory pathway is hypoactive upon bacterial exposure in the newborn CF pig airway and apoptotic pathway is hyperactive125. Future studies will help clarify this discrepancy in the literature.

The damage occurring because of infection and inflammation of the lung, leads to a decrease in pulmonary function over time122,127. The inability of the CF lung to clear infection, is thought to be associated with inflammation and inability to clear the pathogen. There are a number of factors which makes pathogen clearance difficult for the CF lung:

(1) Decrease in functional expression of CFTR120,128 (fluid secretory molecule) and increase in the functional expression of ENaC129,130 (fluid absorptive molecule), leads to a decrease in ASL height, and formation of inspissated mucus18. This results in impairment of ciliary function, and inability to clear pathogens131,132.

(2) Acidic ASL pH of the CF epithelium inactivates host antimicrobial peptides133. It also results in mucus adherence123,134,135 and further impairment in ciliary function131.

(3) Systemic inflammation because of immune dysregulation – (a) CF affected naïve T-helper cells have an intrinsic propensity to be differentiated towards pro-inflammatory Th-17 subtype, thereby enhancing inflammation136. (b) Hyper-responsiveness and enhanced cytokine production by CF affected macrophages upon exposure to pathogen also results in heightened inflammation137. (c) Increased neutrophil degranulation138, defective neutrophil phagolysosomal chlorination139, increased neutrophil elastase activity140 in the CF lung and the exaggerated effect of neutrophil elastase on the CF epithelium141 are responsible for innate immune dysfunction in CF.

The decrease in lung function over time is the most common concern for morbidity and mortality in CF122, and chronic infections with high bacterial counts associate with worsening pulmonary pathology142. As shown in Figure 1.3.2, the vicious cycle of infection and inflammation is one of the key pathophysiological features of disease progression in CF20,124,125. With evolving inflammation and disease progression, the pathogens adapt and evolve, resulting in a severe chronic infection with time143.

Figure 1.3.2: Inflammation and infection cycle in CF. The CF airway epithelium is predisposed to inflammation125, and this results in various pulmonary infections as shown in the bar graph (adapted from The Canadian CF registry 2014). The height of the bar graph represents prevalence of the bacterial species from the airways of CF patients across all ages. These infections further enhance inflammation in the CF lung, resulting in a vicious cycle of inflammation and infection.

1.4 CF animal models

Soon after the discovery of the CFTR gene in 198914,15, the first mouse model of CF was generated in C57Bl/6 mouse by Snouwaert et al in 1992144. This model was a germline knock- out of Cftr, created by putting an in-frame stop codon in exon 10 (S489X)144. Following this, several additional murine models of CF emerged145-147, with the first F508del CF murine model being developed in 1995 by two groups148,149.

1.4.1 Murine CF intestinal phenotype

The classic germline Cftr knock-out (Cftrtm1UNC) resulted in decreased survival after birth, with most of any surviving Cftr (-/-) offspring dying post-weaning144. The decreased survival of CF mice post-weaning can be improved by providing liquid diet to these mice150,151. This suggested that the loss of functional CFTR in the intestinal tissue of these mice144, resulted from a fluid secretory defect152, and the ensuing severe intestinal phenotype was responsible for decreased survival in CF mice150,151. The intestine of these mice also displayed intra-luminal obstruction by inspissated fecal material and mucus144. Histologically, both the small and the large intestine show characteristic pathology in CF mouse models153. The intestinal crypts are dilated with mucus accumulation, presence of eosinophilic concretions, goblet cell hyperplasia and increased crypt-villus axis height154,155.

The intestine of some CF mice also shows enlargement of the circular smooth muscle layer which is dysfunctional156. However, smooth muscle enlargement is not consistent among different CF mouse models, suggesting a role of modifier genes156. It has been hypothesized that increase in smooth muscle thickness is due to increased microbial load in the CF intestine157 and due to presence of intestinal obstructions causing mechanical stress on the intestinal wall158. Translationally, CF patients do experience intestinal obstruction and abdominal cramps159. In fact, distal intestinal obstruction syndrome (DIOS) caused by impaction of viscid fecal material

in the distal ileum and proximal colon, is a common complication in CF, with a prevalence of 5- 12 episodes/1000 patients/year in children160. Though there is no definitive evidence of intestinal smooth muscle enlargement in human CF intestines, there is however thickening of airway smooth muscle in CF human airways161,162.

1.4.2 Growth of CF mice

Although CF mice are born at Mendelian ratios, a significant number of deaths occur post- weaning144,146,148. The weight and size of CF mice remained low in most strains, well before death144. The decreased weight gain in CF mice has been attributed to malabsorption163, intestinal pathology154 and a defective secretion of Insulin like Growth Factor-1 (IGF-1)164. This defect in weight gain cannot be attributed to any differences in food intake164. Also post- weaning, the surviving CF mice continue to grow but never reach the size of their Wt littermates148.

1.4.3 Murine pulmonary CF phenotype

CF mice do not display an obvious pulmonary phenotype144. Although an inbred congenic strain of CF mouse has shown histological changes in the lung, no evidence for spontaneous infection was evident165. This limits its use in studying the pulmonary CF phenotype. However, to model CF lung disease, other approaches have been insightful. This includes the infection model, wherein instillation of Pseudomonas aeruginosa in the lungs of Cftr knock-out mice, using agar beads, leads to exacerbation of infection and increased mortality166. This finding has been replicated using an agar free instillation of a mucoid strain of Pseudomonas167. Another, model of CF-like lung disease, is the Nedd4 knock-out mouse model, which overexpresses the epithelial sodium channel (ENaC)24. This model shows airway mucus obstruction, inflammation, goblet cell hyperplasia and early death24.

1.4.4 Other CF phenotypes in mice

CF pancreatic disease is prevalent in the patient population168, however the pancreas is minimally affected in the CF mouse model144. It has been argued that this is due to the presence of residual CFTR in the pancreatic ducts of Cftr(F508del/F508del) mice and the presence of calcium activated chloride channels169.

CF mice do not develop a prominent liver disease, but they do show changes in the gall bladder, with hyperplastic bile duct epithelium170 and organ distension144.

Although male infertility is seen in CF patients171, CF mice do not exhibit this phenotype, again attributing this difference to the presence of the calcium activated chloride channels in the murine epididymis and seminal vesicles172.

CF patients show a decrease in bone mass which has been attributed to Vitamin D deficiency secondary to malabsorption173. CF mice do show severe osteopenia which is thought to be due to decreased bone formation and increased activity of osteoclasts174.

1.4.5 Other CF animal models

Drawbacks of the murine model of CF include: (1) absence of overt pulmonary phenotype144 (2) presence of alternative chloride channels165 (3) lack of submucosal glands in the murine airways115 (4) absence of ATP12A in the murine airway epithelium (discussed in Section 1.5)117. The absence of the pulmonary phenotype in the CF mouse necessitated the generation of larger animal models for CF. Currently there are two CF animal models in large animals – ferret175 and pig176, and the CFTR amino acid sequence identity compared to humans is ~91% and ~92% in these animals respectively177. Also, both animals display a pulmonary phenotype and have proved to be valuable in the understanding of CF pathophysiology175,176. However, given the cost of maintaining the pig and ferret models and the high incidence of death with MI at birth, the murine model remains attractive for complex genetic studies like the study of modifier genes177. Also, the small size, the short gestation period and the presence of the intestinal phenotype argues for the relevance of murine model in the study of modifier genes.

1.5 Modifier genes in CF

As described above, CF is caused by mutations in the CFTR gene, which results in its decreased functional expression. There are over 2000 mutations in the CFTR gene that are linked to disease (http://www.genet.sickkids.on.ca/), with F508del being the most common (CF Canada Registry 2014). However, patients homozygous for the F508del CFTR mutation, show considerable variation in their lung function, measured as forced expiratory volume in the first second

(FEV1)5,122. This variation in disease severity could be due to environmental or genetic factors. It is thought that modifier genes account for about 50% variation in lung function178. However, the strong evidence for the role of modifier genes in CF came from a classic genetic study, wherein the same CFTR gene knock-out allele was examined in different mice strains. Although failure to thrive occurs on most genetic backgrounds, Cftr-deficient mice showed a wide range in survival from a few days to many months on some backgrounds, implying that modifier genes modulate disease severity155.

Over the past 21 years since the first demonstration of modifier genes in CF, several strategies have led to their discovery. Broadly speaking, these involved either gene agnostic, or candidate gene approaches, wherein candidacy was based on biochemical or pathophysiological relation to CFTR or CF phenotypes. Specifically, these approaches involve: (1) Genetic analysis – which are gene agnostic except for hypothesis driven gene/region specific sequencing. (2) Systems Biology approach – which is also gene agnostic using data mining to discover new modifier gene candidates. (3) Molecular and Biochemical studies (candidate gene approach) – using proteomic approaches and studying protein-protein interactions. (4) Pathophysiologic analyses -candidate gene approach based on CF pathophysiology (Figure 1.3.1), discovering the role of molecules that modulate epithelial fluid secretion, chloride flux, pH, inflammation or infection in CF.

Figure 1.5.1: Modifier gene discovery approaches and translation into clinic. Since the discovery of CFTR gene in 198914-16, several approaches have been used to identify genetic modifiers of CF. The first column lists the approaches used, and the second column lists the genes identified using those approaches7,24,29,33-39,102,117,179-216. The third column indicates diagnostic and therapeutic translation of modifier gene knowledge base.

Gene Alias Hypothesis Citation Cytokines (Promotes inflammation) IL-8 Interleukin-8 Furlan et al 2016189 LTA Lymphotoxin alpha Corvol et al 2012217 MIF Macrophage migration Plant et al 2005218 inhibitory factor TGFβ1 Transforming Growth Factor Inflammation and Arkwright et al beta 1 fibrosis 2000219, Drumm et al 2005220 TNF Tumor Necrosis Factor Corvol et al 2012217, Laki et al 2006221 TNFA Tumor Necrosis Factor Yarden et al 2005222, Alpha Coutinho et al 2014223, Sanchez- Dominguez 2014224 Immune regulators C3 complement factor 3 Park et al 2011225 CD14 Cluster of Differentiation 14 Receptor for LPS Faria et al 2009226 (highly immunogenic) FCGR2 Fc Fragment Of IgG Neutrophil Fcγ De Rose et al Receptor IIa receptor II involved in 2005227 host defense HLA-DR4/7 Human Leukocyte Antigen Modifies lung disease Aron et al 1999228 DR4 and DR7 RAGE Receptor For Advanced Corvol et al 2012217, Glycosylation End Products Laki et al 2006221 TNFRSF1A TNF Receptor Superfamily Promotes Stanke et al 2006229 Member 1 inflammation Inflammation CEACAM3 Carcinoembryonic Antigen Host defense Stanke et al 2010230 Related Cell Adhesion Molecule 3 CEACAM6 Carcinoembryonic Antigen Host defense Stanke et al 2010230 Related Cell Adhesion Molecule 6 EDNRA Endothelin Receptor Type A Pro-inflammatory Darrah et al 2010231 peptide GCLC Glutamate-cysteine ligase Antioxidant effect - McKone et al catalytic subunit Glutathione 2005232, Marson et al 201443 GSTM1 Glutathione S-Transferase Antioxidant effect - Baranov et al M1 Glutathione 1996233

GSTM3 Glutathione S-Transferase Antioxidant effect - Flamant et al M3 Glutathione 2004234 GSTT1 Glutathione S-Transferase Antioxidant effect - Marson et al 201443 Theta 1 Glutathione HFE Hemochromatosis Oxidative stress Pratap et al 2010235 MRP-1 Multidrug Resistance- Antioxidative defense Mafficini et al Associated Protein 1 mechanism 2011236 Infection ABO ABO Blood Group Predicts airway Taylor-Cousar et al epithelial glycosylation 2009237 FUT2 Fucosyltransferase 2 patterns for bacterial Taylor-Cousar et al binding 2009237 FUT3 Fucosyltransferase 3 Taylor-Cousar et al 2009237 HMOX1 Heme Oxygenase 1 Cytoprotective during Park et al 2011225 infection BPIFA1 BPI fold containing family A, Antimicrobial protein Saferali et al 2015238 member 1 in upper airways BPIFB1 BPI fold containing family B, Antimicrobial protein Saferali et al 2015238 member 1 in upper airways Chaperones HSP70-2 Heat Shock Protein 70-2 Corvol et al 2012217, Laki et al 2006221 Phosphorylation ADRB2 Adrenoceptor Beta 2 CFTR phosphorylation Büscher et al 2002239 GUCY2C Guanylate Cyclase 2C CFTR activation Romi et al240 PPP2R1A Protein Phosphatase 2 CFTR Gisler et al 2013241 Regulatory Subunit 1 Alpha dephosphorylation PPP2R4 Protein Phosphatase 2 CFTR Gisler et al 2013241 Regulatory Subunit 4 dephosphorylation Other ACE Angiotensin I Converting Vasoconstrictor effect Arkwright et al Enzyme leading to Pulmonary 2003242 Dysfunction and portal hypertension in CF DCTN4 Dynactin 4 Modifies lung disease Viel et al 2016243 KRT8 Keratin 8 Cytoskeletal protein Stanke et al 2011244 for CFTR trafficking KRT19 Keratin 19 Cytoskeletal protein Gisler et al 2013241 for CFTR trafficking NOS1 Neuronal Nitric oxide Increases CFTR Grasemann et al Synthase functional expression 2000245 NOS3 Endothelial Nitric oxide Increases CFTR Grasemann et al Synthase functional expression 2003246 SERPINA1 Serpin Peptidase Inhibitor Normal α1-Antitrypsin Mahadeva et al Clade A Member 1 important for bacterial 1998216 killing

SLC26A9 Solute Carrier 26A9 Apical fluid secretion Soave et al 2014247, Li et al 2014 248, Miller et al 2015249 SLC9A3 Sodium/Hydrogen Apical fluid secretion Dorfman et al Exchanger 3 2011250 SFTPA1 Surfactant Protein A1 Pulmonary host Choi et al 2006251 defense SFTPA2 Surfactant Protein A2 Pulmonary host Choi et al 2006251 defense SNAP23 Synaptosome Associated Enhances residual Gisler et al 2013241 Protein 23 CFTR function

Table 1.5.1: Candidate modifier gene studies in patient population. Sequence variants in genes could modulate CF pathophysiology have been extensively studied. Additionally, as seen above genes which can regulate CFTR functional expression have also been implicated as CF disease modifiers.

Figure 1.5.1 summarizes various approaches for the discovery of modifier gene candidates. Some of these targets are discovered through multiple approaches. The above approaches should ideally converge on physiological validation of the target. The ultimate goal of modifier gene discovery should be to better understand disease pathophysiology and disease process, and thereby use that information to design novel therapeutics targeting those genes.

The Genetic approach permits large scale agnostic studies181-183, but the targets identified need to be replicated and causality of association need to be biologically validated using molecular, biochemical or physiologic approaches. However, once the genetic targets are identified and validated in in vitro or in vivo systems, they would serve as potential and easy to use indicators of disease severity or therapeutic response. SLC26A9 serves as an interesting example of such a translational approach– wherein a mutation or “tag” SNP (rs7512462) in SLC26A9 could be used to predict the response to use of VX-770 in CF patients carrying the gating mutation G551D, and also in the most common CF mutant F508del after rescue to the cell surface with VX-809252. As shown in Figure 1.5.1, several other modifiers have been discovered using this approach. Of these, SLC6A14 was of particular interest to us, given that it was discovered using the notably heritable phenotype of meconium ileus (MI), it being the first modifier gene which is an amino-acid transporter, and its role in CF was not being studied. SLC6A14 and its role in disease is further discussed in section 3.2 of this thesis.

The in silico or Systems Biology approach to identify modifier genes is relatively new in CF191. In silico approaches have been used in other fields of biology, and the need for biologic validation from genetic to physiologic levels would remain. Currently there are no fully validated targets that have been identified with this approach, but if successful in the future, the identified candidate modifier genes may well provide insight into CF.

The molecular and biochemical approaches have been used for over two decades in identifying molecules that may modify the CF phenotype. This would not have been possible without the Human Genome Project253, and the resulting increase in our understanding of genes and genetic regulation. This in conjunction with developments in structural biology254-257, led to an expansion in our knowledge of protein structures and their functions at the molecular level258. The mammalian cell is estimated to have 2-4 million proteins per cubic micron259, leading to about 307,707 known protein-protein interactions (http://thebiogrid.org). In context of CF, there are over 600 proteins that interact directly or indirectly with CFTR192. Of the verified interactions, they appear primarily mediated by the regulatory (R) domain84 or the PDZ motif on the C- terminal of CFTR194-196. Figure 1.5.1 shows various molecules that have been shown to interact with CFTR to modulate its functional expression.

The pathophysiologic approach has been extensively used to identify molecular targets. The key pathophysiological processes are outlined in Figure 1.3.1. In fact, the biologic validation of modifier genes from any of the above approaches, requires us to understand the mechanisms by which they modify the key pathophysiologic feature(s) of CF. This would not only allow us to extend our knowledge in context of the new molecular player, but also help us develop novel CF diagnostics and therapeutics based on these modifier findings. Towards this, the epithelial sodium channel (ENaC) serves as an excellent example. ENaC has been described as a modifier of CF at the pathophysiologic level24, before its identification at the genetic level186. Over- expression of ENaC in the murine lung has been shown to increase spontaneous infection24, as seen in the CF population. Given the genetic and physiologic data relating to ENaC, there are various therapies targeting ENaC in the clinical trial pipeline (shown in Figure 1.5.1).

There are various pathologies seen in CF which could individually be modified by tissue relevant molecules:

(1) The chloride secretory defect is the primary defect seen in CF because of decreased functional expression of the CFTR chloride channel on the apical surface1. This started the search for an alternative chloride channel which could be therapeutically targeted in CF, leading to the characterization of the calcium activated chloride channel (CaCC or TMEM16A)260. In fact, TMEM16A deficient mice were shown to die within one month of birth261 and have a chloride secretory defect leading to impaired epithelial fluid homeostasis and mucus accumulation in the airways262. However, the presence of an alternative chloride channel in a CF affected tissue is not enough to rationalize its role in apical fluid secretion. This is highlighted by the fact that even though CLC-2 was proposed as an alternative chloride conduction pathway in the airway epithelium263, it does not appear to modify CF at the genetic264 and physiologic levels265.

(2) The fluid secretory defect is downstream of the basic defect in chloride secretion1,17. However, this defect can be modified by molecules other than chloride channels. Towards this, ENaC is an excellent example where an apically expressed sodium channel is important for fluid absorption18. As described above, ENaC is a validated CF modifier gene24,186. However, the presence of an epithelial fluid regulating molecule is not enough to rationalize its role in apical fluid secretion – as seen with NKCC. This sodium-potassium-chloride co-transporter (NKCC) is important in fluid secretion in the context of cardiogenic pulmonary edema266, but it does not modify the CF intestinal phenotype in mice267.

(3) The bicarbonate transport function of CFTR and its role in pH homeostasis is well established110,114,116,209,268. The effect of loss of functional expression of CFTR, on epithelial pH, can be mitigated by targeting other proteins which also modulate pH homeostasis on the apical surface of the epithelium. This has led to the discovery of apical constituents like SLC26A9269, SLC9A3209,268 and ATP12A117, which can modify CF disease severity. With the advent of the porcine CF model, the role of CFTR in pH regulation and infection became clear118,132. This led to the discovery that the absence of a pulmonary phenotype in CF mice was related to the absence of the epithelial H+/K+ ATPase (ATP12A), which acidifies the ASL in humans117. Additionally, SLC9A3 which is a Na+/H+ exchanger is also known to acidify the apically secreted fluid across the epithelium. SLC9A3 deficient mice improve the CF intestinal phenotype by affecting fluid and pH homeostasis209,250. As described above, later GWAS showed that SLC9A3 is a genetic modifier of CF181. Interestingly, it has been shown that loss of function

in SLC9A3 alone can lead to the CF phenotype of CBAVD in the Taiwanese population, in absence of major CFTR mutations270.

(4) Inflammation and infection are hallmark pathophysiologic features of CF. There are several genetic modifiers of inflammation and infection in CF (Table 1.5.1). Pharmacological modulation of inflammation has been considered in CF, and the non-steroidal anti-inflammatory drug - Ibuprofen has been shown to decrease lung function decline in these CF patients with mild lung disease271. As shown in Figure 1.5.1, there are therapies in development that target inflammation in CF.

(5) Modifiers of other diseases can cause tissue specific modulation of CF disease severity3. As listed in Figure 1.5.1, modifiers of Type II Diabetes Mellitus and liver disease can modulate phenotypic severity of CF related diabetes and CF liver disease respectively. This highlights that some modifier genes can influence disease severity for various diseases affecting the same organ or that impinge upon similar pathophysiology. For example, TGFβ1 is a potential modifier of hemochromatosis associated hepatic fibrosis272 and also a modifier of lung disease in CF220.

1.6 Cystic Fibrosis Diagnostic tests

High salt (sodium and chloride) concentration in the sweat has been used to diagnose CF for decades273. The current diagnostic criteria for CF involves the use of various laboratory tests including sweat chloride concentration, nasal potential difference measures or a genetic testing for the presence of CFTR mutations10. The sweat chloride testing has improved over the decades from the classic Gibson and Cooke method273 to the development of a nanoduct sweat test system274 measuring chloride.

To assess and monitor pulmonary function, spirometry is usually performed which measures the forced expiratory volume in the first second (FEV1), and the forced vital capacity (FVC)275. However, given the variability of the test and the difficulty of use in preschool children, the development of other measures of pulmonary function, including a lung clearance index (LCI)276, can also be used, and maybe more suitable for young children. Given the advancements in CF therapy, the Ontario Newborn screening programme includes screening for CF, with a

two-stage procedure similarly as for many US states and European countries. This involves an initial measure of immunoreactive trypsinogen (IRT) followed by gene assessment for the most common CFTR mutations, if the IRT measure was high (www.newbornscreening.on.ca).

1.7 CFTR directed therapeutics

The basic defect in CF is caused by mutations in the CFTR gene14,15, and so the goal of CFTR directed therapeutics is to enhance the functional expression of CFTR. As shown in Figure 1.2.2.2, depending upon the mutation class, different drugs are needed for effective CFTR- directed therapies. Currently, there are two FDA approved therapies for CF – KALYDECOTM (VX-770) and ORKAMBITM (VX-809). KALYDECOTM which is potentiator of CFTR channel function, was originally approved for use for the gating mutation G551D-CFTR277,278. Thereafter its use was expanded to 33 CFTR gating mutations (https://www.cff.org/News/News- Archive/2017/FDA-Approves-Ivacaftor-for-23-Additional-CFTR-Mutations/). The average improvement in lung function (FEV1) for patients bearing at least one G551D CFTR allele was ~10% with treatment with Ivacaftor (STRIVE and ENVISION trials)279. However, robust responses were not seen in all patients, with about ~25% not responding279, indicating personalized therapies must go beyond CFTR mutation specific approaches.

ORKAMBITM has been FDA approved for use in the most common mutation F508del280. It is a combination of two drugs – VX-809 which is a chemical chaperone (corrector), and VX-770280. However, only 40-60% patients showed a clinically relevant response of increase in FEV1 by 3- 4% to ORKAMBITM 280. This has warranted the discovery of newer drugs for CF. Currently there are various academic laboratories and companies working on developing novel CF therapies (https://www.cff.org/Trials/pipeline). These efforts have led to a new class of drug called amplifiers, which stabilize the CFTR mRNA, in attempts to improve protein abundance281. For CFTR stop mutations, one of the therapeutic strategies is to develop read-through drugs like Ataluren. However, given the diversity in types of mutations and the inadequate responses in patients to ataluren282, development of other strategies are also underway. Towards this, antisense oligonucleotide (ASO) therapies are being developed for splicing and stop mutations283-285, which target specific gene regions to disfavor aberrant splicing. Additionally, gene therapy efforts are also underway286,287 and have led to the development of the first in-man

gene therapy trial287. Recent developments in nanotechnology, have also led to the development of nanoparticle based miRNA therapy for increasing CFTR expression in vitro288. These therapeutic approaches aim to treat CF patients with previously untreatable CF mutations in a precise fashion.

1.8 Thesis Rationale

CF is a complex multi-system disease, caused by mutations in the CFTR gene1,14,15. There are over 2000 disease causing mutations in the CFTR gene which cause CF with varying severity. For patients with the same CFTR genotype F508del on both alleles, there is a wide range of variation in disease severity5,122. Also, as described above the response to therapeutics is variable in CF patients, making it difficult to predict which patient will respond to which therapy. This necessitates the development of a personalized in vitro phenotypic readout, which can help in predicting the response to currently available CF therapies in individual CF patients. Towards this, we developed an in vitro CFTR functional assay for studying individualized patient specific drug responses to various patient-derived tissues, paving the path to personalized medicine in CF.

The variation in CF disease severity can partly be explained by sequence variations in modifier genes. GWAS have discovered that sequence variants in the putative promoter region of SLC6A14 could be partly responsible for this phenotypic variation181. However, there is limited biologic evidence showing the role of amino-acid transporter SLC6A14 in CF. Understanding this link, would be useful in understanding CF pathophysiology and its possible use as an alternative drug target in CF. Given the milder impact of modifier genes compared to the primary disease-causing CFTR gene, we used a knock-out murine model of Slc6a14 to study its impact on the CF phenotype. These studies describe the biologic role of SLC6A14 in CF, exploring it as a possible drug target, which might be useful especially in patients resistant to currently available CFTR modulators.

1.9 Hypothesis

The differences in patient specific responses in CFTR function to CF therapeutics, can be measured using a membrane potential based medium-throughput in vitro assay.

Ion transport function by mutant CFTR epithelium is variable and this is due in part to the activity of modifier genes. The genetic modifier of Cystic Fibrosis SLC6A14 is a broadly specific secondary active neutral/cationic amino-acid transporter, which modifies the fluid secretory capacity of the CF affected intestinal epithelium.

Chapter 2 Patient specific in vitro assessment of CFTR modulators: Path towards personalized medicine in CF

Content presented in this chapter has been published in the following research article, and reproduced under a Creative Commons Attribution 4.0 International License (License: http://creativecommons.org/licenses/by/4.0/)

Saumel Ahmadi, Zoltan Bozoky, Michelle Di Paola, Sunny Xia, Canhui Li, Amy P. Wong, Leigh Wellhauser, Steven V. Molinski, Wan Ip, Hong Ouyang, Julie Avolio, Julie D. Forman- Kay, Felix Ratjen, Jeremy A. Hirota, Johanna Rommens, Janet Rossant, Tanja Gonska, Theo J. Moraes & Christine E. Bear.

Phenotypic profiling of CFTR modulators in patient-derived respiratory epithelia. NPJ Genom Med. 2017 Apr 14;2:12. doi: 10.1038/s41525-017-0015-6. (PMID: 28649446).

Contributions: Data was generated and analyzed by Saumel Ahmadi, with help of the authors of the abovementioned research article. Study was designed by S.A., C.E.B. based on discussions with T.J.M., F.R., J.H., J.Rommens, J. Rossant and J.F.K. Functional studies of CFTR by FLIPR in epithelial cell lines were designed by C.L., M.D., S.A. and C.E.B.; S.A. and Z.B. developed the software for analyses. H.O., T.J.M. generated nasal cultures. J.A., T.G., T.J.M. did the nasal brushing. W.I. and T.G. performed the Ussing chamber experiments. Expression studies were conducted by K.D., S.X., A.P.W., M.D., S.V.M. and L.W. A.P.W., L.W. generated lung from iPSCs. Manuscript was written by S.A. and C.E.B. with inputs from all the authors.

2.1 Summary

Since the discovery of the CFTR gene in 1989, more than 2000 disease causing CFTR mutations being discovered. Depending on the types of mutations, mild to severe CF can ensue. An additional complexity is the presence of modifier gene sequence variations which can influence disease severity in different organs even in patients carrying the same mutations on both CFTR alleles. This has led to the realization of need for personalized medicine, warranting the development of in vitro patient specific phenotypic assays, for assessment of CFTR modulators in individual patients. Towards this, we developed a fluorescence based assay to measure CFTR function in respiratory epithelial cultures from CF patients. We found that in vitro patient responses to CFTR modulators is variable. Nonetheless, the patient specific responses in this assay are reproducible and correlate with the gold standard but relatively low-throughput Ussing chamber studies. In proof of concept studies, we also validated the use of this platform in measuring drug responses in lung cultures differentiated from CF-iPS cells. Taken together, we show that this medium through-put assay of CFTR activity has the potential to stratify CF patient specific responses to approved drugs and investigational compounds in-vitro in primary and iPS cell derived airway cultures.

2.2 Introduction

Pulmonary disease is the major cause of morbidity and mortality in CF patients 3,122. More than 2000 different CFTR mutations have been identified to cause CF (www. genet.sickkids.on.ca). As described in section 1.7, most patients bearing CFTR mutations that lead to defects in channel activation, or “gating mutations” exhibit a positive response to the “potentiator” called Ivacaftor or VX-770, a compound that acts directly to increase phosphorylation dependent CFTR channel opening277,278,289. G551D-CFTR is one such “gating mutant” and as it is a rare CF-causing mutation, patients bearing this mutation are generally heterozygous with another mutation on the other allele. Although most patients bearing G551D exhibit improved lung function following treatment with Ivacaftor279, approximately 25% fail to show a positive response – a failure which may reflect a number of factors including the influence of the other CFTR mutation or other tissue specific factors290.

The major CF causing mutation, F508del causes misfolding, misassembly and mistrafficking of CFTR120,291,292. In-vitro studies in primary airway epithelial cultures obtained at the time of lung

transplantation showed that in combination, the corrector compound: Lumacaftor plus the potentiator compound: Ivacaftor, were effective in partially rescuing the mistrafficking defect of F508del-CFTR and enhancing its channel activity respectively94,293-295. Recently, the combination, ORKAMBITM, was approved by the FDA after Phase III clinical trial data showing that treatment led to modest but significant improvement in lung function. There was a 3-4% increase in FEV1280 in the population of CF patients tested, all homozygous for F508del. However, the clinical trial revealed considerable heterogeneity in patient responses with close to 30-40% of patients failing to show a significant increase in FEV1296.

Research efforts to discover the next generation of mutation targeted therapies have escalated rapidly297-299 and in parallel, a need to predict patient specific responses. However, the inaccessibility of in-vitro, medium or high-throughput platforms that enables profiling of patient- specific responses to emerging compounds constitutes a major barrier to translation into the clinic. While rectal organoids are being explored as a means to predict patient specific responses to approved drugs300,301, such tissues may not fully recapitulate the context of the most severely affected tissue in CF- the airway epithelium. Here we describe a mid-throughput method that can be applied to study pharmacological modulation of mutant CFTR in patient-derived, primary nasal epithelial cultures and lung derived from induced pluripotent stem cells (iPS-cells).

In “proof of concept” studies, we showed that a novel adaptation of a fluorescence-based method for detecting ion channel activity was effective in measuring CFTR activity in primary nasal epithelial cultures and iPS-cell derived lung epithelium grown at the air/liquid interface- two new cultures systems with the promise of providing a renewable source of relevant tissue for personalized therapy development. Further, an exploratory trial of this in-vitro platform revealed its potential impact in defining individuals (homozygous for F508del) with variable responses to the ORKAMBITM therapy. This exploratory trial also highlighted the potential benefit of introducing a companion therapy to augment the response to Ivacaftor in patients who are heterozygous for G551D.

2.3 Materials and Methods

2.3.1 iPSC derived airway epithelial cell generation

The human ESC line CA1 was differentiated into lung epithelium as previously described in detail302,303. In brief, CA1 cells adapted to single cell passage were plated on a 10 cm (2.5x106 cells) dish coated with matrigel 24 hours prior to the induction of definitive endoderm (DE) according to the manufacturer’s protocol (DE Kit, StemDiff, Vancouver, Canada). DE cells were plated on transwells of a 12 well plate coated with human placenta collagen IV at a density of 5x105 cells/well and pushed towards anterior foregut endoderm using high concentrations of FGF2 in the presence of SHH. Differentiation into lung progenitors and immature lung cells were performed as previously described303. The epithelium was polarized using air-liquid- interface (ALI) and the cells were used after three to five weeks in ALI culture.

2.3.2 Nasal cell culture method

As described previously286, nasal brushings were performed (REB # 1000044783 at the Hospital for Sick Children, Toronto), and the cells were cultured in basal epithelial growth media (BEGM, Lonza, Walkersville, MD). After 2 passages, cells were plated on a collagen-coated 96 well transwells to differentiate them, and basal differentiation media (PneumaCult, StemCell Tech., Vancouver, Canada) was used. After 21 days of growth under air-liquid interface (ALI), they were used to measure function using the membrane potential assay.

2.3.3 Western blotting

Cells from transwell or regular multi-well plates were lysed using the modified radioimmunoprecipitation assay buffer (50 mM Tris-HCl, 150 mM NaCl, 1 mM EDTA, pH 7.4, 0.1% (v/v) SDS, and 1% (v/v) Triton X-100) containing a protease inhibitor cocktail (Roche) for 5-10 min. The lysates were then spun down at max speed (>10,000 rpm) for 5 min using a table- top centrifuge. The supernatant was then collected in a separate tube and laemmli sample buffer was added (1/5 dilution), and then sample was run on a SDS-PAGE gel. The protein from the gel was then transferred to a nitrocellulose membrane and blocked with 5% (w/v) milk. After blocking, the membrane was incubated with one of the following antibodies: human tight junction protein 1 (ZO-1, 1:5000, Life Technologies), human cytokeratin (pan clone AE1/AE3, 1:500, Zymed), human CFTR (UNC 596, CFFT), human Na+/K+ ATPase (mouse monoclonal

a6F, DSHB, Developmental Studies Hybridoma Bank at the University of Iowa) and human calnexin (1:10000, Sigma). After incubation with any of the above, the membrane was incubated with horseradish peroxidase conjugated secondary antibody raised in goat (against mouse or rabbit primary antibody, 1:2000 dilution), and after multiple washes chemi-luminescence signal was detected using the Li-Cor Odyssey Fc image acquisition system. The images were exported in tag image file format (TIFF), and analyzed using ImageJ 1.42 Q software (National Institutes of Health).

2.3.4 Immunofluorescence

Samples were fixed with ice-cold methanol in -20 ⁰C for 15 minutes. They were then washed and blocked in 4% BSA in PBS for 30 min and incubated with the appropriate primary antibodies against CFTR (Abcam), ZO-1 (Thermo Fisher Scientific, Abcam), Pancytokeratin (Abcam), and DAPI (Thermo Fisher Scientific) overnight at room temperature. Primary antibody was washed away with PBS and samples were incubated with monoclonal or polycolonal secondary antibodies (Life Technologies) for 1 hour93. Samples were imaged using Olympus IX81 Quorum Spinning Disk Confocal Microscope and Volocity 6.3.

2.3.5 qRT-PCR

RNA extraction was performed according to manufacturer’s protocol (Qiagen Micro or Mini Kit). Briefly, cells were lysed and after RNA extraction, concentration was measured using NanoDrop 2000. Only samples with a concentration >100 ng/μL were used, with a 260/280 ratio between 1.8-2.1. cDNA synthesis was performed using reverse transcriptase (iSCRIPT cDNA synthesis kit – Biorad) or without reverse transcriptase (negative control). Quantitative real-time PCR was performed using Eva green (Ssofast Evagreen – Biorad) fluorophore in 96 well plates (Biorad). The primers used for amplification are listed in the table below. Gene expression was normalized to GAPDH.

Table 2.3.5: Primers employed for qRT-PCR

Gene Forward Reverse

CFTR 5'-GCATTTGCTGATTGCACAGT-3' 5'-CTGGATGGAATCGTACTGCC-3'

GAPDH 5'-AAGAGCACAAGAGGAAGAGAG- 5'-TACATGGCAACTGTGAGGAG -3' 3'

FOXJ1 5'-GAGCGGCGCTTTCAAGAAG-3' 5'-GGCCTCGGTATTCACCGTC-3'

MRP4 5′-GGACAAAGACAACTGGTGTGCC- 5′-AATGGTTAGCACGGTGCAGTGG 3′ -3′

2.3.6 Ussing chamber

Nasal epithelial cells were studied in a non-perfused Ussing chamber (Physiologic Instruments,

San Diego, CA). The buffer (126 mM NaCl, 24 mM NaHCO3, 2.13 mM K2HPO4, 0.38 mM o KH2PO4, 1 mM MgSO4, 1 mM CaCl2 and 10 mM glucose) was maintained at pH 7.4 and 37 C 304 and continuously gassed with 5% CO2 / 95% O2. mix . The transepithelial potential (Vte) was recorded and the baseline resistance (Rte) was measured following repeated, brief short-circuit current pulses (1 µA every 30 sec). The results are presented as equivalent transepithelial current (Ieq), which was calculated using Ohm’s law. CFTR function was determined after inhibition of the epithelial sodium channel (ENaC) with amiloride (100µM, Spectrum Chemical, Gardena, CA) and cAMP activation with forskolin (10 µM, Sigma-Aldrich, US). CFTR activity was calculated as Ieq difference following CFTR inhibition with CFTRInh-172 (5 µM, EMD Millipore Corp. US). For drug rescue experiments nasal cell were treated with the corrector VX- 809 (6 µM) for 48 hours and acutely with the potentiator VX-770 (1 µM).

2.3.7 Membrane potential assay

Cells were grown on regular 96 well plates or transwell plates (individual transwell – 24 well plates, or HTS 96 well plates). Sources and format of plates were as follows: (1) nasal brushings from patients or healthy volunteers and plated on 96 well plate HTS transwells (2) CF and non- CF iPSC derived airway epithelial cells were plated on 24 well individual transwells or 96 well HTS transwell plates.

If the plate consisted the use of a transwell, the basal and apical solutions were kept different. The basal side had Hanks’ buffered solution containing chloride, and the apical solution

contained chloride free buffer (150 mM NMDG-Gluconate, 3 mM K.Gluconate, 10 mM HEPES, pH 7.35, osmolarity 300 mOsm)305,306. The blue membrane potential dye (Molecular devices) was loaded in the apical compartment only. However, if the plate was a regular multi-well plate (not a transwell plate), then a single solution of chloride free buffer (NMDG gluconate buffer) was used, and the blue membrane potential dye was dissolved in it at a concentration of 0.5 305 mg/ml . After 40 min of loading the dye at 37 ⁰C, 5% CO2 and humidified air, the plate was transferred to the microplate reader (Molecular devices i3 multi-well microplate reader). The microplate reader was set up at least 15 min before the experiment. Briefly, the reader was heated to 37 ⁰C and the settings for fluorescence whole well scan were turned on, with multiple points being read around the center. Excitation wavelength was set at 530 nm and emission at 560 nm. Upon start of the experiment, baseline reads of at least 3-4 scans were made, followed by addition of drugs (2.5 µL/well). After addition of each drug again 3-4 scans were done. The experiment usually ended by addition of inhibitor (CFTRinh-172 10 μM in this case). Since the membrane potential dye can work bi-directionally, the inhibitor response was prominent, as it caused change in chloride conductance across the apical plasma membrane. Upon completion of experiment the data was exported in a tab delimited format and analyzed using our analysis platform.

2.3.8 Chemical reagents

All of the following were dissolved in DMSO to make a 1000 fold concentrated stock solutions - Forskolin (Enzo lifesciences); Correctors of CFTR – VX-809, VX-661 (Selleck); Potentitor – VX-770 (Selleck), Inhibitor of CFTR – CFTRinh-172 (Cystic Fibrosis Foundation Therapeutics - CFFT); Calcein AM (Affymetrix eBioscience).

2.3.9 Statistical Analysis

One-way ANOVA with Tukey’s multiple comparison test was performed on all data with more than two data-sets for comparison, and SD was calculated using data from biological replicates. Unpaired two-tailed t-test was performed on data with two data sets. Kolmogorov-Smirnov test was used to determine normal distribution of data for ACC assay and Ussing chamber studies. Both Pearson and Spearman’s rank correlation tests were performed on Ussing chamber and ACC assays. P < 0.05 was considered statistically significant. Statistical analyses were performed using GraphPad Prism 6.01.

2.4 Results

2.4.1 Fluorescence-based assay of CFTR channel activity in patient derived respiratory epithelial cultures

The CF drug discovery landscape has recently expanded with the identification of multiple novel modulators of mutant CFTR278,280,307,308. Therefore, a priority for the field is to rank the efficacy of emerging compounds relative to approved drugs such as Ivacaftor (VX-770) in relevant tissues such as the airway epithelium. Furthermore, given the well-documented variability in drug responses amongst individuals with an identical CFTR genotype296, it is also essential to rank functional rescue by these new compounds in tissues from multiple individuals. Therefore, our primary goal was to develop and test methods for profiling a panel of compounds on patient specific airway epithelial cultures.

The classical method for testing the efficacy of modulators involves measuring the functional expression of mutant CFTR as a cyclic AMP regulated chloride channel in the apical membrane of primary airway epithelial cultures in the Ussing chamber309,310. This method, while providing direct electrophysiological parameters, is low-throughput. The FLIPR® Membrane Potential Plate Reader Assay is commonly used to screen chemical libraries for modulators of normal or mutant ion channels, over-expressed in fibroblast-like HEK-293 cells306,311,312. We adapted the use of the membrane potential sensitive FLIPR® dye to monitor apical chloride conductance mediated by normal and mutant CFTR in airway epithelium. First, we showed that the pharmacological rescue of F508del-CFTR could be measured using the membrane potential sensitive FLIPR® dye in a well-studied CF bronchial epithelial cell line (CFBE41o-)313. In these studies, the cells were grown to 5 days post-confluency at low temperature (27°C for 48 hours), to ensure partial rescue of the primary trafficking defect exhibited by F508del51. In Figure 2.4.1.1a, we show that apical conduction conferred by F508del-CFTR in rescued (r) CFBE41o- can be detected as membrane depolarization using the FLIPR® membrane potential dye assay after stimulation by forskolin, an agonist of cyclic AMP dependent protein kinase A (PKA), and Ivacaftor (VX-770), a potentiator of CFTR channel activity. In this study, the sensitivity of this assay was optimized by imposing an outward chloride gradient across the apical membrane. Further, to remove confounding effects of the PKA-sensitive apical sodium channel ENaC314,315, sodium was replaced with the non-permeant cation, N-methyl glucamine. The CFTR channel inhibitor (CFTRinh-172)316 was effective in inhibiting this fluorescence response, supporting the

specificity of the depolarization-mediated increase in fluorescence as reporting CFTR channel activity. The specificity of this response for CFTR channel function was confirmed in studies employing a bronchial epithelial cell line in which CFTR was completely disrupted by CRISPR- Cas9 (HBE-CFTR(-/-)). A western blot confirming absence of CFTR in HBE-CFTR (-/-) cells is shown in Figure 2.4.1.1b.

Figure 2.4.1.1: Functional measurement of CFTR chloride conductance in CFBE41o- F508del CFTR using the apical chloride conductance assay (ACC). (a) CFBE41o- cells over- expressing F508del CFTR were rescued with low temperature (27 ⁰C) for 48 hours and fluorescence based membrane potential assay conducted. CFTR activation by forskolin (10 µM) and VX-770 (1 µM), caused an increase in chloride conductance leading to membrane depolarization (red line) whereas additions of vehicle alone (DMSO) caused a minor deviation (teal line). This conductance decreased upon addition of CFTRinh-172 (10 µM), resulting in membrane repolarization. Error bars reflect SD across a 96 well plate (where n = 16 individual wells). HBE CFTR knock-out cell line is used as a negative control. Black line represents the effect of CFTR agonist followed by CFTRinh-172 on HBE CFTR(-/-) cell line. (b) Western blotting of lysates from human bronchial epithelial (HBE) cells, were loaded in decreasing amount (first 3 lanes on the left). Lysate from HBE-CFTR(-/-) cell line, was ran in the last lane. Electrophoretic pattern consistent with Wt CFTR, seen as a lower band B and a higher band C, is seen in the first three lanes (HBE Wt CFTR(+/+). Both bands are absent in the lysates derived from the HBE-CFTR(-/-) cell line. Na+/K+ ATPase was used as loading control.

Having optimized conditions for studying CFTR mediated depolarization in a CF bronchial epithelial cell line, we then developed conditions to measure CFTR-dependent membrane potential changes across the apical membrane of differentiated primary respiratory epithelia obtained from a CF lung transplant patient. In contrast to the epithelial cell lines grown on

plastic, such primary airway tissues express endogenous levels of F508del-CFTR, are cultured on a semi-permeable filter in a transwell chamber and exposed to an air/ liquid interface (Figure 2.4.1.2a). While transwells provide the optimal substrate for generating properly differentiated airway cultures, the filter confers significant background fluorescence, necessitating the development of a novel analytical technique to derive the specific signal conferred by CFTR channel opening on the apical membrane or apical CFTR conductance (ACC).

This analytical method involves the implementation of a thresholding function, the principles and application of which are shown in Figures 2.4.1.2a and b. First, the FLIPR® dye fluorescence intensities associated with the transwell chamber alone versus fluorescence intensities associated with live primary bronchial epithelial cell cultures can be described as a bimodal distribution and fitted with two Gaussian equations, peak #1 and peak #2, respectively. These two Gaussian functions are clearly separated permitting the definition of a threshold fluorescence intensity value (x axis intercept) beyond which the FLIPR® dye intensities can be attributed to properties of the tissue (peak #2). FLIPR® dye fluorescence intensity conferred by CFTR channel activation in the apical membrane of the primary airway cultures increases following activation of CFTR channel activity - corresponding to peak #3 (red). This method enabled better resolution of CFTR specific function in primary bronchial epithelial cultures derived from a CF lung transplant donor (Figure 2.4.1.2c). Further, as changes in fluorescence with activation (red peak in Figure 2.4.1.2b) are specifically normalized to resting cells (blue peak in Figure 2.4.1.2b) this method reports CFTR function that is normalized and cell number- independent. In Figure 2.4.1.2c, we show the kinetics of F508del-CFTR mediated changes in the apical CFTR conductance (ACC) assay after correction with VX-809 for 48 hours. The ACC response to forskolin activation and potentiation with VX-770 is shown as the red line after normalization to this response in cells pretreated with DMSO. As expected for F508del-CFTR channel activation, the peak response is reached within five to ten minutes and can subsequently be inhibited by CFTRinh-172. In summary, we developed a fluorescence-based method with the potential to detect potentiation of rescued F508del-CFTR on the apical surface of differentiated airway epithelium grown on transwell filters.

Figure 2.4.1.2: Application of Apical Chloride Conductance (ACC) assay to measure F508del CFTR function in primary bronchial tissue (a) Cartoon shows components of fluorescence based assay of CFTR mediated membrane potential changes. Airway epithelial cells, differentiated in air-liquid interface (ALI) on filters in transwell inserts are loaded with membrane potential sensitive dye following application to the apical surface. Resting apical membrane protein is measured in the presence of vehicle, and membrane potential changes mediated by increased apical CFTR conductance (ACC) determined following the addition of agonist (forskolin). (b) All fluorescence pixels from the well are plotted as a histogram and Gaussian curves are fit to values with low and high fluorescence peaks. Peak #1 represents background fluorescence conferred by regions on the filter not populated with living cells and Peak #2, corresponds to FLIPR® dye intensities conferred by the tissue. The fluorescence corresponding to background is removed by setting a threshold which corresponds to the tail of the Gaussian curve describing Peak #2– or 20% of the maximum fluorescence described by this peak. Peak #3, corresponds to FLIPR® dye fluorescence intensity conferred by CFTR channel activation after agonist addition in the apical membrane of the primary airway cultures increases following activation of CFTR corresponding to peak #3 (red). (c) FLIPR® based ACC assay and analytical function applied to the study of CFTR activation in primary bronchial epithelial cell cultures from a CF patient. Cultures from this F508del homozygous patient were pretreated with

CFTR corrector VX-809 or DMSO control. All cultures were acutely stimulated with CFTR agonist forskolin (10 µM) and VX-770 (1 µM), followed by CFTRinh-172 (10 µM). The analytical function (as described in Figures 1 a-b) was applied before calculating mean fluorescence intensity at each time point. The line graph represents change in fluorescence relative to baseline (ΔF/ F0), and this ratio was normalized to the vehicle (DMSO) treated well. The error bars reflect SD (n = 3 biological replicates).

2.4.2 ACC assay of patient-specific responses in nasal epithelial cultures is accurate and reproducible

We were then prompted to determine if this new method was accurate in reporting in-vitro patient-specific responses to pharmacological interventions in the nasal epithelium, a source of respiratory epithelium that is relatively accessible and of interest for testing patient-specific drug responses. In order to assess accuracy and reproducibility, we compared ACC responses to known interventions in primary nasal cultures derived from three CF patients and differentiated on transwells on three separate occasions (experiments 1-3). Nasal cultures from the F508del homozygous patients were treated for 48 hours with the corrector compound: VX-809 or vehicle alone. Each box of Figure 2.4.2.1a shows the individual ACC trace for a separate culture, at the time at which forskolin plus potentiator was added and the time at which CFTRinh-172 was added. The cultures derived from the patient heterozygous for G551D (G551D/2622+1G>A) were similarly treated. Quantitation of epithelial cell marker proteins (ZO-1 and pancytokeratin C) showed that the cultures were similarly differentiated (Figure 2.4.2.1c and Supplementary Figure 1).

It is apparent from Figure 2.4.2.1a that the ACC traces are different for each patient- with the stimulation with forskolin and VX-770 causing a larger peak response in the G551D/2662+1G>A cultures than the cultures from F508del/F508del cultures (without the corrector: VX-809), as expected. Importantly, the ACC responses measured for each donor were similar for experiments 1 through 3 as shown in Figure 2.4.2.1b, highlighting the accuracy and reproducibility of this assay. Reproducibility of ACC measurements between two different platings was also observed in a larger number of patient derived cultures (n=19, Figure 4.4.2.1d, donor genotype and interventions listed in Appendix Table 1, pointing to the utility of this method in reporting patient-specific responses to CFTR modulators.

Figure 2.4.2.1: ACC Assay is accurate measure of mutant CFTR function and responses to interventions in patient-specific primary nasal epithelial cultures (a) Primary nasal cultures from 3 different CF patients were analyzed in 3 separate experiments for CFTR function using the ACC assay. 2 patients were homozygous for F508del CFTR and 1 patient had G551D on one allele and 2622+1G>A on the other. Cultures from each patient were pretreated with CFTR corrector VX-809 or control: DMSO for 48 hours. As shown in the magnified well (upper, right), all cultures were acutely treated with CFTR agonist fsk (Forskolin: 10 µM) and VX-770 (1 µM) followed by CFTRinh-172 (10 µM). (b) Scatter plot represents maximum percentage change in fluorescence (ΔF) after addition of CFTR agonist and potentiator, relative to baseline (F0) measurements prior to addition. (c) Consistent expression of epithelial differentiation markers: ZO-1, pancytokeratin (PanCK) relative to loading control Calnexin (CNX) in each of the wells (i) Densitometry analyses of bands corresponding to ZO-1, normalized to loading control CNX (western blot shown in Supplementary Figure 1). One-way ANOVA and Tukey’s multiple comparison tests show no significant (ns) difference in the expression of ZO-1 across

the 3 patients (n = 6). (ii) Densitometry analyses for PanCK expression normalized to CNX (blots in supplementary Figure 1). One-way ANOVA and Tukey’s multiple comparison tests show no significant (ns) difference in the expression of PanCK across the 3 patients (n = 6). (d) Reproducibility between two biological replicates using the ACC assay is shown. Values derived from each experiment are listed in Appendix Table 1. There is a significant Pearson and Spearman correlation r (p < 0.0001, n = 4 subjects).

2.4.3 ACC assay correlates with Ussing chamber measurements of patient-specific responses to interventions.

Paired studies of the peak ACC response to forskolin and/or VX-770 and peak transepithelial currents as measured in Ussing chambers were conducted in order to assess the correlation between ACC and the gold standard method for assessing functional rescue of F508del-CFTR in tissues derived from patients homozygous for this mutation. As in our previous studies, we focused on studies of patient-derived primary nasal epithelial cultures. In this study, nasal cultures from six different F508del/F508del patients were plated in a 96 transwell plate format for the ACC assay. Nine replicate ACC measurements were obtained from each well by imposing a 3 by 3 matrix as shown in Figure 2.4.3.1a. The peak ACC responses within each well were colour-coded according to the attached scale bar, with cobalt blue being a low response and red- the maximum within the plate. As expected, the peak responses to forskolin- /+ VX-770 were modest in all of the patient-derived nasal cultures unless pretreated with VX- 809. In comparative Ussing chamber studies of nasal cultures from the same individuals (a representative tracing shown in Supplementary Figure 2), we found that there was an excellent correlation (Figure 2.4.3.1b, p= 0.002) between the magnitude of the CFTR channel activity as measured in the Ussing chamber and measured in the ACC assay.

Figure 2.4.3.1: ACC Assay correlates with Ussing chamber measurements of patient- specific responses to interventions in primary nasal epithelial cultures (a) Heat map visualization of apical CFTR chloride conductance measured in nasal epithelia cultured on a 96 transwell plate. Cultures from 6 F508del homozygous patients subjected to correction – chronic 48 hour treatment with VX-809 (or vehicle –DMSO) and acute agonist treatment with FSK and/or VX-770 with DMSO as control. Colour scale reflects range of depolarization responses on this plate with red representing the maximum response and blue, the minimum. (b) Correlation plot of ACC assay and Ussing chamber studies. CFTR mediated chloride conductance determined in ACC assay (ΔF/F0) correlated with CFTR mediated chloride conductance measured in Ussing chamber studies (ΔIeq) for cultures derived from seven patients homozygous for F508del. The values derived from each assay for each culture are listed in Appendix Table 2. A representative Ussing chamber tracing is shown in Supplementary Figure 2. The diamond shaped points represent CFTR mediated forskolin responses after correction with VX-809 and potentiation with VX-770 with the filled circles representing forskolin responses in the absence of the VX compounds. There is a significant Pearson and Spearman correlation , between these assays (n = 8 patients, pre- and post-treatment).

2.4.4 ACC assay enables profiling of existing and novel interventions in patient-specific primary nasal epithelial cultures

We generated a 96 transwell array of nasal cultures from multiple subjects in order to evaluate the potential of our new method to compare patient-specific drug responses and to profile different modulatory compounds.

Figure 2.4.4.1a shows the ACC responses to interventions tested on nasal cultures generated from multiple patients, five CF and 2 non-CF subjects (one to two columns per subject). Of the CF patients studied, three are homozygous for F508del and two heterozygous for G551D. The traces show experimental data similar to that displayed in Figure 2.4.2.1a, in which, over time,

43 nasal tissues responded to CFTR channel stimulation with depolarization (upward deflection). Subsequently, repolarization was induced after the addition of CFTRinh-172 (downward deflection). Nine technical replicates for responses within each transwell were generated by subdividing each well according to a 3 by 3 matrix and we provide a “zoomed-in” image of the nine replicate traces in two wells in Supplementary Figure 3. As for the studies of cultures in a 24 transwell plate shown in Figure 2.4.4.1a, we confirmed that the cell density and CFTR mRNA expression was consistent across the 96 transwell plate (Supplementary Figure 4).

Figure 2.4.4.1: ACC Assay enables profiling of existing and novel interventions in patient- specific primary nasal epithelial cultures (a) Nasal brushings from 5 CF and 2 non-CF patients were differentiated together in a 96 transwell plate. Patients bearing the F508del mutation on both alleles were tested after rescue (48-hour treatment) with corrector VX-809 or vehicle. During the assay CFTR was activated with forskolin (10 µM) plus or minus compounds/drugs listed on the right axis with vehicle (DMSO) as control. CFTRinh-172 was added to terminate the response and assess CFTR specificity. Changes in fluorescence in response to CFTR modulators were monitored overtime simultaneously in multiple regions in each well (n=9). Hence, 9 overlapping traces were depicted in each rectangle (magnification shown in Supplementary Figure 3). Supplementary Figure 4 shows that there is consistent CFTR mRNA expression amongst the wells and the protein expression of markers of epithelial differentiation is consistent (b) The data in panel a) was also represented in the form of a heat map. Blue and red represents minimal and maximal CFTR stimulations after forskolin +/- potentiator for the plate. (c) Bar graphs show peak forskolin mediated responses for nasal cultures from non-CF individuals and peak forskolin responses +/- modulators for cultures from CF subjects. Error bars represent SD. Differences were assessed using two-way ANOVA followed by a multiple comparison test. Asterisks represents statistical differences of p < 0.01. Bars with white hatched lines represent responses to ORKAMBITM for each of the F508del homozygous patients, and the differences between them were again analyzed using two-way ANOVA with multiple comparison test. The hash (#) sign represents statistical difference (p < 0.02) in response to ORKAMBITM in patient 2 compared to patient 1 and 3.

It is clear from Figures 2.4.4.1a and b (the colour-coded heat map of peak responses for the same plate) that the columns containing differentiated, nasal epithelial cultures from non-CF (Wt)

individuals exhibit robust depolarization responses after addition of the CFTR channel activator forskolin (fsk) and/or potentiator drugs (VX-770 and four distinct modulators, to be discussed). On the other hand, for columns containing transwells seeded with nasal epithelial cultures derived from CF patients, the amplitude of the response is less, as expected.

As previously mentioned, nasal cultures from patients homozygous for F508del were rescued with the pharmacological corrector VX-809 or DMSO as control for 48 hours prior to measuring CFTR channel function. After rescue, F508del-CFTR was stimulated acutely with forskolin with or without VX-770 or other investigational modulators (ie., PSK21 or PSK22). Both PSK potentiators were derived from in-house screens and P5 was obtained from Cystic Fibrosis Foundation Therapeutics since it had been previously described as a potentiator for various mutant CFTR proteins317. In certain cases, the multidrug resistance protein 4 (MRP4) inhibitor, MK-571 was tested318,319. Inhibition of MRP4 was previously suggested to prevent cAMP efflux within cellular microdomains containing CFTR and thereby augmenting CFTR phosphorylation when added concurrently with agonists of cAMP319. As seen in the bar graph (Figure 2.4.4.1c), FLIPR responses of cultures from Wt-CFTR subjects after stimulation with forskolin (white bars) were approximately four-fold greater than the forskolin responses from untreated nasal cultures from CF patients (light grey bars). With Lumacaftor pretreatment and acute potentiation with Ivacaftor and forskolin, cultures generated from all three patients homozygous for F508del exhibited a statistically significant response (p<0.01). However, absolute functional responses post-treatment were variable for the three individuals. We found that Pt #2 had a significantly lower response to in-vitro ORKAMBITM treatment, compared to Pt #1 and Pt # 3 (p < 0.02, Fig 2.4.4.1c).

Also, the absolute responses to various potentiators after correction with Lumacaftor were variable for the three individuals homozygous for F508del. Together with forskolin addition, acute addition of PSK22 induced a modest improvement relative to VX-770 in Pt #3 (*, p<0.05 or p<0.01 respectively). The MRP4 inhibitor, MK-571, did not augment the effect of correction with VX-809 and potentiation with VX-770 in nasal cultures from any of the F508del homozygous patients. Therefore, this platform revealed patient-specific responses to existing and investigational modulators of F508del-CFTR.

This platform also revealed variable responses by two CF patients heterozygous for G551D to the same panel of compounds (Figure 2.4.4.1c). PSK22 was effective in enhancing the functional expression of G551D for one patient, (G551D/2622+1G>A, *, p<0.01) relative to Ivacaftor, pointing to the value of this platform in ranking patient specific responses to investigational modulators. Interestingly, addition of MK-571 augmented the potentiation mediated by VX-770 observed in nasal cultures derived from both of these patients. We confirmed that primary nasal epithelial cultures express MRP4 using RT-PCR (Supplementary Figure 5), supporting the claim that MK-571 is mediating this rescue effect by inhibiting MRP4. This is the first evidence in patient-derived tissues supporting a role for MRP-4 inhibition as a therapeutic target in CF patients who are heterozygous for gating mutations such as G551D.

Examination of CFTR protein abundance in the same nasal cultures provides a potential explanation for the differential effect of MK-571 in rescuing the functional expression of CFTR in cultures from patients bearing G551D relative to F508del (Supplementary Figure 6). The abundance of mature and thus plasma membrane localized G551D-CFTR protein (C band) is close to that detected in non-CF cultures. On the other hand, the abundance of F508del-CFTR protein is low and the mature form visualized as a diffuse, weak band even following Lumacaftor treatment. According to the model first proposed by Naren and colleagues, the MRP-4 transporter requires expression within a macromolecular complex with CFTR on the cell surface in order to exert a modulatory effect319. Hence, our platform provides the first evidence that MRP-4 may augment Ivacaftor responses in patients bearing mutations that do not impair CFTR processing.

2.4.5 FLIPR® based ACC assay quantifies CFTR channel activity and pharmacological rescue of F508del in iPS cell-derived lung

CF researchers have been encouraged by the recent progress in differentiating patient-derived stem cells (induced pluripotent cells) into CF affected epithelial tissues as these have the capacity for infinite expansion for patient-specific drug-profiling platforms. Further, in proof of concept studies, CF patient (F508del)-derived iPS cells, differentiated into proximal lung have been shown to recapitulate the primary defect in the functional expression of CFTR mediated chloride channel activity302,303. Such functional studies have been performed using iodide efflux, patch

clamp or Ussing chamber studies of iPS cell-derived lung epithelium300,302,320,321. To date, none of these functional assays in stem cell derived tissues have been adapted to the mid-throughput format necessary for patient-specific drug profiling. Hence, we tested the potential of our new method to measure the primary defect caused by F508del and its utility in quantifying the pharmacological rescue of F508del-CFTR in iPS cell-derived lung cultures.

Figure 2.4.5.1: ACC Assay reports primary defect and pharmacological rescue of major CF mutant in lung cultures differentiated from iPSCs (a) Embryonic stem (ES) cell from Wt (CA1) and iPSC derived from F508del CF patient (GM00997) were differentiated to airway epithelia as previously described303. Apical CFTR conductance (ACC) is measured in these epithelia. Airway tissue generated from Wt CA1 ES cells show a robust response to CFTR stimulation by forskolin (FSK) using the fluorescence based detection method and the analysis shown in Figure 2.4.1.2. The changes in fluorescence in activated cultures were normalized to fluorescence measurements in vehicle (DMSO) treated cultures. The traces (upper panel) are representative of 3 biological replicates (or three transwells) wherein >50 regions within each transwell were monitored over time. iPSC derived airway epithelium from CF affected individuals were rescued with corrector VX-661 (1 µM) or DMSO as control (lower panel). CFTR channels were activated in all cultures by forskolin and VX-770 (1 μM). (b) Bar graphs

represent maximum percentage change in fluorescence (ΔF) after addition of CFTR agonist, relative to baseline (F0) measurements prior to agonist addition (time=10 minutes). Asterisks indicate statistical significance using one-way ANOVA and Tukey’s multiple comparison tests for the 3 distinct differentiations (*p<0.05, **p<0.01, ***p<0.001).

We measured the functional response of Wt-CFTR expressed in ES cell derived proximal lung grown on semi-permeable supports at air/liquid interface using our ACC method (Figure 2.4.5.1). Forskolin evoked a robust depolarization (ie. CFTR activation) in these airway cultures and CFTR inhibition using CFTRinh-172 led to a subsequent repolarization. In contrast, in airway cultures differentiated from iPS cell lines derived from a patient homozygous for F508del-CFTR and corrected with VX-661 (chemically related to VX-809), the forskolin response (together with VX-770) was modest, but significant, as shown in Figure 2.4.5.1. These results recapitulate those previously published using the time-consuming iodide efflux methods using the same iPS cell line302,303 and show the potential utility of the ACC method for studying modulation of mutant CFTR responses in iPS cell-derived lung tissue. Taken together, CFTR function and pharmacological rescue of F508del-CFTR can be assessed using our ACC technique in lung cultures differentiated from patient-specific iPS cells.

2.5 Discussion

CF is a complex genetic disorder, and disease severity depends on the CFTR mutations43 present in the patient, as well as sequence variation in modifier genes3,155. This has warranted the development of patient specific phenotypic assays. Medium-throughput platforms for testing patient tissue-dependent responses to approved drugs and investigational compounds will enhance progress in effectively treating CF. Previously, the feasibility of such “CF clinical trials in a dish” was limited by the paucity of patient-derived lung tissue and the lack of higher- throughput methods to profile modulation of mutant CFTR channel activity by a panel of lead compounds. Recently, the Beekman group showed the utility of functional assays of mutant CFTR-mediated swelling of patient-derived rectal organoids in informing potential treatment strategies301. The fluorescence-based ACC assay described here provides a tool for profiling responses to multiple treatments on patient-derived airway epithelia.

Importantly, we showed that this fluorescence-based method for measuring pharmacological rescue of mutant CFTR is accurate in reproducibly reporting patient-specific responses to CFTR modulators across different platings. It also recapitulates the relative patient-specific responses

to treatments observed in the “gold-standard”, but lower through-put Ussing chamber system. In principle, the 24 transwell based, Transepithelial Chloride Conductance assay (TECC) developed by Robert Bridges and used in recent papers by Mutyam et al. and Vu et al.322,323 measures CFTR channel function in a mid-throughput format. However, scalability is difficult and also there have been no published accounts of the accuracy of the TECC assay in reporting responses by patient-derived tissues to particular interventions across different biological replicates as we show in Figure 2.4.2.1. Further, unlike the TECC assay, the ACC assay doesn’t require the formation of a tight epithelial resistance as it measures apical membrane potential changes. This difference is particularly relevant in measurements of CFTR function in iPS cell-derived lung cultures, as these cultures are not homogeneously differentiated302,303.

In its current form as a mid-throughout platform, we showed the potential for the ACC assay to identify a novel potential treatment. We showed that the MRP4 inhibitor: MK-571 enhanced the potentiation caused by Ivacaftor of nasal epithelial cell cultures derived from CF patients heterozygous for the G551D mutation (i.e, G551D/E585X or G551D/2622+1G>A). Like CFTR, MRP4 is a member of the ABC superfamily of membrane proteins318. It functions as a pump that transports a diverse range of substrates including lipophilic drugs324,325, glutathione326 and cyclic AMP325,327. Hence, augmentation of Ivacaftor mediated potentiation by MK-571 could occur through induction of local elevations in cytosolic cAMP as suggested by Naren and colleagues319, enhanced accumulation of VX-770 or another unknown substrate of MRP4 that modulates CFTR activity. Our platform provided evidence that MRP4 modulation will not be effective in augmenting the response to ORKAMBITM in airway cultures derived from patients homozygous for the F508del mutation (Figure 2.4.4.1). This initial observation needs to be confirmed in a larger cohort of patients using this platform.

Our longer term goal is to optimize the ACC assay such that large chemical libraries can be screened on patient derived nasal cultures and iPSC-derived epithelium. To date, such an assay does not exist and there is an urgent need, given the number of rare CF-causing mutations for which there are no therapeutic options. Currently, the Z prime score associated with the ACC assay of CFTR function in a 96 transwell plate of primary nasal epithelial cultures is promising at close to 0.2, but not yet suitable for a robust chemical library screen (Supplementary Figure 7). The generation of more uniform cultures in the 96 transwell format will constitute a vital step in scaling this assay for high-throughput screens. More efficient protocols are being developed for

the differentiation of conditionally-reprogrammed nasal cells to differentiated nasal epithelium328 or iPS cells to CF lung302,303. These innovations will enable the generation of uniform cultures of patient-specific tissues on 96 transwell plates–platforms suitable for defining the best intervention for each patient.

Finally, we have yet to test the utility of the ACC assay of patient-derived respiratory tissues in predicting the clinical outcome to existing and emerging therapies. Currently, we are testing the predictive power of this platform in collaboration with clinical scientists as patients (F508del homozygotes) are being enrolled for treatment with ORKAMBITM and providing drug naïve nasal epithelium and blood samples for iPS cell generation.

Chapter 3 SLC6A14 modulates epithelial fluid secretion

Content presented in this chapter has been submitted for publication:

Saumel Ahmadi, Sunny Xia, Yu-Sheng Wu, Michelle Di Paola, Randolph Kissoon, Catherine Luk, Fan Lin, Kai Du, Zeynep Baskur, Lisa Strug, Johanna M. Rommens and Christine E. Bear.

Genetic modifier of Cystic Fibrosis regulates the primary defect in fluid secretion.

Contributions: Data was generated and analyzed by Saumel Ahmadi, with help of the authors of the abovementioned research article. Study was designed by C.E.B., S.A. based on discussions with J.R., C.L., S.X., M.D., Y-S.W. and R.K. Murine organoids were generated by S.X., mouse genotyping was performed by C.L., qRT-PCR was performed by M.D., Y-S.W., F.L. ELISA assays were performed by R.K., S.A., analysis of histology and morphology was performed by R.K. and Y-S.W., live cell fluorescence was performed by S.X. and K.D., survival statistics were performed by Z.B., L.S. and ex-vivo closed loop assays were performed by S.A., M.D. and Y- S.W. Manuscript was written by C.E.B. and S.A. with input from all the authors.

Acknowledgements:

We are thankful to the staff at the Laboratory Animal Services (LAS) at The Hospital for Sick Children for their help in maintaining the mouse colony, the Toronto Center for Phenogenomics (TCP), NORCOMM project and Dr. Lauryl Nutter for helping generate the Slc6a14 knock-out mice. We are grateful to CFFT (Cystic Fibrosis Foundation Therapeutics) for providing the CFTR modulators and to Dr. Bob Scholte, Erasmus Medical Center Rotterdam, The Netherlands for providing the FVB CftrF508del mice. We are also grateful to Hayley Craig-Barnes and the Analytical Facility for Bioactive Molecules for help with mass spectrometry, to members of Dr. Nicola Jones’ laboratory for their help with organoid culture, to the Imaging Facility at the Hospital for Sick Children for their technical help with imaging, and to Dr. Hartmut Grasemann, Dr. Tanja Gonska, Dr. Lisa Strug and Dr. Jaques Belik for their helpful discussions. We are thankful to all members of Dr. Christine E Bear’s laboratory and to all members of Dr. Johanna Rommens’ laboratory for their helpful suggestions.

3.1 Summary

CF patients bearing the common F508del-CFTR mutation on both alleles, exhibit tremendous variation in the severity of their disease5,329. GWAS to investigate the intestinal phenotype of Meconium Ileus (MI), has suggested that SNPs in the putative promoter region of SLC6A14 influence susceptibility181. However, the biologic basis of this modification is not known. To understand this, we used the CF associated Cftr(F508del/F508del) and Slc6a14 knock-out alleles to generate double mutant mice with loss of function of both Cftr and Slc6a14 genes, for comparison with Cftr(F508del/F508del) mice. SLC6A14 is abundantly expressed in the murine colonic epithelium and is a major transporter of amino acids such as arginine across the epithelial apical surface. We also found that the fluid secretory capacity of the colonic epithelium is modulated by SLC6A14 and that its loss led to a worsening of CF intestinal phenotype, consistent with SLC6A14 being a modifier of disease presentation in CF patients.

3.2 Introduction

As discussed in section 1.3 of this thesis, one approach to discover modifier genes is to do a genomes-wide association study (GWAS)330. A recent large study of nearly 4000 CF patients considered the readily diagnosed gastrointestinal phenotype of MI. MI corresponds to an intestinal obstruction requiring immediate intervention and occurs in ~18% of CF cases. It is associated with severe CFTR mutations including the common F508del mutation. 181. This study looked at an easy to diagnose GI phenotype of meconium ileus (MI). In contrast to the progressive pulmonary phenotypes3, MI is a highly heritable (~80% heritability)181 and presents at birth, thereby engaging minimal environmental influence181. This study showed strong association between occurrence of MI and SNPs in the putative promoter region of SLC6A14 (minimum P = 1.28 × 10−12 at rs3788766). Consequent studies have demonstrated pleiotropic effect of SLC6A14 with the same risk alleles being associated with increased severity of lung disease as well as early age of onset of first P. aeruginosa infection in paediatric patients182,248.

3.2.1 Intestinal amino-acid transporters

There are several amino-acid transporters expressed in the large and small intestines namely: SLC16A10331,332; SLC7 family including SLC7A1,2,3 and 4/SLC3A2333, SLC7A6 and 7/SLC3A2333,334, SLC7A8/SLC3A2335, SLC7A9/SLC3A2333; SLC43A2336; SLC38 family337,338;

SLC36A1339; SLC36A4339; SLC6 family: SLC6A19340 and SLC6A14341. Of these, the SLC6 transporter expressed in the colon - SLC6A14 is unique in having a broad specificity and a sodium driven amino-acid transport mechanism, allowing it to concentrate amino-acids within cells342,343. In the intestine, amino-acid transporters are engaged in several roles including amino- acid absorption344, nitric-oxide production333, cell proliferation345, epithelial cell restitution maintainence346 and inflammation347.

Figure 3.2.1: Amino-acid transporters in the a) ileum and b) colon. The panels shows amino- acid transporters in the apical and basal membranes of ileal and colonic epithelia. (AA = amino- acid, superscripts: 0 is for neutral amino-acids, + for positively charged amino-acids, - for negatively charged amino-acids, zw for zwitterions).

3.2.2 SLC6A14 (ATB0,+)

3.2.2.1 Introduction

The SLC (Solute Carrier) families include about 300 transporter genes (HUGO – Human Genome Organization database), which encode channels, ion coupled transporters and exchangers. SLC6A14 is a member of SLC6 family and is maps on chromosome Xq23-24 (Pubgene, Gene ID: 11254). It is an electrogenic sodium and chloride-dependent neutral and basic amino acid transporter348. It transports two sodium ions, one chloride ion and a basic/neutral amino acid into the cell from the apical surface with each transport cycle348. It has a relatively high affinity for amino-acids (EC50 in the micromolar range)342, and it uses secondary

active transport for neutral/cationic amino-acid transport into the cell, against their concentration gradient342. SLC6A14 has the higher affinity for non-polar amino-acids (EC50 of 5-141 µM except proline), compared to polar amino-acids (EC50 38-695 µM). It can also transport D- isomers of amino-acids such as D-serine349, as well as L-carnitine350 and amino-acid based prodrug molecules351,352. It is known to be expressed in CF affected tissues including colon and lungs, and in oocytes, blastocyst and other organs341,353. As shown in Figure 3.2.2.1, in colonic epithelial cells, it is thought to be expressed at the apical membrane354.

Figure 3.2.2.1: SLC6A14 is expressed in CF affected tissues like colon and lungs. There is evidence that SLC6A14 is expressed on the apical surface of colonic epithelial cells354. (Graphic courtesy Dr. Stan Pasyk)

Currently, the SLC6 family consists of 20 structurally related symporters/co-transporters. They use the electrochemical gradient to transport substrate into the cell. Many of the members of the SLC6 family are neurotransmitter re-uptake transporters, and are drug targets for neuro- psychiatric disorders342. For example, SLC6A4, also known as SERT (Serotonin transporter) is a heavily studied transporter, and is important in various neuro-psychiatric disorders such as anxiety, depression, obsessive-compulsive disorder and ADHD (Attention deficit hyperactivity

disorder)355. The most commonly used antidepressant SSRI (Selective Serotonin Reuptake Inhibitors) binds to SLC6A4 and inhibits its transport activity356.

The SLC6A14 gene (Gene ID: 11254) is 24 kbp long and consists of 14 exons. The 4.579 kb mRNA transcript encodes a 642 amino-acid polypeptide. The upstream region of SLC6A14 gene, is known to have steroid binding elements among other transcription factor binding sites, and is sensitive to estrogen levels343. SNPs in this region show association to CF disease severity181.

The murine Slc6a14 gene (Gene ID: 56774) is 27 kbp long and also consists of 14 exons. The 3.433 kb mRNA transcript encodes for a 638 amino-acid polypeptide.

Structurally, SLC6A14 protein is predicted to have 12 transmembrane domains with a relatively large extracellular loop between TM3 and TM4. They are also predicted to have 8 N- glycosylation sites (proteinmodelportal.org). These predictions are based on the structure of Drosophila Dopamine Transporter (dDAT), with which it shares about 48% sequence identity357. Together with the structure of related bacterial transporter LeuT358, it is has been proposed that the binding of one Na+ atom to the substrate (amino-acid in this case), is important for binding to the substrate binding pocket in the protein342.

SLC6A14 is predicted to have 9 PKC phosphorylation sites359. However, the role of PKC mediated phosphorylation on the functional expression of SLC6A14 is controversial. PKC is known to enhance the surface expression and function of SLC6A14 in an over-expression system using HEK293 cells360. However, it is also shown that carnitine transport into the brain which is controlled by SLC6A14 expressed in the brain capillary endothelial cells, causes a decrease in its function upon PKC phosphorylation361.

3.2.2.2 SLC6A14 and tissue growth

From the physiologic perspective, SLC6A14 has a broad specificity for amino-acid transport343,362. Combining this with high concentrating ability of transporting amino-acids against their concentration gradients, SLC6A14 would provide amino-acid sufficiency and nutrient stimulated cell growth363. One of the important cellular signals for cell growth is the activation of mammalian target of rapamycin (mTOR)364, a complex which senses nutritional status of the cell and is important for cell proliferation363. The availability of nitrogen sources

like asparagine365, glutamine and leucine366, and their availabilities are key to regulation of mTOR345. Since, SLC6A14 can transport amino-acids into the cell, it can activate mTOR and thereby promote growth353,354. Similarly, deletion of Slc6A14 results in decreased mTOR signalling and decreased cell proliferation354.

The role of amino-acid transporters in cell proliferation and tissue growth has been established (Figure 3.2.2.2). Apart from SLC6A14, plasma membrane amino-acid transporters SLC1A5 and SLC7A5/SLC3A2 can also activate mTOR345. Amino-acid transporters not only help make important amino-acids available in the cytoplasm345, but can also activate mTOR directly367. The amino-acid availability sensing and downstream activation of mTOR is mediated by the lysosomal sodium coupled amino-acid transporter SLC38A9367,368. This cytosolic amino-acid sensing function mediated via SLC38A9 is dependent on there being adequate levels of arginine within the cell, to activate mTOR368.

Figure 3.2.2.2: Role of amino-acid (AA) transporters in mTOR activation. Plasma membrane amino-acid transporters like SLC6A14 and other amino-acid concentrating transporters (SLC1A5, SLC7A5, A8/SLC3A2 SLC36 and SLC38), bring amino-acids into the cell thereby maintaining cytosolic amino-acid availability. This sufficiency of amino-acids is sensed by mTOR in the lysosome, via the amino-acid transporter SLC38A9. (EAA = essential amino-acids, AA(0) = neutral amino-acids, Gln = Glutamine).

3.2.2.3 Disease associations with SLC6A14 There are several SNPs in SLC6A14 that have been associated with human diseases. As mentioned above, SNPs in SLC6A14 are associated with severity of CF features181,182,248. In context of the CF pathophysiology, there are several associations that relate SLC6A14 to phenotypic aspects of CF and other diseases, as follows:

(1) Activation of CFTR and upregulation of SLC6A14 are seen in the gut during the acute phase of Vibrio cholera infection leading to diarrhea369.

(2) Upregulation of SLC6A14 expression is seen in the rectal biopsies of patients with the chronic inflammatory condition of Ulcerative colitis370-372.

(3) SNPs in SLC6A14 are associated with changes in nutritional status in children373, and concerns for poor nutritional status remain as a serious morbidity for CF patients374.

(4) SNPs in SLC6A14 are associated with idiopathic infertility in Persian men375; a situation that may reflect the phenotype of male infertility in CF patients4.

Less obviously linked with CF aspects, other disease associations with SLC6A14 include:

(1) Relation to autism376 as SLC6A14 transports carnitine350,362, where inadequate carnitine levels in the brain lead to abnormalities of mitochondrial function.

(2) Relation to cancers as expression of SLC6A14 is found to be upregulated in epithelial tumors including those from colon377, breast343,354, pancreatic378 and cervical379 tissues.

An increased understanding of this amino-acid transporter could be beneficial for CF and many other diseases contributed to by SLC6A14.

3.2.3 Murine CF phenotype

The role of modifier genes in modifying CF disease severity has been studied extensively using CF mouse models155,209,268,269,380,381. Deletion of the Cftr gene or knock-in of the mutant F508del Cftr gene generates significant intestinal pathology146,153,382,383. CF mice have growth retardation compared to their Wt littermates, that has been attributed to malabsorption and decreased

secretion of IGF-1154,383,384. Histologically, the intestine of CF mice exhibits mucus accumulation, inflammation and goblet cell hyperplasia in the epithelial layers146,382, and circular smooth muscle hypertrophy in the muscularis externa156. This increase in smooth muscle thickness of the intestinal wall is variable in CF mice on different backgrounds156,157, and modifier genes have been attributed to these differences.

In CF mouse models, although failure to thrive occurs on most genetic backgrounds, Cftr- deficient mice live for many months on some backgrounds, and these differences in CF disease severity have been attributed to modifier genes155. Since, we hypothesized that deletion of SLC6A14 would worsen the CF phenotype, we used a CF mouse model with a mild CF phenotype, so that worsening of the disease could be studied. Towards this we utilized the F508del CF mouse on FVB background, which expresses a mild phenotype383. These mice have residual Cftr function as observed in Ussing chamber studies in ex vivo ileal and gall bladder epithelia, and in vivo using nasal potential difference383. They do not have lethal intestinal obstruction post-weaning. However, they do have growth retardation (F508del mice: 17.9 g ± 4.4; Non-CF Wt mice: 23.4 g ± 2.6), and a histological phenotype in the intestine, which is consistent with CF383. Taken together, the CF phenotype in the FVB strain is mild, making it amenable for the study of a modifier gene, which has a positive impact on CF.

To date, the biological role of SLC6A14 in modifying the CF phenotype has not been interrogated. The aim of the current study is to determine the impact of disrupting Slc6a14 expression in CF mice harbouring the major CF causing mutation: F508del.

3.3 Materials and Methods

3.3.1 Mice

All animal experiments were carried out under the animal use protocol (AUP) number 1000023276 at the Laboratory Animal Services (LAS) at the Hospital for Sick Children. The Slc6a14(-/y) on C57Bl/6N background was generated at the Toronto Center for Phenogenomics (TCP) under the KOMP2 project. This mouse line was backcrossed with the FVB strain obtained from Dr. Bob Scholte’s laboratory at Erasmus Medical Center Rotterdam, The Netherlands.

Mice were sacrificed with CO2 inhalation.

3.3.2 Slc6a14 mouse genotyping

DNA was extracted from clipped tails after overnight digestion with Proteinase K. Polymerase chain reaction (PCR) was performed using the KAPA HotStart mouse genotyping kit, for 35 cycles. 2 sets of primers were used for amplification - first set amplified the region in intron 5 of Slc6a14 gene which is present in Wt mice; second set of primers amplifies the lacZ cassette present in the Slc6a14 knock-out mice. The primer sequence of the two sets are in Table 3.3.1.

Table 3.3.1: Genotyping primers

Primer Set Forward (5’-3’) Reverse (5’-3’)

1 TTCAAGTCTCTCTAGCTTCAGGTC TTATCTGGTAGCTTCCTGTGACTC

2 CCATTACCAGTTGGTCTGGTGTC AAGGTGCTTATTTGAACTGATGG CGAGC

3.3.3 Quantitative real-time PCR

Murine organs were collected and immediately placed in RNAlater (Ambion) at room temperature. Samples were later stored at 4°C until further use. Within 1 week, samples were homogenized using Rotor-TissueRuptor (Qiagen) and mRNA was extracted using RNeasy® Plus Mini Kit (Qiagen) following listed instructions. After determining the purity and yield of the RNA spectrophotometrically, all samples were immediately stored at -80°C. Samples were used with a 260/280 ratio of 2.0-2.2 and concentrations higher than >100 ng/µL. 1µg of total mRNA from each sample was reverse transcribed into cDNA using iScriptTM cDNA Synthesis Kit (BioRad) per the manufacturer’s instructions. For real-time PCR (qRT-PCR), expression levels were measured using primers as listed in Table 2.3.2 on the CFX96 TouchTM Real-Time PCR Detection System (BioRad) using EvaGreen fluorophore (SsoFast EvaGreen Supermix with Low Rox, BioRad).

Table 3.3.2: Primers used for qRT-PCR

Gene Forward (5’-3’) Reverse (5’-3’)

Cftr CGGAGTGATAACACAGAAAGT CAGGAAACTGCTCTATTACAGAC

Tbp CAAACCCAGAATTGTTCTCCTT ATGTGGTCTTCCTGAATCCCT

Gusb CCGATTATCCAGAGCGAGTATG CTCAGCGGTGACTGGTTCG

Slc6a14 GCTTGCTGGTTTGTCATCACTCC TACACCAGCCAAGAGCAACTCC

3.3.4 Mouse colonic organoid culture

As described previously300,301,385, murine colonic tissue was obtained from mice on FVB or C57BL/6N background, aged 7-8 weeks old. 3-4 cm of colonic tissue was cut into pieces approximately 0.5 cm2 in size and washed with phosphate buffered saline (PBS). Colonic crypts were then incubated in PBS containing 10 mM EDTA for 30 minutes at 4⁰C. Colonic crypts were removed from the epithelium through mechanical stress and pelleted through centrifugation. Pelleted crypts were then washed once with ice cold PBS to remove excess debris from the primary colonic tissue. Pelleted crypts were re-suspended in 100% matrigel and seeded at approximately 10 crypts/μl of matrigel per well. The matrigel dome was allowed to dry for 30 minutes at 37⁰C. Once solidified, growth factor conditioned medium was added. Growth factor conditioned media was changed every second day for optimal organoid formation.

3.3.5 Transfection of organoids

Organoids were grown for 7 days as above and pelleted through centrifugation. Pelleted organoids were resuspended into 1 mL of Gentle Cell Dissociation Reagent (GCDR) (STEMCELL Technologies, Canada) and incubated at room temperature for 10 minutes. Organoids were then broken into fragments through vigorous pipetting with p1000 pipette and pelleted through centrifugation (500g for 5 minutes and 4°C). The pelleted organoid fragments were transfected with pcDNA 6.2 containing human SLC6A14-gfp with X-tremeGENE 9 (Sigma

Aldrich) for 30 minutes. Transfected fragments were pelleted through centrifugation and plated as described above with transfection reagent present for 24 hours after initial plating of the organoids. Organoids were cultured and imaged after 5 days. Prior to imaging, organoids were incubated with live cell nuclear dye, NucBlue® Live ReadyProbes® Reagent (ThermoFisher Scientific, USA) for 30 minutes at 37°C, and imaging was performed with Olympus Quorum Spinning Disk Confocal Microscope (SickKids Imaging Facility).

3.3.6 Studies of CFTR and F508del-CFTR mediated swelling of colonic organoids

Organoids were pre-incubated with a fluorescent live-cell marker, Calcein-AM (20μM) for 30 minutes and were removed from the matrigel matrix by washing the matrigel dome with ice cold buffer. Organoids were plated into 96-well plate and were centrifuged at 500g and 4⁰C to settle. Organoids were imaged for 5 mins to establish baseline. CFTR was stimulated with 1 μM forskolin (FSK). CFTR-mediated fluid secretion measured as FSK Induced Swelling (FIS) was captured by imaging for 30 minutes at 5 minute intervals, following the addition of FSK, with bright field and fluorescence microscopy (Nikon Epifluorescence/Histology Microscope), as previously described300. Analysis was performed using Cell Profiler v2.1. Fluorescent images extracted from FIS videos were exported as TIFF (Tagged Image File Format) files and a pipeline using CellProfiler (Carpenter Lab) was created to analyze the video files. Following manual thresholding (0.1) of background fluorescence and identifying and tracking primary objects (>100 organoids/well) within the diameter limit of 40-150 pixels, the CellProfiler pipeline generated each identified objects’ area, perimeter, diameter, and radius changes. The change in area (ΔA) relative to baseline (A0) was calculated, and the maximum change in ΔA/A0 within 30 minutes of FSK addition was used as a measure of FIS300.

3.3.7 Mass spectrometry

Whole tissue extracts were lysed using a buffer containing 50 mM Tris-HCl, pH 7.4, 150 mM NaCl, 1 mM EDTA, containing 0.1% (vol/vol) SDS, 0.1% (vol/vol) Triton X-100, 2% (vol/vol) protease inhibitor mixture (Amresco) and 1x phosphatase inhibitor cocktail (Roche) and then incubated at 4 °C for 15 min. Samples were centrifuged for 15 minutes at 18,000 xg and supernatants were collected for analysis, as previously described386. From these samples, amino

acids were quantified by LC/MS/MS at the Analytical Facility for Bioactive Molecules (The Hospital for Sick Children, Toronto, Canada).

3.3.8 IGF-1 level measurements

Serum samples were obtained from adult mice and kept at room temperature for 2 hours. The sera were collected following centrifugation and frozen immediately at -80°C until analysis.

The Quantikine ELISA Mouse/Rat IGF-1 Kit (MG100, R&D Systems, USA) was used, which utilizes a double-antibody sandwich enzyme-linked immunosorbent assay (ELISA) to determine the level of mouse insulin growth factor-1 (IGF-1). Assay procedure was followed as specified by the manufacturer. The absorbance was measured at 450 nm using the I3x-Spectrophotometer (Molecular Devices, USA). Data was analyzed using GraphPad Prism v6.01. The standard curve for the assay is shown in Figure 3.3.7.

Figure 3.3.7: Standard curve for ELISA based detection of IGF-1 levels. Purified IGF-1 was used generate the standard curve. Sigmoidal 4PL curve fit was used as recommended by the manufacturer. Goodness of fit as measured by R2 was 0.9999.

3.3.9 Histology analysis

Samples were fixed in formalin before being submitted for standard histological processing. Paraffin-embedded sections were stained with Alcian blue for mucous, at the Mount Sinai Toronto Center for Phenogenomics (TCP) Pathology Department. These slides were scanned (SickKids Imaging Facility) and sections were analyzed using 3DHISTECH software.

Annotations were drawn along the perceived borders of tissue layers (Figure 3.3.8). This was repeated for a total of 3 sections per slide per mouse. All the algorithms used by the software had fixed parameters for all replicates and conditions. Data was exported in excel file and statistics was performed using GraphPad Prism v.6.01.

Figure 3.3.8: Representative intestinal section with annotations made in 3DHISTECH software are shown. These annotations were used to calculate relevant areas of tissue layers.

3.3.10 Morphometric Measurements

Mice weights were measured (American Weigh Scales Black Blade Digital Pocket Scale) and tracked 2-3 times a week for 8 weeks, starting from day 9. Length of the body (tip of nose to base of tail - crown rump length) was also tracked for all mice.

3.3.11 In vivo fluid secretory assay

All animal studies were performed after ethics approval of Animal use protocol. C57BL/6N and FVB mice strains were used for the experiment. Protocol was adapted and modified from a previously described method387. Mice were fasted 15 hours before the experiment and given colyte + 5% dextrose. Next day, they were anesthetized using 2-3% of isoflurane for maintenance; up to 4-5% for induction, with oxygen from a precision vaporizer. Mice were given buprenorphine (0.01 mg/kg) as an analgesic before surgery. A heating pad was used to maintain

body temperature. Mice were placed on dorsal recumbent position throughout the procedure. The ventral abdominal wall was shaved using a surgical electric shaver. Antiseptic scrubbing of the abdominal wall was done with 70% ethanol followed by a solution of 10% povidone-iodine. After sterile draping, a ventral incision along the left midclavicular line was made through the skin, subcutaneous tissue and anterior abdominal muscle layers. This was followed by an incision through the peritoneum, to gain access to the peritoneal cavity. After visual inspection, caecum was identified and a ligature was placed, using a perma-hand silk (5-0; black braided; Ethicon) near the caeco-colic junction. Another ligature was made about 1 cm from the 1 st ligature, to form a closed loop of 1 cm in length. A 3rd and 4th ligature was made 1 cm from the 2nd ligation, to form 2 closed loops side by side. Care was taken to prevent damage to mesenteric vessels and ischemia of the intestinal loop. Loops were injected with test drugs dissolved in 50 uL of HBSS (without glucose). The test drugs included: control (DMSO), forskolin (10 μM) and IBMX (100 μM). Mice were properly sutured with coated Vicryl suture (5-0; undyed braided; Ethicon) and let recovered in a clean housing cage under a warm pad. Mice were monitored closely during recovery period, for any signs of distress. Mice were checked every 30 min during the 1-hour time, after which they were euthanized using CO2 and the loops were removed at the end of the 1 hour. Loop length and weight were measured to quantify fluid secretion.

3.3.12 Ex vivo closed loop amino-acid uptake assay

Mice were housed in a pathogen-free environment in the SickKids Laboratory Animal Services (LAS) facility. All experiments were performed under the SickKids Animal Care and Use Committee approved protocols. Male mice between the ages of 6-8 weeks were sacrificed. Colon or ileum of mice were isolated and three closed loops each measuring 1.5 cm were formed using silk sutures (Ethicon). Each loop was injected with 100 µl of buffer (25 mM HEPES, 140 mM sodium chloride, 5.4 mM potassium chloride, 1.8 mM calcium chloride, 0.8 mM magnesium sulfate and 5 mM glucose; pH 7.4; 300 mOsm) supplemented with 1 µCi/mL of L-[2,3-3H]- arginine (specific activity of 54.6 Ci/mmol), and 100 µM or 20 mM cold L-arginine.

After waiting 15 min, the loops were opened and the apical surfaces were flushed with ice-cold buffer (as above) supplemented with 20 mM arginine and lysed with 0.5M NaOH. Following lysis on ice for approximately 1 hour, samples were spun down at 18,000 x g for 10 minutes.

12.5 µL of each lysate was added to 2 mL of EcoScint A Scintillation Fluid (Diamed, Switzerland) along with the appropriate controls and counts were read using a Beckman Scintillation Counter (LS-6000IC). Total lysate protein was determined using a Bio-Rad protein assay. Amount of protein was determined by absorbance (at 595 nm) of a Coomassie Brilliant Blue G-250 dye. 10 µl of diluted protein samples (diluted 1:10 and 1:20 in ddH2O) was added to 200 µl of dye, and protein concentration was interpolated using a bovine serum albumin standard curve. Normalized arginine uptake was calculated by dividing scintillation counts per minute (CPM) by the determined protein concentration. The final calculation units used were in CPM per mg/mL of protein.

3.3.13 Statistical Analysis

One-way ANOVA with Tukey’s multiple comparison test was performed on all data with more than two data-sets for comparison, and SD or SEM was calculated using data from biological replicates. For mice survival data analysis, Log Rank test was used. Unpaired two-tailed t-test was performed on data with two data sets and a paired t-test was performed where the data was paired from littermates. P < 0.05 was considered statistically significant. Statistical analyses were performed using GraphPad Prism 6.01.

3.4 Results

3.4.1 Slc6a14 is a major apical amino-acid transporter in the colon

Quantification of the relative mRNA expression by qRT-PCR revealed that Slc6a14 is expressed predominantly in the mouse colon (Figure 3.4.1.1 a). In order to define SLC6A14 protein localization in colonic epithelium, we transfected mouse colonic organoids with SLC6A14-gfp and examined localization by confocal microscopy. We found that SLC6A14 was localized at the apical pole on the apical surface as expected (Figure 3.4.1.1 b). The subsequent series of experiments were designed to interrogate the biological role of SLC6A14 in mediating amino acid transport in the mouse colon using mice in which SLC6A14 expression was abolished. First, we confirmed that the Slc6a14 gene was disrupted and expression abrogated in the knock- out (KO) mouse (C57BL/6N) created by the NORCOMM Consortium (Figure 3.4.1.1 c,d). Slc6a14 is located on the X chromosome, with null male mice used for all analyses.

Figure 3.4.1.1: Expression of Slc6a14 in CF affected tissues from Wild type C57Bl/6 mice tissue (a) mRNA expression levels of Slc6a14 normalized to housekeeping gene Gusb and relative to its expression in the lung. Expression levels were measured in various CF affected tissues using qRT-PCR. Bars represent mean ± SEM. One-way ANOVA with Tukey’s multiple comparison test was performed (*P < 0.0001, n ≥ 3 mice). (b) Colonic organoids derived from a Wt C57BL/6N mouse were transfected with human SLC6A14-gfp, showing apical localization of SLC6A14. (c) For genotype confirmation, DNA was extracted from Wt and Slc6a14(-/y) mice tails and PCR was performed (35 cycles) with two sets of primers to indicate Wt (225 bp), and Slc6a14 Knock-out (386 bp) alleles. Both amplicons are evident in female heterozygous mice. (d) Slc6a14 mRNA expression in the colon of Wt and Slc6a14-KO mice as measured by RT- PCR with (+) and without (-) reverse transcriptase (RT) (30 cycles). Tbp was used as housekeeping gene.

There are many known apically expressed amino-acid transporters in the colonic epithelium341,344,388, hence, we asked if deletion of Slc6a14 had any impact on amino acid uptake across the apical membrane of this epithelium. To address this, we measured 3H-arginine uptake across the apical surface of the colonic epithelium, since arginine is an important substrate for epithelial health347,386,389,390. We used the ex-vivo closed loop assay to measure arginine uptake in these mice388. We performed this assay on the colon from C57BL/6N Wt and Slc6a14(-/y) mice. Interestingly, we found that the Slc6a14(-/y) mice exhibited a ~75% reduction in apical arginine transport (Figure 3.4.1.2 a) indicating that Slc6a14 constitutes a major apical arginine uptake pathway in the mouse colon.

Since, SLC6A14 played a major role in apical arginine flux, we asked if the steady state levels of arginine were affected in the Slc6a14 knock-out murine epithelium. We performed mass spectrometry using freshly lysed tissue and found that the steady state levels of arginine were indeed lower in the Slc6a14 knock-out than in wild type colonic epithelium (Figure 3.4.1.2 b). Thus, we found that Slc6a14 is a major apically expressed amino-acid transporter in the colon, that contributes to the maintenance of the steady state levels of arginine in the colonic epithelium.

Figure 3.4.1.2: SLC6A14 is a major arginine transporter on the apical colonic epithelium. (a) Ex vivo closed loop assay was performed after injecting buffer containing 3[H]-Arginine supplemented with 100 µM (Km of SLC6A14) or 20 mM (saturated levels) of cold Arginine.

After 15 minutes epithelium was lysed and intracellular 3[H]-Arginine levels measured. Bar graph represents mean ± SD of arginine uptake by the epithelium (counts per minute – CPM), normalized to total protein in the lysate (CPM/[protein]). Unpaired t-test was performed (**P = 0.0082; n = 3 biological replicates). (b) Bar graphs represent arginine bioavailability in freshly lysed colonic tissue – mean ± SD. Arginine bioavailability is defined as the ratio of arginine to citrulline plus ornithine391. Unpaired t-test was performed (*P = 0.047; n ≥ 3 biological replicates).

3.4.2 Disruption of Slc6a14 influence growth of FVB mice modeling the major CF mutation: F508del

To study the impact of Slc6a14 knock-out on the CF affected epithelium, we generated double mutant mice. It is known that the F508del mutation in mice results in intestinal features of varying severity, which is dependent on the genetic background of the mouse155. We hypothesized that Slc6a14 knock-out would worsen the CF phenotype, hence, we chose the Cftr(F508del/F508del) mouse on the FVB strain as it exhibits a relatively mild CF intestinal phenotype and would permit quantification of a deleterious effect383. We backcrossed the C57BL/6N Slc6a14 knock-out mouse to the FVB Cftr F508del heterozygous mouse, for >7 generations to generate a congenic mouse strain carrying a double mutation of Cftr(F508del/F508del) Slc6a14(-/y). The double mutant mice of the FVB strain exhibit Mendelian inheritance in this congenic strain (Table 3.4.2.1).

Table 3.4.2.1: Slc6a14 deletion in CF mice maintains Mendelian inheritance. Table shows observed and expected distribution of mice with different genotypes.

Genotype Observed Distribution (n=100) Expected Distribution Sex Cftr Slc6a14 Male F508del/F508del (‐/y) 7/100 1/16 (6.25%) Male F508del/F508del (+/y) 5/100 1/16 (6.25%) Male F508del/WT (‐/y) 13/100 1/8 (12.5%) Male F508del/WT (+/y) 15/100 1/8 (12.5%) Male WT/WT (‐/y) 6/100 1/16 (6.25%) Male WT/WT (+/y) 6/100 1/16 (6.25%) Female F508del/F508del (+/‐) 5/100 1/16 (6.25%) Female F508del/WT (+/‐) 12/100 1/8 (12.5%) Female F508del/WT (+/+) 14/100 1/8 (12.5%) Female F508del/F508del (+/+) 5/100 1/16 (6.25%) Female WT/WT (+/+) 8/100 1/16 (6.25%) Female WT/WT (+/‐) 4/100 1/16 (6.25%)

It has been previously reported that CF mice exhibit reduced mass relative to their wild-type siblings after the weaning period383 and we confirmed this defect in the current experiment at day 34 after birth in the Cftr(F508del/F508del) mice (Figure 3.4.2.1). As shown in Figure 3.4.2.1, disruption of Slc6a14 in CF mice (Cftr(F508del/F508del) Slc6a14(-/y)) led to a further reduction in weight gain and BMI post weaning (day 34), relative to age matched controls, (Cftr(F508del/F508del) Slc6a14(+/y)) (Figure 3.4.2.1a,b,c). This defect in weight gain was temporary in nature with the double mutant mice recovering to the same weight as the F508del mice by 56 days in age. Further, there was no difference serum amino-acid levels between any of the genotypes at this time (Figure 3.4.2.2, values in Appendix Table 3).

Figure 3.4.2.1: Disruption of Slc6a14 in F508del CF mice leads to decrease in weight gain post weaning. (a) Dorsal view of CF (CftrF508del/F508del) mice and double mutant (CftrF508del/F508del; Slc6a14(-/y)) mice. Both mice were Day 30 male mice. (b) Bar graph represents weights of Wt, CftrF508del/F508del, Slc6a14(-/y) and double mutant – (CftrF508del/F508del; Slc6a14(-/y)) mice at day 34 - mean ± SEM. Unpaired t-test was performed (*P = 0.0315; n > 5 mice for each genotype). (c) Bar graph represents Body Mass Index – BMI of mice at day 34 (mean ± SEM). Unpaired t-test

was performed (*P = 0.0261; n > 5 mice for each genotype). (d) The graph represents Kaplan- Meier survival curves for the four genotypes of mice. Double mutant mice were susceptible to death post-weaning. There was no statistical difference in the survival, between Wt and F508del- Cftr mice, However, the survival of double mutant mice was significantly lower than Slc6a14(-/y) mice (Log Rank test, P = 0.0036; Wt mice n = 14, Slc6a14(-/y) n = 15, CftrF508del/F508del n = 10, double mutant - CftrF508del/F508del; Slc6a14(-/y) n = 11).

Figure 3.4.2.2: Deletion of Slc6a14 in CF mice does not cause a change in serum amino-acid levels. Serum was collected from Wt, CF (CftrF508del/F508del), Slc6a14(-/y) and double mutant (CftrF508del/F508del; Slc6a14(-/y)) mice on FVB background. Stacked bar graph represents levels of essential amino-acids (a) or non-essential amino-acids (b) in the serum (mean ± SD). Two-way ANOVA with Tukey’s multiple comparison test was performed. No relevant significant changes were observed (n ≥ 4 for each genotype).

Interestingly, the transient defect in weight gain post-weaning measured in the double mutant mice occurred despite an increase in circulating levels of IGF-1 (Figure 3.4.2.3). Importantly- while the post-weaning period was not associated with increased mortality in the Cftr(F508del/F508del) mice (consistent with previous reports of this mice in the FVB background383, the double mutant mice did exhibit increased mortality in (Figure 3.4.2.1d) relative to their Wt littermates. Hence, Slc6a14 would maintain survival in CF mice bearing the major mutation at weaning.

Figure 3.4.2.3: Deletion of Slc6a14 in CF mice does not cause a change in serum IGF-1 levels. ELISA based IGF-1 detection assay was performed on serum samples obtained from 56 day old mice (a) or 30 day old mice (b), and actual concentrations interpolated from the standard curve. Bar graphs represent mean ± SD. Serum was collected from Wt, CF (CftrF508del/F508del), Slc6a14(-/y) and double mutant (CftrF508del/F508del; Slc6a14(-/y)) mice on FVB background. One-way ANOVA with Tukey’s multiple comparison test was performed (*P = 0.048, # P = 0.015, ** P = 0.0422, n ≥ 6 for each genotype for Day 56 mice, n ≥ 3 for Day 30 CF and double mutant mice).

3.4.3 Loss of Slc6a14 worsens defective F508del-Cftr mediated secretion in murine colonic epithelium

The mechanism of growth retardation in CF mice has been attributed to a number of defects including reduced IGF1 secretion384 and lipid metabolism392. It has been suggested that decreased fluid secretory capacity154,156,164,393 by intestinal epithelia of CF mice will also contribute to obstruction and morbidity, particularly at weaning153. Hence, we were prompted to determine if disruption of Slc6a14 led to worsening of this defect. We first confirmed that the in-vivo closed loop assay described by Verkman and colleagues387 is effective in reporting CFTR mediated fluid secretion in colonic tissue (Figure 3.4.3.1a). As expected, CFTR mediated fluid accumulation was decreased in segments derived from mice, homozygous for F508del-Cftr (Figure 3.4.3.1b). Interestingly, there was a negative effect of Slc6a14 disruption on net fluid accumulation after stimulation by cAMP agonists, forskolin and IBMX relative to Wt (C57BL/6N and FVB background) siblings (Figure 3.4.3.1b, c). This decrease in fluid secretion was not due to change in Cftr mRNA expression (Figure 3.4.3.1d).

Importantly, the impaired secretory capacity of the “loops” derived from CftrF508del/F508del mice was further impaired in mice lacking Slc6a14 expression (Figure 3.4.3.1b). Together, these studies suggest that Slc6a14 expression exerts a positive effect on Wt-CFTR and residual F508del-CFTR function in fluid secretion.

Figure 3.4.3.1: Loss of Slc6a14 worsens defective F508del Cftr mediated secretion in murine colonic epithelium. (a) In vivo closed loop assay performed on Wt mice. Each loop was injected with CFTR cAMP agonist forskolin (FSK 10 µM) and IBMX (100 µM), or DMSO vehicle. Weight relative to length was determined for both the loops and used as a measure of fluid

secretion. (b) Bar graph represents fluid secretory capacity of in vivo colonic closed loops stimulated with Cftr cAMP agonist FSK (10 µM) and IBMX (100 µM), relative to DMSO alone (mean ± SD). Experiments were performed on Wt and Slc6a14(-/y) mice on C57BL/6N background. Fluid secretion is represented as weight/length (w/l) for each loop stimulated with FSK and IBMX, normalized to w/l for vehicle DMSO. Paired t-test was performed (*P = 0.0125, n = 4 mice for each genotype). (c) Bar graph represents fluid secretion in loops stimulated with CFTR cAMP agonist FSK and IBMX, relative to DMSO alone (mean ± SD). Fluid secretion is represented as weight/length for each loop. Wt mice showed significantly higher fluid secretion than CftrF508del/F508del (unpaired t-test, P = 0.0036, n = 3 for each genotype). Double mutant (CftrF508del/F508del; Slc6a14(-/y)) mice showed significantly lower fluid secretion than CF CftrF508del/F508del mice - unpaired t-test was performed (*P = 0.0428, n≥3 biological replicates for each genotype). (d) Bar graph depicts Cftr mRNA expression relative to housekeeping gene Tbp, from freshly lysed colon of Wt and Slc6a14(-/y) mice on C57BL/6N background (mean ± SD). Unpaired t-test was performed (ns = not significant, n = 4 for each genotype).

We pursued validation of these findings using organoids 300,385, a stem cell derived model that reports CFTR mediated fluid secretion 394,395. We generated colonic organoids from double mutant and CftrF508del/F508del mice, and studied CFTR function in these organoids using the forskolin induced swelling (FIS) assay 300. As expected from the above studies, the FIS response was significantly lower in the colonic organoids from double mutant mice (CftrF508del/F508del; Slc6a14(-/y)), compared to the F508del-CF mice (Figure 3.4.3.2a, b).

Figure 3.4.3.2: Loss of Slc6a14 worsens defective F508del Cftr mediated secretion in murine colonic organoids. (a) Representative fluorescence images of murine colonic organoids derived

from CftrF508del/F508del and double mutant – (CftrF508del/F508del; Slc6a14(-/y)) mice, and their responses after 30 minutes of stimulation with FSK (1 µM). (b) Bar graph represents FSK induced swelling (FIS) after 30 minutes of stimulation, in both CftrF508del/F508del and double mutant – (CftrF508del/F508del; Slc6a14(-/y)) murine organoids (mean ± SD). FIS is measured as change in area of the organoid after 30 minutes of FIS (ΔA) relative to baseline before stimulation (A0). Unpaired t-test was performed (****P < 0.0001, n > 4 biological replicates for each genotype).

Histological analyses of the large intestine confirmed previous reports of increased smooth muscle thickness of intestinal segments of CftrF508del/F508del mice relative to Wt- mice 154,156,164,393.This increase in smooth muscle thickness is thought to be secondary to the primary defect in fluid secretion. Therefore, we reasoned that the worsened secretory defect caused by disruption of Slc6a14 in the CftrF508del/F508del mice would be associated with even greater increase in smooth muscle thickness. The smooth muscle thickness in sections from CftrF508del/F508del; Slc6a14(-/y) mice was significantly enhanced in both the distal ileum and the colon relative to sections from CftrF508del/F508del mice (Figure 3.4.3.3). Together with the studies of secretion above, these findings support our hypothesis that Slc6a14 facilitates normal intestinal secretion and smooth muscle morphology.

Figure 3.4.3.3: Slc6a14 disruption results in an increase in smooth muscle thickness in the ileum and colon. (a) Representative distal ileal sections stained with Alcian blue, for Wt, CftrF508del/F508del and double mutant – (CftrF508del/F508del; Slc6a14(-/y)) mice. (b) Representative colonic sections stained with Alcian blue, for Wt, CftrF508del/F508del and double mutant – (CftrF508del/F508del; Slc6a14(-/y)) mice. (c) Bar graph represents ileal smooth muscle area relative to

total section area for each genotype (mean ± SD). One-Way ANOVA with Tukey’s multiple comparison test was performed (**P = 0.007, **** P <0.0001, n ≥ 4 mice for each genotype). (d) Bar graph represents colon smooth muscle area relative to total section area for each genotype (mean ± SD). One-Way ANOVA with Tukey’s multiple comparison test was performed (*P = 0.042, n ≥ 4 mice for each genotype).

3.5 Discussion

As described in section 3.2.1, multiple amino-acid transporters are expressed in the colon. So, we investigated the functional significance of SLC6A14. We found that SLC6A14 is a major apical amino-acid transporter in the colonic epithelium responsible for apical arginine flux (Figure 3.4.1.2a) as well as maintaining the steady state levels of arginine (Figure 3.4.1.2b) in the epithelium.

SLC6A14 has been identified as a genetic modifier of multiple CF disease outcomes, including meconium ileus susceptibility181, lung disease severity182 and the age of first infection by P. aeruginosa248. This pleiotropy supports that SLC6A14 modifies the basic defect responsible for disease in multiple organs, namely deficiency in anion conduction and fluid secretion. The work in this study focused on the impact of disrupting Slc6a14 on the CF phenotype in the mouse intestine because this phenotype has been well described by multiple laboratories153-155,383,396. In support of the above hypothesis, we found that the intestinal phenotypes associated with the primary CF defect in mice, namely, defective secretory capacity with thickened smooth muscle were worsened by disruption of Slc6a14 expression. An alternate model organism or system, that better recapitulates CF airway disease would be necessary to study airway and lung effects, although parallels can be drawn given a similar pathophysiology of fluid secretory defect in the human airway epithelium18.

Interestingly- disruption of SLC6A14 expression in mice bearing Wt-CFTR, did not cause profound changes in intestinal function and mouse survival. These investigations are consistent with the findings reported by Ganapathy and colleagues indicating that in an independently derived Slc6a14(-/y) mouse, the loss of the functional allele did not lead to obvious changes in mouse health354. However-disruption of Slc6a14 did impair arginine uptake across the apical membrane of colonic epithelium in Wt-Cftr mice (Figure 3.4.1.2) and impaired cAMP mediated fluid secretion (Figure 3.4.3.1c). These findings suggest that Slc6a14 expression exerts a

modulatory role for Cftr mediated fluid secretion and this modulatory role only becomes essential for normal growth and intestinal morphology in the context of function- reducing mutations of Cftr.

All of our studies of the CF phenotype employed transgenic mice engineered to express the F508del-Cftr mutation. In contrast to the consequences of the F508del mutation in the human CFTR protein, the negative consequences of this mutation in folding and assembly of the mouse CFTR protein are somewhat less severe, hence there is residual epithelial expression and function being detected in this organism153,383,396. As previously published, this reflects a less deleterious effect of the mutation on the folding, assembly and maturation such that a significant amount of the mutant mouse protein will reach the surface and function albeit- this function is reduced compared to Wt-CFTR153,155,383,397. This residual function was detected as cyclic AMP dependent fluid secretion in colonic organoids derived from F508del mice (Figure 3.4.3.2). Although we show a modulatory role of Slc6a14 in F508del-Cftr mediated swelling in mouse derived organoids –our future studies will focus on evaluating this modulatory role in CF patient- derived organoids and predict that this will depend, in part, on the extent of residual F508del- CFTR and SLC6A14 expression in each patient-derived tissue.

Chapter 4 Mechanistic insight into SLC6A14 mediated modulation of CFTR anion conductance

Content presented in this chapter has been submitted for publication:

Saumel Ahmadi, Sunny Xia, Yu-Sheng Wu, Michelle Di Paola, Randolph Kissoon, Catherine Luk, Fan Lin, Kai Du, Johanna M. Rommens and Christine E. Bear.

Genetic modifier of Cystic Fibrosis regulates the primary defect in fluid secretion.

Contributions: Data was generated and analyzed by Saumel Ahmadi, with help of the authors of the abovementioned research article. Study was designed by C.E.B., S.A. based on discussions with J.R., C.L., S.X., M.D., Y-S.W. and R.K. Murine organoids were generated by S.X., mouse genotyping was performed by C.L., qRT-PCR was performed by M.D., Y-S.W., F.L. live cell fluorescence was performed by S.X. and K.D., nitric-oxide measurements were performed by Y- S.W. Manuscript was written by C.E.B. and S.A. with inputs from all the authors.

4.1 Summary

SLC6A14 is a genetic modifier of CF which is apically expressed on the colonic epithelium. We found that it is a major apical transporter for arginine, and enhances cAMP mediated fluid secretion in the colonic epithelium. So, we hypothesized that SLC6A14 could modulate CFTR anion conductance. To test this, we developed a novel method to assess the functional expression of apical constituents in murine organoids. This allowed us to show that SLC6A14 indeed facilitates CFTR chloride conductance, and this is caused by an increase in nitric-oxide mediated PKG phosphorylation of the mutant F508del CFTR protein. Taken together, we define the biologic role of SLC6A14 in the CF affected epithelium.

4.2 Introduction

CF is caused by mutations in the CFTR gene14,15, which results in decreased functional expression of the CFTR anion channel1,120. The CF intestinal phenotype of impaired fluid secretion is evident in the murine CF mice153 leading to the development of the in vitro organoids (mini-guts) to model385 and test the effect of CFTR modulators300,301. The generation of organoids allows a relatively indefinite propagation of epithelial cultures398. Recently, it has also been hypothesized that drug responses in these organoids in vitro can predict patient specific responses301. We thus used the murine organoid model to study mechanism by which Slc6a14 could modify the fluid secretory capacity of the CF epithelium.

Being an apically expressed channel, CFTR can mediate vectoral fluid transport in epithelia from basal to apical surface300. In the context of a three-dimensional organoid, the apical surface forms a relatively closed compartment, which can expand upon CFTR stimulation with cAMP agonists, forming the forms the basis of the Forskolin induced swelling (FIS) assay and provide a measure of CFTR function300.

As described in section 1.5, there are several genetic modifiers of CF, many of which are co- expressed with CFTR at the apical surface. To directly modulate these apically localized proteins, we developed a method to gain apical access by splitting open the organoids. This allowed us to study the effect of the genetic modifier of SLC6A14 on CFTR chloride conductance.

4.2.1 Arginine nitric-oxide pathway

L-arginine is considered to be a semi-essential amino-acid399 in adults, wherein its endogenous production is adequate in healthy states, but it becomes an essential amino-acid in catabolic states like trauma and inflammation400 among others399,401. Apart from being important from the nutritional perspective, L-arginine is also been used as a supplement399, and has shown benefit in CF402,403, Congestive Heart Failure (CHF)404, pulmonary hypertension405, in growth hormone secretion406, burn patients407, senile dementia408 and appears to improve insulin sensitivity in diabetes mellitus409. The major molecular mechanism which rationalized the use of arginine relates to its role in the production of nitric-oxide399,410,411.

To produce sufficient nitric-oxide, the cell has to transport arginine from extracellular compartment via membrane transporters333,347. This coupling of one or more membrane arginine transporter to the metabolic process of NO production is thought to constitute individual membrane transport metabolons412.

The conversion of arginine to nitric-oxide (NO) is dependent on the constitutively expressed NOS enzymes including neuronal NOS (nNOS or NOS1), endothelial NOS (eNOS), and on the inducible NOS enzyme (iNOS)413. The intestine expresses all three genes414, however under induced conditions such as inflammation415, NO production via iNOS can far exceed that of nNOS and eNOS416. In the context of CF, it has been shown that the exhaled concentrations of NO in the breath of CF patients is low417. This has been attributed to an increase in the activity of the enzyme arginase, which causes substrate depletion for NO production389. The arginase and NOS pathways are depicted in Figure 4.2.1.

Figure 4.2.1: Arginine nitric oxide metabolism. Membrane arginine transporters bring arginine into the cell, which is then converted into nitric-oxide by NOS (mainly iNOS for inflamed epithelial tissues)415. Alternatively, arginine can be converted by the enzyme arginase into ornithine and polyamines389.

4.2.2 Rationale and Hypothesis

The role of arginine nitric-oxide pathway in CF has been well studied. CF patients have been shown to have a lower FeNO (Forced expiration of Nitric-oxide)417, which can be enhanced by inhalation of arginine402. Also, at the molecular level, it is known that nitric-oxide mediated PKG phosphorylation can modulate CFTR channel function59,88,98. Given that, SLC6A14 is a major arginine transporter in the intestinal epithelium, we tested the hypothesis that its activation could enhance epithelial nitric-oxide levels, thereby increasing CFTR channel function (Figure 4.2.2).

Figure 4.2.2: Hypothesis of SLC6A14 contribution to CFTR function. Model depicting the hypothesis that SLC6A14 mediated arginine uptake across the apical surface of the epithelium would lead to an increase an intracellular nitric-oxide levels, thereby causing an increase in cGMP levels, which in turn would increase CFTR channel function via Protein Kinase G (PKG) mediated phosphorylation.

4.3 Materials and Methods

4.3.1 Mice

All animal experiments were carried out under the animal use protocol (AUP) number 1000023276 at the Laboratory Animal Services (LAS) at the Hospital for Sick Children. The Slc6a14(-/y) on C57Bl/6N background was generated at the Toronto Center for Phenogenomics (TCP) under the KOMP2 project. This mouse line was backcrossed with the FVB strain (Cftrtm1Eur; www.informatics.jax.org) obtained from Dr. Bob Scholte’s laboratory at Erasmus 383 Medical Center Rotterdam, The Netherlands . Mice were sacrificed with CO2 inhalation.

4.3.2 Generation of split open organoids

Organoids were generated as previously described300,385 and as in section 3.3.3. These organoids were suspended in 10 ml of DMEM/F12 (Dulbecco’s Modified Eagle Medium with nutrient

mixture F12) medium and spun down at 500 x g for 5 minutes. The pelleted organoids were then resuspended in conditioned media (containing Wnt, Noggin, Rspondin, EGF, FGF10, Y-27632, N-Actylcysteine, B27, N2, A-83-01, Nicotinamide)300 and plated on 96 well clear bottom plates coated with Poly-L-Lysine. Briefly, 0.01% L-Lysine solution was added to each well (40 µL/well) with shaking for 5 minutes with an orbital shaker at room temperature. The excess fluid was then removed and wells were washed with PBS (Phosphate Buffered Solution). The treated plates were then allowed to dry in a sterile environment for 2 hours. Thereafter, 200 µL of the resuspended organoid culture was pipetted into each well. After 24 hours (Day 2), 100 µL of media was removed from each well and replaced with 100 µL of fresh conditioned media. On Day 3 fluorescence based assays for Nitric-oxide production or CFTR channel function were performed using DAF-FM or ACC assay respectively.

4.3.3 Apical Chloride Conductance (ACC) assay for CFTR in “split-open” organoids

Murine organoids were generated as previously described300,385. After growth on poly-lysine coated plates for 48 hours, in absence of Matrigel, split-open murine intestinal organoids had grown upright as two dimensional (2D lawns) giving access to their apical membranes (Figure 4.3.3). FLIPR based ACC assay could then be performed essentially as previously descirbed305,418. Briefly, 0.5 mg/ml of the blue membrane potential dye, was dissolved in buffer containing – 112.5 mM NMDG- Gluconate, 36.25 mM NaCl, 2.25 mM K.Gluconate, 0.75 mM KCl, 0.75 mM CaCl2, 0.5 mM MgCl2 and 10 mM HEPES; Osmolarity 300 mOsm, pH 7.35. After 30-minute incubation, the plate was transferred to a micro-plate reader (Molecular Devices Paradigm). Fluorescence was read using excitation wavelength of 530 nm and emission wavelength of 560 nm. Multiple points in each well were read, and after capturing at least three baseline reads, pharmacological modulators were added (2.5 µL/well). Data was analyzed using previously described algorithm418, and statistical analyses were performed using GraphPad Prism v6.01.

Figure 4.3.3: Split-open organoid concept. Diagram depicts the concept of gaining apical access to the epithelium by splitting open a 3D organoid, there by resulting in patches of split- open 2D lawns, which can then be studied using fluorescence based assays.

4.3.4 Nitric oxide measurements in “split-open” organoids

Colonic organoids from mice were grown in a three-dimensional matrix and then acutely (72 hours) split-open in 96 well plates as described above. The organoids were incubated in DMEM- Dulbecco’s Modified Eagle Medium (ThermoFisher Scientific, USA) with 1mM L-Arginine (Sigma-Aldrich, USA). To prepare the organoids for nitric oxide measurements, 200 µL of NaCl

buffer (145 mM NaCl, 3 mM KCl, 3 mM CaCl2, 2 mM MgCl2, 10 mM HEPES; pH 7.35 osmolarity 300 mOsm) containing 7 µM DAF-FM Diacetate (4-Amino-5-Methylamino-2',7'- Difluorofluorescein Diacetate, Cayman Chemical, USA) and 10 µM Calcein Blue AM (ThermoFisher Scientific, USA) was added to each well for 1 hour. The organoids were then washed 3 times with phosphate buffered saline (PBS) to wash out excess dye. The organoids were then kept in the aforementioned NaCl buffer at 37°C for 20 min. Fluorescence was measured using the i3x-Spectrophotometer (Molecular Devices, USA), with excitation and emission wavelengths of 322 nm and 435 nm respectively. After reading baseline fluorescence, 1mM L-Arginine was added to acquire fluorescence on live NO production. The fluorescence measurements were expressed as the change in fluorescence (ΔF) relative to the fluorescence measurement just before Arginine addition (F0), after applying a thresholding function described previously418 to select for regions in the well which contain cells. To account for the heterogeneity of the organoid cultures, final readout of DAF-FM fluorescence (Excitation wavelength 490 nm/Emission wavelength 515 nm) was calculated over Calcein Blue AM

86 fluorescence. Each condition was repeated with at least 4 technical replicates on the sample plate and a total of at least 4 biological replicates per condition. Analysis was performed as previously described418.

Figure 4.3.4: Standard curve for Nitric-Oxide (NO) measurement. Split-open organoids from Wt mice were used to measure NO levels with DAF-FM fluorescence as described in the method. Increasing levels of NO in the epithelium were achieved by addition of known NO donor (Proli NONOate). One-phase association was used to fit the data. Goodness of the fit was R2 = 0.868 (n = 3 biological replicates).

4.3.5 Statistical analysis

As described in the previous chapter, One-way ANOVA with Tukey’s multiple comparison test was performed on all data with more than two data-sets for comparison, and SD or SEM was calculated using data from biological replicates. Unpaired two-tailed t-test was performed on data with two data sets and a paired t-test was performed where the data was paired from littermates. P < 0.05 was considered statistically significant. Statistical analyses were performed using GraphPad Prism 6.01.

4.4 Results

4.4.1 Direct measurement of CFTR channel function in split-open organoids

The culturing of three-dimensional intestinal organoids in the absence of non-epithelial niches, has allowed the development of self-renewing epithelium385,398 sources, which have widespread applications419. Although, these organoids represent the normal intestinal epithelium385, accessibility to inward facing apical surface remains limited particularly for study of apical membrane constituents420. This has led to the development of complete 2D monolayers from these organoids, enabling study of apical transporters and bacterial-host ineraction394,420-422. Using a modification of this strategy, we generated 2D split-open lawns from these organoids instead of complete monolayers, to gain apical access. These lawns are non-confluent, however CFTR function could be measured in them using a previously described membrane potential assay418 (details in chapter 2). As shown in Figure 4.4.1, both Wt and mutant F508del CFTR function can be measured in these split-open organoids.

Figure 4.4.1: CFTR channel function in split-open murine organoids. (a) Line graph represents change in ACC fluorescence from baseline (ΔF/F0), as a measure of CFTR channel function. Upon cAMP stimulation (FSK 10 µM), CFTR stimulation is seen, more in Wt than mutant F508del CFTR, at 37 °C. The grey hatched line represents F508del CFTR function after low temperature rescue (27 °C). This increase in CFTR function could be inhibited by addition of CFTR specific inhibitor CFTRinh-172. (b) Bar graph represents maximum change in ACC fluorescence from baseline (ΔF/F0) after acute addition of FSK, in split-open murine organoids

(mean ± SEM). One-way ANOVA with Tukey’s multiple comparison test was performed (*P ≤ 0.0001, n ≥ 3 biological replicates for each genotype).

4.4.2 Slc6a14 is important for epithelial NO production

We first confirmed that Slc6a14 is a major apical arginine transporter in the colonic epithelium of Wt and F508del mice on the FVB background Cftrtm1Eur (Figure 4.4.2.1). Hence, our next goal was to determine if the loss of an arginine transporter would modify the channel function of CFTR, given our understanding of arginine and NO synthase, and the ability of NO to stimulate guanylate cyclase and PKG mediated phosphorylation (Figure 4.2.2), which in turn is known to stimulate CFTR function80,88,423. We investigated the hypothesis that if loss of Slc6a14 – mediated arginine transport affected steady state levels of NO in 2D colonic epithelial tissue. In these studies, a fluorescence based assay was used to measure intracellular NO levels (DAF-FM- AM). First, we generated colonic organoids from Wt mice, and used the split-open method to gain apical access (Figure 4.3.2). The details of this method are provided in section 4.3.2 but briefly, after dissociation of the cysts from the matrigel and plating in liquid culture on poly-L- lysine coated 96 well plates- the cysts open up to expose the apical membrane.

Figure 4.4.2.1: Slc6a14 is a major apical arginine transporter in the colonic epithelium. Ex vivo intestinal closed loop assay was performed in mice on FVB background across four genotypes: Wt (Cftr(+/+)), CF (CftrF508del/F508del), Slc6a14(-/y) and double mutant (CftrF508del/F508del; Slc6a14(-/y)). Each loop for each mouse was injected with buffer containing 3[H]-Arginine supplemented with 100 µM or 20 mM cold Arginine. Bar graph represents arginine uptake by the epithelium (counts per minute – CPM), normalized to total protein in the lysate

(CPM/[protein]) - mean ± SD. Two-way ANOVA with Tukey’s multiple comparison test was performed for colon and ileum (**P = 0.0013, *** P = 0.0007, **** P <0.0001, ns = not significant, n ≥ 4 for each genotype of mouse).

A standard curve for change in DAF-FM fluorescence with increasing concentrations of NO was generated (Figure 4.3.3). As shown in Figure 4.4.2.2a, colonic epithelial tissue derived from Slc6a14 knock-out mice exhibited constitutively lower NO levels. Further, there was significantly less increase in NO production evoked by acute addition of arginine (1 mM) to the apical surface of split-open colonic organoids obtained from Slc6a14(-/y) mice relative to their Wt littermates (Figure 4.4.2.2b, c). Altogether, these data show that Slc6a14 mediated arginine uptake is important for NO synthesis and maybe able to affect CFTR function.

Figure 4.4.2.2: Slc6a14 contributes to NO production in colonic epithelial cells. (a) Epithelial NO levels were measured using DAF-FM fluorophore, in split open colonic organoids derived from Wt and Slc6a14(-/y) mice. Bar graph represents mean ± SEM. Unpaired t-test was performed (*P = 0.022, n = 5 biological replicates). (b) Line graph represents change in DAF-FM fluorescence from baseline (ΔF/F0) upon acute addition of L-arginine (1 mM), in split-open colonic organoids derived from Wt and Slc6a14(-/y) mice. Two-way ANOVA with Sidak’s multiple comparison test was performed (*P < 0.05, n = 5 for each genotype, for t = 15 min P = 0.0191, for t = 20 min P = 0.0216, for t = 25 min P = 0.0121, for t = 30 min P = 0.0103) (c) Bar graph represents maximum change in DAF-FM fluorescence from baseline (ΔF/F0) upon acute addition of L-arginine (1 mM), in split-open colonic organoids from Wt and Slc6a14(-/y) mice (mean ± SD). Unpaired t-test was performed (*P = 0.0133, n = 10 biological replicates for each genotype).

4.4.3 Slc6a14 enhances F508del-CFTR mediated chloride conductance

To test the hypothesis that SLC6A14 enhances mutant CFTR mediated anion conductance, we used the split-open organoid model. This model, enables direct assessment of CFTR channel function and CFTR mediated apical membrane depolarization using fluorescence-based assays102,305,418. The apical chloride conductance (ACC) through forskolin-activated CFTR can be measured using the ACC method as previously published418, and as described in detail in Chapter 2 of this thesis. Colonic organoids from F508del mice were split open and CFTR channel function measured using the ACC assay, after 24 hour incubation of cultures at 27° Celsius, just prior to the assay (Figure 4.4.3a,b). This low temperature rescue at 27° Celsius, is an experimental maneuver, well known to enhance the trafficking of F508del to the cell surface51. Importantly, we found that acute pre-addition of L-arginine (1 mM) on the apical surface, increased the residual function of F508del-CFTR in the Slc6a14(+/y) mice at 27° C (Figure 4.4.3a and b). This arginine dependent effect was not apparent in cells obtained from double mutant mice (Figure 4.4.3).

Figure 4.4.3: Slc6a14 activation with L-Arginine potentiates F508del CFTR function. (a) Split-open colonic organoids from CF (CftrF508del/F508del) and double mutant (CftrF508del/F508del; Slc6a14(-/y)) mice were studied for CFTR channel function using the previously described membrane potential based ACC (Apical Chloride Conductance) assay. Line graph represents change in fluorescence relative to baseline (ΔF/F0) as a measure of F508del- CFTR function after low temperature rescue (27 °C) of the mutant protein. After capturing baseline fluorescence

reads, cells were acutely treated with L-Arginine (1 mM) to activate SLC6A14 or vehicle, followed by CFTR activation with cAMP agonist Forskolin (FSK – 10 µM) or vehicle DMSO. Thereafter, CFTRinh-172 (10 µM) was added to all the wells. (b) Bar graph represents maximum change in ACC fluorescence from baseline (ΔF/F0) after acute addition of FSK, following low temperature (27 °C) rescue of F508del- CFTR protein in split-open murine organoids (mean ± SEM). Paired t-test was performed (*P = 0.045, ns = not significant, n = 4 mice for each genotype).

4.4.4 Loss of Slc6a14 and arginine-mediated NO generation contributes to worsening of defective epithelial fluid secretion

Since we used the FVB mouse strain with Cftr(F508del alleles, we anticipated exhibiting a mild intestinal phenotype, and also tested mutant F508del-CFTR channel using the ACC assay at 37° C. Consistently, residual F508del-CFTR function was evident upon FSK stimulation. As above, this residual mutant CFTR function, was enhanced by pre-addition of L-arginine on the apical surface in the Cftr(F508del/F508del) cells, but not in the double mutant Cftr(F508del/F508del); Slc6a14(-/y) cells (Figure 4.4.4.2a, b). Since iNOS is expressed in the murine intestinal epithelium (Figure 4.4.4.1), we hypothesized that the L-arginine mediated enhancement of mutant CFTR channel function occurred via nitric-oxide-PKG mediated CFTR activation. We tested this by measuring F508del-CFTR function in split-open murine organoids after pre-treatment with iNOS (inducible nitric oxide synthase) inhibitor 1400W (Cayman Chemicals, 100 µM) at 37° C. We found that mutant CFTR could be activated by FSK in presence of 1400W, but no arginine mediated enhancement of F508del-CFTR function occurred (Figure 4.4.4.2c, d). Together with the data in Figure 4.4.2.2, these findings suggest that Slc6a14 –mediates arginine uptake in colonic epithelia, enhances NO synthesis and cyclic-AMP dependent activation of F508del-CFTR.

Figure 4.4.4.1 Murine colonic epithelial tissue expresses inducible nitric oxide synthase (iNOS). (a) PCR amplification for detecting iNOS expression was performed, showing that iNOS is expressed in murine colonic tissue.

Figure 4.4.4.2: Slc6a14 causes NO mediated enhancement of F508del CFTR channel function (a) ACC assay performed on split-open colonic organoids from CF (CftrF508del/F508del) and double mutant (CftrF508del/F508del; Slc6a14(-/y)) mice for CFTR channel function at physiological temperature (37 °C). As above, Slc6a14 was activated with L-arginine (1 mM) or vehicle followed by CFTR stimulation by FSK (10 µM) or vehicle DMSO. All wells received CFTRinh-172 (10 µM) as indicated. (b) Bar graph represents maximum change in ACC fluorescence from baseline (ΔF/F0) after acute addition of FSK, at physiological temperature (37 °C) in F508del-CFTR split-open murine organoids (mean ± SEM). Paired t-test was performed (****P <0.0001, ns = not significant, n = 3 mice for each genotype). (c) Split-open colonic organoids from CF (CftrF508del/F508del) mice were studied using the ACC assay at physiological temperature (37 °C), with an iNOS inhibitor. Line graph represents change in fluorescence

relative to baseline (ΔF/F0) as a measure of F508del- CFTR, after pre-incubation with vehicle or iNOS inhibitor 1400W (100 µM for 30 minutes before starting fluorescence reads). (d) Bar graph represents maximum change in ACC fluorescence from baseline (ΔF/F0) after acute addition of FSK, at physiological temperature (37 °C) in F508del-CFTR split-open murine organoids (mean ± SEM), after 30-minute pre-incubation with vehicle or iNOS inhibitor 1400W (100 µM). (**P = 0.006, ns = not significant, n = 4 biological replicates for each genotype).

Next, we were prompted to determine if the reduced F508del-CFTR function measured in the double mutant mice could be overcome by bypassing the need for arginine uptake and directly modifying F508del-CFTR protein by protein kinase G dependent phosphorylation, a signaling pathway that is stimulated by NO. The ACC assay was used to measure F508del-CFTR function in colonic split-open organoids from the double mutant mice, after pre-treatment with 8B-cGMP (Sigma-Aldrich) or vehicle (control). At 27 °C, pretreatment of split open organoids from double mutant mice with cGMP, enhanced forskolin-activated F508del-CFTR channel activity compared to that observed in tissues from F508del mice with native Slc6a14 expression (Figure 4.4.4.3a,b). Importantly, this partial rescue effect induced by the cGMP analog pretreatment was recapitulated at 37° C in the organoid swelling assay (Figure 4.4.4.3c,d). Specifically, pretreatment of organoids from double mutant mice with cGMP increased forskolin induced swelling (FIS). Together, these findings support the hypothesis that the arginine transport activity of SLC6A14 and its downstream signaling via cyclic GMP normally contributes to the regulation of F508del-CFTR channel activity.

Figure 4.4.4.3: Mutant F508del- CFTR function in double mutant murine organoids can be enhanced by enhancing PKG mediated phosphorylation. (a) Split-open organoids from congenic double mutant (CftrF508del/F508del; Slc6a14(-/y)) mice were used to perform ACC assay, after low temperature (27 °C) rescue of the mutant F508del CFTR protein. Line graph represents change in fluorescence from baseline (ΔF/F0) relative to DMSO vehicle addition. After capturing baseline reads, cells were acutely treated with 8B-cGMP (10 µM) or vehicle, followed by addition of CFTR cAMP agonist FSK (10 µM). All wells received CFTRinh-172 (10 µM) at the end. (b) Bar graph represents maximum change in ACC fluorescence from baseline (ΔF/F0) after acute addition of FSK, following low temperature (27 °C) rescue of F508del CFTR protein in double mutant spilt-open organoids (mean ± SEM). Stippled line represents ACC response in CF (CftrF508del/F508del) split-open murine organoids. Paired t-test was performed (**P = 0.0029, n = 5 mice for each genotype). (c) Organoids from double mutant (CftrF508del/F508del; Slc6a14(-/y)) mice

were stained with Calcein green and after capturing baseline images, organoids were acutely treated with 8Bromo-cGMP (8B-cGMP 10 µM) for 15 minutes, followed by FSK (1 µM) for 30 minutes. Representative images of Forskolin Induced Swelling (FIS) are shown for organoids pre-treated with 8B-cGMP or vehicle (Control). Experiments were performed at physiological temperature of 37 °C. (d) Bar graph represents FIS response in double mutant (CftrF508del/F508del; Slc6a14(-/y)) murine organoids, as measured by change in area from baseline after 30 minutes of F508del/F508del FSK addition (ΔA/A0). Stippled line represents FIS response in CF (Cftr ) organoids. Graph represents mean ± SEM. Experiments were performed at physiological temperature of 37° C. Unpaired t-test was performed (***P = 0.0005, n = 4 mice).

4.5 Discussion

SLC6A14 was discovered as a modifier of the intestinal phenotype of CF. The previous chapter, highlights the importance of Slc6a14 in modulating intestinal secretion, and at the pathophysiologic level, regulated the fluid secretory capacity of the CF affected colonic epithelium. Here, we seek to understand the mechanism behind this modulation.

We considered the murine intestinal organoid model300,385, but since, Slc6a14 is an apically expressed protein, the use of this model was limited as the apical surface is facing inwards and is in a closed compartment424. This limits the direct modulation of apical proteins424. To overcome this problem, in a parallel situation, monolayers have been generated from enteric organoids for the study of apical protein SLC9A3, and for host-microbe interactions394,395,420-422. However, this system involves generation of a confluent monolayer over days, and is time-consuming thereby not allowing for high-throughput assessment of apical constituents. We devised a split-open approach (Figure 4.3.2), wherein the organoids are plated on coated 96 well plates for generating 2D lawns which could then be assessed using a fluorescence based method, such as the ACC assay418 (Figure 4.4.1). We used this system to understand how Slc6a14 may affect intestinal secretion and provide insight into its role in modifying the CF phenotyopes.

Our findings suggest that the modulation of F508del-CFTR by SLC6A14 is dependent, at least in part, on arginine mediated nitric oxide production and PKG activation. The role of arginine-nitric oxide pathway has been well studied in models of CF100,386,389-391,417,425. At the molecular level, it is known that the downstream regulators of this pathway - cGMP and PKG can enhance Wt and mutant CFTR channel function and trafficking98,100,426,427. Previous studies of mouse derived intestinal organoids showed that agonists of PKG are effective in enhancing CFTR mediated

organoid swelling in the absence of agonists of the canonical regulation of CFTR, such as PKA59,98,99,428. Our studies confirm the functional significance of this regulatory pathway in murine colon and further, our findings suggest that PKG mediated phosphorylation exerts a “priming” effect on the canonical PKA mediated activation of CFTR. There are five sites at which CFTR can be phosphorylated by PKG (NetPhosK v3, consensus strength >0.5). Two are localized to the regulatory “R” domain at amino-acid residues at positions 788 and 795 and overlap with consensus sites for PKA phosphorylation93. Interestingly, the strength of the consensus sites for PKG at position 788 is predicted to exceed that of PKA and may be preferentially phosphorylated by PKG. There are three other sites, at residues 1178, 1387 and 1456 and their functional significant have yet to be determined. Future studies are required to understand the molecular basis for the “priming” effect that we detect following stimulation of PKG in CF tissues.

Both of our assays of regulated F508del-CFTR channel function in colonic tissue lacking Slc6a14 showed that cyclic GMP partially rescued the defects in cAMP induced channel activation and organoid swelling. However, while significant, this augmentation did not fully restore these functions to those observed in tissues obtained from the F508del-mice (expressing native Slc6a14). Since, the secretory function of the double -mutant organoids do not fully recover upon acute cGMP treatment, there could be other, as yet unidentified signaling pathway(s) involved in Slc6a14 mediated CFTR modulation. Alternatively, the doubly mutated mice may harbor secondary, chronic effects of loss of Cftr function (including the abnormal muscle wall) and we will assess such consequences in future studies.

The current findings suggest that upregulation of Slc6a14 expression and corresponding arginine uptake will ameliorate the disease phenotype associated with Cftr mutation. Interestingly, studies by Grasemann and colleagues that showed that patients with CF have a decreased concentration of exhaled NO417, related to decreased nitric-oxide synthase (iNOS) activity, increased cellular arginase activity and decreased arginine bioavailability386,389,390. The mechanism underlying decreased arginine bioavailability in CF is unknown, but we propose that this deficit worsens the channel function of the already low, residual level of F508del-CFTR and that increased SLC6A14 expression could offset this deficit.

An increase in SLC6A14 mediated arginine uptake would be expected to overcome decreased arginine bioavailability in CF tissues, enhance NO and cyclic GMP-mediated “priming” of F508del-CFTR channel activity. Graseman and Ratjen did conduct clinical trials, testing the efficacy of inhaled arginine in ameliorating CF airway disease402,403. While there is no clear consensus whether L-arginine inhalation improved lung function (FEV1), there was a significant benefit in exhaled nitric oxide (FeNO) after L-arginine inhalation. Given our studies showing a relationship between the relative amount of residual F508del-CFTR at the cell surface and response to arginine pre-treatment, it would be important to pursue a clinical trial of inhaled arginine in a larger patient population receiving CFTR modulatory therapies, such as ORKAMBITM. In this case, arginine (or perhaps a direct activator of SLC6A14) may serve as a companion therapy by complementing the activity of the CFTR modulatory therapy.

Chapter 5 Discussion and Future Directions 5.1 Path towards personalized medicine in CF

5.1.1 Summary

CF disease severity is variable between patients, as are responses to CFTR-directed treatments (discussed in sections 1.7 and 2.2). Using a membrane potential based assay, we developed a method for the phenotypic profiling of patient derived respiratory epithelia, based on their in vitro CFTR channel function. A modification of this method was also applied on assessing CFTR function in split-open intestinal organoids. The ability to assess heterogenous cell types using this method, will allow for the broad application of the system for the assessment of ion channels from patient derived tissues. In the context of CF, this and other emerging methods, will allow for the advancement of personalized medicine in coming years.

5.1.2 Improvisation of technology

The ACC (Apical CFTR conductance) assay described in this thesis, like other developments need improvisation and constant innovation. One of the caveats of the current work flow is the inability to generate large numbers of consistent patient derived cultures. Automation and use of robotic handling of cultures in a highly controlled environment will help minimize human error and variation429,430, thereby increasing the consistency of these cultures. In parallel, improvisation of the current method would involve the use of currently available high-speed cameras which can detect the fluorescence of membrane potential dyes, across multiple wells in lesser time. This would generate larger data-sets and the analysis work flow would need to be adjusted to accommodate these gains.

5.1.3 Improving clinical trial design for personalized medicine

The development of in vitro assays to measure drug responses in patient derived tissues301,418 and the ongoing efforts to correlate specific patient responses to CF therapeutics are important steps towards improving patient care. This will help find drugs most suited to the genetic make-up of the patient, and together with the advent of stem cell derived tissues, boost personalized drug discovery. Currently, only 5% of drugs that enter phase I clinical trial get FDA approved (PBF Attrition Data - KMR Group www.kmrgroup.com Press release 2012). One of the reasons for this high failure is that most of these trials are basket trials, where the candidate drug is expected

to work across entire patient populations. However, the paradigm is shifting towards conducting clinical trials that are taking genetic make-up into account, paving the way for one-person trials431. Knowing which drug would work in an individual has multiple advantages. It is also hoped that a larger portion of drugs from pre-clinical or clinical trials can be used for patients.

5.1.4 Patient tissue on chip technologies for CF

With recent advancements in stem cell biology, it is now possible to convert stem cells from individual patients (iPSCs) into various CF affected tissues like the lung302,303, intestine432, pancreas433 and cholangiocytes321. The knowledge and approaches of CF patient-derived organs, cultures or organoids can also be applied for the study of CF modifier genes. These technologies combined with the methods for high-throughput phenotypic profiling, at the genomic, transcriptomic, proteomic and functional level, will prompt personalized medicine for CF.

5.1.5 Personalized medicine in CF

With the development of more rapid whole genome sequencing434 which is becoming less expensive435, one-time analysis of DNA will soon become possible for every patient. This will help in the discovery of known and unknown variations in the CFTR gene, as well as in modifier genes, for guiding therapeutic strategies and predicting prognosis for CF patients. However, given the role of dynamic non-genetic factors436, as well as factors which can affect epigenetic gene regulation, one time DNA sequencing needs to be combined with longitudinal follow-up of the transcriptomic, proteomic (biochemical and functional), microbiome and metabolomic profiles. The usefulness of such an approach referred to as integrative personal omics profile (iPOP) has recently been shown437. The large amount of data generated using these technologies requires the use of machine learning438 for creating individualized profiles for disease risk439, prognosis440 and therapeutic benefit441. The application of these technologies in CF is underway442, and will hopefully benefit all CF patients in the future.

100

5.2 Role of genetic modifier SLC6A14 in Cystic Fibrosis

5.2.1 Summary

CF disease severity is variable among patients bearing the same F508del CFTR mutation on both alleles5,122. This variation has been attributed, in part, to sequence variants which affect the CF phenotype155. It has been shown that such genetic factors contribute to ~50% variation in lung function in CF patients, and environmental factors contribute to the remaining ~50%178,443. To discover which modifier genes, contribute to this variation, GWAS was recently performed, which looked at the intestinal phenotype of meconium ileus (MI), which pointed to SLC6A14 as a modifier of MI presentation181. However, modifier genes have a smaller effect on the overall disease severity compared to the primary disease-causing gene (CFTR in this case)444. Also, SLC6A14 accounts for nearly 5% variation of the CF phenotype of MI181. We generated a complete genetic knock-out mouse model of Slc6a14, to determine its role in CF. In our murine studies, we found that deletion of Slc6a14 worsened the CF phenotype. Indeed, at the molecular level, we found that SLC6A14 enhanced CFTR anion conductance. The mechanism for this enhancement was studied, wherein SLC6A14 mediated arginine transport and increased nitric- oxide mediated PKG phosphorylation of CFTR. Taken together, we discovered a biologic role of Slc6a14 in the intestine and how modification of CF phenotypes by Slc6a14 occur. We conclude that activation of SLC6A14 may help guide future diagnostic and therapeutic strategies.

5.2.2 Effect of SLC6A14 on CFTR channel gating

The enhancement by SLC6A14 appears to at least partially result from PKG phosphorylation of F508del CFTR. The effect of SLC6A14 mediated enhancement of NO and its role in potentiation of F508del CFTR function, needs further validation. Towards this, future experiments should be focussed on using the split-open organoid model to examine the effect of CFTR potentiation by increasing intracellular NO levels, and likewise rescue mutant F508del function in double mutant organoids by enhancing NO levels.

One important finding of our study was that PKG mediated phosphorylation in split-open murine organoids, primed CFTR for activation by PKA. The molecular basis of this priming effect is still unknown. Possible strategies to understand this effect would involve using a reductionist approach with over-expression of various CFTR phosphorylation insensitive mutants together

101

with SLC6A14, to measure modulation of channel function of these CFTR mutants by SLC6A14. This would help understand which phosphorylation sites on CFTR are primarily affected by SLC6A14. Additionally, a similar platform can be used to measure enhancement of functional expression of other disease-causing CFTR mutations by activation of SLC6A14. These studies would help us better understand the molecular basis of the effect of SLC6A14 on CFTR channel gating, and define a new theratype for CF3.

5.2.3 Role of SLC6A14, arginine and nitric-oxide in CF

Roles for NO and arginine in CF have been well studied by Grasemann and colleagues. They have shown that the NO (FeNO) levels in expired air of patients with CF is low417. Additionally, they have also shown that this defect may be corrected, at least partially by inhalational arginine treatment in CF patients402, with marginal indications of improved lung function386,402,403. Further, drugs which increase the level of downstream cGMP have been shown to improve the CF phenotype100. These strategies should be further developed and translated in the clinic. Since, SLC6A14 appears to affect cellular NO levels, the improvement in FeNO levels and the modest FEV1 gains to inhaled arginine, should be studied in the context of SLC6A14 variants. This could help us stratify patients who respond most robustly to arginine inhalation, and recommend treatment as a complement strategy.

5.2.4 Validation and translation to human intestinal tissue

Our mouse models, helped us understand the pathophysiologic role of SLC6A14 in CF, from the functional and molecular perspectives. These studies should now be translated to human primary tissues. For example, the intestinal phenotype of impaired fluid secretion can be recapitulated in vitro using the human organoid model generated from rectal biopsies300, where this model has been shown to have some degree of in vivo correlation with patient responses301. Our murine findings should be validated in this model system both from diagnostic and therapeutic perspectives. Based on GWAS, the impact of SLC6A14 sequence variants on CFTR function should be measured in human organoids from CF patients bearing the same CFTR genotype of F508del on both alleles. SLC6A14 sequence variants which are associated with a severe CF phenotype, should display decreased CFTR function. Next, the pathophysiologic role of SLC6A14 in the colonic epithelium should be explored by deleting SLC6A14 in rectal organoids from F508del homozygous CF patients. Finally, to assess if pharmacological enhancement of

102

SLC6A14 could be beneficial in treatment resistant patients, we need to over-express SLC6A14 in these rectal organoids to see if the CF phenotype of impaired fluid secretion can be at least partially ameliorated, in this model system.

5.2.5 Role of SLC6A14 in other CF affected tissues

Since SLC6A14 was initially discovered as a modifier of an intestinal phenotype of CF181, we used the CF mouse model with its prominent and well studied intestinal secretion phenotypes consistent with CF153. However, from our studies, it was clear that Slc6a14 was also expressed in other CF affected tissues such as the lung (Figure 3.4.1.1a). Population studies have also suggested that SLC6A14 is a modifier of CF pulmonary phenotypes182,248. Since progressive pulmonary disease is most concerning cause of morbidity and mortality in CF, future studies are focussed on understanding the role of SLC6A14 in modifying this phenotype, using the human primary bronchial epithelial models.

5.2.6 Translation to the clinic

There are over 2000 disease causing mutations in the CFTR gene which can result in mild to severe disease. The compilation of phenotypic data and its correlation to various CFTR mutations, can help the clinician predict disease severity and prognosis based on a molecular diagnosis of CF45 (https://www.cftr2.org/). However, the CFTR genotype- CF phenotype correlation is not always strong, as evident by the variability in disease severity in patients carrying the same mutations5,122. This makes it necessary to increase the complexity of this model by involving modifier genes in predicting the phenotype.

With the advent of CFTR modulators and the approval of CF drugs that act directly on the CFTR molecule, KALYDECOTM (Ivacaftor) and ORKAMBITM (combination of Lumacaftor and Ivacaftor), it is becoming apparent that these drugs will not work in all patients bearing the respective gating G551D445 or the trafficking F508del280,296 mutations. This has led researchers to further realize that modifier genes may also be responsible for modulating patient specific drug responses. One study has found that sequence variations in SLC26A9 can modify KALYDECOTM response in patients, where effects on CFTR function can be directly measured in patient cells252. In the future, it would be important to correlate disease severity and responses to treatment, with sequence variations in the SLC6A14 gene and other genetic modifiers of CF.

103

5.2.7 Genetics of CF disease and application to other genetic disorders

The discovery of the CFTR gene in 1989 was one of the first instances of using positional cloning for gene discovery14,15. Even though CF is considered to be a Mendelian single gene disorder446, there have been instances where CF like disease occurs in absence of mutations in the CFTR gene. The reason for that is thought to be due to mutations in other genes like ENaC447-449. The impact of modifier genes has been implicated in a large number of genetic disorders444,450, and its role in CF has been extensively studied. This has led to the hypothesis depicted in Figure 5.2.7. It highlights the importance of both genetic background and the impact of mutation severity in the primary disease-causing gene (CFTR in this case).

Figure 5.2.7: Hypothetical model for the role of primary and modifier genes in genetic disorders. In the case of CF, loss of function mutations in CFTR result in disease. Depending on the type and class of CFTR mutation, the disease severity can vary tremendously. Second source

104

of variability in disease presentation is the presence of modifier gene variations. A combination of these two sources of genetic variability is plotted in the figure. As shown above the horizontal axis represents severity of CFTR mutation, and the vertical axis represents increasing detrimental effect of modifier gene variations. The heat map (white to black) depicts the severity of CF disease, with white being less severe disease and black being more severe disease. The lower horizontal base of the triangle would thus represent CF disease severity which is purely dependent on the severity of CFTR gene mutation. The circle represents individual modifier gene and the size of the circle represents disease severity conferred by sequence variants in that modifier gene, which are detrimental to the CF phenotype.

5.2.8 Significance

There is a significant CF population which is not responsive to currently available CF therapeutics279,280,296. With the development of the ACC method to measure CFTR channel function in individual patient derived epithelial tissues418, it would be possible to predict therapeutic responses in each CF patient. To validate that, we are currently performing studies to see if these in vitro experiments on patient derived tissues, correlates with in vivo responses to therapies in CF patients. Individuals who are not responsive to currently available CF therapeutic strategies would be good candidates for companion therapies targeting modifier genes like SLC6A14. Taken together, these studies will allow in vitro personalized trials in CF patient derived tissues, before using them clinically in patients.

105

Appendix for Tables

Appendix Table 1: Reproducibility of apical CFTR conductance (ACC) assay between two different platings of nasal cultures. Maximum change in fluorescence relative to baseline

(ΔF/F0) from one plating was plotted against the ΔF/F0 from another plating of nasal cultures derived from the same four different subjects as shown in Figure 2.4.2.1d.

106

Appendix Table 2: Correlation of fluorescence based assay of apical CFTR activation (∆RFU) versus Ussing chamber measurements of CFTR activation (∆ Ieq µA/cm2) in nasal cultures from the same patients (homozygous for F508del CFTR). Paired measurements were performed for nasal cultures pretreated with VX-809 (3 µM) or vehicle (DMSO) for 48 hours and acutely stimulated with either forskolin alone or forskolin plus VX-770.

107

Appendix Table 3: Serum amino-acid levels measured in FVB mice that are 56 days old, across four genotypes - Wt, CF (CftrF508del/F508del), Slc6a14(-/y) and double mutant (CftrF508del/F508del; Slc6a14(-/y)). Mean and standard deviation (SD) are shown (n ≥ 4 for each genotype).

Wt CFTR (F508del) Double Mutant SLC6A14 (KO) Amino-acid MEAN (μM) SD (μM) MEAN (μM) SD (μM) MEAN (μM) SD (μM) MEAN (μM) SD (μM) Pser 33.28 27.66 40.56 33.61 13.93 20.96 Asp 56.76 12.04 47.68 36.98 38.52 13.73 62.50 42.30 Glu 210.42 69.09 146.18 83.06 128.79 56.32 217.88 108.28 Aad 3.64 1.47 6.08 3.21 4.62 2.31 4.96 3.06 Hypro 21.06 4.60 23.08 5.08 25.73 6.43 19.95 6.66 Ser 210.67 38.46 213.70 44.42 248.22 66.41 221.77 53.94 Asn 68.58 10.63 68.05 11.11 76.91 21.08 74.76 14.01 Gly 379.29 90.72 402.48 117.02 462.34 134.54 405.99 74.47 Gln 1056.90 204.45 1074.24 75.40 1249.97 241.83 1132.21 210.23 B-ala 7.08 1.02 6.03 1.98 5.97 2.74 7.71 2.73 Tau 1280.87 414.68 1096.71 291.77 1324.00 400.01 1563.46 422.23 His 74.16 10.66 66.06 8.06 81.16 8.07 83.52 21.21 Arg 17.56 22.90 14.00 10.46 11.74 6.23 19.46 9.36 Cit 134.50 34.32 111.30 28.15 133.13 35.39 103.78 11.63 Thr 224.43 34.01 242.85 50.63 262.20 71.61 244.45 49.70 Ala 705.63 145.49 724.92 110.21 836.93 229.96 776.83 183.07 NH3+Baib 76.66 33.67 43.65 26.05 56.49 35.27 31.07 30.62 Carn 104.90 44.51 133.26 88.74 130.15 88.30 193.97 102.64 Pro 206.58 46.92 180.37 20.32 217.38 107.30 200.50 78.04 1M-His 18.43 11.35 18.32 8.29 15.99 11.56 16.06 7.95 Ans 8.71 5.70 3.52 2.97 7.96 4.68 5.95 6.39 3M-His 14.79 13.38 13.05 10.64 14.53 9.68 27.26 8.36 Crea 20.11 4.06 30.43 10.08 19.74 5.07 23.70 7.00 Aab 7.84 1.94 8.06 1.07 8.04 2.71 7.56 2.84 Tyr 126.82 21.34 115.57 10.97 138.89 36.35 119.81 18.04 Val 353.55 70.73 306.24 28.30 373.34 85.48 339.39 53.41 Met 126.42 16.95 141.73 34.28 138.02 53.26 122.09 33.26 Cysta1 2.70 0.07 3.66 0.33 2.30 0.91 3.45 1.55 Cysta2 42.98 24.23 40.58 16.33 41.08 25.32 48.28 16.17 Cys 5.35 4.15 5.30 0.71 4.16 1.78 9.51 7.60 Ile 157.30 42.47 131.87 18.00 167.16 42.37 139.03 28.79 Leu 310.46 76.47 259.56 30.17 323.97 88.76 279.48 60.39 Phe 163.99 49.62 129.18 4.80 162.60 54.02 53.71 54.97 Trp 63.63 12.01 69.46 11.99 70.52 13.21 68.78 6.72 Orn 142.24 16.18 93.12 38.69 84.52 39.87 155.68 69.55 Lys 277.34 23.72 269.35 53.46 300.57 73.31 282.37 65.56

108

Supplementary Figures

Supplementary Figure 1: Nasal cultures from three CF patients have no significant differences in expression of epithelial cell differentiation markers. Nasal epithelial cells from three different CF patients were differentiated to measure CFTR function using the ACC assay (Figure 2.4.2.1a). These cultures were tested for markers of epithelial differentiation – ZO-1 and pancytokeratin (PanCK). Analyses of these western blots is shown in Figure 2.4.2.1c, showing no significant differences in these markers across the three CF patients.

109

Supplementary Figure 2: Representative Ussing chamber studies of bronchial epithelial cultures. (a) Bronchial epithelial cells from CF patient homozygous for F508del CFTR were studied using the Ussing chamber. The mutant F508del protein was rescued with either corrector VX-809 or vehicle DMSO. As seen in the figure, acute stimulation of CFTR with potentiator VX-770 (1 µM) followed by cAMP agonist forskolin (10 µM) led to a greater response in the VX-809 rescued cultures compared to vehicle treated. Also, the response to CFTR specific inhibitor CFTRinh-172 was higher after maximal stimulation in the VX-809 rescued cultures compared to control. (b) Bar graph represents maximal CFTR stimulation achieved with forskolin and potentiator VX-770, when mutant CFTR is rescued with VX-809 or vehicle alone. Error bars represent SEM from 3 biological replicates. Asterisk indicates statistical significance using unpaired two-tailed t-test (*p = 0.029).

110

Supplementary Figure 3: Representative ACC traces of F508del-CFTR (-/+ VX-809 pretreatment) from nasal epithelial cultures grown on 96 transwell plate. As shown in Figure 2.4.4.1b, each transwell of a 96 well transwell plate was monitored for change in fluorescence in response to CFTR modulators, using the ACC assay. Responses were monitored overtime simultaneously in multiple regions in each well (n=9). The grey lines depict change in fluorescence relative to baseline for each region of the well analyzed. The black line represents mean value at each time point and red lines indicate standard deviation. The line graph shows the response to CFTR agonist forskolin (10 µM) + VX-770 (1 µM) followed by CFTRinh-172 (10 µM). The responses were measured in nasal cultures derived from F508del homozygous patient, rescued with corrector VX-809 or vehicle.

111

Supplementary Figure 4: Nasal cultures grown on 96 transwell plate express epithelial differentiation markers. (a) Nasal cultures grown on 96 transwell plate were analyzed for CFTR function using the ACC assay. qRT-PCR studies were performed on them. The expression levels of CFTR and FOXJ1 across all patients was within a 2 fold range, confirming that the cultures were similarly differentiated. (b) Western blotting analyses of ZO-1 and pancytokeratin (PanCK) protein also showed comparable levels of epithelial differentiation. Calnexin (CNX) analysis assessed protein loading.

112

Supplementary Figure 5: MRP4 expression in primary nasal cultures. PCR amplification for detecting expression of MRP4 was performed, and it showed that, MRP4 is expressed in differentiated nasal cultures. TBP (TATA binding protein) was used as loading control.

Supplementary Figure 6: CFTR protein expression in primary nasal cultures. CFTR protein expression assessed using western blotting of nasal cultures from non-CF (Wt), G551D/2622+1G>A and F508del homozygous patient (rescued with VX-809 or vehicle DMSO). As seen on the blot, mature band C is abundantly present in cultures derived from Wt and G551D subjects and is relatively low in cultures from the F508del homozygous patient, with partial rescue in presence of corrector VX-809.

113

Supplementary Figure 7: Z prime test of ACC on 96 transwell plate of nasal epithelial cultures. Nasal epithelial cells derived from a non-CF (Wt) individual were differentiated on a 96 well transwell plate. Each alternating column was stimulated with CFTR agonist Forskolin (10 µM) and VX-770 (1 µM), or vehicle (DMSO) alone. The mean and standard deviation was calculated for both positive control (Forskolin + VX-770 stimulated transwells) and for negative control (DMSO). Calculated Z prime score was 0.16. There was also a statistically significant difference between negative and positive controls, using unpaired t-test (p < 0.0001).

114

References

1 Quinton, P. M. Cystic fibrosis: a disease in electrolyte transport. FASEB journal : official publication of the Federation of American Societies for Experimental Biology 4, 2709- 2717, (1990).

2 Kerem, E. et al. Clinical and genetic comparisons of patients with cystic fibrosis, with or without meconium ileus. The Journal of pediatrics 114, 767-773, (1989).

3 Cutting, G. R. Cystic fibrosis genetics: from molecular understanding to clinical application. Nature reviews. Genetics 16, 45-56, (2015).

4 Gilljam, M. et al. Clinical manifestations of cystic fibrosis among patients with diagnosis in adulthood. Chest 126, 1215-1224, (2004).

5 Kerem, E. & Kerem, B. Genotype-phenotype correlations in cystic fibrosis. Pediatric pulmonology 22, 387-395, (1996).

6 Weller, P. H. & Williams, J. Clinical features, pathogenesis and management of meconium ileus equivalent. Journal of the Royal Society of Medicine 79 Suppl 12, 36-37, (1986).

7 Blackman, S. M. et al. Relative contribution of genetic and nongenetic modifiers to intestinal obstruction in cystic fibrosis. Gastroenterology 131, 1030-1039, (2006).

8 Fakhoury, K., Durie, P. R., Levison, H. & Canny, G. J. Meconium ileus in the absence of cystic fibrosis. Archives of disease in childhood 67, 1204-1206, (1992).

9 Keats, T. E. & Smith, T. H. Meconium Ileus: a demonstration of the ileal meconium mass by barium enema examination. Radiology 89, 1073-1074, (1967).

10 Boyle, M. P. Nonclassic cystic fibrosis and CFTR-related diseases. Current opinion in pulmonary medicine 9, 498-503, (2003).

11 Farrell, P. M. et al. Guidelines for Diagnosis of Cystic Fibrosis in Newborns through Older Adults: Cystic Fibrosis Foundation Consensus Report. The Journal of pediatrics 153, S4-S14, (2008).

12 Accurso, F. J., Sontag, M. K. & Wagener, J. S. Complications associated with symptomatic diagnosis in infants with cystic fibrosis. The Journal of pediatrics 147, S37- 41, (2005).

13 Kobelska-Dubiel, N., Klincewicz, B. & Cichy, W. Liver disease in cystic fibrosis. Przegla̜ d Gastroenterologiczny 9, 136-141, (2014).

14 Riordan, J. R. et al. Identification of the cystic fibrosis gene: cloning and characterization of complementary DNA. Science 245, 1066-1073, (1989).

115

15 Rommens, J. M. et al. Identification of the cystic fibrosis gene: chromosome walking and jumping. Science 245, 1059-1065, (1989).

16 Kerem, B. et al. Identification of the cystic fibrosis gene: genetic analysis. Science 245, 1073-1080, (1989).

17 Boucher, R. C. Cystic fibrosis: a disease of vulnerability to airway surface dehydration. Trends in molecular medicine 13, 231-240, (2007).

18 Tarran, R., Trout, L., Donaldson, S. H. & Boucher, R. C. Soluble mediators, not cilia, determine airway surface liquid volume in normal and cystic fibrosis superficial airway epithelia. The Journal of general physiology 127, 591-604, (2006).

19 Konstan, M. W., Hilliard, K. A., Norvell, T. M. & Berger, M. Bronchoalveolar lavage findings in cystic fibrosis patients with stable, clinically mild lung disease suggest ongoing infection and inflammation. American journal of respiratory and critical care medicine 150, 448-454, (1994).

20 Heijerman, H. Infection and inflammation in cystic fibrosis: a short review. Journal of cystic fibrosis : official journal of the European Cystic Fibrosis Society 4 Suppl 2, 3-5, (2005).

21 Frizzell, R. A. & Hanrahan, J. W. Physiology of Epithelial Chloride and Fluid Secretion. Cold Spring Harbor perspectives in medicine 2, a009563, (2012).

22 Schroeder, B. C., Cheng, T., Jan, Y. N. & Jan, L. Y. Expression cloning of TMEM16A as a calcium-activated chloride channel subunit. Cell 134, 1019-1029, (2008).

23 Huang, F. et al. Studies on expression and function of the TMEM16A calcium-activated chloride channel. Proceedings of the National Academy of Sciences of the United States of America 106, 21413-21418, (2009).

24 Kimura, T. et al. Deletion of the ubiquitin ligase Nedd4L in lung epithelia causes cystic fibrosis-like disease. Proceedings of the National Academy of Sciences of the United States of America 108, 3216-3221, (2011).

25 Ellsworth, R. E. et al. Comparative genomic sequence analysis of the human and mouse cystic fibrosis transmembrane conductance regulator genes. Proceedings of the National Academy of Sciences of the United States of America 97, 1172-1177, (2000).

26 Li, C. et al. The cystic fibrosis mutation (delta F508) does not influence the chloride channel activity of CFTR. Nature genetics 3, 311-316, (1993).

27 Kelley, K. A., Stamm, S. & Kozak, C. A. Expression and chromosome localization of the murine cystic fibrosis transmembrane conductance regulator. Genomics 13, 381-388, (1992).

116

28 Rogers, C. S. et al. Production of CFTR-null and CFTR-ΔF508 heterozygous pigs by adeno-associated virus–mediated gene targeting and somatic cell nuclear transfer. The Journal of clinical investigation 118, 1571-1577, (2008).

29 Matthews, R. P. & McKnight, G. S. Characterization of the cAMP Response Element of the Cystic Fibrosis Transmembrane Conductance Regulator Gene Promoter. Journal of Biological Chemistry 271, 31869-31877, (1996).

30 MOUCHEL, N. et al. HNF1alpha is involved in tissue-specific regulation of CFTR gene expression. Biochemical Journal 378, 909-918, (2004).

31 McCarthy, V. A. & Harris, A. The CFTR gene and regulation of its expression. Pediatric pulmonology 40, 1-8, (2005).

32 Brouillard, F. et al. NF-κB Mediates Up-regulation of CFTR Gene Expression in Calu-3 Cells by Interleukin-1β. Journal of Biological Chemistry 276, 9486-9491, (2001).

33 Hassan, F. et al. MiR-101 and miR-144 regulate the expression of the CFTR chloride channel in the lung. PloS one 7, e50837, (2012).

34 Oglesby, I. K., Chotirmall, S. H., McElvaney, N. G. & Greene, C. M. Regulation of cystic fibrosis transmembrane conductance regulator by microRNA-145, -223, and -494 is altered in DeltaF508 cystic fibrosis airway epithelium. Journal of immunology (Baltimore, Md. : 1950) 190, 3354-3362, (2013).

35 Oglesby, I. K., Agrawal, R., Mall, M. A., McElvaney, N. G. & Greene, C. M. miRNA- 221 is elevated in cystic fibrosis airway epithelial cells and regulates expression of ATF6. Molecular and cellular pediatrics 2, 1, (2015).

36 Oglesby, I. K. et al. miR-17 overexpression in cystic fibrosis airway epithelial cells decreases interleukin-8 production. The European respiratory journal 46, 1350-1360, (2015).

37 Kumar, P. et al. miR-16 rescues F508del-CFTR function in native cystic fibrosis epithelial cells. Gene therapy 22, 908-916, (2015).

38 Endale Ahanda, M. L. et al. The hsa-miR-125a/hsa-let-7e/hsa-miR-99b cluster is potentially implicated in Cystic Fibrosis pathogenesis. Journal of cystic fibrosis : official journal of the European Cystic Fibrosis Society 14, 571-579, (2015).

39 Montanini, L. et al. miR-146a, miR-155, miR-370, and miR-708 Are CFTR-Dependent, Predicted FOXO1 Regulators and Change at Onset of CFRDs. The Journal of clinical endocrinology and metabolism 101, 4955-4963, (2016).

40 Saayman, S. M. et al. Long Non-coding RNA BGas Regulates the Cystic Fibrosis Transmembrane Conductance Regulator. Molecular therapy : the journal of the American Society of Gene Therapy 24, 1351-1357, (2016).

117

41 Welsh, M. J. & Smith, A. E. Molecular mechanisms of CFTR chloride channel dysfunction in cystic fibrosis. Cell 73, 1251-1254, (1993).

42 MacDonald, K. D., McKenzie, K. R. & Zeitlin, P. L. Cystic fibrosis transmembrane regulator protein mutations: 'class' opportunity for novel drug innovation. Paediatric drugs 9, 1-10, (2007).

43 Marson, F. A. L., Bertuzzo, C. S. & Ribeiro, J. D. Classification of CFTR mutation classes. The Lancet Respiratory Medicine 4, e37-e38.

44 De Boeck, K. & Amaral, M. D. Progress in therapies for cystic fibrosis. The Lancet Respiratory Medicine 4, 662-674, (2016).

45 Sosnay, P. R. et al. Defining the disease liability of variants in the cystic fibrosis transmembrane conductance regulator gene. Nature genetics 45, 1160-1167, (2013).

46 Hollenstein, K., Dawson, R. J. & Locher, K. P. Structure and mechanism of ABC transporter proteins. Current opinion in structural biology 17, 412-418, (2007).

47 Gadsby, D. C., Vergani, P. & Csanady, L. The ABC protein turned chloride channel whose failure causes cystic fibrosis. Nature 440, 477-483, (2006).

48 Cheng, S. H. et al. Functional activation of the cystic fibrosis trafficking mutant delta F508-CFTR by overexpression. The American journal of physiology 268, L615-624, (1995).

49 Trezise, A. E. & Buchwald, M. In vivo cell-specific expression of the cystic fibrosis transmembrane conductance regulator. Nature 353, 434-437, (1991).

50 O'Riordan, C. R., Lachapelle, A. L., Marshall, J., Higgins, E. A. & Cheng, S. H. Characterization of the oligosaccharide structures associated with the cystic fibrosis transmembrane conductance regulator. Glycobiology 10, 1225-1233, (2000).

51 Denning, G. M. et al. Processing of mutant cystic fibrosis transmembrane conductance regulator is temperature-sensitive. Nature 358, 761-764, (1992).

52 Bear, C. E. et al. Purification and functional reconstitution of the cystic fibrosis transmembrane conductance regulator (CFTR). Cell 68, 809-818, (1992).

53 Gee, H. Y., Noh, S. H., Tang, B. L., Kim, K. H. & Lee, M. G. Rescue of DeltaF508- CFTR trafficking via a GRASP-dependent unconventional secretion pathway. Cell 146, 746-760, (2011).

54 Piao, H. et al. Sec16A is critical for both conventional and unconventional secretion of CFTR. Scientific reports 7, 39887, (2017).

55 ZAMAN, K., PALMER, L. A., DOCTOR, A., HUNT, J. F. & GASTON, B. Concentration-dependent effects of endogenous S-nitrosoglutathione on gene regulation by specificity proteins Sp3 and Sp1. Biochemical Journal 380, 67-74, (2004).

118

56 Zaman, K. et al. S-Nitrosylating Agents: A Novel Class of Compounds That Increase Cystic Fibrosis Transmembrane Conductance Regulator Expression and Maturation in Epithelial Cells. Molecular Pharmacology 70, 1435-1442, (2006).

57 Wood, J. & Garthwaite, J. Models of the diffusional spread of nitric oxide: Implications for neural nitric oxide signalling and its pharmacological properties. Neuropharmacology 33, 1235-1244, (1994).

58 Marozkina, N. V. et al. Hsp 70/Hsp 90 organizing protein as a nitrosylation target in cystic fibrosis therapy. Proceedings of the National Academy of Sciences 107, 11393- 11398, (2010).

59 Picciotto, M. R., Cohn, J. A., Bertuzzi, G., Greengard, P. & Nairn, A. C. Phosphorylation of the cystic fibrosis transmembrane conductance regulator. The Journal of biological chemistry 267, 12742-12752, (1992).

60 Wang, W. et al. Reversible Silencing of CFTR Chloride Channels by Glutathionylation. The Journal of general physiology 125, 127-141, (2005).

61 Jilling, T., Haddad, I. Y., Cheng, S. H. & Matalon, S. Nitric oxide inhibits heterologous CFTR expression in polarized epithelial cells. American Journal of Physiology - Lung Cellular and Molecular Physiology 277, L89-L96, (1999).

62 Zeitlin, P. L. Is It Go or NO Go for S-Nitrosylation Modification-Based Therapies of Cystic Fibrosis Transmembrane Regulator Trafficking? Molecular Pharmacology 70, 1155-1158, (2006).

63 Liu, L. et al. Essential roles of S-nitrosothiols in vascular homeostasis and endotoxic shock. Cell 116, 617-628, (2004).

64 Barraud, N. et al. Involvement of Nitric Oxide in Biofilm Dispersal of Pseudomonas aeruginosa. Journal of Bacteriology 188, 7344-7353, (2006).

65 Xue, C., Botkin, S. J. & Johns, R. A. Localization of endothelial NOS at the basal microtubule membrane in ciliated epithelium of rat lung. The journal of histochemistry and cytochemistry : official journal of the Histochemistry Society 44, 463-471, (1996).

66 Li, D., Shirakami, G., Zhan, X. & Johns, R. A. Regulation of ciliary beat frequency by the nitric oxide-cyclic guanosine monophosphate signaling pathway in rat airway epithelial cells. American journal of respiratory cell and molecular biology 23, 175-181, (2000).

67 Fowler, C. J. et al. Abnormal nasal nitric oxide production, ciliary beat frequency, and Toll-like receptor response in pulmonary nontuberculous mycobacterial disease epithelium. American journal of respiratory and critical care medicine 187, 1374-1381, (2013).

119

68 Ballard, S. T. & Spadafora, D. FLUID SECRETION BY SUBMUCOSAL GLANDS OF THE TRACHEOBRONCHIAL AIRWAYS. Respiratory physiology & neurobiology 159, 271-277, (2007).

69 Gaston, B. et al. Endogenous nitrogen oxides and bronchodilator S-nitrosothiols in human airways. Proceedings of the National Academy of Sciences 90, 10957-10961, (1993).

70 Palmer, R. M., Ferrige, A. G. & Moncada, S. Nitric oxide release accounts for the biological activity of endothelium-derived relaxing factor. Nature 327, 524-526, (1987).

71 Snyder, A. H. et al. Acute effects of aerosolized S-nitrosoglutathione in cystic fibrosis. American journal of respiratory and critical care medicine 165, 922-926, (2002).

72 Dawson, R. J. & Locher, K. P. Structure of a bacterial multidrug ABC transporter. Nature 443, 180-185, (2006).

73 Dawson, R. J. & Locher, K. P. Structure of the multidrug ABC transporter Sav1866 from Staphylococcus aureus in complex with AMP-PNP. FEBS letters 581, 935-938, (2007).

74 Mornon, J. P., Lehn, P. & Callebaut, I. Molecular models of the open and closed states of the whole human CFTR protein. Cellular and molecular life sciences : CMLS 66, 3469- 3486, (2009).

75 Mornon, J. P., Hoffmann, B., Jonic, S., Lehn, P. & Callebaut, I. Full-open and closed CFTR channels, with lateral tunnels from the cytoplasm and an alternative position of the F508 region, as revealed by molecular dynamics. Cellular and molecular life sciences : CMLS 72, 1377-1403, (2015).

76 Zhang, Z. & Chen, J. Atomic Structure of the Cystic Fibrosis Transmembrane Conductance Regulator. Cell 167, 1586-1597.e1589, (2016).

77 Liu, F., Zhang, Z., Csanady, L., Gadsby, D. C. & Chen, J. Molecular Structure of the Human CFTR Ion Channel. Cell 169, 85-95.e88, (2017).

78 Li, C. et al. ATPase activity of the cystic fibrosis transmembrane conductance regulator. The Journal of biological chemistry 271, 28463-28468, (1996).

79 Bear, C. E. et al. Coupling of ATP hydrolysis with channel gating by purified, reconstituted CFTR. Journal of bioenergetics and biomembranes 29, 465-473, (1997).

80 Gadsby, D. C. & Nairn, A. C. Control of CFTR channel gating by phosphorylation and nucleotide hydrolysis. Physiological reviews 79, S77-S107, (1999).

81 Ford, B. CFTR structure: lassoing cystic fibrosis. Nature structural & molecular biology 24, 13-14, (2017).

120

82 Gee, H. Y., Tang, B. L., Kim, K. H. & Lee, M. G. Syntaxin 16 binds to cystic fibrosis transmembrane conductance regulator and regulates its membrane trafficking in epithelial cells. The Journal of biological chemistry 285, 35519-35527, (2010).

83 Rich, D. P. et al. Effect of deleting the R domain on CFTR-generated chloride channels. Science 253, 205-207, (1991).

84 Bozoky, Z. et al. Regulatory R region of the CFTR chloride channel is a dynamic integrator of phospho-dependent intra- and intermolecular interactions. Proceedings of the National Academy of Sciences of the United States of America 110, E4427-E4436, (2013).

85 Luo, J., Pato, M. D., Riordan, J. R. & Hanrahan, J. W. Differential regulation of single CFTR channels by PP2C, PP2A, and other phosphatases. The American journal of physiology 274, C1397-1410, (1998).

86 Dahan, D. et al. Regulation of the CFTR channel by phosphorylation. Pflugers Archiv : European journal of physiology 443 Suppl 1, S92-96, (2001).

87 Cheng, S. H. et al. Phosphorylation of the R domain by cAMP-dependent protein kinase regulates the CFTR chloride channel. Cell 66, 1027-1036, (1991).

88 Tien, X. Y., Brasitus, T. A., Kaetzel, M. A., Dedman, J. R. & Nelson, D. J. Activation of the cystic fibrosis transmembrane conductance regulator by cGMP in the human colonic cancer cell line, Caco-2. The Journal of biological chemistry 269, 51-54, (1994).

89 Jia, Y., Mathews, C. J. & Hanrahan, J. W. Phosphorylation by protein kinase C is required for acute activation of cystic fibrosis transmembrane conductance regulator by protein kinase A. The Journal of biological chemistry 272, 4978-4984, (1997).

90 Hegedus, T. et al. Role of individual R domain phosphorylation sites in CFTR regulation by protein kinase A. Biochimica et biophysica acta 1788, 1341-1349, (2009).

91 Antigny, F., Norez, C., Becq, F. & Vandebrouck, C. CFTR and Ca Signaling in Cystic Fibrosis. Frontiers in pharmacology 2, 67, (2011).

92 Billet, A. & Hanrahan, J. W. The secret life of CFTR as a calcium-activated chloride channel. The Journal of physiology 591, 5273-5278, (2013).

93 Pasyk, S. et al. The major cystic fibrosis causing mutation exhibits defective propensity for phosphorylation. Proteomics 15, 447-461, (2015).

94 Molinski, S. et al. Functional Rescue of F508del-CFTR Using Small Molecule Correctors. Frontiers in pharmacology 3, 160, (2012).

95 Neville, D. C. et al. Evidence for phosphorylation of serine 753 in CFTR using a novel metal-ion affinity resin and matrix-assisted laser desorption mass spectrometry. Protein science : a publication of the Protein Society 6, 2436-2445, (1997).

121

96 Csanady, L. et al. Preferential phosphorylation of R-domain Serine 768 dampens activation of CFTR channels by PKA. The Journal of general physiology 125, 171-186, (2005).

97 Alcolado, N., Conrad, D. J., Rafferty, S., Chappe, F. G. & Chappe, V. M. VIP-dependent increase in F508del-CFTR membrane localization is mediated by PKCepsilon. American journal of physiology. Cell physiology 301, C53-65, (2011).

98 Vaandrager, A. B. et al. cGMP stimulation of cystic fibrosis transmembrane conductance regulator Cl- channels co-expressed with cGMP-dependent protein kinase type II but not type Ibeta. The Journal of biological chemistry 272, 4195-4200, (1997).

99 Seidler, U. et al. A functional CFTR protein is required for mouse intestinal cAMP-, cGMP- and Ca2+-dependent HCO3− secretion. The Journal of physiology 505, 411-423, (1997).

100 Poschet, J. F. et al. Pharmacological modulation of cGMP levels by phosphodiesterase 5 inhibitors as a therapeutic strategy for treatment of respiratory pathology in cystic fibrosis. American journal of physiology. Lung cellular and molecular physiology 293, L712-719, (2007).

101 Kongsuphol, P. et al. Mechanistic insight into control of CFTR by AMPK. The Journal of biological chemistry 284, 5645-5653, (2009).

102 Bozoky, Z. et al. Synergy of cAMP and calcium signaling pathways in CFTR regulation. Proceedings of the National Academy of Sciences of the United States of America 114, E2086-e2095, (2017).

103 Billet, A. & Hanrahan, J. W. The secret life of CFTR as a calcium-activated chloride channel. The Journal of physiology 591, 5273-5278, (2013).

104 Billet, A., Luo, Y., Balghi, H. & Hanrahan, J. W. Role of Tyrosine Phosphorylation in the Muscarinic Activation of the Cystic Fibrosis Transmembrane Conductance Regulator (CFTR). Journal of Biological Chemistry 288, 21815-21823, (2013).

105 Liang, L., Woodward, O. M., Chen, Z., Cotter, R. & Guggino, W. B. A novel role of protein tyrosine kinase2 in mediating chloride secretion in human airway epithelial cells. PloS one 6, e21991, (2011).

106 Park, S. et al. Irbit mediates synergy between ca(2+) and cAMP signaling pathways during epithelial transport in mice. Gastroenterology 145, 232-241, (2013).

107 McCarty, N. A. & Zhang, Z. R. Identification of a region of strong discrimination in the pore of CFTR. American journal of physiology. Lung cellular and molecular physiology 281, L852-867, (2001).

108 Hwang, T.-C. & Kirk, K. L. The CFTR Ion Channel: Gating, Regulation, and Anion Permeation. Cold Spring Harbor perspectives in medicine 3, a009498, (2013).

122

109 Linsdell, P. Anion conductance selectivity mechanism of the CFTR chloride channel. Biochimica et biophysica acta 1858, 740-747, (2016).

110 Hug, M. J., Tamada, T. & Bridges, R. J. CFTR and Bicarbonate Secretion to Epithelial Cells. Physiology 18, 38-42, (2003).

111 Tang, L., Fatehi, M. & Linsdell, P. Mechanism of direct bicarbonate transport by the CFTR anion channel. Journal of Cystic Fibrosis 8, 115-121, (2009).

112 Kartner, N. et al. Expression of the cystic fibrosis gene in non-epithelial invertebrate cells produces a regulated anion conductance. Cell 64, 681-691, (1991).

113 Tabcharani, J. A., Chang, X. B., Riordan, J. R. & Hanrahan, J. W. Phosphorylation- regulated Cl- channel in CHO cells stably expressing the cystic fibrosis gene. Nature 352, 628-631, (1991).

114 Coakley, R. D. et al. Abnormal surface liquid pH regulation by cultured cystic fibrosis bronchial epithelium. Proceedings of the National Academy of Sciences of the United States of America 100, 16083-16088, (2003).

115 Verkman, A. S., Song, Y. & Thiagarajah, J. R. Role of airway surface liquid and submucosal glands in cystic fibrosis lung disease. American Journal of Physiology - Cell Physiology 284, C2-C15, (2003).

116 Song, Y., Salinas, D., Nielson, D. W. & Verkman, A. S. Hyperacidity of secreted fluid from submucosal glands in early cystic fibrosis. American Journal of Physiology - Cell Physiology 290, C741-C749, (2006).

117 Shah, V. S. et al. Airway acidification initiates host defense abnormalities in cystic fibrosis mice. Science 351, 503-507, (2016).

118 Tang, X. X. et al. Acidic pH increases airway surface liquid viscosity in cystic fibrosis. The Journal of clinical investigation 126, 879-891, (2016).

119 Wang, F., Zeltwanger, S., Hu, S. & Hwang, T. C. Deletion of phenylalanine 508 causes attenuated phosphorylation-dependent activation of CFTR chloride channels. The Journal of physiology 524 Pt 3, 637-648, (2000).

120 Lukacs, G. L. et al. The delta F508 mutation decreases the stability of cystic fibrosis transmembrane conductance regulator in the plasma membrane. Determination of functional half-lives on transfected cells. The Journal of biological chemistry 268, 21592- 21598, (1993).

121 Cholon, D. M., O'Neal, W. K., Randell, S. H., Riordan, J. R. & Gentzsch, M. Modulation of endocytic trafficking and apical stability of CFTR in primary human airway epithelial cultures. American Journal of Physiology - Lung Cellular and Molecular Physiology 298, L304-L314, (2010).

123

122 Kerem, E. et al. The relation between genotype and phenotype in cystic fibrosis--analysis of the most common mutation (delta F508). The New England journal of medicine 323, 1517-1522, (1990).

123 Garcia, M. A., Yang, N. & Quinton, P. M. Normal mouse intestinal mucus release requires cystic fibrosis transmembrane regulator-dependent bicarbonate secretion. The Journal of clinical investigation 119, 2613-2622, (2009).

124 Cantin, A. M., Hartl, D., Konstan, M. W. & Chmiel, J. F. Inflammation in cystic fibrosis lung disease: Pathogenesis and therapy. Journal of Cystic Fibrosis 14, 419-430.

125 Bartlett, J. A. et al. Newborn Cystic Fibrosis Pigs Have a Blunted Early Response to an Inflammatory Stimulus. American journal of respiratory and critical care medicine 194, 845-854, (2016).

126 Keiser, N. W. et al. Defective innate immunity and hyperinflammation in newborn cystic fibrosis transmembrane conductance regulator-knockout ferret lungs. American journal of respiratory cell and molecular biology 52, 683-694, (2015).

127 Khan, T. Z. et al. Early pulmonary inflammation in infants with cystic fibrosis. American journal of respiratory and critical care medicine 151, 1075-1082, (1995).

128 Lukacs, G. L. et al. Conformational maturation of CFTR but not its mutant counterpart (delta F508) occurs in the endoplasmic reticulum and requires ATP. The EMBO journal 13, 6076-6086, (1994).

129 Hobbs, C. A. et al. Identification of the SPLUNC1 ENaC-inhibitory domain yields novel strategies to treat sodium hyperabsorption in cystic fibrosis airway epithelial cultures. American Journal of Physiology - Lung Cellular and Molecular Physiology 305, L990- L1001, (2013).

130 Myerburg, M. M. et al. Airway Surface Liquid Volume Regulates ENaC by Altering the Serine Protease-Protease Inhibitor Balance: A MECHANISM FOR SODIUM HYPERABSORPTION IN CYSTIC FIBROSIS. Journal of Biological Chemistry 281, 27942-27949, (2006).

131 Hoegger, M. J. et al. Assessing mucociliary transport of single particles in vivo shows variable speed and preference for the ventral trachea in newborn pigs. Proceedings of the National Academy of Sciences of the United States of America 111, 2355-2360, (2014).

132 Shah, V. S. et al. Relationships among CFTR expression, HCO3- secretion, and host defense may inform gene- and cell-based cystic fibrosis therapies. Proceedings of the National Academy of Sciences of the United States of America 113, 5382-5387, (2016).

133 Pezzulo, A. A. et al. Reduced airway surface pH impairs bacterial killing in the porcine cystic fibrosis lung. Nature 487, 109-113, (2012).

134 De Lisle, R. C. Pass the bicarb: the importance of HCO3- for mucin release. The Journal of clinical investigation 119, 2535-2537, (2009).

124

135 Gustafsson, J. K. et al. Bicarbonate and functional CFTR channel are required for proper mucin secretion and link cystic fibrosis with its mucus phenotype. The Journal of experimental medicine 209, 1263-1272, (2012).

136 Kushwah, R., Gagnon, S. & Sweezey, N. B. Intrinsic predisposition of naive cystic fibrosis T cells to differentiate towards a Th17 phenotype. Respiratory research 14, 138, (2013).

137 Bruscia, E. M. et al. Macrophages Directly Contribute to the Exaggerated Inflammatory Response in Cystic Fibrosis Transmembrane Conductance Regulator−/− Mice. American journal of respiratory cell and molecular biology 40, 295-304, (2009).

138 Koller, D. Y., Urbanek, R. & Götz, M. Increased degranulation of eosinophil and neutrophil granulocytes in cystic fibrosis. American journal of respiratory and critical care medicine 152, 629-633, (1995).

139 Painter, R. G. et al. CFTR Expression in Human Neutrophils and the Phagolysosomal Chlorination Defect in Cystic Fibrosis. Biochemistry 45, 10260-10269, (2006).

140 Birrer, P. et al. Protease-antiprotease imbalance in the lungs of children with cystic fibrosis. American journal of respiratory and critical care medicine 150, 207-213, (1994).

141 Nakamura, H., Yoshimura, K., McElvaney, N. G. & Crystal, R. G. Neutrophil elastase in respiratory epithelial lining fluid of individuals with cystic fibrosis induces interleukin-8 gene expression in a human bronchial epithelial cell line. Journal of Clinical Investigation 89, 1478-1484, (1992).

142 Ulrich, M. et al. Relative contribution of Prevotella intermedia and Pseudomonas aeruginosa to lung pathology in airways of patients with cystic fibrosis. Thorax 65, 978-984, (2010).

143 Folkesson, A. et al. Adaptation of Pseudomonas aeruginosa to the cystic fibrosis airway: an evolutionary perspective. Nat Rev Micro 10, 841-851, (2012).

144 Snouwaert, J. N. et al. An Animal Model for Cystic Fibrosis Made by Gene Targeting. Science 257, 1083-1088, (1992).

145 Dorin, J. R. et al. Cystic fibrosis in the mouse by targeted insertional mutagenesis. Nature 359, 211-215, (1992).

146 Ratcliff, R. et al. Production of a severe cystic fibrosis mutation in mice by gene targeting. Nature genetics 4, 35-41, (1993).

147 O'Neal, W. K. et al. A severe phenotype in mice with a duplication of exon 3 in the cystic fibrosis locus. Human molecular genetics 2, 1561-1569, (1993).

148 Zeiher, B. G. et al. A mouse model for the delta F508 allele of cystic fibrosis. Journal of Clinical Investigation 96, 2051-2064, (1995).

125

149 Colledge, W. H. et al. Generation and characterization of a [Delta]F508 cystic fibrosis mouse model. Nature genetics 10, 445-452, (1995).

150 Kent, G. et al. Phenotypic abnormalities in long-term surviving cystic fibrosis mice. Pediatric research 40, 233-241, (1996).

151 Eckman, E. A., Cotton, C. U., Kube, D. M. & Davis, P. B. Dietary changes improve survival of CFTR S489X homozygous mutant mouse. The American journal of physiology 269, L625-630, (1995).

152 Gabriel, S. E., Brigman, K. N., Koller, B. H., Boucher, R. C. & Stutts, M. J. Cystic Fibrosis Heterozygote Resistance to Cholera Toxin in the Cystic Fibrosis Mouse Model. Science 266, 107-109, (1994).

153 Scholte, B. J., Davidson, D. J., Wilke, M. & De Jonge, H. R. Animal models of cystic fibrosis. Journal of cystic fibrosis : official journal of the European Cystic Fibrosis Society 3 Suppl 2, 183-190, (2004).

154 Canale-Zambrano, J. C., Poffenberger, M. C., Cory, S. M., Humes, D. G. & Haston, C. K. Intestinal phenotype of variable-weight cystic fibrosis knockout mice. American journal of physiology. Gastrointestinal and liver physiology 293, G222-229, (2007).

155 Rozmahel, R. et al. Modulation of disease severity in cystic fibrosis transmembrane conductance regulator deficient mice by a secondary genetic factor. Nature genetics 12, 280-287, (1996).

156 Risse, P. A. et al. Ileal smooth muscle dysfunction and remodeling in cystic fibrosis. American journal of physiology. Gastrointestinal and liver physiology 303, G1-8, (2012).

157 Bazett, M. et al. Cystic fibrosis mouse model-dependent intestinal structure and gut microbiome. Mammalian genome : official journal of the International Mammalian Genome Society 26, 222-234, (2015).

158 Zhao, J., Liao, D., Yang, J. & Gregersen, H. Biomechanical remodelling of obstructed guinea pig jejunum. Journal of biomechanics 43, 1322-1329, (2010).

159 Robertson, M. B., Choe, K. A. & Joseph, P. M. Review of the abdominal manifestations of cystic fibrosis in the adult patient. Radiographics : a review publication of the Radiological Society of North America, Inc 26, 679-690, (2006).

160 Colombo, C. et al. Guidelines for the diagnosis and management of distal intestinal obstruction syndrome in cystic fibrosis patients. Journal of cystic fibrosis : official journal of the European Cystic Fibrosis Society 10 Suppl 2, S24-28, (2011).

161 Hays, S. R., Ferrando, R. E., Carter, R., Wong, H. H. & Woodruff, P. G. Structural changes to airway smooth muscle in cystic fibrosis. Thorax 60, 226-228, (2005).

126

162 Regamey, N. et al. Increased airway smooth muscle mass in children with asthma, cystic fibrosis, and non-cystic fibrosis bronchiectasis. American journal of respiratory and critical care medicine 177, 837-843, (2008).

163 Bijvelds, M. J. et al. Fat absorption in cystic fibrosis mice is impeded by defective lipolysis and post-lipolytic events. American journal of physiology. Gastrointestinal and liver physiology 288, G646-653, (2005).

164 Rosenberg, L. A., Schluchter, M. D., Parlow, A. F. & Drumm, M. L. Mouse as a Model of Growth Retardation in Cystic Fibrosis. Pediatric research 59, 191-195, (2006).

165 Kent, G. et al. Lung disease in mice with cystic fibrosis. The Journal of clinical investigation 100, 3060-3069, (1997).

166 Gosselin, D. et al. Impaired ability of Cftr knockout mice to control lung infection with Pseudomonas aeruginosa. American journal of respiratory and critical care medicine 157, 1253-1262, (1998).

167 Hoffmann, N. et al. Novel Mouse Model of Chronic Pseudomonas aeruginosa Lung Infection Mimicking Cystic Fibrosis. Infection and immunity 73, 2504-2514, (2005).

168 Kopelman, H. et al. Impaired chloride secretion, as well as bicarbonate secretion, underlies the fluid secretory defect in the cystic fibrosis pancreas. Gastroenterology 95, 349-355, (1988).

169 Gray, M. A., Winpenny, J. P., Porteous, D. J., Dorin, J. R. & Argent, B. E. CFTR and calcium-activated chloride currents in pancreatic duct cells of a transgenic CF mouse. The American journal of physiology 266, C213-221, (1994).

170 Stotland, P. K., Radzioch, D. & Stevenson, M. M. Mouse models of chronic lung infection with Pseudomonas aeruginosa: models for the study of cystic fibrosis. Pediatric pulmonology 30, 413-424, (2000).

171 Sokol, R. Z. Infertility in men with cystic fibrosis. Current opinion in pulmonary medicine 7, 421-426, (2001).

172 Leung, A. Y., Wong, P. Y., Yankaskas, J. R. & Boucher, R. C. cAMP- but not Ca(2+)- regulated Cl- conductance is lacking in cystic fibrosis mice epididymides and seminal vesicles. The American journal of physiology 271, C188-193, (1996).

173 Aris, R. M. et al. Guide to Bone Health and Disease in Cystic Fibrosis. The Journal of Clinical Endocrinology & Metabolism 90, 1888-1896, (2005).

174 Dif, F., Marty, C., Baudoin, C., de Vernejoul, M. C. & Levi, G. Severe osteopenia in CFTR-null mice. Bone 35, 595-603, (2004).

175 Sun, X. et al. Disease phenotype of a ferret CFTR-knockout model of cystic fibrosis. The Journal of clinical investigation 120, 3149-3160, (2010).

127

176 Rogers, C. S. et al. Disruption of the CFTR Gene Produces a Model of Cystic Fibrosis in Newborn Pigs. Science 321, 1837-1841, (2008).

177 Fisher, J. T., Zhang, Y. & Engelhardt, J. F. Comparative Biology of Cystic Fibrosis Animal Models. Methods in molecular biology (Clifton, N.J.) 742, 311-334, (2011).

178 Collaco, J. M., Blackman, S. M., McGready, J., Naughton, K. M. & Cutting, G. R. Quantification of the relative contribution of environmental and genetic factors to variation in cystic fibrosis lung function. The Journal of pediatrics 157, 802-807.e801- 803, (2010).

179 Zielenski, J. et al. Detection of a cystic fibrosis modifier locus for meconium ileus on human chromosome 19q13. Nature genetics 22, 128-129, (1999).

180 Henderson, L. B. et al. Variation in MSRA Modifies Risk of Neonatal Intestinal Obstruction in Cystic Fibrosis. PLOS Genetics 8, e1002580, (2012).

181 Sun, L. et al. Multiple apical plasma membrane constituents are associated with susceptibility to meconium ileus in individuals with cystic fibrosis. Nature genetics 44, 562-569, (2012).

182 Corvol, H. et al. Genome-wide association meta-analysis identifies five modifier loci of lung disease severity in cystic fibrosis. Nature communications 6, 8382, (2015).

183 Gu, Y. et al. Identification of IFRD1 as a modifier gene for cystic fibrosis lung disease. Nature 458, 1039-1042, (2009).

184 Wright, F. A. et al. Genome-wide association and linkage identify modifier loci of lung disease severity in cystic fibrosis at 11p13 and 20q13.2. Nature genetics 43, 539-546, (2011).

185 Emond, M. J. et al. Exome Sequencing of Phenotypic Extremes Identifies CAV2 and TMC6 as Interacting Modifiers of Chronic Pseudomonas aeruginosa Infection in Cystic Fibrosis. PLoS Genet 11, e1005273, (2015).

186 Agrawal, P. B. et al. Epithelial Sodium Channel ENaC is a Modifier of the Long Term Non-progressive Phenotype Associated with F508del CFTR Mutations. American journal of respiratory cell and molecular biology, (2017).

187 von Kanel, T. et al. Clinical and molecular characterization of the potential CF disease modifier syntaxin 1A. European journal of human genetics : EJHG 21, 1462-1466, (2013).

188 Texereau, J. et al. Nitric oxide synthase 1 as a potential modifier gene of decline in lung function in patients with cystic fibrosis. Thorax 59, 156-158, (2004).

189 Furlan, L. L. et al. IL8 gene as modifier of cystic fibrosis: unraveling the factors which influence clinical variability. Human genetics 135, 881-894, (2016).

128

190 Drumm, M. & Weiler, C. Genetic Influences on Cystic Fibrosis Lung Disease Severity. Frontiers in pharmacology 4, (2013).

191 Trouvé, P., Génin, E. & Férec, C. In silico search for modifier genes associated with pancreatic and liver disease in Cystic Fibrosis. PloS one 12, e0173822, (2017).

192 Pankow, S. et al. ∆F508 CFTR interactome remodelling promotes rescue of cystic fibrosis. Nature 528, 510-516, (2015).

193 Ko, S. B. H. et al. Gating of CFTR by the STAS domain of SLC26 transporters. Nature cell biology 6, 343-350, (2004).

194 Raghuram, V., Mak, D.-O. D. & Foskett, J. K. Regulation of cystic fibrosis transmembrane conductance regulator single-channel gating by bivalent PDZ-domain- mediated interaction. Proceedings of the National Academy of Sciences of the United States of America 98, 1300-1305, (2001).

195 Moyer, B. D. et al. A PDZ-interacting domain in CFTR is an apical membrane polarization signal. The Journal of clinical investigation 104, 1353-1361, (1999).

196 Cushing, P. R., Fellows, A., Villone, D., Boisguerin, P. & Madden, D. R. The Relative Binding Affinities of PDZ Partners for CFTR: A Biochemical Basis for Efficient Endocytic Recycling. Biochemistry 47, 10084-10098, (2008).

197 Liang, X. et al. Phosphorylation-dependent 14-3-3 protein interactions regulate CFTR biogenesis. Molecular biology of the cell 23, 996-1009, (2012).

198 Rennolds, J. et al. Cystic fibrosis transmembrane conductance regulator trafficking is mediated by the COPI coat in epithelial cells. The Journal of biological chemistry 283, 833-839, (2008).

199 Wang, X. et al. COPII-dependent export of cystic fibrosis transmembrane conductance regulator from the ER uses a di-acidic exit code. The Journal of cell biology 167, 65-74, (2004).

200 Meacham, G. C. et al. The Hdj-2/Hsc70 chaperone pair facilitates early steps in CFTR biogenesis. The EMBO journal 18, 1492-1505, (1999).

201 Alberti, S., Bohse, K., Arndt, V., Schmitz, A. & Hohfeld, J. The cochaperone HspBP1 inhibits the CHIP ubiquitin ligase and stimulates the maturation of the cystic fibrosis transmembrane conductance regulator. Molecular biology of the cell 15, 4003-4010, (2004).

202 Yang, Y., Janich, S., Cohn, J. A. & Wilson, J. M. The common variant of cystic fibrosis transmembrane conductance regulator is recognized by hsp70 and degraded in a pre- Golgi nonlysosomal compartment. Proceedings of the National Academy of Sciences of the United States of America 90, 9480-9484, (1993).

129

203 Klein, H. et al. Investigating CFTR and KCa3.1 Protein/Protein Interactions. PloS one 11, e0153665, (2016).

204 Bilan, F. et al. Syntaxin 8 impairs trafficking of cystic fibrosis transmembrane conductance regulator (CFTR) and inhibits its channel activity. J Cell Sci 117, 1923- 1935, (2004).

205 Stanke, F. et al. The CF-modifying gene EHF promotes p.Phe508del-CFTR residual function by altering protein glycosylation and trafficking in epithelial cells. European journal of human genetics : EJHG 22, 660-666, (2014).

206 Veit, G. et al. Ribosomal Stalk Protein Silencing Partially Corrects the DeltaF508-CFTR Functional Expression Defect. PLoS biology 14, e1002462, (2016).

207 Caputo, A. et al. TMEM16A, a membrane protein associated with calcium-dependent chloride channel activity. Science 322, 590-594, (2008).

208 Sala-Rabanal, M., Yurtsever, Z., Nichols, C. G. & Brett, T. J. Secreted CLCA1 modulates TMEM16A to activate Ca(2+)-dependent chloride currents in human cells. eLife 4, (2015).

209 Bradford, E. M., Sartor, M. A., Gawenis, L. R., Clarke, L. L. & Shull, G. E. Reduced NHE3-mediated Na+ absorption increases survival and decreases the incidence of intestinal obstructions in cystic fibrosis mice. American journal of physiology. Gastrointestinal and liver physiology 296, G886-898, (2009).

210 Sun, H. et al. Tgf-beta downregulation of distinct chloride channels in cystic fibrosis- affected epithelia. PloS one 9, e106842, (2014).

211 Harris, W. T. et al. Myofibroblast differentiation and enhanced TGF-B signaling in cystic fibrosis lung disease. PloS one 8, e70196, (2013).

212 Chalmers, J. D., Fleming, G. B., Hill, A. T. & Kilpatrick, D. C. Impact of mannose- binding lectin insufficiency on the course of cystic fibrosis: A review and meta-analysis. Glycobiology 21, 271-282, (2011).

213 Gravina, L. P. et al. Mannose-binding lectin gene as a modifier of the cystic fibrosis phenotype in Argentinean pediatric patients. Journal of cystic fibrosis : official journal of the European Cystic Fibrosis Society 14, 78-83, (2015).

214 Blackman, S. M. et al. Genetic modifiers of cystic fibrosis-related diabetes. Diabetes 62, 3627-3635, (2013).

215 Bradley, G. M., Blackman, S. M., Watson, C. P., Doshi, V. K. & Cutting, G. R. Genetic modifiers of nutritional status in cystic fibrosis. The American Journal of Clinical Nutrition 96, 1299-1308, (2012).

216 Mahadeva, R., Stewart, S., Bilton, D. & Lomas, D. A. Alpha-1 antitrypsin deficiency alleles and severe cystic fibrosis lung disease. Thorax 53, 1022-1024, (1998).

130

217 Corvol, H. et al. Ancestral haplotype 8.1 and lung disease severity in European cystic fibrosis patients. Journal of cystic fibrosis : official journal of the European Cystic Fibrosis Society 11, 63-67, (2012).

218 Plant, B. J. et al. Cystic fibrosis, disease severity, and a macrophage migration inhibitory factor polymorphism. American journal of respiratory and critical care medicine 172, 1412-1415, (2005).

219 Arkwright, P. D. et al. TGF-beta(1) genotype and accelerated decline in lung function of patients with cystic fibrosis. Thorax 55, 459-462, (2000).

220 Drumm , M. L. et al. Genetic Modifiers of Lung Disease in Cystic Fibrosis. New England Journal of Medicine 353, 1443-1453, (2005).

221 Laki, J. et al. The 8.1 ancestral MHC haplotype is associated with delayed onset of colonization in cystic fibrosis. International immunology 18, 1585-1590, (2006).

222 Yarden, J. et al. Association of tumour necrosis factor alpha variants with the CF pulmonary phenotype. Thorax 60, 320-325, (2005).

223 Coutinho, C. A. et al. TNF-alpha polymorphisms as a potential modifier gene in the cystic fibrosis. International journal of molecular epidemiology and genetics 5, 87-99, (2014).

224 Sanchez-Dominguez, C. N. et al. The tumor necrosis factor alpha (-308 A/G) polymorphism is associated with cystic fibrosis in Mexican patients. PloS one 9, e90945, (2014).

225 Park, J. E. et al. Cystic fibrosis modifier genes related to Pseudomonas aeruginosa infection. Genes and immunity 12, 370-377, (2011).

226 Faria, E. J. et al. Association of MBL2, TGF-beta1 and CD14 gene polymorphisms with lung disease severity in cystic fibrosis. Jornal brasileiro de pneumologia : publicacao oficial da Sociedade Brasileira de Pneumologia e Tisilogia 35, 334-342, (2009).

227 De Rose, V. et al. Fcgamma receptor IIA genotype and susceptibility to P. aeruginosa infection in patients with cystic fibrosis. European journal of human genetics : EJHG 13, 96-101, (2005).

228 Aron, Y. et al. HLA-DR polymorphism in allergic bronchopulmonary aspergillosis. Journal of Allergy and Clinical Immunology 104, 891-892.

229 Stanke, F. et al. The TNFalpha receptor TNFRSF1A and genes encoding the amiloride- sensitive sodium channel ENaC as modulators in cystic fibrosis. Human genetics 119, 331-343, (2006).

230 Stanke, F. et al. Hierarchical fine mapping of the cystic fibrosis modifier locus on 19q13 identifies an association with two elements near the genes CEACAM3 and CEACAM6. Human genetics 127, 383-394, (2010).

131

231 Darrah, R. et al. EDNRA variants associate with smooth muscle mRNA levels, cell proliferation rates, and cystic fibrosis pulmonary disease severity. Physiological Genomics 41, 71-77, (2010).

232 McKone, E. F. et al. Variants in the Glutamate-Cysteine-Ligase Gene Are Associated with Cystic Fibrosis Lung Disease. American journal of respiratory and critical care medicine 174, 415-419, (2006).

233 Baranov, V. S. et al. Proportion of the GSTM1 0/0 genotype in some Slavic populations and its correlation with cystic fibrosis and some multifactorial diseases. Human genetics 97, 516-520, (1996).

234 Flamant, C. et al. Glutathione-S-transferase M1, M3, P1 and T1 polymorphisms and severity of lung disease in children with cystic fibrosis. Pharmacogenetics and Genomics 14, 295-301, (2004).

235 Pratap, U., Quinn, S., Blizzard, L. B. & Reid, D. W. Population-based study of cystic fibrosis disease severity and haemochromatosis gene mutations. Respirology 15, 141-149, (2010).

236 Mafficini, A. et al. Impact of polymorphism of Multidrug Resistance-associated Protein 1 (ABCC1) gene on the severity of cystic fibrosis. Journal of cystic fibrosis : official journal of the European Cystic Fibrosis Society 10, 228-233, (2011).

237 Taylor-Cousar, J. L. et al. Histo-blood group gene polymorphisms as potential genetic modifiers of infection and cystic fibrosis lung disease severity. PloS one 4, e4270, (2009).

238 Saferali, A. et al. Polymorphisms associated with expression of BPIFA1/BPIFB1 and lung disease severity in cystic fibrosis. American journal of respiratory cell and molecular biology 53, 607-614, (2015).

239 Büscher, R. et al. β2 adrenoceptor gene polymorphisms in cystic fibrosis lung disease. Pharmacogenetics and Genomics 12, 347-353, (2002).

240 Romi, H. et al. Meconium Ileus Caused by Mutations in GUCY2C, Encoding the CFTR- Activating Guanylate Cyclase 2C. American journal of human genetics 90, 893-899, (2012).

241 Gisler, F. M., von Kanel, T., Kraemer, R., Schaller, A. & Gallati, S. Identification of SNPs in the cystic fibrosis interactome influencing pulmonary progression in cystic fibrosis. European journal of human genetics : EJHG 21, 397-403, (2013).

242 Arkwright, P. D. et al. End-organ dysfunction in cystic fibrosis: association with angiotensin I converting enzyme and cytokine gene polymorphisms. American journal of respiratory and critical care medicine 167, 384-389, (2003).

132

243 Viel, M., Leroy, C., Hubert, D., Fajac, I. & Bienvenu, T. ENaCbeta and gamma genes as modifier genes in cystic fibrosis. Journal of cystic fibrosis : official journal of the European Cystic Fibrosis Society 7, 23-29, (2008).

244 Stanke, F., Hedtfeld, S., Becker, T. & Tummler, B. An association study on contrasting cystic fibrosis endophenotypes recognizes KRT8 but not KRT18 as a modifier of cystic fibrosis disease severity and CFTR mediated residual chloride secretion. BMC Med Genet 12, 62, (2011).

245 Grasemann, H. et al. Airway nitric oxide levels in cystic fibrosis patients are related to a polymorphism in the neuronal nitric oxide synthase gene. American journal of respiratory and critical care medicine 162, 2172-2176, (2000).

246 Grasemann, H. et al. Endothelial nitric oxide synthase variants in cystic fibrosis lung disease. American journal of respiratory and critical care medicine 167, 390-394, (2003).

247 Soave, D. et al. Evidence for a causal relationship between early exocrine pancreatic disease and cystic fibrosis-related diabetes: a Mendelian randomization study. Diabetes 63, 2114-2119, (2014).

248 Li, W. et al. Unraveling the complex genetic model for cystic fibrosis: pleiotropic effects of modifier genes on early cystic fibrosis-related morbidities. Human genetics 133, 151- 161, (2014).

249 Miller, M. R. et al. Variants in Solute Carrier SLC26A9 Modify Prenatal Exocrine Pancreatic Damage in Cystic Fibrosis. The Journal of pediatrics 166, 1152-1157.e1156, (2015).

250 Dorfman, R. et al. Modulatory effect of the SLC9A3 gene on susceptibility to infections and pulmonary function in children with cystic fibrosis. Pediatric pulmonology 46, 385- 392, (2011).

251 Choi, E. H., Ehrmantraut, M., Foster, C. B., Moss, J. & Chanock, S. J. Association of common haplotypes of surfactant protein A1 and A2 (SFTPA1 and SFTPA2) genes with severity of lung disease in cystic fibrosis. Pediatric pulmonology 41, 255-262, (2006).

252 Strug, L. J. et al. Cystic fibrosis gene modifier SLC26A9 modulates airway response to CFTR-directed therapeutics. Human molecular genetics, (2016).

253 Venter, J. C. et al. The Sequence of the Human Genome. Science 291, 1304-1351, (2001).

254 Congreve, M., Murray, C. W. & Blundell, T. L. Keynote review: Structural biology and drug discovery. Drug Discovery Today 10, 895-907, (2005).

255 Heck, A. J. Native mass spectrometry: a bridge between interactomics and structural biology. Nature methods 5, 927-933, (2008).

133

256 Shi, Y. A Glimpse of Structural Biology through X-Ray Crystallography. Cell 159, 995- 1014.

257 Bai, X.-c., McMullan, G. & Scheres, S. H. W. How cryo-EM is revolutionizing structural biology. Trends in Biochemical Sciences 40, 49-57.

258 Burley, S. K. et al. Structural genomics: beyond the human genome project. Nature genetics 23, 151-157, (1999).

259 Milo, R. What is the total number of protein molecules per cell volume? A call to rethink some published values. BioEssays : news and reviews in molecular, cellular and developmental biology 35, 1050-1055, (2013).

260 Romio, L. et al. Characterization of a murine gene homologous to the bovine CaCC chloride channel. Gene 228, 181-188, (1999).

261 Rock, J. R., Futtner, C. R. & Harfe, B. D. The transmembrane protein TMEM16A is required for normal development of the murine trachea. Developmental biology 321, 141- 149, (2008).

262 Rock, J. R. et al. Transmembrane Protein 16A (TMEM16A) Is a Ca(2+)-regulated Cl(–) Secretory Channel in Mouse Airways. The Journal of biological chemistry 284, 14875- 14880, (2009).

263 Schwiebert, E. M. et al. Analysis of ClC-2 channels as an alternative pathway for chloride conduction in cystic fibrosis airway cells. Proceedings of the National Academy of Sciences of the United States of America 95, 3879-3884, (1998).

264 Blaisdell, C. J. et al. CLC-2 single nucleotide polymorphisms (SNPs) as potential modifiers of cystic fibrosis disease severity. BMC Medical Genetics 5, 26-26, (2004).

265 Zdebik, A. A., Cuffe, J. E., Bertog, M., Korbmacher, C. & Jentsch, T. J. Additional Disruption of the ClC-2 Cl- Channel Does Not Exacerbate the Cystic Fibrosis Phenotype of Cystic Fibrosis Transmembrane Conductance Regulator Mouse Models. Journal of Biological Chemistry 279, 22276-22283, (2004).

266 Solymosi, E. A. et al. Chloride transport-driven alveolar fluid secretion is a major contributor to cardiogenic lung edema. Proceedings of the National Academy of Sciences of the United States of America 110, E2308-2316, (2013).

267 Grubb, B. R., Lee, E., Pace, A. J., Koller, B. H. & Boucher, R. C. Intestinal ion transport in NKCC1-deficient mice. American Journal of Physiology - Gastrointestinal and Liver Physiology 279, G707-G718, (2000).

268 Walker, N. M. et al. Down-regulated in adenoma Cl/HCO3 exchanger couples with Na/H exchanger 3 for NaCl absorption in murine small intestine. Gastroenterology 135, 1645- 1653.e1643, (2008).

134

269 Liu, X. et al. Loss of Slc26a9 anion transporter alters intestinal electrolyte and HCO3(-) transport and reduces survival in CFTR-deficient mice. Pflugers Archiv : European journal of physiology 467, 1261-1275, (2015).

270 Wang, Y. Y. et al. Loss of SLC9A3 decreases CFTR protein and causes obstructed azoospermia in mice. PLoS Genet 13, e1006715, (2017).

271 Konstan , M. W., Byard , P. J., Hoppel , C. L. & Davis , P. B. Effect of High-Dose Ibuprofen in Patients with Cystic Fibrosis. New England Journal of Medicine 332, 848- 854, (1995).

272 Wood, M. J., Powell, L. W. & Ramm, G. A. Environmental and genetic modifiers of the progression to fibrosis and cirrhosis in hemochromatosis. Blood 111, 4456-4462, (2008).

273 Gibson, L. E. & Cooke, R. E. A test for concentration of electrolytes in sweat in cystic fibrosis of the pancreas utilizing pilocarpine by iontophoresis. Pediatrics 23, 545-549, (1959).

274 Vernooij-van Langen, A. et al. Clinical evaluation of the Nanoduct sweat test system in the diagnosis of cystic fibrosis after newborn screening. European journal of pediatrics 174, 1025-1034, (2015).

275 Kerem, E., Corey, M., Gold, R. & Levison, H. Pulmonary function and clinical course in patients with cystic fibrosis after pulmonary colonization with Pseudomonas aeruginosa. The Journal of pediatrics 116, 714-719, (1990).

276 Aurora, P. et al. Multiple-Breath Washout as a Marker of Lung Disease in Preschool Children with Cystic Fibrosis. American journal of respiratory and critical care medicine 171, 249-256, (2005).

277 Van Goor, F. et al. Rescue of CF airway epithelial cell function in vitro by a CFTR potentiator, VX-770. Proceedings of the National Academy of Sciences of the United States of America 106, 18825-18830, (2009).

278 Yu, H. et al. Ivacaftor potentiation of multiple CFTR channels with gating mutations. Journal of cystic fibrosis : official journal of the European Cystic Fibrosis Society 11, 237-245, (2012).

279 Ramsey , B. W. et al. A CFTR Potentiator in Patients with Cystic Fibrosis and the G551D Mutation. New England Journal of Medicine 365, 1663-1672, (2011).

280 Wainwright, C. E. et al. Lumacaftor-Ivacaftor in Patients with Cystic Fibrosis Homozygous for Phe508del CFTR. The New England journal of medicine, (2015).

281 Molinski, S. V. et al. Orkambi(R) and amplifier co-therapy improves function from a rare CFTR mutation in gene-edited cells and patient tissue. EMBO molecular medicine, (2017).

135

282 Kerem, E. et al. Ataluren for the treatment of nonsense-mutation cystic fibrosis: a randomised, double-blind, placebo-controlled phase 3 trial. The Lancet Respiratory Medicine 2, 539-547, (2014).

283 Friedman, K. J. et al. Correction of aberrant splicing of the cystic fibrosis transmembrane conductance regulator (CFTR) gene by antisense oligonucleotides. The Journal of biological chemistry 274, 36193-36199, (1999).

284 Igreja, S., Clarke, L. A., Botelho, H. M., Marques, L. & Amaral, M. D. Correction of a Cystic Fibrosis Splicing Mutation by Antisense Oligonucleotides. Human mutation 37, 209-215, (2016).

285 Aznarez, I., Zielenski, J., Rommens, J. M., Blencowe, B. J. & Tsui, L. C. Exon skipping through the creation of a putative exonic splicing silencer as a consequence of the cystic fibrosis mutation R553X. Journal of Medical Genetics 44, 341-346, (2007).

286 Cao, H. et al. Testing gene therapy vectors in human primary nasal epithelial cultures. Molecular therapy. Methods & clinical development 2, 15034, (2015).

287 Alton, E. W. et al. Preparation for a first-in-man lentivirus trial in patients with cystic fibrosis. Thorax 72, 137-147, (2017).

288 McKiernan, P. J., Cunningham, O., Greene, C. M. & Cryan, S. A. Targeting miRNA- based medicines to cystic fibrosis airway epithelial cells using nanotechnology. International journal of nanomedicine 8, 3907-3915, (2013).

289 Eckford, P. D., Li, C., Ramjeesingh, M. & Bear, C. E. Cystic fibrosis transmembrane conductance regulator (CFTR) potentiator VX-770 (ivacaftor) opens the defective channel gate of mutant CFTR in a phosphorylation-dependent but ATP-independent manner. The Journal of biological chemistry 287, 36639-36649, (2012).

290 Clancy, J. P. et al. Clinical Pharmacogenetics Implementation Consortium (CPIC) guidelines for ivacaftor therapy in the context of CFTR genotype. Clinical pharmacology and therapeutics 95, 592-597, (2014).

291 Du, K., Sharma, M. & Lukacs, G. L. The DeltaF508 cystic fibrosis mutation impairs domain-domain interactions and arrests post-translational folding of CFTR. Nature structural & molecular biology 12, 17-25, (2005).

292 Sharma, M. et al. Misfolding diverts CFTR from recycling to degradation: quality control at early endosomes. The Journal of cell biology 164, 923-933, (2004).

293 Eckford, P. D. et al. VX-809 and Related Corrector Compounds Exhibit Secondary Activity Stabilizing Active F508del-CFTR after Its Partial Rescue to the Cell Surface. Chemistry & biology, (2014).

294 Kopeikin, Z., Yuksek, Z., Yang, H. Y. & Bompadre, S. G. Combined effects of VX-770 and VX-809 on several functional abnormalities of F508del-CFTR channels. Journal of

136

cystic fibrosis : official journal of the European Cystic Fibrosis Society 13, 508-514, (2014).

295 Van Goor, F. et al. Correction of the F508del-CFTR protein processing defect in vitro by the investigational drug VX-809. Proceedings of the National Academy of Sciences of the United States of America 108, 18843-18848, (2011).

296 Boyle, M. P. et al. A CFTR corrector (lumacaftor) and a CFTR potentiator (ivacaftor) for treatment of patients with cystic fibrosis who have a phe508del CFTR mutation: a phase 2 randomised controlled trial. The Lancet. Respiratory medicine 2, 527-538, (2014).

297 Almaça, J. et al. in Cystic Fibrosis Vol. 742 Methods in Molecular Biology (eds Margarida D. Amaral & Karl Kunzelmann) Ch. 15, 249-264 (Humana Press, 2011).

298 Trzcinska-Daneluti, A. M. et al. High-content functional screen to identify proteins that correct F508del-CFTR function. Molecular & cellular proteomics : MCP 8, 780-790, (2009).

299 Galietta, L. V., Jayaraman, S. & Verkman, A. S. Cell-based assay for high-throughput quantitative screening of CFTR chloride transport agonists. American journal of physiology. Cell physiology 281, C1734-1742, (2001).

300 Dekkers, J. F. et al. A functional CFTR assay using primary cystic fibrosis intestinal organoids. Nature medicine 19, 939-945, (2013).

301 Dekkers, J. F. et al. Characterizing responses to CFTR-modulating drugs using rectal organoids derived from subjects with cystic fibrosis. Science translational medicine 8, 344ra384, (2016).

302 Wong, A. P. et al. Directed differentiation of human pluripotent stem cells into mature airway epithelia expressing functional CFTR protein. Nature biotechnology 30, 876-882, (2012).

303 Wong, A. P. et al. Efficient generation of functional CFTR-expressing airway epithelial cells from human pluripotent stem cells. Nature protocols 10, 363-381, (2015).

304 Galietta, L. J. et al. An electrogenic amino acid transporter in the apical membrane of cultured human bronchial epithelial cells. The American journal of physiology 275, L917- 923, (1998).

305 Molinski, S. V., Ahmadi, S., Hung, M. & Bear, C. E. Facilitating Structure-Function Studies of CFTR Modulator Sites with Efficiencies in Mutagenesis and Functional Screening. Journal of biomolecular screening 20, 1204-1217, (2015).

306 Maitra, R., Sivashanmugam, P. & Warner, K. A rapid membrane potential assay to monitor CFTR function and inhibition. Journal of biomolecular screening 18, 1132- 1137, (2013).

137

307 Amaral, M. D. Novel personalized therapies for cystic fibrosis: treating the basic defect in all patients. Journal of internal medicine 277, 155-166, (2015).

308 Veit, G. et al. From CFTR biology toward combinatorial pharmacotherapy: expanded classification of cystic fibrosis mutations. Molecular biology of the cell 27, 424-433, (2016).

309 Ussing, H. H. & Zerahn, K. Active transport of sodium as the source of electric current in the short-circuited isolated frog skin. Acta physiologica Scandinavica 23, 110-127, (1951).

310 Li, H., Sheppard, D. N. & Hug, M. J. Transepithelial electrical measurements with the Ussing chamber. Journal of Cystic Fibrosis 3, 123-126, (2004).

311 Molinski, S., Ahmadi, S., Hung, M. & Bear, C. Facilitating Structure-Function Studies of CFTR Modulator Sites with Efficiencies in Mutagenesis and Functional Screening. Journal of biomolecular screening, (2015).

312 Van Goor, F. et al. Rescue of DeltaF508-CFTR trafficking and gating in human cystic fibrosis airway primary cultures by small molecules. American journal of physiology. Lung cellular and molecular physiology 290, L1117-1130, (2006).

313 Kunzelmann, K. et al. An immortalized cystic fibrosis tracheal epithelial cell line homozygous for the delta F508 CFTR mutation. American journal of respiratory cell and molecular biology 8, 522-529, (1993).

314 Itani, O. A. et al. Glucocorticoid-stimulated lung epithelial Na(+) transport is associated with regulated ENaC and sgk1 expression. American journal of physiology. Lung cellular and molecular physiology 282, L631-641, (2002).

315 Thomas, C. P., Campbell, J. R., Wright, P. J. & Husted, R. F. cAMP-stimulated Na+ transport in H441 distal lung epithelial cells: role of PKA, phosphatidylinositol 3-kinase, and sgk1. American journal of physiology. Lung cellular and molecular physiology 287, L843-851, (2004).

316 Caci, E. et al. Evidence for direct CFTR inhibition by CFTR(inh)-172 based on Arg347 mutagenesis. The Biochemical journal 413, 135-142, (2008).

317 Veit, G. et al. Some gating potentiators, including VX-770, diminish DeltaF508-CFTR functional expression. Science translational medicine 6, 246ra297, (2014).

318 Kruh, G. D. & Belinsky, M. G. The MRP family of drug efflux pumps. Oncogene 22, 7537-7552, (2003).

319 Li, C. et al. Spatiotemporal coupling of cAMP transporter to CFTR chloride channel function in the gut epithelia. Cell 131, 940-951, (2007).

320 Huang, S. X. L. et al. Efficient generation of lung and airway epithelial cells from human pluripotent stem cells. Nat Biotech 32, 84-91, (2014).

138

321 Ogawa, M. et al. Directed differentiation of cholangiocytes from human pluripotent stem cells. Nature biotechnology 33, 853-861, (2015).

322 Mutyam, V. et al. Discovery of Clinically Approved Agents That Promote Suppression of Cystic Fibrosis Transmembrane Conductance Regulator Nonsense Mutations. American journal of respiratory and critical care medicine 194, 1092-1103, (2016).

323 Vu, C. B. et al. Fatty Acid Cysteamine Conjugates as Novel and Potent Autophagy Activators That Enhance the Correction of Misfolded F508del-Cystic Fibrosis Transmembrane Conductance Regulator (CFTR). Journal of medicinal chemistry, (2016).

324 Deeley, R. G., Westlake, C. & Cole, S. P. Transmembrane transport of endo- and xenobiotics by mammalian ATP-binding cassette multidrug resistance proteins. Physiological reviews 86, 849-899, (2006).

325 Russel, F. G., Koenderink, J. B. & Masereeuw, R. Multidrug resistance protein 4 (MRP4/ABCC4): a versatile efflux transporter for drugs and signalling molecules. Trends in pharmacological sciences 29, 200-207, (2008).

326 Rius, M., Nies, A. T., Hummel-Eisenbeiss, J., Jedlitschky, G. & Keppler, D. Cotransport of reduced glutathione with bile salts by MRP4 (ABCC4) localized to the basolateral hepatocyte membrane. Hepatology (Baltimore, Md.) 38, 374-384, (2003).

327 Gold, M. J. et al. Mucosal production of uric acid by airway epithelial cells contributes to particulate matter-induced allergic sensitization. Mucosal immunology 9, 809-820, (2016).

328 Gentzsch, M. et al. Pharmacological Rescue of Conditionally Reprogrammed Cystic Fibrosis Bronchial Epithelial Cells. American journal of respiratory cell and molecular biology, (2016).

329 Rosenstein, B. J. Genotype-phenotype correlations in cystic fibrosis. Lancet 343, 746- 747, (1994).

330 Saito, A. & Kamatani, N. Strategies for genome-wide association studies: optimization of study designs by the stepwise focusing method. Journal of human genetics 47, 360-365, (2002).

331 Kim, D. K. et al. Expression cloning of a Na+-independent aromatic amino acid transporter with structural similarity to H+/monocarboxylate transporters. The Journal of biological chemistry 276, 17221-17228, (2001).

332 Mariotta, L. et al. T-type amino acid transporter TAT1 (Slc16a10) is essential for extracellular aromatic amino acid homeostasis control. The Journal of physiology 590, 6413-6424, (2012).

333 Verrey, F. et al. CATs and HATs: the SLC7 family of amino acid transporters. Pflugers Archiv : European journal of physiology 447, 532-542, (2004).

139

334 Kanai, Y. et al. Expression Cloning and Characterization of a Transporter for Large Neutral Amino Acids Activated by the Heavy Chain of 4F2 Antigen (CD98). Journal of Biological Chemistry 273, 23629-23632, (1998).

335 Rossier, G. et al. LAT2, a New Basolateral 4F2hc/CD98-associated Amino Acid Transporter of Kidney and Intestine. Journal of Biological Chemistry 274, 34948-34954, (1999).

336 Bodoy, S. et al. Identification of LAT4, a Novel Amino Acid Transporter with System L Activity. Journal of Biological Chemistry 280, 12002-12011, (2005).

337 Bröer, S. The SLC38 family of sodium–amino acid co-transporters. Pflügers Archiv - European Journal of Physiology 466, 155-172, (2014).

338 Pochini, L., Scalise, M., Galluccio, M. & Indiveri, C. Membrane transporters for the special amino acid glutamine: structure/function relationships and relevance to human health. Frontiers in Chemistry 2, (2014).

339 Thwaites, D. T. & Anderson, C. M. The SLC36 family of proton-coupled amino acid transporters and their potential role in drug transport. British journal of pharmacology 164, 1802-1816, (2011).

340 Kleta, R. et al. Mutations in SLC6A19, encoding B0AT1, cause Hartnup disorder. Nature genetics 36, 999-1002, (2004).

341 Sloan, J. L. & Mager, S. Cloning and Functional Expression of a Human Na+and Cl−- dependent Neutral and Cationic Amino Acid Transporter B0+. Journal of Biological Chemistry 274, 23740-23745, (1999).

342 Rudnick, G., Kramer, R., Blakely, R. D., Murphy, D. L. & Verrey, F. The SLC6 transporters: perspectives on structure, functions, regulation, and models for transporter dysfunction. Pflugers Archiv : European journal of physiology 466, 25-42, (2014).

343 Karunakaran, S. et al. SLC6A14 (ATB0,+) protein, a highly concentrative and broad specific amino acid transporter, is a novel and effective drug target for treatment of estrogen receptor-positive breast cancer. The Journal of biological chemistry 286, 31830- 31838, (2011).

344 Broer, S. Amino acid transport across mammalian intestinal and renal epithelia. Physiological reviews 88, 249-286, (2008).

345 Nicklin, P. et al. Bidirectional transport of amino acids regulates mTOR and autophagy. Cell 136, 521-534, (2009).

346 Rhoads, J. M. et al. Arginine stimulates intestinal cell migration through a focal adhesion kinase dependent mechanism. Gut 53, 514-522, (2004).

347 Coburn, L. A. et al. L-arginine supplementation improves responses to injury and inflammation in dextran sulfate sodium colitis. PloS one 7, e33546, (2012).

140

348 Karunakaran, S. et al. Interaction of tryptophan derivatives with SLC6A14 (ATB0,+) reveals the potential of the transporter as a drug target for cancer chemotherapy. The Biochemical journal 414, 343-355, (2008).

349 Hatanaka, T. et al. Transport of D-serine via the amino acid transporter ATB(0,+) expressed in the colon. Biochemical and biophysical research communications 291, 291- 295, (2002).

350 Nakanishi, T. et al. Na+- and Cl--coupled active transport of carnitine by the amino acid transporter ATB(0,+) from mouse colon expressed in HRPE cells and Xenopus oocytes. The Journal of physiology 532, 297-304, (2001).

351 Bhutia, Y. D., Babu, E., Prasad, P. D. & Ganapathy, V. The amino acid transporter SLC6A14 in cancer and its potential use in chemotherapy. Asian Journal of Pharmaceutical Sciences 9, 293-303, (2014).

352 Ganapathy, M. E. & Ganapathy, V. Amino Acid Transporter ATB0,+ as a delivery system for drugs and prodrugs. Current drug targets. Immune, endocrine and metabolic disorders 5, 357-364, (2005).

353 Gonzalez, I. M. et al. Leucine and arginine regulate trophoblast motility through mTOR- dependent and independent pathways in the preimplantation mouse embryo. Developmental biology 361, 286-300, (2012).

354 Babu, E. et al. Deletion of the amino acid transporter Slc6a14 suppresses tumor growth in spontaneous mouse models of breast cancer. The Biochemical journal, (2015).

355 Hahn, M. K. & Blakely, R. D. The functional impact of SLC6 transporter genetic variation. Annual review of pharmacology and toxicology 47, 401-441, (2007).

356 Tatsumi, M., Groshan, K., Blakely, R. D. & Richelson, E. Pharmacological profile of antidepressants and related compounds at human monoamine transporters. European journal of pharmacology 340, 249-258, (1997).

357 Wang, K. H., Penmatsa, A. & Gouaux, E. Neurotransmitter and psychostimulant recognition by the dopamine transporter. Nature 521, 322-327, (2015).

358 Yamashita, A., Singh, S. K., Kawate, T., Jin, Y. & Gouaux, E. Crystal structure of a bacterial homologue of Na+/Cl--dependent neurotransmitter transporters. Nature 437, 215-223, (2005).

359 Samluk, L., Czeredys, M. & Nalecz, K. A. Regulation of amino acid/carnitine transporter B 0,+ (ATB 0,+) in astrocytes by protein kinase C: independent effects on raft and non- raft transporter subpopulations. Journal of neurochemistry 115, 1386-1397, (2010).

360 Samluk, L., Czeredys, M., Skowronek, K. & Nalecz, K. A. Protein kinase C regulates amino acid transporter ATB(0,+). Biochemical and biophysical research communications 422, 64-69, (2012).

141

361 Michalec, K. et al. Protein kinase C restricts transport of carnitine by amino acid transporter ATB(0,+) apically localized in the blood-brain barrier. Archives of biochemistry and biophysics 554, 28-35, (2014).

362 Anderson, C. M., Ganapathy, V. & Thwaites, D. T. Human solute carrier SLC6A14 is the beta-alanine carrier. The Journal of physiology 586, 4061-4067, (2008).

363 Hara, K. et al. Amino Acid Sufficiency and mTOR Regulate p70 S6 Kinase and eIF-4E BP1 through a Common Effector Mechanism. Journal of Biological Chemistry 273, 14484-14494, (1998).

364 Kim, D. H. et al. mTOR interacts with raptor to form a nutrient-sensitive complex that signals to the cell growth machinery. Cell 110, 163-175, (2002).

365 Krall, A. S., Xu, S., Graeber, T. G., Braas, D. & Christofk, H. R. Asparagine promotes cancer cell proliferation through use as an amino acid exchange factor. 7, 11457, (2016).

366 Jewell, J. L. et al. Metabolism. Differential regulation of mTORC1 by leucine and glutamine. Science 347, 194-198, (2015).

367 Rebsamen, M. et al. SLC38A9 is a component of the lysosomal amino acid sensing machinery that controls mTORC1. Nature 519, 477-481, (2015).

368 Wang, S. et al. The amino acid transporter SLC38A9 is a key component of a lysosomal membrane complex that signals arginine sufficiency to mTORC1. Science (New York, N.Y.) 347, 188-194, (2015).

369 Flach, C. F. et al. Differential expression of intestinal membrane transporters in cholera patients. FEBS letters 581, 3183-3188, (2007).

370 Sadlack, B. et al. Ulcerative colitis-like disease in mice with a disrupted interleukin-2 gene. Cell 75, 253-261, (1993).

371 Flach, C. F. et al. Detection of elafin as a candidate biomarker for ulcerative colitis by whole-genome microarray screening. Inflammatory bowel diseases 12, 837-842, (2006).

372 Eriksson, A., Flach, C. F., Lindgren, A., Kvifors, E. & Lange, S. Five mucosal transcripts of interest in ulcerative colitis identified by quantitative real-time PCR: a prospective study. BMC gastroenterology 8, 34, (2008).

373 Miranda, R. C. et al. SLC6A14 and 5-HTR2C polymorphisms are associated with food intake and nutritional status in children. Clinical biochemistry 48, 1277-1282, (2015).

374 Shoff, S. M., Tluczek, A., Laxova, A., Farrell, P. M. & Lai, H. J. Nutritional status is associated with health-related quality of life in children with cystic fibrosis aged 9- 19years. Journal of cystic fibrosis : official journal of the European Cystic Fibrosis Society, (2013).

142

375 Siasi, E. & Aleyasin, A. Four Single Nucleotide Polymorphisms in INSR, SLC6A14, TAS2R38, and OR2W3 Genes in Association with Idiopathic Infertility in Persian Men. The Journal of reproductive medicine 61, 145-152, (2016).

376 Beaudet, A. L. Brain carnitine deficiency causes nonsyndromic autism with an extreme male bias: A hypothesis. BioEssays : news and reviews in molecular, cellular and developmental biology 39, (2017).

377 Gupta, N. et al. Upregulation of the amino acid transporter ATB0,+ (SLC6A14) in colorectal cancer and metastasis in humans. Biochimica et biophysica acta 1741, 215- 223, (2005).

378 Coothankandaswamy, V. et al. Amino acid transporter SLC6A14 is a novel and effective drug target for pancreatic cancer. British journal of pharmacology, (2016).

379 Gupta, N. et al. Up-regulation of the amino acid transporter ATB0,+ (SLC6A14) in carcinoma of the cervix. Gynecologic oncology 100, 8-13, (2006).

380 Hillesheim, J. et al. Down regulation of small intestinal ion transport in PDZK1- (CAP70/NHERF3) deficient mice. Pflugers Archiv : European journal of physiology 454, 575-586, (2007).

381 Singh, A. K. et al. Molecular transport machinery involved in orchestrating luminal acid- induced duodenal bicarbonate secretion in vivo. The Journal of physiology 591, 5377- 5391, (2013).

382 Grubb, B. R. & Gabriel, S. E. Intestinal physiology and pathology in gene-targeted mouse models of cystic fibrosis. The American journal of physiology 273, G258-266, (1997).

383 van Doorninck, J. H. et al. A mouse model for the cystic fibrosis delta F508 mutation. The EMBO journal 14, 4403-4411, (1995).

384 Rogan, M. P. et al. Pigs and humans with cystic fibrosis have reduced insulin-like growth factor 1 (IGF1) levels at birth. Proceedings of the National Academy of Sciences of the United States of America 107, 20571-20575, (2010).

385 Sato, T. et al. Single Lgr5 stem cells build crypt-villus structures in vitro without a mesenchymal niche. Nature 459, 262-265, (2009).

386 Grasemann, H. & Ratjen, F. Nitric oxide and L-arginine deficiency in cystic fibrosis. Current pharmaceutical design 18, 726-736, (2012).

387 Sonawane, N. D., Hu, J., Muanprasat, C. & Verkman, A. S. Luminally active, nonabsorbable CFTR inhibitors as potential therapy to reduce intestinal fluid loss in cholera. FASEB journal : official publication of the Federation of American Societies for Experimental Biology 20, 130-132, (2006).

143

388 Ugawa, S. et al. Characterization of a mouse colonic system B⁰⁺ amino acid transporter related to amino acid absorption in colon. American Journal of Physiology - Gastrointestinal and Liver Physiology 281, G365-G370, (2001).

389 Grasemann, H., Schwiertz, R., Matthiesen, S., Racke, K. & Ratjen, F. Increased arginase activity in cystic fibrosis airways. American journal of respiratory and critical care medicine 172, 1523-1528, (2005).

390 Grasemann, H. et al. Asymmetric dimethylarginine contributes to airway nitric oxide deficiency in patients with cystic fibrosis. American journal of respiratory and critical care medicine 183, 1363-1368, (2011).

391 Scott, J. A. et al. Plasma arginine metabolites reflect airway dysfunction in a murine model of allergic airway inflammation. Journal of applied physiology (Bethesda, Md. : 1985) 118, 1229-1233, (2015).

392 Bederman, I. et al. Altered de novo lipogenesis contributes to low adipose stores in cystic fibrosis mice. American journal of physiology. Gastrointestinal and liver physiology 303, G507-518, (2012).

393 De Lisle, R. C., Meldi, L. & Mueller, R. Intestinal smooth muscle dysfunction develops postnatally in cystic fibrosis mice. Journal of pediatric gastroenterology and nutrition 55, 689-694, (2012).

394 Foulke-Abel, J. et al. Human Enteroids as a Model of Upper Small Intestinal Ion Transport Physiology and Pathophysiology. Gastroenterology 150, 638-649.e638, (2016).

395 Saxena, K. et al. Human Intestinal Enteroids: a New Model To Study Human Rotavirus Infection, Host Restriction, and Pathophysiology. Journal of virology 90, 43-56, (2015).

396 Wilke, M. et al. Mouse models of cystic fibrosis: phenotypic analysis and research applications. Journal of cystic fibrosis : official journal of the European Cystic Fibrosis Society 10 Suppl 2, S152-171, (2011).

397 French, P. J. et al. A delta F508 mutation in mouse cystic fibrosis transmembrane conductance regulator results in a temperature-sensitive processing defect in vivo. The Journal of clinical investigation 98, 1304-1312, (1996).

398 Sato, T. et al. Long-term expansion of epithelial organoids from human colon, adenoma, adenocarcinoma, and Barrett's epithelium. Gastroenterology 141, 1762-1772, (2011).

399 Appleton, J. Arginine: Clinical potential of a semi-essential amino acid. Alternative medicine review : a journal of clinical therapeutic 7, 512-522, (2002).

400 De Betue, C. T., Van Waardenburg, D. A., Joosten, K. F. & Deutz, N. E. Inflammation causes arginine to become an essential amino acid in critically ill children. Critical Care 15, P384-P384, (2011).

144

401 Flynn, N. E., Meininger, C. J., Haynes, T. E. & Wu, G. The metabolic basis of arginine nutrition and pharmacotherapy. Biomedicine & Pharmacotherapy 56, 427-438, (2002).

402 Grasemann, H., Kurtz, F. & Ratjen, F. Inhaled L-arginine improves exhaled nitric oxide and pulmonary function in patients with cystic fibrosis. American journal of respiratory and critical care medicine 174, 208-212, (2006).

403 Grasemann, H., Tullis, E. & Ratjen, F. A randomized controlled trial of inhaled L- arginine in patients with cystic fibrosis. Journal of cystic fibrosis : official journal of the European Cystic Fibrosis Society 12, 468-474, (2013).

404 Rector, T. S. et al. Randomized, double-blind, placebo-controlled study of supplemental oral L-arginine in patients with heart failure. Circulation 93, 2135-2141, (1996).

405 Schulze-Neick, I. et al. L-arginine and substance P reverse the pulmonary endothelial dysfunction caused by congenital heart surgery. Circulation 100, 749-755, (1999).

406 Besset, A., Bonardet, A., Rondouin, G., Descomps, B. & Passouant, P. Increase in sleep related GH and Prl secretion after chronic arginine aspartate administration in man. Acta endocrinologica 99, 18-23, (1982).

407 Yu, Y. M. et al. Arginine and ornithine kinetics in severely burned patients: increased rate of arginine disposal. American journal of physiology. Endocrinology and metabolism 280, E509-517, (2001).

408 Ohtsuka, Y. & Nakaya, J. Effect of oral administration of L-arginine on senile dementia. The American journal of medicine 108, 439, (2000).

409 Wascher, T. C. et al. Effects of low-dose L-arginine on insulin-mediated vasodilatation and insulin sensitivity. European journal of clinical investigation 27, 690-695, (1997).

410 Lubec, B., Hayn, M., Kitzmuller, E., Vierhapper, H. & Lubec, G. L-Arginine reduces lipid peroxidation in patients with diabetes mellitus. Free radical biology & medicine 22, 355-357, (1997).

411 Muriel, P. Nitric oxide protection of rat liver from lipid peroxidation, collagen accumulation, and liver damage induced by carbon tetrachloride. Biochemical pharmacology 56, 773-779, (1998).

412 Moraes, T. F. & Reithmeier, R. A. Membrane transport metabolons. Biochimica et biophysica acta 1818, 2687-2706, (2012).

413 Fritz, J. H. Arginine Cools the Inflamed Gut. Infection and immunity 81, 3500-3502, (2013).

414 Colgan, S. P. Nitric oxide and intestinal epithelia: Just say NO. Gastroenterology 114, 601-603.

145

415 Singer, II et al. Expression of inducible nitric oxide synthase and nitrotyrosine in colonic epithelium in inflammatory bowel disease. Gastroenterology 111, 871-885, (1996).

416 Nathan, C. & Xie, Q. W. Nitric oxide synthases: roles, tolls, and controls. Cell 78, 915- 918, (1994).

417 Grasemann, H., Michler, E., Wallot, M. & Ratjen, F. Decreased concentration of exhaled nitric oxide (NO) in patients with cystic fibrosis. Pediatric pulmonology 24, 173-177, (1997).

418 Ahmadi, S. et al. Phenotypic profiling of CFTR modulators in patient-derived respiratory epithelia. npj Genomic Medicine 2, 12, (2017).

419 Fatehullah, A., Tan, S. H. & Barker, N. Organoids as an in vitro model of human development and disease. Nature cell biology 18, 246-254, (2016).

420 Foulke-Abel, J. et al. Human enteroids as an ex-vivo model of host–pathogen interactions in the gastrointestinal tract. Experimental Biology and Medicine 239, 1124-1134, (2014).

421 In, J. G. et al. Human mini-guts: new insights into intestinal physiology and host- pathogen interactions. Nat Rev Gastroenterol Hepatol 13, 633-642, (2016).

422 In, J. et al. Enterohemorrhagic Escherichia coli reduce mucus and intermicrovillar bridges in human stem cell-derived colonoids. Cellular and molecular gastroenterology and hepatology 2, 48-62.e43, (2016).

423 Kibbe, M. R. et al. Inducible nitric oxide synthase (iNOS) expression upregulates p21 and inhibits vascular smooth muscle cell proliferation through p42/44 mitogen-activated protein kinase activation and independent of p53 and cyclic guanosine monophosphate. Journal of vascular surgery 31, 1214-1228, (2000).

424 Bryant, D. M. & Mostov, K. E. From cells to organs: building polarized tissue. Nature reviews. Molecular cell biology 9, 887-901, (2008).

425 Oliynyk, I., Hussain, R., Amin, A., Johannesson, M. & Roomans, G. M. The effect of NO-donors on chloride efflux, intracellular Ca(2+) concentration and mRNA expression of CFTR and ENaC in cystic fibrosis airway epithelial cells. Experimental and molecular pathology 94, 474-480, (2013).

426 Golin-Bisello, F., Bradbury, N. & Ameen, N. STa and cGMP stimulate CFTR translocation to the surface of villus enterocytes in rat jejunum and is regulated by protein kinase G. American journal of physiology. Cell physiology 289, C708-716, (2005).

427 Leier, G., Bangel-Ruland, N., Sobczak, K., Knieper, Y. & Weber, W. M. Sildenafil acts as potentiator and corrector of CFTR but might be not suitable for the treatment of CF lung disease. Cellular physiology and biochemistry : international journal of experimental cellular physiology, biochemistry, and pharmacology 29, 775-790, (2012).

146

428 Pattison, A. M. et al. Intestinal Enteroids Model Guanylate Cyclase C-Dependent Secretion Induced by Heat-Stable Enterotoxins. Infection and immunity 84, 3083-3091, (2016).

429 Kempner, M. E. & Felder, R. A. A Review of Cell Culture Automation. JALA: Journal of the Association for Laboratory Automation 7, 56-62, (2002).

430 Gómez-Sjöberg, R., Leyrat, A. A., Pirone, D. M., Chen, C. S. & Quake, S. R. Versatile, Fully Automated, Microfluidic Cell Culture System. Analytical Chemistry 79, 8557- 8563, (2007).

431 Schork, N. J. Personalized medicine: Time for one-person trials. Nature 520, 609-611, (2015).

432 Spence, J. R. et al. Directed differentiation of human pluripotent stem cells into intestinal tissue in vitro. Nature 470, 105-109, (2011).

433 Hohwieler, M. et al. Human pluripotent stem cell-derived acinar/ductal organoids generate human pancreas upon orthotopic transplantation and allow disease modelling. Gut, (2016).

434 Kircher, M. & Kelso, J. High-throughput DNA sequencing – concepts and limitations. BioEssays : news and reviews in molecular, cellular and developmental biology 32, 524- 536, (2010).

435 Heather, J. M. & Chain, B. The sequence of sequencers: The history of sequencing DNA. Genomics 107, 1-8, (2016).

436 Schechter, M. S. Nongenetic influences on cystic fibrosis outcomes. Current opinion in pulmonary medicine 17, 448-454, (2011).

437 Chen, R. et al. Personal Omics Profiling Reveals Dynamic Molecular and Medical Phenotypes. Cell 148, 1293-1307.

438 Ghahramani, Z. Probabilistic machine learning and artificial intelligence. Nature 521, 452-459, (2015).

439 Zhanpeng, J., Oresko, J., Shimeng, H. & Cheng, A. C. in 2009 IEEE/NIH Life Science Systems and Applications Workshop. 80-83.

440 Cruz, J. A. & Wishart, D. S. Applications of Machine Learning in Cancer Prediction and Prognosis. Cancer Informatics 2, 59-77, (2006).

441 Menden, M. P. et al. Machine Learning Prediction of Cancer Cell Sensitivity to Drugs Based on Genomic and Chemical Properties. PloS one 8, e61318, (2013).

442 Tufail, S. in The Alan Turing Institute Newsroom (2017).

147

443 Vanscoy, L. L. et al. Heritability of Lung Disease Severity in Cystic Fibrosis. American journal of respiratory and critical care medicine 175, 1036-1043, (2007).

444 Cutting, G. R. Modifier genes in Mendelian disorders: the example of cystic fibrosis. Annals of the New York Academy of Sciences 1214, 57-69, (2010).

445 Flume, P. A. et al. Ivacaftor in subjects with cystic fibrosis who are homozygous for the F508del-CFTR mutation. Chest 142, 718-724, (2012).

446 Tsui, L.-C. & Dorfman, R. The Cystic Fibrosis Gene: A Molecular Genetic Perspective. Cold Spring Harbor perspectives in medicine 3, a009472, (2013).

447 Sheridan, M. B. et al. Mutations in the beta-subunit of the epithelial Na+ channel in patients with a cystic fibrosis-like syndrome. Human molecular genetics 14, 3493-3498, (2005).

448 Azad, A. K. et al. Mutations in the amiloride-sensitive epithelial sodium channel in patients with cystic fibrosis-like disease. Human mutation 30, 1093-1103, (2009).

449 Fajac, I., Viel, M., Gaitch, N., Hubert, D. & Bienvenu, T. Combination of ENaC and CFTR mutations may predispose to cystic fibrosis-like disease. The European respiratory journal 34, 772-773, (2009).

450 Nadeau, J. H. Modifier genes in mice and humans. Nature reviews. Genetics 2, 165-174, (2001).