Understanding the Molecular Pathobiology of Acid Ceramidase Deficiency

Fabian Yu

A thesis submitted in conformity with the requirements for the degree of Doctor of Philosophy

Institute of Medical Science University of Toronto

Understanding the Molecular Pathobiology of Acid Ceramidase Deficiency

Fabian Yu

Doctor of Philosophy

Institute of Medical Science University of Toronto

2018

Abstract

Farber disease (FD) is a devastating Lysosomal Storage Disorder (LSD) caused by mutations in ASAH1, resulting in acid ceramidase (ACDase) deficiency. ACDase deficiency manifests along a broad spectrum but in its classical form patients die during early childhood.

Due to the scarcity of cases FD has largely been understudied. To circumvent this, our lab previously generated a mouse model that recapitulates FD. In some case reports, patients have shown signs of visceral involvement, retinopathy and respiratory distress that may lead to death.

Beyond superficial descriptions in case reports, there have been no in-depth studies performed to address these conditions. To improve the understanding of FD and gain insights for evaluating future therapies, we performed comprehensive studies on the ACDase deficient mouse.

In the visual system, we reported presence of progressive uveitis. Further tests revealed cellular infiltration, lipid buildup and extensive retinal pathology. Mice developed retinal dysplasia, impaired retinal response and decreased visual acuity. Within the pulmonary system, lung function tests revealed a decrease in lung compliance. Mice developed chronic lung injury that was contributed by cellular recruitment, and vascular leakage. Additionally, we report impairment to lipid homeostasis in the lungs.

To understand the liver involvement in FD, we characterized the pathology and performed transcriptome analysis to identify gene and pathway changes. We revealed progressive liver injury, inflammation and fibrosis. RNAseq analyses on hepatocytes revealed activation of pathways in inflammation response and cellular recruitment and deactivation of pathways related to lipid metabolism.

MCP-1 is an inflammatory chemokine that is dramatically elevated in ACDase deficient mice and humans. To understand the role of MCP-1 in FD, we created and characterized the

Asah1P361R/P361R;MCP-1-/- double mutant mice. Ablation of MCP-1 attenuated disease and provided a modest extension of life from ~9 weeks to ~14 weeks of age. The greatest reduction in inflammation was detected in the lung. Decreased inflammation in the Asah1P361R/P361R;MCP-

1-/- mice also resulted in less ceramide accumulation in the lung and liver. Taken together, targeting MCP-1 though not effective in all organs may provide a benefit in combination with other therapies until a cure is available for this debilitating disorder.

iii

Acknowledgments

I would like to first and foremost thank my supervisor Dr. Jeffrey Medin, for being encouraging and giving me intellectual freedom to pursue my projects. Thanks to my committee members Dr.

Razqallah Hakem and Dr. Mingyao Liu for their guidance and intellectual input. I would also like to acknowledge all my colleagues and collaborators for their scientific and intellectual contributions.

I am thankful for the generous funding to study and travel to conferences provided during my

PhD from the Institute of Medical Science, the University of Toronto School of Graduate

Studies, the University Health Network Office of Research Trainees, the WORLD Symposium, and the Midwest Athletes Against Childhood Cancer Fund.

For my supportive Parents.

Facta, non verba

Contributions

Chapter 3

Dr. Iris Kassem performed slit-lamp analyses (Figure 9, Figure 15). Ben Sajdak and

Alexander Salmon performed fundus photography, OCT, cSLO and analyzed OCT images for retinal thickness (Figure 9, Figure 10, Figure 11 and Figure 17). Dr. Daniel Lipinski and I performed ERG analyses (Figure 12 and Figure 13). Dr. Jakub Sikora and Jirí Gurka prepared tissue for electron microscopy and performed ultrastructure analyses (Figure 20, Figure 21 and

Figure 22). Children’s Hospital of Wisconsin (CHW) Children’s Research Institute (CRI)

Histology Core performed immunohistochemistry staining (Figure 18, Figure 19, Figure 21 and

Figure 23). Medical University of South Carolina (MUSC) Lipidomic Core identified sphingolipids in the retina (Figure 25).

Chapter 4

Figures and text adapted from Yu et al., 2017 with permission from American Journal of

Physiology- Lung Cellular and Molecular Physiology. Diana Islam performed the lung mechanics tests and blood oxygenation analyses (Figure 26 and Figure 27). Dr. Wolfgang

Kuebler assisted with pulse oximeter measurements (Figure 27). The Centre for Modeling

Human Disease (CMHD) performed the histological and immunohistochemistry staining (Figure

28, Figure 31 and Figure 36). Dr. Jakub Sikora and Jirí Gurka prepared lungs for electron microscopy and performed ultrastructure analyses (Figure 30 and Figure 31). Lucía López-

Vásquez and I performed the Evans Blue dye experiment (Figure 33). Dr. Shaalee Dworski and

I performed the cytokine analyses on BALF samples (Figure 34). Dr. Thierry Levade and I identified the phospholipids and sphingolipids in BALF and lung tissue (Figure 37 and Figure

38).

Chapter 5

The Centre for Modeling Human Disease (CMHD) performed the histological and immunohistochemistry staining (Figure 40, Figure 44 and Figure 45). Dr. Jakub Sikora performed TEM analyses (Figure 41, Figure 42 and Figure 43). Dr. Patricia Turner performed liver injury assessment (Table 10). Medical University of South Carolina (MUSC) Lipidomics

Core identified sphingolipids liver and hepatocytes (Figure 46, Figure 47 and Figure 49). Dr.

Shauna Rasmussen, Dr. Salvatore Molino and I, performed the liver perfusion and hepatocyte isolation. Everett Tate performed FACS analyses (Figure 48, Figure 49, Figure 50, Figure 51 and Figure 52). Dr. Salvatore Molino performed RNA extraction, and qPCR. Transcriptome analyses was performed by the MCW Sequencing library. Dr. Salvatore Molino performed bioinformatic and pathway analyses (Figure 50, Figure 51 and Figure 52).

Chapter 6

Figures and text adapted from Yu et al., 2018 with permission from Scientific Reports.

Dr. Shaalee Dworski and I performed cytokine analyses on serum samples (Figure 66 and

Figure 67), and flow cytometry analyses (Figure 57, Figure 58 and Figure 59). CMHD and CHW-

CRI performed the histological and immunohistochemistry staining (Figure 57, Figure 58, Figure

59, Figure 60 and Figure 62).

Chapter 7

The List of ASAH1 mutations (Table 12) were curated by Dr. Samuel Amintas, Dr.

Thierry Levade and myself. The predicted 3D structure of the mouse and human ACDase

(Figure 68 was developed by Zi Jian Xiong). I annotated it based on color-coded mutations on

PyMOL.

Table of Contents

Acknowledgments ...... iv

Contributions ...... v

Table of Contents ...... vii

List of Abbreviations ...... xvi

List of Tables ...... xxiii

List of Figures ...... xxiv

Chapter 1 Literature Review ...... 1

1.1 Lysosome and Lysosomal Storage Disorders ...... 1

1.1.1 Discovery of the lysosome ...... 1

1.1.2 Function of the lysosome ...... 2

1.1.3 Overview of Lysosomal Storage Disorders ...... 5

1.1.4 Prevalence and affected populations ...... 8

1.1.5 Disease screening ...... 9

1.1.6 Diagnosis and screening ...... 10

1.1.7 Overview of treatment and therapy ...... 11

1.1.8 Hematopoietic stem cell transplantation ...... 12

1.1.9 Enzyme replacement therapy ...... 13

1.1.10 Substrate reduction therapy ...... 15

1.1.11 Gene therapy ...... 16

1.2 Sphingolipids ...... 19

1.2.1 Overview of sphingolipids ...... 19

1.2.2 Structure ...... 19

1.2.3 Source and metabolism ...... 20

1.2.4 Overview of ceramide ...... 22

vii

1.2.5 Ceramide species ...... 23

1.2.6 Complexity in sphingolipid research ...... 24

1.3 Ceramidase ...... 26

1.3.1 Overview ...... 26

1.3.2 Neutral ceramidase ...... 26

1.3.3 Alkaline ceramidase ...... 27

1.3.4 Acid ceramidase ...... 29

1.3.4.1 Characterization of acid ceramidase...... 29

1.3.4.2 Acid ceramidase in cancer and proliferation ...... 30

1.3.4.3 Acid ceramidase in other complex diseases ...... 31

1.4 Acid Ceramidase Deficiency ...... 32

1.4.1 Introduction ...... 32

1.4.2 Traditional classification of FD ...... 32

1.4.3 The diverse signs and symptoms in ACDase deficiency ...... 34

1.4.3.1 Cardinal triad symptoms of FD ...... 35

1.4.3.2 Hematologic findings in FD ...... 37

1.4.3.3 Neurological findings in FD ...... 38

1.4.3.4 Pulmonary findings in FD ...... 39

1.4.3.5 Ophthalmic findings in FD ...... 40

1.4.3.6 Gastrointestinal findings in FD ...... 41

1.4.3.7 Hepatic findings in FD ...... 41

1.4.3.8 Bone findings in FD ...... 42

1.4.3.9 Dermatological findings in FD ...... 43

1.4.3.10 Hydrops fetalis in FD ...... 43

1.4.3.11 SMA-PME ...... 44

1.4.3.12 Keloid formation ...... 45

viii

1.4.4 Phenotypic variability in ACDase deficiency ...... 45

1.4.5 Genetics and diagnosis ...... 47

1.4.5.1 Prevalence of ACDase deficiency ...... 47

1.4.5.2 Genetics and mutations ...... 47

1.4.5.3 Clinical diagnosis ...... 50

1.4.5.4 Biochemical and genetic diagnostics ...... 51

1.4.5.5 Genetic testing ...... 53

1.4.5.6 Biomarkers ...... 53

1.4.6 Research, treatment and future therapy ...... 54

1.4.6.1 Animal models ...... 54

1.4.6.2 Current treatment ...... 56

1.4.6.3 Gene therapy for ACDase deficiency ...... 57

1.4.6.4 Enzyme replacement therapy for ACDase deficiency ...... 59

1.4.7 Concluding thoughts on FD ...... 60

Chapter 2 Aims/Hypotheses ...... 61

2.1 Aim 1: To elucidate the effects of FD in the respiratory and visual systems...... 61

2.2 Aim 2: To elucidate the hepatic manifestations, and corresponding gene expression

changes in FD ...... 62

2.3 Aim 3: To understand the role of MCP-1 in the pathogenesis of FD ...... 62

Chapter 3 Ocular Pathology and Visual Impairment in Acid Ceramidase Deficiency ...... 64

3.1 Abstract ...... 64

3.2 Introduction ...... 65

3.3 Material and Methods ...... 67

3.3.1 Animal use, breeding and genotyping ...... 67

3.3.2 Slit-lamp ...... 67

3.3.3 Fundus imaging and confocal scanning laser ophthalmoscopy ...... 68

3.3.4 Optical coherence tomography ...... 68

3.3.5 Electroretinogram ...... 69

3.3.6 Visual cliff behavioral test ...... 70

3.3.7 Histopathology and eye measurements ...... 71

3.3.8 Retinal dysplasia scoring ...... 71

3.3.9 Immunohistochemistry and immunofluorescence ...... 72

3.3.10 Electron microscopy ...... 73

3.3.11 Sphingolipid mass spectrometry ...... 73

3.3.12 Statistical analyses ...... 74

3.4 Results ...... 76

3.4.1 Non-invasive imaging reveals ocular pathology ...... 76

3.4.2 Impaired retinal and visual function ...... 80

3.4.3 Increased retinal thickness ...... 84

3.4.4 Variable penetrance of retinal dysplasia ...... 87

3.4.5 Inflammation and retinal pathology ...... 89

3.4.6 Optic neuropathy and storage pathology ...... 93

3.4.7 Presence of Astrogliosis ...... 96

3.4.8 Neuroaxonal dystrophy and reduced axonal density ...... 98

3.4.9 Altered sphingolipid profile in the retina ...... 100

3.5 Discussion and conclusion ...... 102

Chapter 4 Chronic Lung Injury and Impaired Pulmonary Function in a Mouse Model of Acid

Ceramidase Deficiency ...... 108

4.1 Abstract ...... 108

4.2 Introduction ...... 110

4.3 Materials and methods...... 112

4.3.1 Animal breeding ...... 112

4.3.2 Pulmonary function test ...... 112

4.3.3 Blood gas, oximetry, and blood counts ...... 112

4.3.4 Lung histopathology and immunohistochemistry ...... 113

4.3.5 Electron Microscopy ...... 114

4.3.6 Alveolar type II cell quantification ...... 114

4.3.7 Western Blots ...... 115

4.3.8 BALF cytospin and differentials ...... 115

4.3.9 BALF turbidity, cytokine analysis and ELISA ...... 116

4.3.10 Vascular permeability and wet-to-dry ratio ...... 116

4.3.11 Phospholipid and sphingolipid analyses ...... 117

4.3.12 Statistical analyses ...... 118

4.4 Results ...... 119

4.4.1 Impaired lung mechanics and decreased blood oxygenation...... 119

4.4.2 Absence of lung fibrosis but presence of lung edema...... 126

4.4.3 Elevation of leukocytes in lung ...... 128

4.4.4 Vascular permeability in multiple organs...... 129

4.4.5 Increased cytokine production in BALF ...... 130

4.4.6 Abnormal lamellar body formation and increased expression of surfactant

proteins ...... 132

4.4.7 Accumulation and disruption of phospholipid composition ...... 134

4.4.8 Ceramide and sphingomyelin accumulation in BALF and lungs ...... 136

4.5 Discussion and conclusion ...... 138

Chapter 5 Liver Injury and Altered Gene Expression Changes in Acid Ceramidase

Deficiency ...... 146

5.1 Abstract ...... 146

5.2 Introduction ...... 148

5.3 Material and Methods ...... 151

5.3.1 Animal use, breeding and genotyping ...... 151

5.3.2 Animal liver weights and collagen assay ...... 151

5.3.3 Serum biochemistry and ELISAs ...... 151

5.3.4 Histopathology and immunohistochemistry ...... 152

5.3.5 Pathologic Evaluation ...... 153

5.3.6 Quantitation of cell death and proliferation ...... 153

5.3.7 Transmission electron microscopy ...... 154

5.3.8 Murine liver perfusion and hepatocyte isolation ...... 154

5.3.9 Flow cytometry analyses of macrophages ...... 156

5.3.10 RNA isolation, library preparation for RNASeq and sequencing ...... 156

5.3.11 Quantitative real time PCR ...... 157

5.3.12 Sphingolipids in liver tissue and hepatocytes ...... 158

5.3.13 Measurement of lipids in liver and serum ...... 159

5.3.14 Statistical analyses ...... 160

5.4 Results ...... 161

5.4.1 Hepatomegaly and liver injury ...... 161

5.4.2 Liver pathology and cellular infiltration ...... 163

5.4.3 Liver storage pathology ...... 165

5.4.4 Progressive liver fibrosis ...... 169

5.4.5 Increased cell proliferation and cell death in the liver ...... 171

5.4.6 Liver injury score ...... 173

5.4.7 Alteration to lipids in plasma and liver tissue ...... 173

5.4.8 Altered sphingolipid profile in the liver ...... 175

5.4.9 Hepatocyte isolation and sphingolipid profile ...... 177

5.4.10 Differential gene expression in hepatocytes ...... 180

xii

5.4.11 Pathway clustering for inflammatory response and lipid homeostasis ...... 180

5.4.12 Perturbed sphingolipid homeostasis ...... 183

5.5 Discussion and conclusion ...... 186

Chapter 6 MCP-1 Impedes Pathogenesis of Acid Ceramidase Deficiency ...... 192

6.1 Abstract ...... 192

6.2 Introduction ...... 193

6.3 Material and Methods ...... 196

6.3.1 Animal use, breeding and genotyping ...... 196

6.3.2 Animal data and organ weights ...... 197

6.3.3 Peripheral blood and serum biochemistry, ELISA, and cytokine analyses ...... 197

6.3.4 BALF turbidity, cytospin & differential ...... 198

6.3.5 Histopathology and immunohistochemistry ...... 199

6.3.6 Flow cytometry of hematopoietic cells ...... 200

6.3.7 Mass spectrometry for sphingolipids ...... 200

6.3.8 Quantification and image analyses ...... 201

6.3.9 Behavioral tests ...... 202

6.3.10 Statistical analyses ...... 202

6.4 Results ...... 203

6.4.1 Deletion of MCP-1 improves the course of FD ...... 203

6.4.2 Peripheral blood cell counts ...... 206

6.4.3 MCP-1 deletion does not normalize hematopoiesis ...... 207

6.4.4 Liver inflammation and injury makers are reduced ...... 211

6.4.5 Reduction of pulmonary infiltrates and protein accumulation ...... 214

6.4.6 Brain and behavioral phenotypes persist in MCP-1 deficient mice...... 217

6.4.7 Behavioral phenotypes unchanged in MCP-1 deficient mice ...... 217

6.4.8 Tissue-specific changes in sphingolipid profiles due to MCP-1 deletion ...... 220

xiii

6.4.9 Unique cytokine signature double mutants ...... 224

6.5 Discussion and conclusion ...... 228

Chapter 7 General Discussion ...... 236

7.1 Evaluating our hypotheses, and the success of our aims ...... 236

7.1.1 Aim 1 ...... 236

7.1.2 Aim 2 ...... 238

7.1.3 Aim 3 ...... 239

7.2 Pathogenesis in the ACDase deficient mouse ...... 241

7.2.1 Phenotypes manifest post weening ...... 241

7.2.2 Pathology is organ-specific ...... 242

7.2.3 Potential role of resident macrophages in FD pathology ...... 243

7.2.4 Limitations of the ACDase deficient murine model ...... 245

7.3 Characterization of sphingolipids in Asah1P361R/P361R mice ...... 248

7.3.1 Sphingolipid accumulation in Asah1P361R/P361R mice ...... 248

7.3.2 Sphingolipids in the respiratory system ...... 248

7.3.3 Sphingolipids in the ocular system ...... 249

7.3.4 Sphingolipids in the hepatic system ...... 251

7.3.5 Complexity and limitations in understanding ceramide metabolism ...... 252

7.3.6 Technical limitations with our lipidomic analysis ...... 254

7.4 Improving understanding of FD ...... 256

7.4.1 Data from FD mouse extends our understanding of human FD ...... 256

7.4.2 FD Diagnosis and screening technology ...... 256

7.4.3 Mutations causing ACDase deficiency ...... 257

7.4.4 Improved FD diagnosis and treatment ...... 262

7.4.5 Current and emerging research in ACDase deficiency ...... 263

Chapter 8 Conclusions ...... 266

xiv

Chapter 9 Future Directions ...... 268

9.1 Pathobiology of the Asah1P361R/P361R mouse ...... 268

9.2 Studies on ACDase protein trafficking and folding ...... 270

9.3 Patient specific induced pluripotent stem cells ...... 272

9.4 Ex vivo gene therapy for FD ...... 273

References ...... 275

List of Abbreviations

AAV9 adeno-associated virus serotype 9

ACDase Acid ceramidase

ACER alkaline Cdase

ACHDNC Advisory Committee on Heritable Disorders in Newborn and Children

ALP alkaline phosphatase

ALT alanine aminotransferase

AMPP aminomethyl phenyl pyridium

ApoA2 apolipoprotein A2

ApoM apolipoprotein M

ARDS acute respiratory distress syndrome

ART automatic real-time

ASAH2 N-acylsphingosine amidohydrolase 2

ASMase acid sphingomyelinase

AST amino aspartate transferase

AUC area under the curve

BA bile acids

BAF blue auto-fluorescence

BALF bronchial alveolar lavage fluid

BCA bicinchoninic acid bFGF basic fibroblast growth factor

BUN blood urea nitrogen

C1P ceramide-1-phosphate

CBC complete blood count

CCR2 chemokine receptor 2

CDase ceramidase

xvi

CerK ceramide kinase

CerS ceramide synthase

Cers2 ceramide synthase 2

Cers4 ceramide synthase 4

CERT cermide transfer protein

CFA Freund's complete adjuvant

CHO chinese hamster ovary

CHW Children's Hospital of Wisconsin

CLN neuronal ceroid lipofuscinosis

CLN2 neuronal ceroid lipofuscinosis type 2

CMHD Centre for Modeling Human Disease

CNS central nervous system

CRI Children's Research Institute

CRP chitotriosidase and C-reactive protein cSLO confocal scanning light opthalmoscope

CTB curved semilinear tubular bodies

Cxcr4 chemokine C-X-C motif receptor 4

DAPI 4’6-diamidino-2-phenylindole

Dgat2 diacylglycerol acyltransferase 2

DSS dextran sulfate sodium

DTIC dacarbazine

EAU experimental autoimmune uveoretinitis

EBD Evans blue dye

EBSS Earle's balanced salt solution

EEG electroencephalogram

EMG electromyogram

ER endoplasmic reticulum

xvii

ERAD ER associated degradation

ERT enzyme replacement therapy

ESI/MS electrospray ionization mass spectrometry

FACS fluorescence-activated cell sorting

FC free cholesterol

FCS fetal calf serum

FD Farber disease

FFA free fatty acid

FiO2 Fraction of inspired O2

FOT frequency force oscillation

GAG glycosaminoglycan

GCL ganglion cell layer

GFAP glial fibrillary acidic protein

GGT gamma-glutamyl transferase

GLD globoid cell leukodystrophy

GM-CSF granulocyte macrophage colony-stimulating factor

H&E hematoxylin and eosin

HDL high-density lipoprotein

Hexβ β-hexosaminidase

HSC hematopoietic stem cells

HSCT hematopoietic stem cell transplantation

IACUC Institutional animal care and use committee

Iba-1 Ionized-calcium-binding adaptor molecule

IF immunofluorescence

IFN-y Interferon gamma

IHC immunohistochemistry

IL interleukin

xviii

ILM inner limiting membrane

INL inner nuclear layer

IP-10 interferon gamma-induced protein 10

IPA Ingenuity pathway analysis

IPL inner plexiform layer iPSC induced pluripotent stem cell

IR infrared

ISI interstimulus internal

JIA juvenile idiopathic arthritis

KC keratinocyte chemoattractant

LC-MS liquid chromatography-mass spectrometry

LFB Luxol fast blue

LSD lysosomal storage disorder

LXR/RXR liver X receptor/retinoid x receptor

M6P mannose-6-phosphate

MALDI matrix-assisted laser desorption ionization

MCHC mean corpuscular hemoglobin concentration

MCOLN1 Mucolipin-1

MCP-1 monocyte chemoattractant protein 1

MCV mean corpuscular volume

MCW Medical College of Wisconsin

MHC monohexosylceramide

MIG monocyte induced by gamma interferon

MIP-1α macrophage inflamamtory protein-1α

MPS Mucopolysaccharidosis

MRM multiple reaction mode

MS mass spectrometry

xix

MS multiple sclerosis

MSV multivesicular bodies mTOR mammalian target of rapamycin

MUSC Medical University of South Carolina

NCDase neutral CDase

NFL nerve fiber layer

NIR near-infrared

NO nitric oxide

NPC Niemann-Pick C ns not significant

OCT optical coherence tomography

ON optic nerve head

ONH optic nerve head

ONL outer nuclear layer

OPL outer plexiform layer

P proline

PaCO2 partial pressure of CO2

PaO2 partial pressure of O2

PAP pulmonary alveolar proteinosis

PBS phosphate buffered saline

PC phosphatidylcholine

PE phosphatidylethanolamine

PEEP positive end expiratory pressure

PFA paraformaldehyde

PI phosphatidylinositol

PIP peak inspiratory pressure

PNS peripheral nervous system

pro-SP-C pro-surfactant protein-C

PS phosphatidylserine

Psap prosaposin

R arginine

RBC red blood cell rd10 retinal degeneration 10 rhACDase recombinant human ACDase

RPE retinal pigmented epithelium

RR respiratory rate

S1P spingosine-1-phosphate

SaO2 arterial oxygen saturation

SAPs sphingolipid activator proteins

SD-OCT spectral domain-optial coherence tomography

SGPL sphingosine 1-phosphate lyase

SM sphingomyelin

SMA-PME spinal muscular atrophy with progressive myoclonic epilepsy

SP surfactant protein

Sph sphingosine

SPT serine palmitoyltransferase

SRT substrate reduction therapy

SSIEM Society for the Study of Inborn Errors

TB total bilirubin

TC total cholesterol

TEM transmission electron microscope

TFEB transcription factor EB

TG triglyceride

TLC total lung capacity

xxi

TNFα tumor necrosis factor alpha

TRPML1 transient receptor potential cation channels, mucolipin subfamily, member 1

TRT total retinal thickness

TUNEL Terminal deoxynucleotidyl transferase dUTP Nick- End Labeling

Ugcg ceramide glucosyltransferase

UHN University Health Network

UPR unfolded protein response

US United States

Vcam1 vascular cell adhesion molecule 1

VEGF vascular endothelial growth factor

VLC very long chain

WBC white blood cell

WES whole-exome sequencing

xxii

List of Tables

Table 1. Partial list of LSDs by class and stored biomolecule ...... 7

Table 2. Selected LSD, with defective enzyme and presenting clinical feature ...... 8

Table 3. Percentage of ERT recipients that report immune reactions...... 15

Table 4. Ceramide synthases, acyl chain length and tissue expression...... 24

Table 5. Ceramidase localization and expression ...... 28

Table 6. ACDase case reports from 1952-2017 by clinical presentation ...... 34

Table 7. Common clinical features in ACDase deficiency ...... 35

Table 8. Number of globes examined ...... 87

Table 9 Yield and quality of RNA for RNA-seq analysis ...... 158

Table 10. Liver Injury Score ...... 171

Table 11. Yield and quality of RNA for RNAseq analysis ...... 180

Table 12. Pathological mutations in ACDase deficiency ...... 260

xxiii

List of Figures

Figure 1. Clathrin-mediated endocytosis...... 3

Figure 2. Schematic of in vivo and ex vivo gene therapy strategies ...... 17

Figure 3. Features in sphingolipids ...... 20

Figure 4. The de novo sphingolipid synthesis pathway ...... 21

Figure 5. Ceramide is an intermediate in sphingolipid metabolism ...... 22

Figure 6. Ceramidase hydrolyzes ceramide into a sphingosine and a free fatty acid...... 26

Figure 7. Main Clinical manifestations of ACDase deficiency...... 37

Figure 8. Structure of the human ASAH1 gene with the protein and distribution of mutations. .. 49

Figure 9. Anterior uveitis and optic nerve pathology in Asah1P361R/P361R mice...... 77

Figure 10. Pathology in fundus of Asah1P361R/P361R mice...... 78

Figure 11. Optical coherence tomography highlights retinal pathology in Asah1P361R/P361R mice. 79

Figure 12. Dark-adapted ERG performed on Asah1P361R/P361R mice demonstrates progressive impairment in rod function in Asah1P361R/P361R mice...... 81

Figure 13. Light-adapted and flicker ERG performed on Asah1P361R/P361R mice demonstrate progressive impairment in cone function...... 82

Figure 14. Visual cliff test reveals perturbed depth perception in Asah1P361R/P361R mice...... 83

Figure 15. Reduced eye opening in Asah1P361R/P361R mice...... 84

Figure 16. No changes in globe diameter, lens diameter, or corneal thickness in Asah1P361R/P361R mice...... 85

Figure 17. Increased retinal thickness in Asah1P361R/P361R mice...... 86

Figure 18. Varying degree of retinal dysplasia in Asah1P361R/P361R mice...... 88 xxiv

Figure 19. Retinal pathology and recruitment of inflammatory cells in Asah1P361R/P361R mice. .... 90

Figure 20. Ultrastructure pathology in ganglion cells and optic nerve disc in Asah1P361R/P361R mice...... 92

Figure 21. Optic nerve pathology in Asah1P361R/P361R mice...... 94

Figure 22. Ultrastructure pathology of the ON of Asah1P361R/P361R mice...... 95

Figure 23. Presence of astrogliosis and activated Müller cells ...... 97

Figure 24. Optic nerve dystrophy and storage pathology ...... 99

Figure 25. Altered sphingolipid species in the retina of Asah1P361R/P361R mice...... 101

Figure 26. Impaired lung mechanics in the Asah1P361R/P361R mice...... 119

Figure 27. Impaired blood oxygenation and polycythemia in the Asah1P361R/P361R mice...... 121

Figure 28. Immune infiltration in Asah1P361R/P361R mice...... 123

Figure 29. Ultrastructural pathology in macrophages and contents of alveolar space in Asah1P361R/P361R mice...... 125

Figure 30. Ultrastructural pathology of pulmonary endothelial cells and respiratory epithelial cells ...... 126

Figure 31. Absence of fibrosis but presence of interstitial edema ...... 127

Figure 32. Immune cell infiltration in bronchoalveolar lavage fluid of Asah1P361R/P361R mice..... 128

Figure 33. Vascular permeability leads to protein leakage in Asah1P361R/P361R mice...... 129

Figure 34. Elevated cytokines in BALF of Asah1P361R/P361R mice...... 130

Figure 35. Unchanged cytokines in BALF of Asah1P361R/P361R mice...... 131

Figure 36. Lamellar body formation and accumulation of surfactant proteins in lungs of Asah1P361R/P361R mice...... 133

Figure 37. Impaired phospholipid homeostasis in BALF of Asah1P361R/P361R mice...... 135

xxv

Figure 38. Significant accumulation of both Ceramide and SM in BALF and lungs of Asah1P361R/P361R mice...... 137

Figure 39. Hepatomegaly, liver injury and perturbed metabolites in Asah1P361R/P361R mice ...... 162

Figure 40. Significant inflammation and pathology in Asah1P361R/P361R mice ...... 164

Figure 41. Storage pathology within the portal fields and liver sinusoid in Asah1P361R/P361R mice ...... 166

Figure 42 Hepatocyte and hepatic stellate cell storage pathology in Asah1P361R/P361R mice ..... 167

Figure 43 Storage pathology in bile duct epithelia of Asah1P361R/P361R mice ...... 168

Figure 44 Progressive liver fibrosis in Asah1P361R/P361R mice ...... 170

Figure 45 Increased cell death and proliferation in liver of Asah1P361R/P361R mice ...... 172

Figure 46 Decreased lipids in serum and liver of Asah1P361R/P361R mice ...... 174

Figure 47 Sphingolipid accumulation in liver of Asah1P361R/P361R mice ...... 176

Figure 48 Reduction of CD11b cells in hepatocyte-enriched cultures from Asah1P361R/P361R mice ...... 177

Figure 49 Sphingolipid accumulation in hepatocyte-enriched cultures from Asah1P361R/P361R mice ...... 179

Figure 50 Differential gene expression and altered pathways in Asah1P361R/P361R mice ...... 181

Figure 51 Misregulated genes in leukocyte extravasation signaling and LXR/RXR activation pathways in Asah1P361R/P361R mice ...... 183

Figure 52 Altered genes in sphingolipid metabolism in Asah1P361R/P361R mice ...... 185

Figure 53. Increased survival and weight gain in Asah1P361R/P361R ;MCP-1-/- mice...... 204

Figure 54. No improvement of disease course in Asah1P361R/P361R ;MCP-1+/- heterozygous mice ...... 205

xxvi

Figure 55. Absolute organ weights of mice are largely unchanged in Asah1P361R/P361R ;MCP-1-/- mice ...... 205

Figure 56. MCP-1 deletion impedes leukocytosis in ACDase-deficiency...... 207

Figure 57. Perturbed hematopoiesis is retained in Asah1P361R/P361R ;MCP-1-/- mice ...... 209

Figure 58. Absence of T-cell population in Asah1P361R/P361R ;MCP-1-/- mice ...... 210

Figure 59. Mild impedance of granulocyte infiltration in Asah1P361R/P361R ;MCP-1-/- mice ...... 211

Figure 60. Delayed signs of liver injury and fibrosis in Asah1P361R/P361R;MCP-1-/- mice ...... 213

Figure 61. Bronchial alveolar lavage fluid (BALF) from Asah1P361R/P361R;MCP-1-/- mice lungs displays lessened signs of pulmonary inflammation and surfactant protein accumulation...... 216

Figure 62. Brain and behavioral deficits persist in Asah1P361R/P361R;MCP-1-/- mice ...... 219

Figure 63. Sphingolipid accumulation and abundance changes are organ specific in the Asah1P361R/P361R;MCP-1-/- mice ...... 222

Figure 64. Monohexosylceramide (MHC) and Ceramide-1-phosphate (C1P) quantification ... 223

Figure 65. Sphingosine (Sph) and sphingosine-1-phosphate (S1P) evaluation ...... 224

Figure 66. Altered cytokines in the serum of Asah1P361R/P361R;MCP-1-/- mice...... 226

Figure 67. Cytokines that were unchanged in serum analyses...... 227

Figure 68. Mutations mapped to predicted ACDase structure ...... 261

xxvii 1

Chapter 1

Literature Review

Section 1.5 of this work is adapted from [Yu et al. 2018] currently in review.

1.1 Lysosome and Lysosomal Storage Disorders

1.1.1 Discovery of the lysosome

Lysosomes are membrane-bound organelles that contain an array of enzymes and serve as the digestive site of a cell. Lysosomes were first identified in 1955 by Christian de

Duve as part of his work on the localization of the blood sugar regulating enzyme glucose-6- phosphatase [Gianetto and De Duve 1955]. This serendipitous discovery came to de Duve and colleagues when they observed higher acid phosphatase activities from cell lysates that were processed with a blender versus a milder technique that kept organelles intact

[Gianetto and De Duve 1955]. De Duve described these new organelle as “saclike structure surrounded by membrane and containing acid phosphatase” [Gianetto and De Duve 1955].

Other hydrolases were later identified within these saclike structures. De Duve gave the new organelle the name “Lysosome” to reflect their lytic function [De Duve and Wattiaux 1966].

De Duve, and his colleagues his colleague Claude, and Palade were awarded the Nobel

Prize in Physiology or Medicine 1974 for this important discovery [de Duve 1975].

1.1.2 Function of the lysosome

Lysosomes are widely abundant in all eukaryotic cell types. They are membrane- enclosed intracellular organelles that measure 0.5 µm in diameter [Luzio et al. 2007].

Lysosomes are primarily responsible for the catabolism of complex lipids, polysaccharides and proteins [Wevers and Gieselmann 2005]. Within the lysosomes there are over 60 different lysosomal hydrolases which include; glycosidases, lipases, proteases, phospholipases, phosphatases, nucleases, and sulfatases [Luzio et al. 2014]. In humans, mutations to these enzymes may cause development of a Lysosomal Storage Disorder

(LSD). In addition to enzymes, lysosomes also contain a variety of membrane proteins, such as proton pumps to regulate the acidic environment (pH 4.6 to 5) within the lumen [Xu and

Ren 2015].

Lysosomal enzymes are first synthesized in the endoplasmic reticulum and then transported to the lysosomes through specialized pathways. Most hydrolases are targeted to the lysosome by a mannose-6-phosphate (M6P) residue signal, and through its respective

M6P receptor [Coutinho et al. 2012]. Degradation of biomaterial in the lysosome can occur through various routes. These include: phagocytosis, macropinocytosis, caveolin-mediated endocytosis, and clathrin-mediated endocytosis [Xu and Ren 2015].

Clathrin-mediated endocytosis (Figure 1) is the most well understood route for endocytosis [Kirchhausen et al. 2014; Bitsikas et al. 2014]. In this pathway biomolecules are first bound to cell specific receptors. After loading of the cargo, fission of the plasma membrane occurs by forming a clathrin-coated pit [Kirchhausen et al. 2014]. Once the vesicle pinches off and uncoats, it will fuse to an early endosome, and after fusion, the endosome undergoes a series of maturation steps resulting in the formation of a late endosome aka, multivesicular bodies (MSVs). After further processing, membrane fusion

3 occurs between the late endosome and lysosome, and degradation of the stored cargo occurs [Le Roy and Wrana 2005].

Recent studies have shown that lysosomes are not solely catabolic organelles, but also contribute to other important roles in a cell. Based on current understanding, lysosomes serve three main functions: 1) degradation, 2) secretion and 3) signaling [Perera and Zoncu

2016]. The catabolic activity of lysosomes largely relies on the various hydrolases for degradation of intracellular and extracellular material, as previously mentioned.

Figure 1. Clathrin-mediated endocytosis.

Schematic diagram of clathrin-mediated endocytosis. The molecule designated for degradation binds to a specific cell surface receptor. Clathrin and other proteins assemble to form a coating around the cargo molecule called the clathrin-coated pit. The clathrin-coated pit grows to form a vesicle where the cargo molecule is internalized. After, the clathrin-coated vesicle uncoats, fuses to the endosome for further processing. Lastly, the cargo is transported into the lysosome for degradation.

Some cells also utilize lysosomes as secretory components for both degradation and storage of proteins [Samie and Xu 2014]. This process called lysosomal exocytosis involves a response to extra stimuli where lysosomes dock at the interior of the cell surface, and fuse to the plasma membrane as a result of Ca2+ elevation [Blott and Griffiths 2002]. The fusion of the lysosome with the plasma membrane has been shown to be an important step for a number of biological processes ranging from innate immunity, bone degradation/resorption by osteoclasts, and coagulation function by platelets [Luzio et al. 2007; Xu and Ren 2015].

While the regulation of how lysosomal exocytosis occurs is still not well understood, this process is in part regulated by the transcription factor EB (TFEB) [Settembre et al. 2011].

TFEB has been demonstrated to transcriptionally regulate lysosomal secretion by inducing release of intracellular Ca2+ by its target gene Mucolipin-1 (MCOLN1) which encodes the transient receptor potential cation channels, mucolipin subfamily, member 1 (TRPML1) [Xu and Ren 2015].

Finally, it is becoming evident that lysosomes are also important signaling hubs for nutrient sensing and maintenance of cellular homeostasis. This role has been shown to involve the kinase complex mammalian target of rapamycin (mTOR), a master growth regulator [Perera and Zoncu 2016]. Studies have shown that when nutrients levels are abundant and stable, mTOR exerts its activity on the lysosomal surface, leading to accumulation of amino acids in the lysosomal lumen to drive mTOR activation [Perera and

Zoncu 2016]. Conversely when nutrient levels are insufficient, mTOR-mediated growth is inhibited is released from the lysosomal surface [Perera and Zoncu 2016].

While studies on the lysosome have highlighted the roles of the lysosme beyond substrate degradation, the remaining section will introduce LSDs, their prevalence, diagnosis, and therapy.

1.1.3 Overview of Lysosomal Storage Disorders

LSDs are a group of more than 50 inherited metabolic disorders (Table 1) [Ferreira and Gahl 2017]. Most LSDs are caused by a deficiency in a single enzyme that is important for proper catabolism of a specific biomolecule such as sphingolipids within the lysosome.

Impaired metabolism will result in the accumulation of the unbroken biomolecule, and in some cases, this can also lead to buildup of other related metabolites (Table 1). While each disorder on its own is rare, as a whole LSDs have a large global impact. LSDs as a group are believed to have an estimated frequency of approximately 1 in 5000 [Fuller et al. 2006].

In addition to defective lysosomal enzyme activity, some LSDs are caused by impaired enzyme trafficking or targeting [Futerman and Van Meer 2004]. This defect can lead to perturbations in the structural proteins or proteins that contribute to proper lysosomal function and biogenesis [Filocamo and Morrone 2011].

The majority of LSDs are inherited in an autosomal recessive manner, however, there are three exceptions: Fabry disease, Dannon disease, and Hunter syndrome

(Mucopolisacaridosis (MPS) type II) which are X-linked recessive [Ferreira and Gahl 2017].

In most cases, hemizygous carriers invariably have 50% enzymatic activity and do not exhibit disease [Wraith 2004]. However, a notable exception is in Fabry disease where some female carriers of the X-linked LSD show symptoms for Fabry disease [Clarke 2007; Wang et al. 2007]. The clinical manifestations of LSDs can vary depending on the disease, however, there are some general features that can be seen in some LSDs (Table 2). These features include hepatosplenomegaly, growth defects, and coarse facial features [Ortolano et al. 2014]. Substrate buildup in the nervous system is prevalent in many LSDs.

Approximately two-thirds of LSDs patients who are untreated will eventually develop some sort of neurological/developmental phenotype [Wraith 2004]. Most LSDs also affect multiple organs and often require the care from a team of physicians and healthcare practitioners.

LSDs are progressive diseases due to the continual accumulation of storage material.

Therefore, if left untreated are fatal and often lead to premature morbidity that can range from infancy to early adulthood depending on the LSD [Parenti et al. 2015].

LSD Classification Disease Stored Biomolecule Glycoproteinosis Aspartylglicosaminuria Aspartylglucosamine Fucosidosis Glycoproteins, glycolipids, fucoside- rich oligosaccharides α-Mannosidosis Mannose-rich oligosaccharides β-Mannosidosis Man(β1-4) GlnNAc Schindler Sialylated/asialo-glycopeptides, glycolipids Sialidosis Oligosaccharides, glycopeptides Glycogenosis Lycogenosis II/Pompe Glycogen Lipidosis Wolman/CESD Cholesterol esters Mucopolysaccha- MPS I (Hurler) Dermatan sulphate, heparan ridosis (MPS) sulphate MPS II (Hunter) Dermatan sulphate, heparan sulphate MPS III A, B, C, D Heparan sulphate (Sanfilippo A, B, C, D) MPS IVA (Morquio A) Keratan sulphate, chondroitin 6- sulphate MPS IV B (Morquio B) Keratan sulphate MPS VI (Maroteaux-Lamy) Dermatan sulphate MPS VII (Sly) Dermatan sulphate, heparan sulphate, chondroitin 6-sulphate MPS IX (Natowicz) Hyaluronan Sphingolipidosis Fabry Globotriaosylceramide Farber Ceramide Gangliosidosis GM1 GM1 ganglioside, keratan sulphate, oligosaccharides, glycolipids Gangliosidosis GM2 (Tay- GM2 ganglioside, oligosaccharides, Sachs) glycolipids Gangliosidosis GM2 GM2 ganglioside, oligosaccharides (Sandhoff) Gaucher Glucosylceramide Krabbe Galactosylceramide Metachromatic Sulphatides leukodystrophy Niemann-Pick Sphingomyelin

Table 1. Partial list of LSDs by class and stored biomolecule Table is adapted from [Filocamo and Morrone 2011].

Disease Defective Enzyme Presenting Signs and symptoms GM1 gangliosidosis Acid β-galactosidase Developmental delay, hepatomegaly, hypotonia, coarse facial features, ataxia, dysarthria, dystonia GM2 gangliosidosis (Tay- Hexosaminidase A Developmental delay, hypotonia, hyperacusis, ataxia, Sachs disease) dystonia, psychoses, psychomotor regression GM2 gangliosidosis Hexaminidase A & B Developmental delay, hypotonia, hyperacusis, ataxia, (Sandhoff disease dystonia, psychoses, psychomotor regression Fabry disease α-Galatosidase A Acroparesthesia, pain cries, corneal opacities, fatigue, angiokeratomas Gaucher disease (type 2 Glucocerebrosidase Developmental delay, hepatosplenomegaly, strabismus, and 3) myoclonus, horizonal supranuclear gaze palsy Niemann-Pick Type A Sphingomyelinase Hepatosplenomegaly, hypotonia, cherry red spot, Krabbe disease Galactocerebrosidase Spastic paraparesis, weakness, burning paresthesia, ataxia, weakness, vision loss, developmental delay MPS I (Hurler) α-L-iduronidase Coarse facial features, developmental delay, dysostosis multiplex, hearing loss, corneal clouding, hernias MPS II (Hunter) Iduronate-2-sulfatase Developmental delay, dysostosis multiplex, hearing loss, coarse facial features, joint stiffness MPS III A (Sanfilippo) Glucosaminase-N- Aggressive behavior, developmental delay, mildly sulfatase coarse facial features, coarse hair, mild dysostosis multiplex Farber Disease Acid ceramidase Painful and swollen joint, nodules, developmental delay, hoarse cry, hypotonia Wolman disease Acid lipase Hepatosplenomegaly, vomiting, diarrhea, anemia, psychomotor regression

Pompe disease α-Glucosidase Hypotonia, developmental delay, cardiac enlargement

Table 2. Selected LSD, with defective enzyme and presenting clinical feature Table Adapted from [Wenger et al. 2003].

1.1.4 Prevalence and affected populations

Prevalence for each individual LSD can be difficult to determine since screening is

only performed for some LSDs and if so only in a handful of regions. Due to limited access

and availability of screening, undiagnosed or misdiagnosed patients is an ongoing concern

[Burke et al. 2011]. Based on a 2006 epidemiological report, the global frequency of LSDs is

9 predicted to be 1 in 5000 [Fuller et al. 2006]. Since LSDs are inherited diseases, the prevalence of each disease is affected by both cultural and geographic variables [Fuller et al.

2006]. One prime example is with Pompe disease, where the frequency in Austria is reported at 1 in 8684 [Mechtler et al. 2012]. Interestingly, in Hungary the adjacent eastern country, the prevalence is twice as high at 1 in 4447 [Wittmann et al. 2012]. Increased incidences in certain ethnicities are common. For instance, Gaucher and Tay-Sachs disease have a high prevalence among the Ashkenazi Jewish population [Triggs-Raine et al. 1990;

Beutler et al. 1993]. Another example is a founder effect for specific mutations associated with Hurler syndrome among people of Scandinavian and Russian decent [Krasnopolskaya et al. 1993; Malm et al. 2008]. Similarly, certain mutations that are linked to clinical manifestation are common, such as the case with Fabry disease. The frequency of the cardiac variant of Fabry disease in Taiwan is reported to as high as 1 in 1600~2450 versus the global prevalence of 1 in 80,000 [Lin et al. 2009; Germain 2010].

1.1.5 Disease screening

Diagnosis and screening for a LSD can occur at different stages of life. Prenatal screening is the earliest form of detection and is usually performed only if parents were previously identified as carriers [Staretz-Chacham et al. 2009]. Screening can also occur right after birth with newborn screening methods, familial screening if a family member has been diagnosed, and lastly when clinical signs manifest [Filocamo and Morrone 2011].

The results of newborn screening have been published by several states in the

United States (US) and have helped to understand individual disease frequency. One example being Washington state, where after screening 110,000 newborns the prevalence of several LSDs were identified. Notably, the report showed the following prevalence data:

Fabry disease affecting males (1 in 7,800), Pompe disease (1 in 27,8000), and Hurler

10 syndrome (1 in 35,500) which was 2-4 times greater than the estimates derived from clinical diagnoses [Scott et al. 2013]. Similar types of reports have been shared by Missouri and most recently Illinois [Hopkins et al. 2015; Burton et al. 2017].

Prevalence of disease reported by screening technology are highly informative but additional considerations of false-positives and false-negatives should be taken into account.

In terms of false-positives, beyond technical factors, many tests are based on chosen cutoffs.

One study has suggested that one of the most important metrics is the measurement of the ratio of mean enzyme activity between non-diseased controls and LSD-affected samples

[Gelb et al. 2015].

In the context of the US, results from the aforementioned reports have been important for organizations like the Advisory Committee on Heritable Disorders in Newborns and Children (ACHDNC) to make recommendations for disorders to be added to the

Recommended Uniform Screening Panel [Ferreira and Gahl 2017]. Although recommendations are made on a national level, implementation is ultimately decided and enforced at the state level. Nonetheless, Pompe disease was added to the RUSP in 2015, and Hurler syndrome in March 2016 [Matern et al. 2015]. A host of other LSDs have been considered as possible conditions, but one main caveat is the lack of a perfected test or effective treatment options for some LSDs [Matern et al. 2015; Ferreira and Gahl 2017].

1.1.6 Diagnosis and screening

A variety of techniques may be used to diagnose LSDs which include 1) genetic sequencing for mutations, 2) direct measurement of enzyme activity by biochemical tests, 3) metabolic testing for unmetabolized substrates (typically in urine or blood) via mass spectrometry and 4) assaying biomarkers [Wenger et al. 2003; Aerts et al. 2011].

LSDs are also very often diagnosed during adulthood. One recent survey conducted by the Society for the Study of Inborn Errors (SSIEM) of the Adult Metabolic Physicians working group stated that patients with LSDs were frequently diagnosed as adults [Sirrs et al.

2015]. Based on their results, the most prevalent adult patients were Fabry disease (8.8%),

Gaucher disease (4.2%), and of all diagnoses mentioned in the report 25% were in the category of LSDs [Sirrs et al. 2015]. Furthermore, of these LSD patients close to 50% of the patients were diagnosed only as adults [Sirrs et al. 2015]. This indicates the importance of knowledge translation in promoting LSD awareness to primary care clinicians.

In regions where LSD are common, carrier screening may be performed. During pregnancy, parents can be screened to identify whether they are carriers for a disease. If the parents are carriers, they may choose to further screen their embryo as part of their genetic counseling [Burke et al. 2011]. Carrier screening was first successfully implemented to evaluate prevalence of Tay-Sachs disease in the Ashkenazi Jewish population [Burke et al. 2011]. Several known mutations in the HEXA gene have been identified for causing Tay-

Sachs disease, a lethal LSD ( ). Individuals of Ashkenazi Jewish descent have been identified as having a higher frequency of being a carrier 1 in 30 versus 1 in 167 for the general population [Blitzer and McDowell 1992]. The success of Tay-Sachs carrier screening has used as a model for other LSDs and inherited disorders such as cystic fibrosis and spinal muscular atrophy [Blitzer and McDowell 1992; Burke et al. 2011].

1.1.7 Overview of treatment and therapy

Currently there is no cure for LSDs. Hematopoietic stem cell transplantation (HSCT) enzyme replacement therapy (ERT), and substrate reduction therapy (SRT) are three treatment approaches that are available for some LSDs. However, for those LSDs with no

12 approved therapy, the main treatment is through the management of patient symptoms

[Cohen et al. 2016].

1.1.8 Hematopoietic stem cell transplantation

The principle of HSCT is that transplantation of donor hematopoietic stem cells (HSC) can replace the hematopoietic system with cells that produce functional enzyme. Between

1980 to 2013, over 2000 patients diagnosed with a LSD have received HSCT [Boelens et al.

2014]. In fact, HSCT is now standard practice in the management of Hurler syndrome (MPS

I), adult onset metachromatic leukodystrophy, and early onset globoid cell leukodystrophy

[Aldenhoven et al. 2014]. Most of the LSD HSCT experience can be derived from Hurler syndrome, where clinicians have had positive outcomes. Not only can HSCT be performed under the age of 2, but after engraftment, patients showed rapid decline in the stored substance glycosaminoglycan (GAG). While the skeletal defects that are common to Hurler syndrome do not change, treated patients have shown significant improvements in cognitive function [Peters et al. 1998; Boelens et al. 2007].

Early intervention is important in the treatment of LSDs. One example is in Niemann-

Pick C (NPC) type 2. Patients who received HSCT at 16 months were able to abolish the respiratory complications that are commonly developed at 5 years of age if untreated [Breen et al. 2013]. Another example would be Hunter syndrome (MPS II) where HSCT can ameliorate and stabilize peripheral motor symptoms but showed minimal improvement to the neurological symptoms [Guffon et al. 2009].

The efficacy of HSCT is linked to the level of enzyme that is produced and secreted by the donor cells. One study showed that children with Hurler syndrome that received cells that had high enzyme expression showed better growth and lessened the need for future

13 surgical intervention compared to those who received cells form low expressing donors

[Aldenhoven et al. 2008].

HSCT indeed has shown positive outcome for the management of HSCT particularly in the recovery of non-neurological symptoms. However, there is always inherent risk with

HSCT, which include anemia, infection, and possible failure of the engraftment [Boelens et al. 2013]. However, as newer transplant methods are being developed, improved safety and utility of HSCT in LSD can be expected.

1.1.9 Enzyme replacement therapy

Enzyme replacement therapy (ERT) was first hypothesized in 1964 when Henri Hers discovered that Pompe disease was caused by a deficient enzyme [Hers 1963]. Five years later, Elizabeth Neufeld and her group showed proof-of-concept by correcting enzyme activity in fibroblasts derived from a Hurler syndrome patient by adding cultured media from cells of a patient with Hunter syndrome, and vice versa [Fratantoni et al. 1968]. The principle of ERT stems from the phenomenon that the fraction of lysosomal enzymes that do not enter a lysosome can be secreted from a cell and then recaptured by other cells through the

M6P or other receptors [Coutinho et al. 2012]. This mechanism of action allows for successful exogenous treatment of enzyme.

ERT has been developed for a handful of LSDs which include: Gaucher, Fabry,

Pompe, MPS I, II, IVA and VI [Parenti et al. 2015; Coutinho and Alves 2016]. ERT is also currently being developed for FD (Section 1.5.6) [He et al. 2017; Gaukel et al. 2018b].

Presently patients who undergo ERT receive weekly or biweekly infusions of enzyme that is generated in vitro. Currently, ERT has been shown to be most successful in the management of Gaucher disease which happens to also be the first LSD where ERT was approved [Brady et al. 1974]. Repeated infusions of enzyme were able to attenuate the

14 progression of disease and correct damage [Barton et al. 1990; Barton et al. 1991]. On the other spectrum ERT for Pompe disease has shown variable success [Banugaria et al. 2011;

Banugaria et al. 2013]. ERT treatment in Pompe has shown inefficient enzyme penetration in neurons, and also rapid development of antibodies against the infused enzyme

[Banugaria et al. 2011; Banugaria et al. 2013].

Despite the efficacy of ERT there are limitations. For example, ERT cannot reverse the effect of disease, it is only a means to manage and stabilize further progression.

Variability in enzyme penetration is different depending on the tissue type. Treatment of central nervous system (CNS) by ERT is also limited as the infused enzyme generally will not cross the blood brain barrier [Beck 2007; Coutinho and Alves 2016]. Furthermore, ERT treatment is short-lived, which means repeated infusions are required for the patient’s life. In addition to decreased quality of life, ERT can also be a financial burden as treatments may cost upward of $300,000 per year [Das et al. 2017].

Lastly, one of the main limitation of ERT is the possibility of rejection due to the development of antibodies against the recombinant enzyme (Table 3). Immune reaction is both disease- and patient-specific [Kishnani et al. 2016]. However, there is data showing that patients with certain LSDs consistently develop antibodies such as in Pompe disease and require immunomodulation along with their regularly enzyme infusion. In contrast, ERT is well received in Gaucher patients and less patients develop antibodies [Kishnani et al.

2016].

While there are certain drawbacks, ERT is still a good treatment modality for other

LSDs. Additionally, ongoing work is being conducted to improve current treatments.

Examples include improving delivery approaches with nanocarriers to better control the release of enzyme and investigating the utility of plant-based cells to produce enzyme [Hsu et al. 2011; Grabowski et al. 2014; Shaaltiel et al. 2015].

% with IgG Generic Compan Recombinant Trade antibody LSD name y enzyme name formatio n Agalsidase Sanofi Fabrazyme 68% beta Genzyme Fabry Α-galactosidase A Agalsidase Replagal Shire 64% alpha Sanofi Imiglucerase Cerezyme 15% Genzyme β- Gauche Velagluceras glucocerebrosidas VPRIV Shire 1.9% r e alpha e Taliglucerase Elelyso Pfizer 53 alfa BioMarin Aldurazyme MPS I α-L-iduronidase Laronidase / Sanofi 97% , Genzyme Iduronate-2- MPS II Idursulfase Elaprase Shire 47% sulfatase Alglucosidase Sanofi Myozyme 95% alfa Genzyme Pompe Acid α-glucosidase Alglucosidase Sanofi Lumizyme 100% alfa Genzyme

Table 3. Percentage of ERT recipients that report immune reactions. Adapted from [Kishnani et al. 2016].

1.1.10 Substrate reduction therapy

The principle of substrate reduction therapy (SRT) was proposed by Norman Radin when he hypothesized that Gaucher patients could theoretically be treated with a drug/compound to slow the synthesis of glucosylceramide, the lipid that is built up [Radin

1996]. Independently, Frances Platt and Terry Butters recognized the ability of the imino sugar N-butyldeoxynojirimycin to inhibit ceramide glucosyltransferase and showed proof of

16 concept in cells [Platt et al. 1994; Aerts et al. 2006]. After pivotal work on animal models and a successful clinical trial [Platt et al. 1994; Cox et al. 2000] N-butyldeoxynojirimycin or

Miglustat has been approved for use in Europe and US for the treatment of Gaucher disease with mild to moderate symptoms [Beck 2007]. Lastly, SRT has also been approved for patients with Niemann Pick Type-C, and infantile nephropathic cystinosis [Platt et al. 2012;

Ortolano et al. 2014].

1.1.11 Gene therapy

Gene therapy is the delivery of genetic material into cells and tissues to treat an acquired or inherited disease. One approach to gene therapy is using viruses such as lentivirus and adenovirus, to introduce a vector containing a normal copy of a gene [Ortolano et al. 2014]. When a vector is delivered to a patient’s own enzyme deficient cells, the corrected copy of the gene should be able to produce physiological levels of enzyme, and if the gene is introduced to normal unaffected cells, these modified cells may be able to produce supraphysiologic levels of enzyme [Naldini 2015]. LSDs are excellent candidates for gene therapy since many of them are single gene disorders, and in some cases only 15-

20% of normal enzyme activity is sufficient for clinical efficacy [Sands and Davidson 2006].

There are two approaches to viral based gene therapy. The first is direct injection of virus into a host known as in vivo gene therapy. The second transduction in donor cells that eventually get transplanted back into a patient defined as ex vivo gene therapy (Figure 2).

Figure 2. Schematic of in vivo and ex vivo gene therapy strategies

In vivo gene therapy occurs by directly treating patients with a therapeutic vector to express a gene of interest. The ex vivo gene therapy approach first involves the collection of patient donor cells, then correcting collected cells with a therapeutic vector that carries the gene of interest, expanding the cells and lastly transplanting the corrected cells back into a patient. Figure is adapted from [Collins and Thrasher 2015].

Early in vivo gene therapy studies have relied on the liver and lung as sites for continuous secretion of enzyme into the bloodstream for cross correlation of other peripheral organs [Beck 2007]. This has been achieved in animal studies with efficacy with the adeno- associated, and retrovirus system [Mango et al. 2004; Ziegler et al. 2002]. However, transducing the liver or surrounding organs only provided enzyme correction to the peripheral organs, as enzyme would not penetrate the blood-brain barrier [Beck 2007].

Since many LSDs have a neurological component, new vector delivery systems have been developed to specifically transfer genes into the central nervous system. One such example

18 is work performed on the LSD neuronal ceroid lipofuscinosis (CLN) using the adeno- associated virus (AAV) system [Passini et al. 2006]. A series of successful studies showed that delivery of AAV expressing the human CLN2 gene into Cln2 mutant mice could reduce storage material, and these positive results prompted a clinical trial [Passini et al. 2006;

Worgall et al. 2008]. In this trial of 10 children with CLN, one patient died 49 days after the procedure. However, the majority of patients demonstrated a significant attenuation in neurological decline when tested 18 months after treatment [Worgall et al. 2008]. While these results are promising, additional work will be required to better understand the long- term efficacy of AAV gene therapy since this is a non-integrating vector system.

Nonetheless AAV gene therapy for CNS related LSDs has prompted great interest and at the date of writing this manuscript there are over 35 AAV based gene therapy trials listed on clinicaltrials.gov categorized under the condition of nutritional and metabolic diseases.

Ex vivo gene therapy is an alternative strategy of gene therapy that may provide a longer lasting treatment. In this approach, patient stem cells (commonly HSCs) are collected, gene augmented, expanded and then transplanted back into the patient [Beck 2007]. When

HSCs are infected with a viral vector that integrates such as a lentivirus, all daughter cells will also carry and express a functional copy of the therapeutic gene [Naldini 2011]. Proof of concept of ex vivo gene therapy has been demonstrated in various mouse models such as metachromatic leukodystrophy and Fabry disease mice [Biffi et al. 2006; Pacienza et al.

2012]. In these studies, full reconstitution of enzyme activity was achieved as well as a reduction in storage product [Biffi et al. 2006; Pacienza et al. 2012]. From these and other similar studies there are now several ongoing clinical trials exploring the treatment option for

LSD patients. Most notably work in our lab has progressed to a phase 1 clinical trial for the correction of Fabry disease with HSC expressing α-galactosidase A indexed in clinicaltrials.gov (ID NCT02800070) [Huang et al. 2017]. The goal of this therapy is to

19 introduce cells that can produce functional enzyme and degrade the sphingolipid (discussed in greater detail in Section 1.2) substrate globotriaosylceramide.

While gene therapy is promising, there are major hurdles before it will become common practice. Rigorous assessment of the safety and efficacy concerns will be imperative before any gene therapy will be fully acceptance by the medical community.

1.2 Sphingolipids

1.2.1 Overview of sphingolipids

Sphingolipids are named after the great Sphinx, the mythical Greek guardian of

Thebes. The use of “sphingo” was coined by Johann Ludwig Wilhem Thudichum in 1884, who discovery this class of lipids [Nyberg et al. 1998]. This reference was made because sphingolipids were enigmatic, and little was known about them [Nyberg et al. 1998].

Since their discovery sphingolipids have been shown to have wide range of functions.

Most notably, sphingolipids are a vital component of the lipid barrier in the cell membranes of eukaryotes. Sphingolipids have also been found present in plants, fungi, and selected prokaryotic organisms [Pruett et al. 2008].

1.2.2 Structure

Sphingolipids are amphipathic molecules that contain both hydrophobic and hydrophilic components. They contain a long chain base (typically an 18-carbon-chain) consisting of either sphinganine or sphingosine that is attached to a fatty acid by an amine bond. The hydrophobic base may consist of either sphingosine, sphinganine, dihydrosphingosine or phytosphingosine and it is attached to a fatty acid by an amine bond to carbon 2 [Futerman and Hannun 2004].

Sphingolipids are differentiated by their hydrophilic head group, hydroxylation, and the length of both the base and fatty acid chain (Figure 3). Therefore, hundreds if not thousands sphingolipid permutations can be synthesized by the different hydrophobic and hydrophilic combinations. To date, 620 unique sphingolipids have been identified in various mammalian cells [Fahy et al. 2009; Quehenberger et al. 2010; Aimo et al. 2015].

Figure 3. Features in sphingolipids

The sphingolipid ceramide-1-phosphate (C1P) with the various annotated features. Sphingolipids can be modified by addition of head groups, degree of saturation and length of fatty acid chain.

1.2.3 Source and metabolism

Sphingolipids can originate from three main routes: The most prevalent is the de novo pathway (Figure 4), followed by conversion from other sphingolipids, and sphingolipids that are obtained through diet [Lai et al. 2016]. De novo sphingolipid synthesis begins at the cytosolic side of the endoplasmic reticulum (ER). Condensation of L-serine and palmitoyl-

CoA occurs to form 3-ketosphinganine. This step is then catalyzed by serine palmitoyltransferase (SPT) and reduced to sphinganine. Afterwards, sphinganine is acylated to create dihydroceramide via dihydroceramide synthases which is also commonly referred to as ceramide synthase (CerS). Following acylation, a double bond is introduced to the sphingoid base to form ceramide. Ceramide is then transported from the ER to the Golgi apparatus by vesicular transport or ceramide transfer protein (CERT) where it can later be converted to more complex sphingolipids. In the reverse, the only known ‘exit’ from the sphingolipid metabolic pathway is through the degradation of sphingosine-1-phosphate (S1P) with S1P lyase [Lai et al. 2016; Hannun and Obeid 2017].

Figure 4. The de novo sphingolipid synthesis pathway

In the ER, condensation of L-serine and palmityol-co-A forms 3-ketosphinganine which is catalyzed by SPT to form sphinganine. Sphinganine is then acylated to form dihydroceramide, and later ceramide. Ceramide can be transported into the Golgi to form more complex glycosphingolipids. Within the ER, ceramide can be degraded into sphingosine, and later processed into S1P. Within the lysosome ceramide can be converted to both sphingosine and sphingomyelin. Adapted from [Ogretmen and Hannun 2004; Deevska and Nikolova- Karakashian 2017]

1.2.4 Overview of ceramide

Ceramide is the central molecule within the sphingolipid metabolism pathway, where it serves as an intermediate when one sphingolipid is metabolized into another (Figure 5).

Ceramide can be converted its metabolites through the addition of various head groups, such as the addition of a phosphocholine group for sphingomyelin (SM), glucose for glucosylceramide, galactose for galactosylceramide, and phosphate for C1P [Hannun and

Obeid 2017].

Figure 5. Ceramide is an intermediate in sphingolipid metabolism

A schematic depicting the metabolites of ceramide. Ceramide can form sphingomyelin (SM), sphingosine (Sph), ceramide-1-phosphate (C1P), and glucosylcerbroside. Consequently, each of the aforementioned sphingolipids can be metabolized to form ceramide.

1.2.5 Ceramide species

The term ceramide is unfortunately deceptive, as it may mislead the reader in thinking that ceramides are a single class of lipids. The reality is that dozens if not hundreds of unique species exist. The diversity of ceramides is in part compounded by the variation in the acyl chain length from their interaction with the six CerS [Levy and Futerman

2010]. Additionally, longer fatty acids may be generated through further reactions with a family of elongases [Guillou et al. 2010]. Taken together, ceramides are not a single species,

24 but should be considered as a diverse family of sphingolipids that contain similarities in structure but may also serve distinct functions [Hannun and Obeid 2008].

From the de novo pathway, ceramides are formed via the addition of an acyl chain to the base sphingosine. The length and amount of saturation on the acyl chain determines the uniqueness of each species of ceramide [Levy and Futerman 2010]. In mammals, the most abundant ceramides are the C16 and C18 species. These ceramides are catalyzed by all of the CerS with the exception of CerS2 [Levy and Futerman 2010].

The function of some ceramides is not well understood, however, form does appear to correlate with function. Namely, CerS3 is expressed in higher levels in the dermis and epidermis, they are responsible for the formation of very long chain (VLC) ceramides.

Functionally these VLC ceramides are found in high concentrations in the skin and appear to play an important role in the formation of lipid bilayer [Feingold 2007].

Different tissues contain distinctive ratios of fatty acids that coincide with the ceramides needed (Table 4). The C22 and C24 fatty acid chain-lengths are the most abundant in eukaryotes, however, ultra-long-chain fatty acids that are C26 and longer are exclusively located in regions where insulation and protective layers are needed. These include the regions in the brain, skin, retina, testes and brain [Sassa and Kihara 2014].

C14 C16 C18 C20 C22 C24 C26 C28 C30 C32 Organ Brain, muscle CerS1 P

Ubiquitous, high in CerS2 P P liver, kidney, lung Skin, testis CerS3 P P P P

Heart, lung, Liver CerS4 P P

Ubiquitous, high in CerS5 P brain, kidney, testis Ubiquitous, high in CerS6 P brain, liver, thymus

Table 4. Ceramide synthases, acyl chain length and tissue expression. The preferred fatty acid chain length catalyzed and tissue distribution of mammalian ceramide synthase (CerS) isozymes. Adapted from Grosch et al., 2012 and Sassa & Kihara, 2014.

1.2.6 Complexity in sphingolipid research

The study of sphingolipids has historically and technically been a struggle. Originally, sphingolipids were only thought of as components of membranes and limited to only providing a physical barrier for organelles and cells [Futerman and Hannun 2004]. However, recently their role in biology as signaling lipids has been acknowledged and there is increased interest in studying them.

From a technical perspective, there has been less innovation in the methods and protocols created for lipid extraction and manipulation compared to those that are available for the study of proteins and DNA in biological research [Glatz 2011]. In terms of techniques, amplification of nucleic acids, expansion of proteins and purification can be routinely performed in many labs. There is no amplification equivalent to PCR for lipids, and antibodies for lipids are both rare and not nearly as specific as those that bind to proteins.

Histological stains for lipids are available for certain categories of lipids. For example, Sudan

Black B, and Oil Red O the most commonly used histological stains are only specific for neutral triglycerides [Kiernan 1999]. There are no specific stains for sphingolipids, let alone

26 ceramides. While a few ceramide specific antibodies have been produced, some of which are available on the market, none are specific to any one particular acyl chain length

[Vielhaber et al. 2001; Kawase et al. 2002; Krishnamurthy et al. 2007; Cowart et al. 2002].

One commercially available alternative is sphingolipid analogues that contain fluorescently- labelled tags that can be used in in vitro trafficking studies [Mason 1999].

Lipidomic analysis with mass spectrometry is currently the most sensitive method for the identification of not only different lipid classes but also the individual species within a lipid category [Bielawski et al. 2010]. Currently lipidomic methods are still in its infancy due to limitation of lipid standards and proper technical protocols [Loizides‐Mangold 2013]. In fact, we have previously published a technique similar to immunohistochemistry for the detection of lipids called matrix assisted laser desorption/Ionization mass spectrometry

(MALDI-MS) [Jones et al. 2014].

One other challenge in the study of sphingolipids is the tight homeostatic balance that exists for sphingolipids to be converted and transformed [Hannun and Obeid 2017]. In normal conditions if one sphingolipid is in abundance it will be converted to another sphingolipid. This is especially true for ceramide since it lies in the central spot in sphingolipid metabolism [Futerman and Hannun 2004]. One exception that exists to this rule is in diseased states such as those seen in LSDs, where lipid storage leads to a host of deleterious downstream effects [Ferreira and Gahl 2017].

In sum, as a result of the technical and biological difficulties, ceramide and other sphingolipids have not been as widely studied as certain proteins. In this thesis, we illustrate the detrimental effects of chronic overabundance of ceramides in different organs through a series of analyses on the acid ceramidase (ACDase) deficient murine model. An introduction to the different mammalian ceramidase follows.

1.3 Ceramidase

1.3.1 Overview

Ceramidase (CDase) (Enzyme commission number 3.5.1.2.3) are a class of hydrolase enzymes that hydrolyze ceramide to produce sphingosine and a free fatty acid

(Figure 6). There are three classes of CDase that are distinguished by their pH for optimal enzyme activity. These three classes include neutral, alkaline and acid CDase.

Figure 6. Ceramidase hydrolyzes ceramide into a sphingosine and a free fatty acid.

1.3.2 Neutral ceramidase

Neutral CDase (NCDase) or formally N-acylsphingosine amidohydrolase 2 (ASAH2) was first isolated and purified from rat brains in 1999 and 1 year later cloned and characterized as a 763-amino acid protein [El Bawab et al. 1999; El Bawab et al. 2000]. NC hydrolyzes ceramide at an optimal pH of 6.5-8.5. ASAH2 is ubiquitously expressed at low levels compared to ASAH1, but is expressed highest in the kidney, small intestine, colon and liver (Table 5) [Coant et al. 2017].

Early studies on NCDase deficient mice report a normal phenotype but were unable to degrade dietary sphingolipids [Kono et al. 2006]. This suggested that NCDase may have a role in the regulation of bioactive sphingolipids within the digestive tract [Kono et al. 2006].

Furthermore, NCDase has shown to potentially have a protective role against inflammation.

One study using the dextran sulphate sodium inflammation model showed that by knocking down ASAH2 it led to an increased level of S1P and increased inflammation in the intestinal tract [Snider et al. 2012]. This observation expands on an earlier study that showed overexpression of ASAH2 was able to block apoptosis caused by TNF-α induced ceramide accumulation in primary hepatocytes [Osawa et al. 2005a].

1.3.3 Alkaline ceramidase

Alkaline CDase (ACER) is a group of three ACER enzymes that were initially identified in yeast [Mao, Xu, Bielawska, Szulc et al. 2000; Mao, Xu, Bielawska and Obeid

2000]. ACER1 is a 264-amino acid protein with a molecular weight of 31 kDa. ACER1 has an optimal pH of around 8.0 and is primarily expressed in the skin and specifically epidermal keratinocytes (Table 5). Studies have shown that ACER1 is primarily localized within the endoplasmic reticulum and is essential for the homeostasis of skin function [Liakath‐Ali et al.

2016].

ACER2 is an enzyme with 275-amino acids and an expected molecular weight of

31.3 kDa [Xu et al. 2006]. ACER2 is ubiquitously expressed albeit at lower levels compared to ACER1 [Sun et al. 2010]. ACER2 is preferentially localized in the Golgi complex (Table

5) and has been shown to have a role in DNA damage response. A study found that DNA damage induced by doxorubicin increased sphingosine levels via upregulation of

ACER2. Cell death from DNA damage was consequently reduced after knocking down

ACER2 [Xu et al. 2016].

ACER3 was cloned in 2001 and is a 267-amino acid protein with a molecular weight of 31.6 kDa. ACER3 is localized in both the ER and Golgi apparatus and is highly expressed in the placenta (Table 5) [Coant et al. 2017]. One study demonstrated that ACER3 preferentially hydrolyzes unsaturated long chain ceramides [Hu et al. 2010]. In the same study, they showed that knockdown of ACER3 reduced cell proliferation and upregulated the cyclin-dependent kinase inhibitor p21 [Hu et al. 2010]. This suggests that ACER3 may participate in cell cycle and proliferation. Additionally, ACER3 may have a role in suppressing inflammation in the digestive tract. One recent study showed that dextran sulfate sodium (DSS)- induced colitis in Acer3 mutant mouse showed aggravated colitis and colitis-associated colorectal cancer [Wang et al. 2016].

Gene Cellular pH mRNA Expression localization optimum ASAH1 Lysosome 4.5-5 Ubiquitous (high)

ASAH2 Plasma 7.0 Ubiquitous (low) membrane Highest in kidney, small intestine, colon and liver ACER1 ER 8.5 High in skin ACER2 Golgi apparatus 9.0 Ubiquitous (low) Highest in placenta ACER3 ER and 9.5 Ubiquitous (high) reticulum/Golgi

Table 5. Ceramidase localization and expression

Summary of the cellular localization, optimal pH activity, and tissue distribution of the five-known ceramidase in mammals. Table adapted from [Mao and Obeid 2008]

1.3.4 Acid ceramidase

1.3.4.1 Characterization of acid ceramidase

ACDase was first identified in 1963 by Gatt in rat brain extracts, where he demonstrated that ACDase was the catalyst for the hydrolysis of the amide bond of ceramide [Gatt 1963]. The optimal pH of ACDase is 4.5-5. Due to the low pH, it has been suggested that the enzyme may have a role in the lysosomal system (Table 5) [De Duve

1959]. The first large purification of the enzyme was not performed until 1995 using human urine samples [Bernardo et al. 1995]. The purified enzyme was later identified as a heterodimer consisting of an α (13 kDa) and a β (40 kDa) subunits. Studies utilizing the first anti-ACDase polyclonal antibody revealed that ACDase is initially synthesized as a precursor polypeptide and then post-transcriptionally modified and processed into the α and

β subunits within the lysosome [Ferlinz et al. 2001]. These studies also revealed that cleavage into its subunits is essential for enzymatic activity. Later studies using rhACDase showed that cleavage of the precursor polypeptide occurs through an autoproteolytic reaction that is dependent on the cysteine residue 143 [He et al. 2003; Shtraizent et al.

2008]. ACDase, like other enzymes, also exhibits a reverse reaction, in which ACDase can use C12 fatty acid and sphingosine to form ceramide at a pH of 6 rather than the lower pH of

4.5 [Okino et al. 2003]. Similar to other acidic hydrolases, ACDase is tagged with a mannose-6-phosphate residue for transport to the lysosomal compartment. ACDase was confirmed to be localized in the lysosome and is ubiquitously expressed but has been found to be highly expressed in the heart and kidney with lower expression in placenta, lung and muscles [Li et al. 1999].

1.3.4.2 Acid ceramidase in cancer and proliferation

Ceramide has been shown to regulate cell cycle arrest, differentiation, and senescence, however, most of the research has been focused on its role in cancer

[Kolesnick and Krönke 1998; Ogretmen and Hannun 2004; Saddoughi et al. 2008].

Commonly, ceramide is attributed as a pro-apoptotic lipid, while S1P lipid that drives pro- survival [Wymann and Schneiter 2008]. Moreover, since ACDase is a key hydrolase in the maintenance of sphingolipid balance, it has also been implicated in cancers.

ACDase has been shown to be involved in a variety of cancer types ranging from head and neck [Mehta et al. 2000], prostate [Seelan et al. 2000], to breast cancer

[Ruckhaberle et al. 2009]. One study showed that over expressing ACDase in a prostate cancer cell line increased autophagy and lysosomal density allowing for a survival advantage [Turner et al. 2011]. Another study showed treatment of human A375 melanoma cells with dacarbazine (DTIC) resulted in a decrease in ACDase expression as a consequence of reactive oxygen species activation by Cathepsin B. Overexpression of

ACDase conferred resistance to DTIC and knockdown was able to sensitize the cells to treatment [Bedia et al. 2011].

These and other similar studies have prompted investigators to target ACDase in the treatment of cancers. One study showed that use of N-oleoylethanolamine to inhibit ACDase was able to increase ceramide levels and presence of cell death in the mouse L929 fibrosarcoma cell line [Strelow et al. 2000]. Another study showed that combined treatment of traditional antitumor drugs along with carmofur on the colon adenocarcinoma SW403 cell line resulted in a synergistic effect (Realini et al., 2013). Lastly, a recent study similarly showed that treatment of ceranib-2, a small molecule inhibitor of ACDase could decrease cell viability in a dose-dependent manner in the MCF7 breast cancer cell line, presumably due to increased ceramide [Vejselova et al. 2016].

1.3.4.3 Acid ceramidase in other complex diseases

In addition to cancer, ACDase also appears to be involved in the pathology of other complex diseases. A series of studies have demonstrated the role of ceramide in mediating insulin resistant-based diabetes [Chaurasia and Summers 2015]. Early studies have demonstrated that ceramide and its metabolites may antagonize insulin signaling [Summers and Nelson 2005]. This was reinforced when a study showed that inhibition of de novo synthesis of ceramide with myriocin (an inhibitor of the enzyme serine palmitoyltransferase) was able to improve glucose tolerance and lower ceramides in gluococortoid-treated rats

[Holland et al. 2007]. Another study demonstrated that the insulin-sensitizing peptide hormone adiponectin stimulated ceramidase activity. Treatment of recombinant adiponectin in obese mice lowered ceramides and resulting in a beneficial metabolic effect [Holland et al.

2011]. A recent study from the same group demonstrated that overexpression of ACDase in the liver of mice fed a high fat diet resulted in reduced ceramide accumulation, better insulin sensitivity, and prevented hepatic steatosis [Xia et al. 2015]. These studies have demonstrated the potential to use ACDase as a therapeutic target in the management of metabolic disorders.

In humans, mutations to ASAH1 can cause a few other diseases. Mutations in

ASAH1 is mainly associated with an ultra-rare lysosomal storage disorder called (Farber disease (FD), which will the focus of this thesis. Additionally, mutations to ASAH1 can also result in a separate disease called spinal muscular atrophy with progressive myoclonic epilepsy (SMA-PME). The following section provides a clinical review of ACDase deficiency.

1.4 Acid Ceramidase Deficiency

This work is adapted from Yu et al., 2018 with permission of Orphanet Journal of rare diseases

1.4.1 Introduction

Dr. Sidney Farber described the first case of “disseminated lipogranulomatosis” in a

14-month-old infant at a Mayo Foundation lecture in 1947. Farber later published a case series of three patients in 1952, as a transaction for the 62nd annual meeting of the American

Pediatric Society [Farber 1952]. He later expanded the descriptions in 1957 [Farber et al.

1957]. Farber originally hypothesized that the disease shared the lipid storage aspects of

Niemann-Pick disease as well as the inflammation observed in Hand-Schüller-Christian disease. Although Farber demonstrated an increase in lipids in his early biochemical studies, the main lipid that accumulates in FD, i.e., ceramide, was not identified until 1967, when it was isolated from a biopsy of a patient’s kidney [Prensky et al. 1967]. ACDase which was first purified in 1963, catalyzes the synthesis and degradation of ceramide into sphingosine and fatty acid [Gatt 1963]. In 1972, Sugita and colleagues established that

ACDase activity was not detectable in post-mortem tissue from a FD patient [Sugita et al.

1972]. In 1996, the ASAH1 gene that encodes ACDase was fully sequenced and characterized [Koch et al. 1996].

1.4.2 Traditional classification of FD

FD (OMIM #228000), also known as Farber’s lipogranulomatosis, is an ultra-rare

LSD. It is caused by mutations in ASAH1, which lead to decreased ACDase activity and in turn, to ceramide accumulation and various pathological manifestations. Moser and colleagues first categorized FD into 5 subtypes in a review in 1989, later adding two other

34 phenotypes [Moser 1989; Levade et al. 2014]. Type 1, also termed the “classical” variant of

FD, includes patients with the cardinal symptoms of subcutaneous nodules, joint contractures, and voice hoarseness. These patients may also develop enlarged liver and spleen along with neurological and respiratory complications [Levade et al. 2014; Beck et al.

2014]. Traditionally, Type 1 FD patients exhibit symptoms during infancy and typically do not live past the age of 2-3 years [Farber et al. 1957; Zetterström 1958]. Types 2 and 3 FD patients have been termed the “intermediate” and “mild” variants, respectively; patients with these phenotypes usually have a longer lifespan due to reduced neurological involvement.

However, Types 2 and 3 FD patients suffer from subcutaneous nodules, joint contractures, and aphonia due to inflammation. Types 4 and 5 FD patients have severe disease manifestations. Type 4 is associated with the “Neonatal-Visceral” variant, wherein neonates experience severe organomegaly and visceral histiocytosis [Willis et al. 2008; Levade et al.

2014]. Type 5 is the “Neurological Progressive” variant, which is manifested by progressive neurological deterioration and seizures. Nodules and joint involvement are present in Type

5; however, they are less severe. Type 6 FD is termed “Combined Farber and Sandhoff

Disease variant.” In this single co-incidental case, the patient had combined Farber and

Sandhoff (OMIM #268800) diseases [Fusch et al. 1989]. The patient presented with clinical signs of FD but demonstrated a deficiency in both ACDase and hexosaminidases A and B

[Fusch et al. 1989]. Finally, Type 7 FD is termed “Prosaposin Deficiency.” This phenotype was identified in one patient and his infant sibling [Harzer et al. 1989]; a mutation was identified in the precursor protein of saposins (i.e., prosaposin, encoded by the PSAP gene)

[Schnabel et al. 1992]. A total of 4 saposins have been identified, and these proteins, along with the GM2 ganglioside activator protein, collectively belong to a group of sphingolipid activator proteins (SAPs). Only a handful of patients with Type 7 FD have been reported

[Hulkova et al. 2001]. Similar to Type 6 FD, these patients often have multiple enzyme deficiencies, such as reduced glucocerebrosidase, galactocerebrosidase and ceramidase

35 activities [Schnabel et al. 1992; Levade et al. 2014]. While patients with prosaposin deficiency may show some biochemical and clinical signs that overlap with FD, it is considered a separate disease (OMIM #176801). Increasingly, many of the more recently reported cases simply identify FD as either the classic childhood or the mild and attenuated form [Solyom, Simonaro et al. 2015; Bao et al. 2017; Schuchman et al. 2017].

1.4.3 The diverse signs and symptoms in ACDase deficiency

Up to the date of writing this chapter, we have identified 174 cases of patients described as having a form of ACDase deficiency. This literature search spans over 70 years of case reports and research articles. I included cases that were published in English, French,

German, Chinese, Russian, and Arabic. While most of the cases we reviewed involved the classical FD phenotype, some were related to the rare motor neuron disease, SMA-PME.

We have summarized these cases in Table 6 and Table 7. Furthermore Figure 7 highlights

Clinical presentation Cases Average age of Average age of last Average age of onset documentation death Classic & Severe FD 97 5.8 ± 4.6 M (70) 1.9 ± 1.8 Y (18) 2.6 ± 6.0 Y (61) Mild & Intermediate 39 1.5 ± 1.4 Y (34) 14.9 ± 18.1 Y (26) 14.3 ± 8.1 Y (8) FD FD (Unspecified) 16 ------

the main clinical manifestations described in case reports and detailed in the subsections below.

SMA-PME 23 5.8 ± 4.2 Y (18) 16.5 ± 5.7 Y (15) 14.4 ± 3.0 Y (5) SMA-PME Like 20Classic and8.9 ± 7.38 (20) Mild 22.68& ± 17.8 SMAY (10)-PME 21.9SMA ± 17.2-PME Y (10) Severe FD Intermediate FD (like) CasesTotal with FD clinical 152 75 35 20 19 Total SMAdetails-PME 43 Nodules 95% 94% 0% 0% Joint Contractures 96% 97% 0% 0% Hoarse Voice 91% 71% 0% 0% Hepatosplenomegaly 40% 3% 0% 0% Neurological and 57% 19% 60% 32% behavioral Respiratory 35% 22% 45% 26% Motor Neuron /Muscle 32% 23% 100% 95% Weakness Ocular 23% 8% 0% 0% Bone 17% 31% 0% 0% Myoclonus & Seizures 17% 11% 100% 100%

Table 6. ACDase case reports from 1952-2017 by clinical presentation

Total number of cases of ACDase deficiency reported from 1952-2017 by clinical presentation, severity, and average ages. Unspecified represents cases in which a diagnosis was made but insufficient clinical information was provided for placement in a clinical category. M - months, Y- years, number in brackets indicates the total number of cases included to calculate the average age and standard deviation. Table Adapted from [Yu et al. 2018]

Table 7. Common clinical features in ACDase deficiency

Percentage representations of common clinical features in the literature for FD and the SMA-PME variant of ACDase deficiency. Table adapted from [Yu et al. 2018]

1.4.3.1 Cardinal triad symptoms of FD

The classical triad of symptoms that manifest in FD is the formation of subcutaneous nodules, painful and swollen joints, and the development of a hoarse voice and aphonia

[Levade et al. 2014]. Subcutaneous nodules are palpable and may cause hyperesthesia; this is often evident within the first few weeks of nodule development in severe cases

[Farber et al. 1957; Zetterström 1958; Fusch et al. 1989]. However, nodule formation may

37 present later in life in attenuated forms of the disease [Abul-Haj et al. 1962; Fiumara et al.

1993; Hugle et al. 2014]. Nodules typically appear on joints and over pressure points. With time, the nodules may thicken and increase in size and number, causing significant swelling.

Joint contractures can manifest in a number of locations, ranging from the interphalangeal, metacarpal, wrist, elbow, knee, ankle and facet joints of the spine [Burck et al. 1985;

Fujiwaki et al. 1992; Mondal et al. 2009; Ekici et al. 2012]. Joint contractures are progressive, and the resulting flexion can severely limit mobility for some patients [Fujiwaki et al. 1992; Y. J. Kim et al. 1998]. The development of a hoarse voice also occurs as a result of nodule formation in the larynx. Infants are often reported to have a weak cry, which progresses to dysphonia and eventually an inability to speak [Burck et al. 1985]. The formation of the nodules in the upper airway may also expand to the epiglottis and cause swelling, which results in feeding and respiratory difficulties [Zetterström 1958; Schultze and

Lang 1960; Ozaki et al. 1978]. If the nodule formation is extreme, tracheostomy may be required [Zetterström 1958; Samuelsson and Zetterström 1971].

While a definitive diagnosis of FD ideally incorporates the measurement of ACDase enzyme activity, accessibility to the assay and/or a reference diagnostic centre is an issue in certain developing countries [Mondal et al. 2009; Erfan et al. 2015; Nasreen et al. 2017]. In these circumstances, diagnosis of FD is made by relying on the triad symptoms and histological analysis.

Figure 7. Main Clinical manifestations of ACDase deficiency.

The typical clinical manifestations by organ type that have been reported in cases of FD and SMA-PME in the published literature. FD symptoms organized by neurological symptoms, ophthalmic symptoms, cardinal triad symptoms, respiratory symptoms, hematopoietic symptoms, gastrointestinal involvement, dermatological manifestations, liver disease, motor neuron and muscle weakness, and bone disease phenotypes. Figure adapted from [Yu et al. 2018]

1.4.3.2 Hematologic findings in FD

Nodule formation and inflammation are ubiquitous within the spectrum of FD. This feature highlights the role that the hematopoietic system may play in the disease. The nodules are composed of foamy histiocytes and macrophages. This distinctive foamy phenotype is caused by the accumulation of storage material [Burck et al. 1985; Fujiwaki et al. 1992; Kattner et al. 1997; Devi et al. 2006; Mondal et al. 2009; Ekici et al. 2012].

Ultrastructural analysis of nodules has revealed the presence of Zebra bodies and curved semi-linear tubular bodies (Farber bodies) [Rauch and Auböck 1983; Cartigny et al. 1985;

Koga et al. 1992]. Bloodwork samples from patients have also revealed an increased leukocyte count, erythrocyte sedimentation rate, moderately elevated plasma chitotriosidase and C-reactive protein (CRP) in severe cases [Schanche et al. 1964; Klingkowski et al.

1998; El-Kamah et al. 2009; Kostik et al. 2013; Torcoletti et al. 2014; Saygi et al. 2015]. The formation of nodules and histiocytic infiltration may extend beyond the extremities and joints, and it has also been observed within the reticuloendothelial system, including the bone marrow, liver, lung, lymph node, and spleen, as well as the thymus and heart, in a number of patients [Antonarakis et al. 1983; Kattner et al. 1997; van Lijnschoten et al. 2000]. In one case, solely the presence of invading histiocytes in a patient’s bone marrow aspirate led to the appropriate clinical identification of FD [Nivaggioni et al. 2016].

Several other hematologic findings have been reported. Enlarged lymph nodes have been noted in autopsy reports [Farber et al. 1957; Moser et al. 1969; Antonarakis et al.

1983; Qualman et al. 1987]. Lymphadenopathy and calcification of the axillary lymph nodes have been detected on X-rays [Mondal et al. 2009]. Finally, anemia, thrombocytopenia and the presence of nucleated red blood cells have also been reported in FD patients

[Antonarakis et al. 1983; Fujiwaki et al. 1992; Mondal et al. 2009].

1.4.3.3 Neurological findings in FD

Neurological manifestations are usually only seen in patients with Type 5 or classical

FD [Levade et al. 2014]; the epileptic picture that is characteristic of SMA-PME is described in a separate paragraph below. Neurological involvement in FD is broad and can affect the central or peripheral nervous systems. Within the brain, hydrocephaly and cortical brain atrophy have been detected by magnetic resonance imaging [Chedrawi et al. 2012;

Muranjan et al. 2012]. Storage pathology has been reported in a variety of neural tissues, including the anterior horns of the spinal cord, brain stem, the cerebral cortex, and cerebellum [Molz 1968; Bierman et al. 1966; Zappatini-Tommasi et al. 1992; Chedrawi et al.

2012; Bao et al. 2017]. A storage pathology has also been reported in the Schwann cells of the peripheral nervous system (PNS), where both myelinating and non-myelinating Schwann cells have large membrane-bound inclusions [Schmoeckel and Hohlfed 1979; Pellissier et al. 1986; Zappatini-Tommasi et al. 1992]. Pathology descriptions suggest that compression of the axonal body may affect proper nerve conduction [Burck et al. 1985; Pellissier et al.

1986; Zappatini-Tommasi et al. 1992]. A number of case reports have documented the occurrence of seizures, and developmental delay leading to intellectual disability [Molz 1968;

Eviatar et al. 1986; Zappatini-Tommasi et al. 1992; Chedrawi et al. 2012]. Due to the pathology in the anterior horn cells and peripheral neuropathy, patients may also present with hypotonia, muscle weakness and atrophy, leading to their requirement for a wheelchair

[Bierman et al. 1966; Molz 1968; Eviatar et al. 1986; Zappatini-Tommasi et al. 1992;

Chedrawi et al. 2012].

1.4.3.4 Pulmonary findings in FD

Beyond the development of the cardinal phenotypes, pulmonary complications are one of the more common occurrences in both classic and attenuated variants of FD [Levade

41 et al. 2014]. Clinical signs may include sternal retraction, expiratory stridor, aphonia, and labored breathing [Farber 1952; Fiumara et al. 1993; Kim et al. 1998; Moser et al. 1969;

Eviatar et al. 1986]. As mentioned above, when nodule formation in the larynx and upper airway is extreme, tracheostomy may be required [Kim et al. 1998; Salo et al. 2003;

Cvitanovic-Sojat et al. 2011]. X-rays have shown presence of consolidation, nodular opacities, and lung atelectasis [Samuelsson and Zetterström 1971; Pellissier et al. 1986;

Fiumara et al. 1993; Y. J. Kim et al. 1998]. Bronchial alveolar lavage and post-mortem analyses of patients have revealed significant inflammation with large lipid-laden macrophages and cellular infiltration throughout the bronchioles and alveoli [Farber et al.

1957; Antonarakis et al. 1984]. The lung tissue of one patient was described as poorly expanded with excessive connective tissue, and the ultrastructural analysis there revealed lung histiocytes containing curvilinear storage bodies [Bierman et al. 1966]. Pulmonary distress, infection and pneumonia are the main causes of mortality [Farber et al. 1957;

Bierman et al. 1966; Ohfu et al. 1987; Fiumara et al. 1993; Ekici et al. 2012; Levade et al.

2014].

1.4.3.5 Ophthalmic findings in FD

Ocular manifestations in FD have mostly been associated with the classic form of FD and those with neurological involvement [Levade et al. 2014]. In Farber’s original description of the disorder, he reported that his second patient was blind; however, limited analysis was performed [Farber et al. 1957]. A variety of ophthalmic findings have been documented in the literature; the most common sign is a cherry red spot [Cogan et al. 1966; Moser et al.

1969; Pellissier et al. 1986; Cvitanovic-Sojat et al. 2011; Zarbin et al. 1988; Saygi et al.

2015]. Additional ocular manifestations include retinal opacification, corneal opacities, and macular degeneration [Zetterström 1958; Cogan et al. 1966; Tanaka et al. 1979; Chandwani

42 and Kuwar 2002]. Other findings related to the eyes have included the presence of xanthoma-like growths in the conjunctiva, poor visual fixation, and nystagmus [Ohfu et al.

1987; Y. J. Kim et al. 1998; Chedrawi et al. 2012]. Post-mortem analyses of the eyes showed no abnormalities in the anterior segment, but the posterior segment contained birefringent lipids within the ganglion cell layer and displayed significant storage pathology in other cell types in the eye [Cogan et al. 1966; Bashyam et al. 2014].

1.4.3.6 Gastrointestinal findings in FD

There are several cases in the literature describing gastrointestinal manifestations of

FD. Persistent diarrhea has occasionally been seen in infants [Fujiwaki et al. 1992; Koga et al. 1992]. One patient also exhibited extensive gastrointestinal lesions with widespread erosion of the gastrointestinal mucosa [Koga et al. 1992]. Another study that biopsied colonic tissue in a patient with severe disease demonstrated an increased level of apoptosis of cells within the crypt of the colon. This study also demonstrated that the caspase-3 positive cells co-localized with cells that were positive for GD3 gangliosides, concluding that colonocyte apoptosis may be triggered by the synthesis of GD3 as a consequence of ceramide accumulation [Farina et al. 2000].

1.4.3.7 Hepatic findings in FD

A palpable liver and hepatomegaly are commonly reported in patients with the classic variant of FD [Farber et al. 1957; Abul-Haj et al. 1962; Samuelsson and Zetterström

1971; Tanaka et al. 1979; Antonarakis et al. 1984]. Zebra bodies and Farber bodies have been observed in hepatocytes, endothelial cells, and Kupffer cells [Hoof and Hers 1968;

Abenoza and Sibley 1987]. The most significant liver pathology seen is in patients with severe type 4 FD [Levade et al. 2014]. Infants have presented with cholestatic jaundice,

43 ascites, liver fibrosis, and elevated liver enzymes [Salo et al. 2003; Willis et al. 2008;

Nowaczyk et al. 1996]. In a unique case, a 6-month-old infant showed significant liver failure and was misdiagnosed with neonatal hepatitis; he/she underwent liver transplantation, which subsequently normalized the liver function [Salo et al. 2003]. FD was properly diagnosed in that instance after the appearance of nodules and histiocytic infiltrates. In these few severe cases, the enlargement of visceral organs and histiocyte formation may mask or precede the appearance of nodules [Salo et al. 2003].

1.4.3.8 Bone findings in FD

When joint involvement is present in FD patients, there may also be juxta-articular bone erosion and demineralization [Abul-Haj et al. 1962; Schultze and Lang 1960;

Samuelsson and Zetterström 1971; Amirhakimi et al. 1976]. In addition to the joints, bone erosion has been observed in long bones, metacarpals, metatarsals, and phalanges

[Schanche et al. 1964; Crocker et al. 1967; Pavone et al. 1980; Antonarakis et al. 1983;

Moritomo et al. 2002]. Osteoporosis is often progressive during the course of disease

[Fujiwaki et al. 1992; Fiumara et al. 1993; Kim et al. 1998]. One patient, a 9-year-old girl, grew a tumorous osseous lesion in her spine, resulting in destruction of the odontoid by inflammatory cells. She underwent two HSCTs, which improved her mobility, but episodes of myoclonic epilepsy were still persistent [Jarisch et al. 2014]. In the milder spectrum, Bonafé et al. presented a case series of three siblings who displayed peripheral osteolysis between the ages of 40-60 years [Bonafé et al. 2016]. The patients all had shortened fingers and toes, as well as redundant skin. One of the siblings had limited movement of his knees and toes [Bonafé et al. 2016]. An unrelated 29-year-old patient also displayed deformities of the hands, displaying shortened fingers and redundant skin [Fiumara et al. 1993]. These

44 patients had longer than average lifespans and were not formally diagnosed with FD until well into adulthood, which indicates that such milder cases may be underrepresented.

1.4.3.9 Dermatological findings in FD

In addition to the formation of subcutaneous nodules, skin lesions and plaques have been reported in some FD patients [Chanoki et al. 1989; Koga et al. 1992; Schmoeckel

1980]. Analyses of dermal biopsies have revealed hyalinized collagen in the dermis, hyperkeratosis, and the presence of large foamy histiocytes [Abenoza and Sibley 1987;

Fujiwaki et al. 1992; Bashyam et al. 2014]. The storage pathology in dermal tissue and histiocytes revealed the presence of Farber bodies [Abenoza and Sibley 1987; Schmoeckel and Hohlfed 1979; Navarro et al. 1999]. A rare presentation featured an infant with clinical signs that overlapped with stiff skin syndrome [El-Kamah et al. 2009]. The infant displayed thick indurated skin since birth, a stiff neck, and scleroderma-like areas; he/she died at around two years of age [El-Kamah et al. 2009].

1.4.3.10 Hydrops fetalis in FD

In the literature to date, there have been two FD patients presenting with hydrops fetalis [Zielonka et al. 2017]. One report is of a 29-week-old stillborn fetus with mild internal hydrops, a well-preserved spleen, and the presence of foamy cells [van Lijnschoten et al.

2000]. The second report is of a 3-day-old neonate with an extreme phenotype of hydrops

[Schafer et al. 1996; Kattner et al. 1997]. The latter infant presented with an enlarged abdomen filled with hemorrhagic ascites, hepatosplenomegaly, and many white nodules on the peritoneal surfaces of the liver, spleen, and other organs [Schafer et al. 1996; Kattner et

45 al. 1997]. These two cases of fetal hydrops represent the shortest-lived patients recorded in the Farber literature.

1.4.3.11 SMA-PME

A new variant of ACDase deficiency has emerged that shares no classical signs and symptoms of FD. These patients have a separate disease called SMA-PME (OMIM

#159950). SMA-PME was first described in 1978 by Jankovic and colleagues. He described patients from a family in Louisiana and Texas who first developed muscle weakness and wasting, which gradually progressed to jerking of the limbs and myoclonus [Jankovic and

Rivera 1978]. Most patients who suffer from SMA typically have a mutation in SMA1 or

SMD2 [Zhou et al. 2012]. However, some patients who have SMA-PME have now been identified to carry mutations in ASAH1 [Zhou et al. 2012; Dyment et al. 2014; Gan et al.

2015; Rubboli et al. 2015; Giráldez et al. 2015; Özkara and Budak 2017]. To the best of our knowledge, there have been 23 confirmed cases of SMA-PME reported in the literature to date (Table 6). Twenty additional cases were reported from 1978 to 2009 to have a SMA-

PME-like clinical presentation, including the original case described by Jankovic [Jankovic and Rivera 1978; Lance and Evans 1984; D'Ecclesia et al. 1985; Taglioli et al. 1990;

Ferlazzo et al. 2009; Haliloglu et al. 2002].

Symptoms of SMA-PME may appear as early as 2 years of age [Rubboli et al. 2015] and include increasing difficulty in walking, sporadic falls, muscle weakness and tremors, and tongue fasciculations [Haliloglu et al. 2002; Zhou et al. 2012]. As the disease progresses, patients experience muscle atrophy and seizures [Haliloglu et al. 2002].

Impaired mobility and difficulty swallowing occur near the end of life. Electromyogram (EMG) and muscle biopsies of patients often reveal denervation, and electroencephalograms (EEG) show generalized spikes and wave discharges in varying areas [Dyment et al. 2014;

Haliloglu et al. 2002; Özkara and Budak 2017]. Death is usually attributed to respiratory failure and has been recorded as early as the teenage years [Zhou et al. 2012; Dyment et al.

2014; Gan et al. 2015; Rubboli et al. 2015; Giráldez et al. 2015].

Generalizations of the clinical picture should be made with caution since the number of identified patients is limited, and most reported cases share the same Thr42Met mutation

(Table 12). However, several cases encourage a broader understanding of SMA-PME. For example, the first description of an adult SMA patient with a mutation in ASAH1 did not present with myoclonic epilepsy [Filosto et al. 2016]. Another patient presented with eyelid myoclonic status epilepticus (SE), in addition to muscle weakness, which has not been previously observed in SMA-PME [Oguz Akarsu et al. 2016].

1.4.3.12 Keloid formation

Susceptibility to the formation of keloids has recently been associated with a heterozygous ASAH1 mutation within a Yoruba family in Nigeria. In this family of 24 members, 9 had keloids and 2 others had hypertrophic or stretched scars [Santos-Cortez et al. 2017]. The Leu401Pro mutation (clinVar ID SCV000538196) was identified through a combination of linkage analyses and exome sequencing [Santos-Cortez et al. 2017]. The appearance of keloids in this family ranged from 2 to 57 years of age. Additionally, the locations of keloid formation varied. Unfortunately, no lipid analysis or enzyme activity was reported. However, this variant nonetheless expands the clinical picture of ACDase deficiency [Santos-Cortez et al. 2017].

1.4.4 Phenotypic variability in ACDase deficiency

ASAH1 mutations seem to result in two separate disorders, demonstrating the broad importance of ACDase for the proper maintenance of health. We have highlighted the diverse clinical spectrum that can be seen in various forms of ACDase deficiency.

Interestingly, phenotypic variability is also seen in reports involving siblings. In one such case, one sibling demonstrated a classic Farber phenotype and died at 6 months of age, whereas the other sibling survived to 12 weeks of age and had extreme histiocytic infiltration throughout the body [Antonarakis et al. 1984]. Surprisingly, post-mortem analyses of liver tissue from both patients revealed a similar level of enzyme activity [Antonarakis et al. 1984].

Another case showed hepatosplenomegaly in a 3-month-old male [Qualman et al. 1987].

While no nodules were noted in this patient, histiocytosis was the dominant phenotype. His sister, who was 5 and one-half months old, displayed a classical phenotype of FD [Qualman et al. 1987]. Fiumara et al. featured two sisters and one female cousin with a mild variant of

FD and significant symptom variability [Fiumara et al. 1993]. Clinically, all three patients displayed nodule formation, joint involvement, and the presence of erosions [Fiumara et al.

1993]. However, variability in symptom onset and longevity was observed. One sister developed symptoms in her second year of life and lived to 30 years of age, whereas the other sister was symptomatic at 20 months of age and died when she was 18 years old.

While ACDase enzyme activity was not reported for the shorter-lived sister, the assay was performed on cells cultured from the cousin who developed symptoms even earlier and died the earliest, at 11-years of age, the long-lived sister, and an established FD control cell line

(FD patient who died at 1.8 years of age). In this enzyme activity assay, both the long-lived sister, the cousin, and the FD control showed enzyme activity between 4-6% of normal controls [Fiumara et al. 1993]. Presumably, the three mild FD patients shared a similar mutation, yet there was obvious variability in symptom onset and patient longevity [Fiumara

48 et al. 1993]. Similarly, enzyme activity for classical patients, who have shorter lifespans, may be comparable to those of patients who are long-lived. Therefore, while enzyme activity is important for diagnosis of FD, there is not a complete correlation between in vitro enzyme activity levels and patient outcomes.

1.4.5 Genetics and diagnosis

1.4.5.1 Prevalence of ACDase deficiency

FD is an ultra-rare disease with a predicted prevalence of <1/1000,000 (Orphanet).

Based on our literature search, we identified 152 reported cases of FD between 1952 and

2017 (Table 6) [Yu et al. 2018]. SMA-PME currently has its own OMIM entry and is usually categorized as a subtype of SMA [Topaloglu and Melki 2016]. We identified 23 cases of

SMA-PME associated with mutations in ASAH1 gene since Zhou and colleagues first reported this finding [Zhou et al. 2012]. Due to the rare nature of both disorders, they originally appeared to be two very separate conditions; however, as more cases of each are characterized, the clinical pictures are starting to overlap. For example, a recent case described a patient who presented with muscle weakness typical of SMA-PME, but who also had joint pain synonymous with FD [Teoh et al. 2016]. Additionally, several cases of FD have shown neurological involvement such as delayed mental development, seizures, and muscle weakness as predominant pathologies [Eviatar et al. 1986; Colamaria et al. 1992;

Ahmad et al. 2009; Levade et al. 2014].

1.4.5.2 Genetics and mutations

The human acid ceramidase gene (ASAH1) is approximately 30 kb in total length. It contains 14 exons that range from 46 to 1200 bp long and is mapped to the short arm of

49 chromosome 8 (8p21.3/22) [Li et al. 1999]. The first mutation identified, c.665C>A

(p.Thr222Lys), was from a patient with a severe form of FD [Koch et al. 1996]. Based on the literature, there are over 60 pathologic mutations leading to FD or SMA-PME [Yu et al.

2018]. Additionally, at the time of this review, more than 120 genetic variants were submitted to the NCBI ClinVar public archive [Landrum et al. 2015]. While a number of these represent published mutations with a pathologic role, most of the variants were submitted by clinical diagnostic testing facilities and did not include confirmed pathological details. Therefore, our curated list is likely an under-representation of all the sequenced pathologic mutations in FD.

Nonetheless, several observations can be extrapolated. Mutations have been identified throughout the ASAH1 gene, but most of the mutations appear to be missense mutations (Figure 8). Among the recorded mutations that result in the diagnosis of FD, the majority of mutations are located within the β-subunit. Eleven patients were identified with a mutation in exon 8, and 8 patients had mutations in exon 13. In contrast, a larger number of mutations in SMA-PME have been identified within the α-subunit. One interesting observation is that the Thr42Ala and Thr42Met mutations in exon 2 accounted for more than half of the total number of reported cases of SMA-PME. While some of these cases are siblings, they have also occurred within independent families [Zhou et al. 2012; Sathe and

Pearson 2014; Rubboli et al. 2015; Johannsen et al. 2015; Giráldez et al. 2015; Filosto et al.

2016]. There is currently no definitive genotype-to-phenotype relationship in the noted mutations, which is especially true based on the observation that one patient with SMA-PME and another with FD had the identical Tyr153Cys mutation [Kernohan et al. 2017; Cozma et al. 2017]. Another patient presented with polyarticular arthritic symptoms synonymous with

FD and later developed muscle weakness with no PME [Teoh et al. 2016]. These examples indicate that mutations in ASAH1 can result in a broad range of phenotypes.

Figure 8. Structure of the human ASAH1 gene with the protein and distribution of mutations.

Mutations in ASAH1. A) ASAH1 genomic structure. B) ASAH1 mature transcript structure. C) Schematic of the ACDase protein with annotations for the signal peptide, α-subunit, and β-subunit. D) Percentages of the reported 65 ASAH1 mutations by type for FD and SMA-PME. E) Frequency of mutations by subunit and reported disease phenotype. Figure adapted from [Yu et al. 2018]

In the same year that the relationship between ASAH1 and SMA-PME was established, another report also demonstrated that Han Chinese patients diagnosed with schizophrenia showed a down regulation of the ASAH1 gene. Furthermore, this study identified two ASAH1 SNPs (rs7830490, and rs3753118) associated with schizophrenia

[Zhang et al. 2012]. This observation was also reported in a separate and larger study that analyzed the exomes of 12,332 Swedish individuals, of which 4,877 were affected by schizophrenia [Genovese et al. 2016]. That study found patients with schizophrenia had a higher abundance of ultra-rare variants, of which 7 SNPs loci were in the ASAH1 gene

(rs781294134, rs759037498, rs761518207, rs13785393, rs764327759, rs757058563, and rs773025886) [Genovese et al. 2016]. One final example of the broad ACDase deficiency

51 phenotype that can occur is the aforementioned case regarding keloid formation and the

Leu401Pro mutation in ASAH1 [Santos-Cortez et al. 2017].

1.4.5.3 Clinical diagnosis

FD is inherited in an autosomal recessive manner. Due to its rarity, prenatal screening/neonatal testing is typically not performed unless an older sibling has been previously diagnosed. Consideration of FD is typically based on the manifestation of the cardinal triad symptoms: 1) subcutaneous nodules, 2) joint pain, and 3) voice hoarseness

[Levade et al. 2014]. Diagnosis of the mild and attenuated variants of FD is more troublesome since one or more of the featured symptoms may be absent or missed at the time of diagnosis. One report describes a patient who had no apparent subcutaneous nodule formation until the age of 12 years [Al Jasmi 2012]. As mentioned, other cases have been misdiagnosed as juvenile idiopathic arthritis (JIA) [Kostik et al. 2013; Hugle et al.

2014]. In fact, one cohort study demonstrated that as many as 71% of FD with mild to intermediate variants of FD were initially misdiagnosed as JIA [Solyom et al. 2017]. Thus, the incidence of FD is likely underestimated. JIA patients who have symptoms consistent with the cardinal triad should also be encouraged to be tested for FD as part of their diagnosis.

In addition to JIA, the differential diagnosis includes rheumatoid arthritis, juvenile hyaline fibromatosis, and multi-centric histiocytosis, due to the similarity in joint and subcutaneous manifestations [Levade et al. 2014]. In severe cases, misdiagnosis may also occur since the main clinical picture is histiocytosis and hepatosplenomegaly [Salo et al.

2003]. In these cases, the cardinal symptoms are often masked or have not yet developed since these severe symptoms usually manifest early in infancy.

While the diagnosis of FD often requires further biochemical and genetic analyses, several case reports originating from developing countries have relied on clinical and histological diagnoses due to limited resources and lack of access to specialized diagnostic centers. Morphologic characterization is often achieved through analyses of subcutaneous nodules or other biopsied tissue. Common features reported include the formation of granulomas and large lipid-laden macrophages. A variety of studies have used ultrastructural analyses to demonstrate the presence of semi-curvilinear inclusions, also known as 'Farber bodies, Banana bodies, and Zebra bodies', in various tissue types [Dustin et al. 1973; Rivel et al. 1977; Tanaka et al. 1979; Schmoeckel and Hohlfed 1979].

1.4.5.4 Biochemical and genetic diagnostics

One method that has been adopted to assist in the diagnosis of FD is a lipid loading test on cultured living cells. In this technique, exogenously labeled sphingolipids are added to patient cells and ceramide turnover is assessed.

A variety of precursors have been used, including [14C] stearic acid–labeled cerebroside sulfate in skin fibroblasts, [3H] sphingomyelin in both patient cultured fibroblasts and transformed lymphocytes, and [14C] serine, a precursor substrate in the de novo ceramide synthesis pathway, to demonstrate impaired ceramide degradation in FD [Kudoh and Wenger 1982; Levade et al. 1995; van Echten-Deckert et al. 1997].

The most common biochemical method in use for a definitive diagnosis of FD is an enzyme activity assay using cultured patient fibroblasts. Enzyme activity in FD cells is typically <10% of normal controls, whereas SMA-PME cells have been reported to have as much as 32% of the activity of controls [Levade et al. 2014; Zhou et al. 2012]. In addition to fibroblasts, the enzyme activity assay has been tested using leukocytes, plasma, post- mortem tissue, and cultured amniocytes from prenatal testing [Sugita et al. 1974; Dulaney et

53 al. 1976; Fensom et al. 1979; Antonarakis et al. 1984; Ben‐Yoseph et al. 1989].

Conventionally, ACDase activity is determined by the use of either radiolabeled ceramides or fluorescent ceramide analogues. Many of these compounds are not water-soluble and require the use of detergents in addition to specialized technical equipment for analyses

[Sugita et al. 1974; Momoi et al. 1982; Ben‐Yoseph et al. 1989; N. Azuma et al. 1994;

Chatelut et al. 1996; Bernardo et al. 1995; Tani et al. 1999; Dagan et al. 2000]. This drawback means that diagnosis is available in only a very limited number of laboratories.

Currently, ACDase activity can be detected with the use of the fluorogenic substrate Rbm14-

12 in a 96-well plate in a high-throughput manner [Bedia et al. 2007; Bedia et al. 2010].

Quantitation of excess ceramides is another method to assist in the diagnosis. The diacylglycerol kinase assay was commonly used in early studies to measure total ceramides, but it was limited because it did not provide information about individual ceramide species [Van Veldhoven et al. 1995]. Later, chromatographic methods such as thin-layer chromatography and high-performance liquid chromatography, were also employed to quantify ceramides [Moser et al. 1969; Iwamori et al. 1979; Cremesti and Fischl

2000]. The major drawback to these methods were the requirement for radiolabeling or fluorophore incorporation. These methods were found to be difficult to perform and provided limited information on individual ceramide species. Mass spectrometry, in particular electrospray ionization mass spectrometry (ESI/MS), is currently the most sensitive method for the discrimination and detection of sphingolipids [Gu et al. 1997; Raith et al. 1998; Raith and Neubert 2000; Sullards 2000; Han 2002; Kasumov et al. 2010]. These methods have been implemented to demonstrate excess ceramide in biopsy samples of subcutaneous nodules, post-mortem liver samples, urine samples, and cultured cells [Moser et al. 1969;

Samuelsson and Zetterström 1971; Dustin et al. 1973; Amirhakimi et al. 1976; Toppet et al.

1978; Ozaki et al. 1978; Chatelut et al. 1996; Levade et al. 2014; Cozma et al. 2017].

1.4.5.5 Genetic testing

The first few mutations in ASAH1 were identified in patient cultured fibroblasts and required amplification of genomic sequences of ASAH1 and a combination of PCR and

Sanger sequencing [Koch et al. 1996; Li et al. 1998]. Whole-exome sequencing (WES) is now commonly performed and, in conjunction with biochemical assays, provides a conclusive diagnosis of ACDase deficiency [Teoh et al. 2016; Gan et al. 2015]. This is particularly informative in patients with non-classical FD, SMA-PME, and in cases in which the symptoms are suggestive of ACDase deficiency but have atypical presentations [Kim et al. 2016; Bonafé et al. 2016; Filosto et al. 2016; Kernohan et al. 2017].

1.4.5.6 Biomarkers

Increased inflammation and the formation of histiocytes are common in many cases of FD. A recent study from our laboratories identified monocyte chemoattractant protein 1

(MCP-1) as a potential biomarker [Dworski et al. 2017]. A multiplex cytokine analysis was performed using plasma obtained from FD, JIA, and FD patients who underwent HSCT. This study demonstrated an elevation of MCP-1 in FD samples, but low levels in JIA and normalized levels in FD patients who underwent HSCT [Dworski et al. 2017]. MCP-1 may thus be a beneficial biomarker and could help address the issue of misdiagnosis in mild cases of FD.

Another potential biomarker for the diagnosis of FD is C26:0 ceramide, which was identified by lipid MS quantification of ceramides from lipids extracted from dried blood spots

[Cozma et al. 2017]. Two isoforms of C26:0 have been described, with isoform 1 being expressed at a significantly higher level in the newborn (0-6 months) cohort versus the juvenile (0.5-4 years) and adult (>17 years) cohorts. No details were provided regarding the

55 clinical phenotypes of these patients, but the application of a platform for bloodspot analysis for newborns could be an important step in earlier diagnosis of ACDase deficiency.

1.4.6 Research, treatment and future therapy

1.4.6.1 Animal models

An ACDase knock-out mouse model was previously generated through insertional mutagenesis into the Asah1 gene. Heterozygous mice (Asah1+/-) did not show any overt changes in phenotype and had a normal life span of at least 1.5 years [Li et al. 2002].

However, analyses of the organs of heterozygous mice 6 months of age and older revealed lipid accumulation and inclusions in the liver, lung, skin, and bone [Li et al. 2002]. The heterozygous liver was the most affected as it became fibrous and pale. While most hepatic cell types were filled with lipids, the most significant effect was observed in Kupffer cells. By

9 months of age, some ceramides were also elevated in the heterozygous animals, in which the greatest accumulation was detected in the liver, with a 1.5-2-fold increase compared with wild-type animals. Homozygous mice (Asah1-/-) were embryonic lethal; none were detected at day E8.5 or later [Li et al. 2002]. A second knock-out mouse was generated via a targeted ES cell clone [Eliyahu et al. 2007]. Analyses of this model demonstrated that homozygous embryos did not survive beyond the 2-cell to the 4-cell stage and underwent apoptotic cell death, highlighting the importance of ACDase as a vital enzyme for early embryonic development.

A tamoxifen-induced conditional Asah1 knock-out mouse has also been developed

[Eliyahu et al. 2012]. Intraperitoneal delivery of tamoxifen in 5-week-old female mice resulted in impaired fertility due to lack of mature follicles in the ovaries. The follicles were not able to fully develop, and apoptosis occurred between the transition from the secondary

56 to the antral stage. This observation supports the essential role of ACDase in ovary maturation and its importance in fertility [Eliyahu et al. 2012]. Tamoxifen injection showed variable penetrance, where 100% Asah1 ablation was reported in skin and 70% in ovaries

[Eliyahu et al. 2012]. Reports using this conditional knockout have focused exclusively on the ovary phenotype. While classical and severe cases of FD do not survive to sexual maturity, data derived from these studies may prove relevant with regard to mild and attenuated FD patients. While no overt FD phenotypes have been reported, this model nonetheless may serve as an important tool for fertility studies.

Finally, a knock-in model has also been developed, in which an ASAH1 patient mutation (P362R) was introduced into the analogous murine locus (P361R), resulting in a mouse that recapitulates many of the phenotypes observed in classical cases of FD

[Alayoubi et al. 2013]. The P362R mutation has been identified in two patients with FD. One patient, who died at 1.5 years of age, had a classical form of FD and was homoallelic for the mutation. The other patient, who died at 8 years of age, was heteroallelic for P362R and

E138V [Li et al. 1999]. Furthermore, this mutation site was selected because it represents the most conserved region of the gene between the species [Alayoubi et al. 2013].

Homozygous (Asah1P361R/P361R) mice have a decreased lifespan and reduced weight. These mice develop a significant inflammatory phenotype and the accumulation of large foamy macrophages in many tissues. Recent studies have also shown that these mice have impaired hematopoiesis, central nervous pathology, abnormal skin development, and impaired lungs [Dworski et al. 2017; Lopez-Vasquez et al. 2016; Sikora et al. 2017; Yu et al.

2017]. The Asah1P361R/P361R model does not develop nodules, but it does exhibit many features that are seen in patients, such as inflammation; enlarged organs, including hepatosplenomegaly; respiratory distress; and neurological and behavioral impairment

[Dworski et al. 2017; Lopez-Vasquez et al. 2016; Sikora et al. 2017; Yu et al. 2017]. Animal studies have thus provided key insights into ACDase biology and Asah1 mutant pathology.

In addition, they have and will continue to serve as important models that will ultimately guide and inform the use of future therapies in patients.

1.4.6.2 Current treatment

There is currently no cure for ACDase deficiency. Current treatment strategies focus on symptom management. Anti-inflammatory medications and physical therapy can help address pain and mobility issues [El-Darouti 2013; Mitchell et al. 2016; Schuchman et al.

2017]. Surgical intervention may occasionally be applied for the removal of nodules in the hands and oral cavity [Haraoka et al. 1997; Moritomo et al. 2002]. In one severe case in which a patient was misdiagnosed with hemangioendothelioma, a series of five surgeries to remove sacrococcygeal masses and three surgeries for scalp masses were performed over the course of a patient’s life. He/she eventually expired at 5 years of age [Lee et al. 2016].

HSCT is another therapeutic option and has been demonstrated to substantially improve mobility and pain in a number of FD patients lacking CNS involvement [Vormoor et al. 2004;

Ehlert et al. 2007]. Early studies in which HSCT was performed in two patients with classical

FD with CNS complications were promising because they showed an elevation in ACDase activity and resolution of voice hoarseness, subcutaneous nodules, and painful joints

[Souillet et al. 1989; Yeager et al. 2000]. However, in both cases, HSCT did not reverse the neurological phenotypes, and the patients deteriorated over time. Despite the scarcity of patient data, HSCT appears to be a promising treatment for mild and attenuated FD.

For SMA-PME, most patients are prescribed anti-epileptic drugs to assist with seizure control, with mixed efficacy [Oguz Akarsu et al. 2016; Özkara and Budak 2017].

Since respiratory complications are progressive, some patients may also require mechanical ventilation and gastric feeding [Rubboli et al. 2015; Oguz Akarsu et al. 2016].

1.4.6.3 Gene therapy for ACDase deficiency

ACDase deficiency is an attractive target for gene therapy because it is caused by a single gene defect. As reviewed earlier, several gene therapies for monogenic lysosomal storage disorders are currently being investigated in clinical trials [Sessa et al. 2016; Tardieu et al. 2017; Huang et al. 2017]. In the context of ACDase deficiency, one early study demonstrated that FD patient cells recovered ACDase activity when the cells were infected with the gamma-retroviral pG1 vector engineered to express human ACDase [Medin et al.

1999]. This study confirmed that the transduced cells had increased ACDase activity and normalized ceramide levels [Medin et al. 1999]. Additionally, the treated cells could also cross-correct untreated cells when supplemented with medium from transduced cells that secreted human ACDase [Medin et al. 1999]. Through the mannose-6-phosphate receptor pathway, the non-infected cells acquired functional enzyme, demonstrating the effect of metabolic co-cooperativity. A later study reproduced this same effect using lentiviral vectors as the delivery vehicle and showed successful gene correction in hematopoietic stem cells

[Ramsubir et al. 2008]. That same study also showed that direct injection of vector into murine neonates could provide long-term expression of ACDase for up to 13 weeks

[Ramsubir et al. 2008]. This same approach was applied to the P361R FD mouse model and demonstrated an increased lifespan from 9-10 weeks to 16.5 weeks of age [Alayoubi et al.

2013].

Ex vivo gene therapy is a treatment strategy that may deliver a longer-lasting therapeutic benefit than traditional HSCT. In this approach, stem, progenitor, or differentiated cells are isolated from a patient or donor, modified by genetic correction, and subsequently transplanted into the patient [Boelens et al. 2014; Naldini 2011]. HSCs are a promising cell type for such gene therapy schemas since they are readily accessible and easily separated from a patient’s blood and can expand/differentiate into long-lived cell types

[Naldini 2015; Morgan et al. 2017]. Ex vivo gene therapy followed by transplant represents an improvement over HSCT alone because the transduced cells express enzyme derived from the therapeutic vector in addition to their endogenous gene expression, which, in theory, allows for increased enzyme production, lysosomal activity, and potential cross- correction.

Many active gene therapy protocols are investigating such ex vivo HSCT transductions/transplantations to treat genetic disorders [Scott and DeFrancesco 2016]. Ex vivo gene therapy followed by transplant may circumvent the limitations of HSCT alone to improve neurological symptoms, as in the case of metachromatic leukodystrophy [Biffi et al.

2013]. In the case of ACDase deficiency, ex vivo transduction followed by HSCT is also a promising option. A series of proof-of-concept studies have demonstrated the successful transduction of the huACDase cDNA into murine CD34+ stem/progenitor cells and later into analogous cells from non-human primates [Ramsubir et al. 2008; Walia et al. 2011]. In the latter study, higher than normal ACDase enzyme activity was detectable in peripheral blood cells, in the bone marrow, the spleen and liver for more than a year [Walia et al. 2011].

Additionally, the animals had decreased ceramide levels [Walia et al. 2011].

At the time that this manuscript was written, a gene therapy trial was initiated for the treatment of SMA type I (clinicaltrials.gov ID NCT02122952). This trial involves the use of adeno-associated virus serotype 9 (AAV9), a non-integrating virus that encodes the SMA1 cDNA, infused through a peripheral vein. While the trial is still ongoing, preliminary data demonstrate a reduced need for pulmonary support, and patients could feed themselves, indicating a potential improvement in swallowing function [Shell et al. 2017; Mendell et al.

2017]. Although these results are for a different type of SMA, it is possible that a similar gene therapy approach may also be promising for patients with the SMA-PME phenotypes.

1.4.6.4 Enzyme replacement therapy for ACDase deficiency

ERT is currently the standard of care for several LSDs. Since early studies demonstrating the efficacy of ERT in Gaucher disease, this treatment strategy has been developed for a wide assortment of LSDs. It has been implemented to treat Pompe disease,

Fabry disease, MPSI, II, and VI, and most recently CLN type 2 (CLN2) [Barton et al. 1991;

Schiffmann et al. 2001; Kakkis et al. 2001; Thurberg et al. 2006; Harmatz et al. 2010]. ERT with rhACDase is currently under development and represents a promising therapy for

ACDase deficiency and several other conditions in which ceramide accumulation is pathologic, such as cystic fibrosis [Schuchman et al. 2015; Pewzner-Jung et al. 2014].

Currently, large volume production of rhACDase is achieved by the amplification and transfection of Chinese hamster ovary (CHO) cells [He et al. 2017]. Overexpression of

ACDase in CHO cells results in the secretion of enzyme into the medium, which is then purified by a series of chromatography steps [He et al. 2017].

A recent proof-of-concept study using the CHO-derived rhACDase as treatment in the P361R FD mouse model has shown promise [He et al. 2017]. Treatment with the recombinant enzyme resulted in decreased ceramide accumulation, less macrophage infiltration, lower MCP-1 expression, and a normalized spleen weight in FD mice [He et al.

2017]. This initial study holds promise for future FD treatments, but further investigations are required to better delineate the dose response in this model and to determine how this effect can be better translated to the human variant of FD or SMA-PME. One limitation of ERT is a reduced ability to cross the blood-brain barrier, which represents an issue for those LSDs that manifest with neurologic components, such as severe cases of FD. However, targeted

CNS administration of enzyme has been observed to circumvent this limitation, and the use of fusion proteins with CNS-targeting moieties is currently being evaluated as a promising method for enzyme delivery to the CNS [Boado et al. 2016].

1.4.7 Concluding thoughts on FD

Over 70 years have passed since Farber’s Mayo Foundation lecture. Included in this historic transition is a brief transcript where Farber states: “The clinical picture I describe may be found to be typical for these 3 cases and may not be encountered in the next 20 or

30. We should, with a disease of this kind, expect to see a number of unrelated clinical pictures in the future” [Farber 1952]. Farber’s comment and insight are highly relevant to this day. ACDase deficiency is a spectrum disorder that includes FD, SMA-PME, and potentially keloid formation or susceptibility to schizophrenia. Even amongst the individual conditions, there is a wide clinical spectrum. In mild cases, a misdiagnosis or a delay in diagnosis could impact the treatment plan and adversely affect the ability to properly manage symptoms

[Zielonka et al. 2017]. A natural history study is currently underway on clinicaltrials.gov (ID

NCT03233841), which aims to gain greater insight into the natural history of ACDase deficiency through retrospective and prospective patient data. It also aims to establish clinical information, biomarkers and other functional data to assess the efficacy of future therapies, such as rhACDase ERT. The establishment of a complete natural history will greatly improve and potentially fill in gaps in the current definition of ACDase deficiency.

Finally, due to the wide spectrum of clinical presentations, the precise number of patients is likely to be underrepresented. An improved understanding of the disease and increasingly effective knowledge translation will allow more patients to be identified, efficiently diagnosed, and effectively managed.

Chapter 2

Aims/Hypotheses

General Hypothesis: Ceramides are accumulated in patients with FD, and case reports have documented a variety of pathological consequences. We hypothesize that our

Asah1P361R/P361R mouse model recapitulates the human disorder. Comprehensive characterization of the pathophysiology of the affected organs in this model will provide insight and mechanisms onto the consequences of ceramide accumulation.

2.1 Aim 1: To elucidate the effects of FD in the respiratory and visual systems.

Hypothesis: ACDase deficiency results in perturbed ocular and respiratory function

ACDase is ubiquitously expressed in all tissues, reduced enzyme activity would presumably affect proper organ function. Due to the scarcity of patients, and the severity of disease, characterization of the pathophysiology has been limited to descriptive case reports. This aim will provide a better understanding of the pathology in FD with the use of the first viable mouse model of FD. Several case reports have demonstrated signs of respiratory distress. Lung mechanics will be exploring along with a comprehensive analysis of the histopathology. Since inflammation is common in the FD, we will study the consequences of the heighted immune response as it pertains to pulmonary function.

Many LSDs contain a neurological and ocular component. Delayed development neurological pathology, and ocular manifestations are also recorded in some FD patient case reports. We have previously described the neuropathology in the FD mouse model.

We aim to extend that original study to also access the ocular function in this model. We will

63 perform non-invasive ocular imaging, histological analyses, and functional tests to assess visual function. Additionally, we aim to characterize the lipidomic profiles in both the pulmonary and ocular systems. This aim will be explored in Chapter 3 and Chapter 4.

2.2 Aim 2: To elucidate the hepatic manifestations, and corresponding gene expression changes in FD

Hypothesis: ACDase deficiency results in liver injury and impaired expression of genes that impact sphingolipid metabolism.

Beyond the cardinal triad of symptoms, inflammation and hepatosplenomegaly are commonly reported in FD. In the severe variants of FD, visceral involvement typically involves significant liver injury and failure. In this aim, we will elucidate the consequences of

ACDase deficiency in the liver. We will assess the onset disease, inflammation, and explore the lipid and sphingolipid profiles in the FD liver. To date, there is no published gene expression profiling on FD. We aim to use whole transcriptome shotgun sequencing to identify candidate gene and pathways that are altered on ACDase deficient hepatocytes.

Results from this study will provide a better understanding of the molecular pathways that drive liver disease and altered metabolic activity in FD. This aim will be explored in Chapter

2.3 Aim 3: To understand the role of MCP-1 in the pathogenesis of FD

Hypothesis: Ablation of MCP-1 will decrease inflammation and hamper pathology

Systemic inflammation is characteristic of FD. MCP-1 has previously been shown to be dramatically elevated in the FD mouse model and patients. We propose that MCP-1 is a

64 key chemokine that recruits macrophages to the various tissues, leading to downstream tissue destruction from macrophages. In this aim, we will cross our FD mouse model with a

MCP-1 knockout mouse to generate a new double mutant. We will characterize the effects of MCP-1 ablation in the hematopoietic, pulmonary, central nervous and hepatic systems.

We will further assess the cytokine profile in normal mutants and double mutants, and also highlight any changes that may occur to the sphingolipid profiles in lung, liver and brain.

This aim is explored in Chapter 6.

Chapter 3

Ocular Pathology and Visual Impairment in Acid Ceramidase

Deficiency

3.1 Abstract

Farber Disease (FD) is a debilitating Lysosomal Storage Disorder (LSD) characterized by severe inflammation, and neurodegeneration. FD is caused by mutations in the ASAH1 gene resulting in deficient acid ceramidase (ACDase) activity. Patients with

ACDase deficiency exhibit a broad clinical spectrum. In classical cases, patients will develop hepatosplenomegaly, nervous system involvement, and childhood morbidity. Like other

LSDs, ocular manifestations have been documented, but the full effect of ACDase deficiency on the visual system has not been studied in detail. We previously developed a mouse model that is orthologous for a known patient mutation in Asah1 that recapitulates human

FD. Herein we report evidence of ocular pathology in Asah1P361R/P361R mice, further asserting their suitability as a phenotypic model. Asah1P361R/P361R mice exhibit progressive retinal and optic nerve pathology. Through non-invasive ocular imaging, and histopathology analyses we revealed progressive inflammation, presence of retinal dysplasia, and significant storage pathology in various cell types in both the retina and optic nerve. Lipidomic analyses of retinal tissues revealed an abnormal accumulation of ceramides and other sphingolipids.

Electroretinograms and behavioral tests showed decreased retinal and visual responses.

Taken together, our data suggests that ACDase deficiency leads to sphingolipid imbalance, inflammation, dysmorphic retinal and optic nerve pathology, and severe visual impairment.

3.2 Introduction

Farber disease (FD) (OMIM #228000) is an ultra-rare LSD caused by an inherited mutation in the ASAH1 gene resulting in deficient ACDase activity [Levade et al. 2014].

ACDase is a key lysosomal enzyme that hydrolyzes the bioactive lipid ceramide into sphingosine (Sph) and a free fatty acid [Schuchman 2016]. Currently there is no cure for FD, and with 152 cases recorded in the literature, obtaining tissues and samples to study this disease has been challenging [Zielonka et al. 2017]. The clinical manifestations of FD are broad, however, patients with the classic variant die early in childhood [Levade et al. 2014].

The cardinal features of FD are presence of subcutaneous nodules, joint stiffness and pain, and aphonia [Levade et al. 2014]. Patients with severe forms will also develop respiratory complications, hepatosplenomegaly, and neurological decline [Ehlert et al. 2007; Bao et al.

2017]. The most frequent ophthalmic manifestation that has been described, is a cherry red spot in the macula [Cogan et al. 1966; Moser et al. 1969; Zarbin et al. 1988; Cvitanovic-

Sojat et al. 2011; Zielonka et al. 2017]. Other reported phenotypes include; corneal opacities, xanthoma-like growth on the conjunctiva, nystagmus, and macular degeneration

[Zetterström 1958; Tanaka et al. 1979; Chandwani and Kuwar 2002].

Ocular manifestations are a common feature in LSDs, and corresponding rodent models have been instrumental in characterizing their pathobiology [Sango et al. 2005;

Dannhausen et al. 2015; Wu et al. 2015; Grishchuk et al. 2016]. We previously reported the first viable model for ACDase deficiency, wherein a known human ASAH1 mutation, proline

(P) 362 to arginine (R), was “knocked-in” to the corresponding locus in murine Asah1

(P361R) [Alayoubi et al. 2013]. Mice homozygous for this mutation mirror many FD patient features including; heightened inflammation, decreased lifespan, and pathology to the hematopoietic, respiratory and neuroglial systems [Alayoubi et al. 2013; Sikora et al. 2017;

Yu et al. 2017].

Impaired ACDase activity leads to systemic ceramide accumulation in FD patients.

Ceramide and other sphingolipids are key components of membranes and play a role in a variety of cellular functions including inflammation, cell proliferation, and apoptosis [Arana et al. 2010]. The balance of ceramide and its metabolites are tightly regulated, and dysregulation will result in disease and potential visual system defects [Chen et al. 2012;

Hannun and Obeid 2017].

In this study, we aim to better understand the consequences of ACDase deficiency by presenting a comprehensive investigation of ocular manifestations in the Asah1P361R/P361R mouse. We apply non-invasive ocular imaging to monitor disease progression, and highlight the abnormal sphingolipids present in the retina. Furthermore, we report that ACDase deficiency leads to perturbed ocular function that is in part due to progressive inflammation, neurodegeneration, and abnormal storage pathology in cells of the visual system.

3.3 Material and Methods

3.3.1 Animal use, breeding and genotyping

To generate homozygous Asah1P361R/P361R mice, we crossed Asah1+/P361R heterozygotes as previously reported 4. Genotypes were confirmed by PCR using genomic

DNA from ear notches. To detect the wild-type Asah1 allele, we used the primers 5-CAG

AAG GTA TGC GGC ATC GTC ATA C-3 and 5-AGG GCC ATA CAG AGA AAC CCT GTC

TC-3, which yielded a 379 bp product. For the Asah1 knock-in allele, we used the primers 5-

TCA AGG CTT GAC TTT GGG GCA C-3 and 5-GCT GGA CGT AAA CTC CTC TTC AGA

CC-3, which amplifies a 469 bp product from the neomycin resistance cassette. All animal procedures were approved and carried out in strict adherence to the policies of the MCW

Institutional Animal Care and Use Committee (IACUC). Animals used for this study were maintained in controlled ambient illumination on a 12-hour light/dark cycle, with an illumination level of 2 to 3 lux. Exposure to bright light was kept to a minimum to all study animals for the duration of this study.

3.3.2 Slit-lamp

Mice were anesthetized with inhaled isoflurane (3% induction, 1-2% maintenance) in

0.6L/min oxygen flow. One eye was dilated and cyclopleged with 1 drop each of topical tropicamide (1%) and phenylephrine hydrochloride (2.5%) (Akron, Inc., Lake Forest, IL), respectively. Eyes were examined and photographed with the Topcon SL-D81 slit-lamp biomicroscope (Topcon Medical Systems Inc, Oakland NJ) with a digital camera (Nikon

D810 36.3MP DSLR Camera). The cornea and the lens were evaluated and photographed prior to dilation with 2.5% phenylephrine hydrochloride, and 1% tropicamide. The lens was

69 then re-evaluated and photographed after dilation. All examinations were performed by a board-certified ophthalmologist.

3.3.3 Fundus imaging and confocal scanning laser ophthalmoscopy

Mice were anesthetized and prepped for imaging as described above, and fundus photographs were taken with the Phoenix Micron IV (Phoenix Research Labs, Pleasanton,

CA). Near-infrared (NIR; 810nm) reflectance imaging and blue auto-fluorescence (BAF; excitation: 486nm, emission filter: 525/50nm) imaging was performed with a customized

Heidelberg Spectralis (Heidelberg Engineering, Heidelberg, Germany) confocal scanning light ophthalmoscope (cSLO). The automatic real-time (ART) composite mode in the

Spectralis software (version 6.6.20) was used to average 40 and 100 frames of the NIR and

BAF images, respectively.

3.3.4 Optical coherence tomography

To perform optical coherence tomography (OCT) imaging, mice were anesthetized and prepped for imaging as described above. Imaging was performed with a Bioptigen

Envisu R2200 spectral domain-optical coherence tomography (SD-OCT) system (Leica

Microsystems, Wetzlar, Germany) equipped with a Superlum Broadlighter T870 light source centered at 878.4 nm with a 186.3 nm bandwidth (Superlum, Cork, Ireland). InVivoVue (ver.

2.4.33) control software and the Bioptigen mouse objective was used for retinal imaging. A customized Bioptigen mouse stage was used to aim the imaging beam to the desired retinal location. GenTeal lubricant eye gel and Systane Ultra lubricating eye drops (Alcon, Fort

Worth, TX) were used as needed to maintain corneal hydration. Dispersion, reference arm position, and light power of the sample arm were optimized iteratively for each animal at each time point. During acquisition, all scans were displayed and acquired in logarithmic

70 intensity mode. Horizontal line scans (1mm, 1000 A-scans/B-scan; 100 repeated scans) of the retina were acquired with the optic nerve head (ONH) centered for each scan. With our system and these scan parameters, the pixel size was calculated to be 1.00 x 1.61µm (xz).

Twenty to fifty B-scans were registered to a manually selected template frame and averaged using custom software described previously [Dubra and Harvey 2010]. The registration was limited to a displacement of ~10 µm to exclude scans acquired at different retinal locations.

The Duke OCT Retinal Analysis Program was used to segment the inner limiting membrane

(ILM) and the retinal pigmented epithelium (RPE) [Chiu et al. 2010], which were clearly visible in all animals. We defined total retinal thickness (TRT) as the optical path length between these boundaries assuming a group refractive index of 1.38. The order of the images was randomized, the boundaries were manually corrected, and thickness was analyzed by an observer (A.E.S) masked to the genotype.

3.3.5 Electroretinogram

Prior to testing, mice were dark-adapted overnight. Apparatus set-up and animal preparations were conducted under dim red illumination. Mice were anesthetized and prepared as they were for imaging. Mice were placed on a heated platform (38°C). A silver- coated nylon contact lens with a custom-made active electrode was positioned on the eye.

To maintain electrical conductivity, two subdermal platinum needle electrodes were positioned in the scruff (reference) and base of the leg (ground). Prepped animals were then positioned inside the Ganzfeld dome of the Espion E2 system (Diagnosys LLC, Cambridge

UK). All recordings were completed in a custom-made Faraday cage. Signals from the ERG were differentially amplified and digitized at a rate of 5 kHz (bandpass filtered 0-100Hz).

Recording sessions started with the dark-adapted flash ERG which consisted of a six-log intensity series (-4 to 1 log cd•s/m2). Twenty responses were collected and averaged for the

-4 and -3 log cd•s/m2 stimulus with an interstimulus internal (ISI) of 5s; 10 responses were collected and averaged for the -2 and -1 log cd•s/m2 stimuli with an ISI of 10s, 5 responses were collected and averaged for the 0 and 1 log cd•s/m2 stimuli with an ISI of 20s. For the light-adapted series, mice were first exposed to a 30 cd•s/m2 white light background for 10 mins for rod saturation, then progressed to a light-adapted flash ERG over a two-log intensity series (0 to 1 log cd•s/m2). Twenty responses were collected and averaged for each condition. Recordings concluded with two flicker ERG tests carried out with continuous 5Hz and 15Hz flicker (25 cd•s/m2). Thirty responses were collected and averaged for each flicker condition.

3.3.6 Visual cliff behavioral test

To evaluate visual perception, we modified a mouse open-field setup to replicate the presence of a visual cliff [Fox 1965]. In brief, a grey circular structure with a diameter of

49.5cm was placed on top of a clear 60 cm x 60 cm plexiglass surface, half of which was hanging off a table. To enhance the edge, the portion of the plexiglass surface on the support table was covered with a checkered pattern. Lamps directed at the floor, which also had a checkered pattern, were used to enhance depth and add illumination. Mice were placed near the middle of the fake cliff and monitored for 5-minutes with a camera linked to the Any-Maze behavior tracking software (Any-Maze, Stoelting, IL). For data analysis, three zones were created; ground, air, and cliff zones. The cliff zone measured 3 cm in depth from the edge of both the ground and air side. This 3 cm area was excluded from data collection since normal animals frequently stretched their body over the cliff edge to assess the area.

For data collection, total distance traveled, and total movement time was used to measure activity. To ensure variability was reduced, we took account of circadian rhythms and

72 performed all experiments between 7 am and 11 am. All tests were performed on the same apparatus, same room, and by the same individual.

3.3.7 Histopathology and eye measurements

Mice were euthanized via CO2 inhalation, and immediately perfused with ice-cold

PBS via cardiac puncture with a 24 G needle. Globes were enucleated with optic nerve intact and fixed in 10% phosphate-buffered formalin or Davidson’s fixative for 24-48 hours.

Whole eye globes and separated optic nerves were dehydrated and embedded in paraffin.

Globes were sectioned sagittally at the midline through the optic nerve (ON) and stained for hematoxylin and eosin (H&E) and Luxol fast blue (LFB). Histology slides were scanned on the Aperio AT2 histology slide scanner (Leica Biosystems, Buffalo Grove, IL) or

NanoZoomer 2.0-HT histology slide scanner (Hamamatsu Photonics, Ichinocho, Japan). All analyses and measurements were performed on Aperio ImageScope analysis software

(Leica Biosystems, Grove, IL). Morphometric analyses of the retinal layers were obtained approximately 2-3-disc diameters away from the optic nerve. Samples that contained retinal folding 2-3-disc diameters from the optic nerve were excluded from measurements. The anterior to posterior globe measurements was measured from the midline of the cornea to the base of the retinal epithelium (Bruch’s membrane). The rest of the globe, lens and corneal measurements were obtained at the anterior/posterior or dorsal/ventral midlines.

3.3.8 Retinal dysplasia scoring

H&E stained retinal sections from Asah1+/+ and Asah1P361R/P361R mice at 3-4 and 8-9 weeks of age were evaluated for retinal dysplasia severity. Our retinal dysplasia scoring

73 system contained three categories; normal, intermediate, and severe. Samples categorized as normal contained no overt folding but may have minor ridges within the retina. The height of each ridge was 20 µm or less and no more than 2 retinal layers were affected. Samples in the intermediate category consisted of animals that developed 1-2-folds/whorls. Each fold/whorl affected up to 3 retinal layers, and the peak height of each fold/whorl was between 20 µm to 100 µm. Finally, for the severe category samples developed more than 2 folds/whorls. The folds/whorls affected more than 3 retinal layers and the peak height of each fold/whorl was over 100 µm. Cases of severe folding may also show signs of retinal detachment.

3.3.9 Immunohistochemistry and immunofluorescence

Eyes were fixed in 10% formalin over night for retinal sectioning as described above.

For immunofluorescence staining, the following primary antibodies were used: rabbit anti- ionized, calcium-binding adapter molecule 1 (Iba-1) at 1:2000 (Wako Chemicals USA,

Cambridge, MA), chicken anti-glial fibrillary acidic protein (GFAP) at 1:2000 (Aves Lab Inc,

Tigard, OR), 4’6-diamidino-2-phenylindole (DAPI) at 1:7000 (Sigma Aldrich) was used for nuclear staining. The following secondary antibodies were used to detect the primary antibodies: goat anti-chicken FITC 1:500 (Aves Lab Inc), donkey anti-Rabbit Cy3 1:500

(Jackson ImmunoResearch USA, West Grove PA). Immunofluorescence microscopy was performed on the Carl Zeiss LSM510 confocal microscope (Carl Zeiss Microscopy, LLC

Thornwood NY) using the Zeiss Aim software (Carl Zeiss Microscopy). For immunohistochemistry (IHC), the following primary, secondary antibodies and reagents were used: rat anti-mouse Mac-2 at 1:8000 (Galectin-3 clone M3/38; Cedarlane, Burlington,

Canada); biotinylated rabbit anti-rat at 1:5000 (Vector Laboratories, Burlingame, CA); avidin-

74 biotin/HRP (Vector Laboratories); DAB kit (Vector Laboratories) and Vectastain ABC Elite

Standard kit (Vector Laboratories).

3.3.10 Electron microscopy

Post-euthanasia, mice were opened and perfused with 4% paraformaldehyde, eye globes with intact ONs were removed and placed in 4% paraformaldehyde for 24-48h. ONs were from the posterior pole of the eye globe and ~2mm transverse sections were cut for

TEM processing. For analysis of the posterior chamber, the cornea was gently cut to expose and remove the lens. The remaining posterior segment structures were used for TEM processing. In brief, samples were post-fixed in 3% glutaraldehyde, then washed and placed in 2% OsO4 in phosphate buffer overnight for contrasting. After dehydration, samples were embedded in Durcupan Epon (Fluka, Hatfield, PA) for polymerization. Ultrathin sections (60 nm) were cut from tissue blocks of ON and posterior eye samples and placed on copper grids (regular Cu 200 Mesh, SPI West, PA). Ultrathin sections were further stained with uranyl acetate and lead citrate. Samples were analyzed with the JEOL 1400+ (JEOL, Tokyo,

Japan) transmission electron microscope (TEM) equipped with an Olympus Veleta CCD camera and Radius software. To assess myelin sheath thickness, the G-ratio (Axon diameter/total outer myelin sheath diameter) was measured on electron micrographs of ON cross sections in 8-9-week-old-mice. Images of the ON were obtained at magnification of x2500. The G-ratio was determined from 300-400 randomly chosen fibers per nerve cross section. Images were analyzed using the freely accessible ImageJ software (National

Institutes of Health, Bethesda, Maryland) and G-ratio measurements were performed with the G-ratio plugin and online source code (http://gratio.efil.de).

3.3.11 Sphingolipid mass spectrometry

After mice were euthanized and perfused as described above, globes were enucleated, and the retina was carefully separated from the globes under a dissection microscope. Retinal tissue was homogenized in 300 µl PBS with the Omni Bead Raptor 24 tissue homogenizer (Omni International, Inc., Kennesaw, GA) using 2.8 mm ceramic beads.

Lipids were extracted from 100 µl of retinal tissue lysate with 400 µl of methanol. The supernatant was reconstituted with 300 µl of water for mass spectrometry analyses as previously described [Yu et al. 2018]. The following internal standards were spiked in each retina homogenate prior to the extraction: 50 ng of deuterated ceramide-1-phosphate

(d18:1/16:0), (d18:1/24:0) and (d18:1/24:1) (Matreya Inc., Pleasant Gap, PA); (Avanti Polar

Lipids Inc.), 50 ng of deuterated ceramide (d18:1/22:0) d4 (Medical University of South

Carolina (MUSC) Lipidomics Core, Charleston, SC), 50 ng of C17 analog of monohexosylceramide (d:18:1/17:0) (AVANTI Polar Lipids Inc., Alabaster, AL), and 500 ng of C17 analog of sphingomyelin (d18:1/17:0) (AVANTI Polar Lipids Inc.). Quality control samples were prepared by pooling some retinal extracts for analytical performance and protocol optimization throughout these analyses. The samples were analyzed on the

Shimadzu 20AD HPLC system using reverse-phase C18 HPLC columns (Agilent Co., Santa

Clara, CA) and a Leap PAL autosampler coupled to a triple quadrupole mass spectrometer

(API-4000: Applied Biosystems, Carlsbad, CA) operated in “Multiple Reaction Mode” at the

MUSC Lipidomics Core. Positive-ion ESI mode was used to detect all sphingolipids. Retinal extraction samples were injected in duplicate for data averaging. The Analyst 1.5.1 software was used for data analysis (Applied Biosystems). Sphingolipid measurements were normalized for individual protein concentrations obtained via a Bicinchoninic acid (BCA) assay (Thermo Scientific Pierce, Waltham, MA) and expressed as fold change over Asah1+/+ mice.

3.3.12 Statistical analyses

Data are expressed as ± standard error and analyzed with a student t-test unless otherwise stated. Statistics were analyzed using GraphPad Prism 5.0 (GraphPad Software

Inc, La Jolla, CA) and MATLAB (MathWorks, Inc Natick, MA). Significant differences are expressed in the figures as *p < 0.05, **p < 0.01, and ***p < 0.001.

3.4 Results

3.4.1 Non-invasive imaging reveals ocular pathology

In comparison to control 8-9-week-old Asah1+/+ mice (Figure 9A), slit-lamp examination of the anterior chamber revealed signs of uveitis in 8-9-week-old

Asah1P361R/P361R mouse eyes (Figure 9B). These included corneal endothelial granulomatous keratic precipitates, Busacca and Koeppe nodules on the iris, and occasional pigmented deposits along the anterior lens capsule (Figure 9B). No corneal epithelial or stromal defects were observed in the slit-lamp examination.

En face infrared (IR) and blue autofluorescence (BAF) cSLO images revealed significant hyperreflectivity and autofluorescence surrounding the periphery of the ONH in 8-

9-week-old Asah1P361R/P361R mice in comparison to controls (Figure 9C,D). Variable patches of hyperreflective and autofluorescent lesions were also observed in the fundus of

Asah1P361R/P361R mice (Figure 9D). Examination of 3-4-week-old mice using IR and BAF cSLO revealed no abnormal reflectivity in both Asah1+/+ and Asah1P361R/P361R mice (data not shown). Colored fundus photographs captured from 8-9-week-old mice appeared to correspond with cSLO images (Figure 10A,B). 8-9-week-old Asah1P361R/P361R mice had significant accumulation of white deposits near the ONH, with occasional spots along the periphery of the fundus of Asah1P361R/P361R mice (Figure 10B).

Figure 9. Anterior uveitis and optic nerve pathology in Asah1P361R/P361R mice.

Slit-lamp photographs of anterior section of 8-9-week-old control Asah1+/+ and Asah1P361R/P361R mice (A,B). Inflammation in the anterior segments of the Asah1P361R/P361R eye are marked with granulomatous keratic precipitates (white arrows), Busacca and Koeppe nodules (Black arrows) and clustered pigment in anterior lens capsule (White arrow heads). Confocal scanning laser ophthalmoscope (cSLO) imaging was performed on 8-9- week-old control Asah1+/+ and Asah1P361R/P361R mice for near-infrared reflectance (IR; 810nm) and blue light auto- fluorescence (BAF; 486nm, emission filter: 525/50nm). Representative IR and BAF images of healthy retinas from Asah1+/+ mice showing no auto-fluorescence (C). Representative IR and BAF images from two Asah1P361R/P361R mice each with varying levels of reflectivity and auto-fluorescence permeating from the optic nerve (D).

Figure 10. Pathology in fundus of Asah1P361R/P361R mice.

Representative fundus photograph of 8-9-week-old Asah1+/+ and Asah1P361R/P361R mice (A,B). White reflective substance is present around the optic nerve head region of Asah1P361R/P361R mice.

To further investigate retinal involvement in Asah1+/+ and Asah1P361R/P361R mice, we performed optical coherence tomography (OCT) imaging at two-time points, first at 3-4 weeks, and again at 8-9 weeks of age. Most OCT images obtained in 3-4-week-old

Asah1P361R/P361R mice normal retinal development. However, some mice displayed less contrast within the individual retinal layers, and on occasion there were hyper-reflective spots within the vitreous (Figure 11C). This was most notable in the regions proximal to the

ONH. OCT at the later time point revealed more pronounced impairment in the retinal layers

(Figure 11D and G). Increased hyperreflectivity was most pronounced along the nerve fiber layer (NFL)/ganglion cell layer (GCL) and presence of hyperreflective specks were observed

81 within the vitreous body of 8-9-week-old Asah1P361R/P361R mouse eyes. (Figure 11G). After

OCT imaging, hematoxylin & eosin (H&E) staining of the corresponding globe confirmed

retinal dysmorphia and inflammation in the 8-9-week-old Asah1P361R/P361R mice (Figure 11H).

Figure 11. Optical coherence tomography highlights retinal pathology in Asah1P361R/P361R mice.

Representative optical coherence tomography (OCT) scans from the same eye at 3-4 and 8-9 weeks of age in a control Asah1+/+ and Asah1P361R/P361R mouse (A-D). Longitudinal OCT scans of Asah1P361R/P361R mice at 3-4 and 8-9-weeks of age show increasing disorganization of retinal morphology with age (C-D). OCT scans with corresponding H&E micrograph in 8-9-week-old Asah1+/+ and Asah1P361R/P361R mice (E-H). Scale bars are 100 µm.

3.4.2 Impaired retinal and visual function

To test retinal function, we performed dark-adapted, light-adapted, and flicker ERG on mice at both 3-4 and 8-9-weeks-of-age. In the dark-adapted condition, A-waves, which measure rod photoreceptor function, were unchanged at lower intensities, but decreased starting at the 1.0 cd·s/m2 flash intensity in 3-4-week-old Asah1P361R/P361R mice (Figure

12A,C). B-waves, which assess rod and bipolar cell function in the inner retina, were found decreased by the 0.001 cd·s/m2 intensity in 3-4-week-old Asah1P361R/P361R mice (Figure

12A,D). Dark-adapted 3-4-week-old Asah1P361R/P361R mice also displayed increased B-wave implicit time at the highest flash intensity (Figure 12F). Dark-adapted ERGs were repeated on the same mice at 8-9-weeks-of-age. There was a significant reduction in signal in the

Asah1P361R/P361R mice (Figure 12B). Reduction in A- and B-wave amplitudes were detected by the 0.001 cd·s/m2 flash intensity, as well as increased implicit times, and negligible response at the 0.0001 cd·s/m2 flash intensity (Figure 12G-J).

In the light-adapted ERG, an inference of cone function, we detected a decrease in both A- and B-wave amplitude as well as an increase in B-wave implicit time at the higher flash intensity in 3-4-week-old Asah1P361R/P361R mice (Figure 13A,C-F). Flicker ERG, another indicator of cone function, revealed a decreased signal in both 5Hz and 15Hz rates of flicker in 3-4-week-old Asah1P361R/P361R mice (Figure 13K-N). Repeating light-adapted ERGs and flicker ERG on the same mice at 8-9-weeks-of-age demonstrated a lack of response from the Asah1P361R/P361R mice (Figure 13B, G-J, L-P).

Figure 12. Dark-adapted ERG performed on Asah1P361R/P361R mice demonstrates progressive impairment in rod function in Asah1P361R/P361R mice.

Dark-adapted ERG response waveforms from 3-4 and 8-9-week-old Asah1+/+ and Asah1P361R/P361R mice (A,B). A and B-wave amplitudes and implicit time from 3-4-week-old mice (C-F). A and B-wave amplitudes and implicit times from 3-4-week-old mice (G-J) n = 10 animals for each genotype. ∗p < 0.05, ∗∗p< 0.01 ∗∗∗p < 0.001.

Figure 13. Light-adapted and flicker ERG performed on Asah1P361R/P361R mice demonstrate progressive impairment in cone function.

Light-adapted ERG response waveforms from 3-4 and 8-9-week-old Asah1+/+ and Asah1P361R/P361R mice (A,B). A and B-wave amplitudes and implicit time from 3-4-week-old mice (C-F). A and B-wave amplitudes and implicit times from 8-9-week-old mice (G-J). Flicker ERG response waveforms from 3-4 and 8-9-week-old Asah1+/+ and Asah1P361R/P361R mice (K,L). Flicker amplitudes (M-P). n = 10 animals for each genotype. ∗p < 0.05, ∗∗p < 0.01 ∗∗∗p < 0.001.

To further validate visual impairment, we performed a modified visual cliff test on 8-9- week-old mice. Control Asah1+/+ mice spent most of their time, and travelled a greater distance, within the ground side (Figure 14A,C-F). Whereas Asah1P361R/P361R mice demonstrated no preference for ground or air zones, spending roughly equal time and travelling an equal total distance within the ground and air side (Figure 14B-F). While we have previously demonstrated other behavioral deficits in Asah1P361R/P361R mice [Sikora et al.

2017], the results from this modified visual cliff test and the ERG data together suggest significant visual impairment.

Figure 14. Visual cliff test reveals perturbed depth perception in Asah1P361R/P361R mice.

Representative movement traces of 8-9-week-old Asah1+/+ and Asah1P361R/P361R mice during the visual cliff test (A,B). Distance traveled in each zone expressed as ground-to-air ratio. Asah1P361R/P361R mice spend equal amounts of time in both ground and air side. (C). Duration spent in each zone expressed as ground-to-air ratio (D). Percentage of distance traveled on the ground side (E). Percentage of time spent on the ground side (F). n = 15 animals for each genotype. ∗∗∗p < 0.001.

3.4.3 Increased retinal thickness

Asah1P361R/P361R mice are smaller than age matched controls and by 4-5 weeks of age, they start to progressively lose weight until they succumb to disease [Alayoubi et al.

2013]. During examination of the eyes, we noticed Asah1P361R/P361R mice had eye openings that were phenotypically smaller and tighter compared to age matched controls (Figure

15B). To assess whether eye development was affected, we performed gross measurements on micrographs obtained from H&E stained globe sections. Measurements revealed no significant differences in globe or lens diameter in 3-4 and 8-9-week-old

Asah1+/+ and Asah1P361R/P361R mice (Figure 16A-D). Measurements of the anterior chamber also revealed no differences in the thickness of the cornea epithelium and stroma between

Asah1+/+ and Asah1P361R/P361R mice at both 3-4 and 8-9 of age (Figure 16E,F).

Figure 15. Reduced eye opening in Asah1P361R/P361R mice.

Representative photograph of 8-9-week-old Asah1+/+ and Asah1P361R/P361R mice taken during slit-lamp analysis (A,B). Asah1P361R/P361R mice display a smaller eye opening (B).

Figure 16. No changes in globe diameter, lens diameter, or corneal thickness in Asah1P361R/P361R mice.

Measurements of globe diameter, lens diameter and cornea thickness analyzed in H&E micrographs of 3-4 and 8-9-week-oldAsah1+/+ and Asah1P361R/P361R mice. Anterior to posterior and ventral to dorsal globe diameters were similar in Asah1+/+ and Asah1P361R/P361R mice (A,B). Anterior to posterior and ventral to dorsal lens diameter measurements were similar in Asah1+/+ and Asah1P361R/P361R mice (C,D). Corneal thickness (epithelium and stroma) was similar in Asah1+/+ and Asah1P361R/P361R mice (E,F). n= 10 animals per group. ns (not significant).

Surprisingly, Asah1P361R/P361R mice displayed increased retinal thickness as early as

3-4 weeks of age when compared to age matched Asah1+/+ mice (Figure 17B). Specifically, by 3-4 weeks of age the outer nuclear layer (ONL) was significantly thicker, and by 8-9 weeks of age increased thickness was measured in both the inner nuclear layer (INL) and

ONL (Figure 17D). To validate our histology measurements, retinal thickness was analyzed from our previous OCT images with OCT retinal analysis software [Dubra and Harvey 2010].

Since some retinal layers from OCT images showed less contract individual layers were not

88 discernable, however total retinal thickness was measured. Analyses of OCT images revealed that 3-4-week-old Asah1P361R/P361R mice exhibited increased nasal and temporal retinal thickness in comparison to age matched Asah1+/+ mice (Figure 17E). The increased thickness persists in the 8-9-week-old Asah1P361R/P361R mice (Figure 17F).

Figure 17. Increased retinal thickness in Asah1P361R/P361R mice.

Morphometric analyses of retinas from 3-4 and 8-9-week-old control Asah1+/+ and Asah1P361R/P361R mice from H&E stained sections (A-D). n= 10 animals per group. ns (not significant), *p<0.05 **p<0.01. Retinal thickness plot generated from OCT B-scans in 3-4 and 8-9-week-old Asah1+/+ and Asah1P361R/P361R mice (E-H). n= 6-10 animals per group. Retinal thickness was significantly increased in the Asah1P361R/P361R mice when we performed an n-way ANOVA for factors that included genotype, age, and retinal location. p < 0.0001 (E-F).

3.4.4 Variable penetrance of retinal dysplasia

To confirm the presence of retinal dysplasia, we examined H&E stained retina to evaluate retinal morphology. All the retinas from 3-4-week-old control Asah1+/+ and 3-4- week-old Asah1P361R/P361R mice showed no dysplasia and were within normal limits (Figure

18A,B and Table 8). Examination of 8-9-week-old Asah1P361R/P361R mice revealed a range in retinal dysplasia (Figure 18D-F and Table 8). Approximately 25% of globes from

Asah1P361R/P361R mice were within normal limits, 45% developed an intermediate degree of dysplasia, and 30% had severe retinal folding and injury (Table 8). Retinal dysplasia was most pronounced in regions most proximal to the ONH. Mice that developed severe retinal disturbance often involved multiple layers, with folds reaching into the vitreous body (Figure

18F).

Within normal Intermediate Severe Group Number of globes limits dysplasia dysplasia

8-9-week-old Asah1+/+ 100% (20) -- -- 20

3-4-week-old 100% (20) -- -- 20 Asah1P361R/P361R

8-9-week-old 25% (10) 45% (18) 30.0% (12) 40 Asah1P361R/P361R

Table 8. Number of globes examined

Percent and sample distribution for retinal dysplasia and injury from 3-4-week-old Asah1+/+, as well as 3-4-week- old and 8-9-week-old Asah1P361R/P361R mice. n= 20 retinal samples for 8-9-week-old Asah1+/+ and 3-4-week-old Asah1P361R/P361R mice were scored. n=40 retinal samples for 8-9-week-old Asah1P361R/P361R mice were scored.

Figure 18. Varying degree of retinal dysplasia in Asah1P361R/P361R mice.

Light micrographs of H&E stained globe cross sections from 3-4 and 8-9-week-old Asah1+/+ and Asah1P361R/P361R mice (A-F). Varying degree of retinal dysplasia is present in 8-9-week-old Asah1P361R/P361R mice (D-F). Scale bars are 500 µm.

3.4.5 Inflammation and retinal pathology

Recruitment of infiltrating cells can be seen as early as 3-4 weeks of age, and become more numerous by 8-9 weeks of age in Asah1P361R/P361R mice (Figure 18B,D-F).

Control Asah1+/+ mice do not display this cell recruitment (Figure 19A). Infiltrating cells can be observed in both the anterior chamber and vitreous humor of 8-9-week-old

Asah1P361R/P361R mice (Figure 19B). An increased abundance of infiltrating cells was notable in the vitreal region above the ONH, presumably where the central retinal vein and artery is located (Figure 19F,G). We performed IHC staining for Mac-2 on globes from 8-9-week-old

Asah1+/+ mice and 8-9-week-old Asah1P361R/P361R mice (Figure 19H-J). Many of these infiltrated cells are macrophages (Figure 19I,J). While many of these Mac-2 positive cells were localized around the ONH, macrophages were also observed within the vessels of the

GCL and along the choroid/pigment epithelium layer in regions where retinal folding occurred (Figure 19I,J). In the 8-9-week-old Asah1P361R/P361R mice that developed significant retinal folding, detachment of the retina often occurred, as well as presence of subretinal hemorrhage (Figure 19G). Examination of the NFL/GCL layers of 8-9-week-old

Asah1P361R/P361R mice revealed increased vacuolization, and occasionally the presence of intracytoplasmic, orange-red eosinophilic granular material (Figure 19D). Of the 8-9-week- old Asah1P361R/P361R mice that had severe retinal dysplasia, the NFL/GCL layer showed degeneration and sloughing (Figure 19G).

Figure 19. Retinal pathology and recruitment of inflammatory cells in Asah1P361R/P361R mice.

Light micrograph of the anterior chamber of 8-9-week-old Asah1+/+ mouse (A). Presence of inflammatory cells in anterior chamber (black arrow) in 8-9-week-old Asah1P361R/P361R mouse (B). Higher magnification light micrograph featuring the retinal layers in 8-9-week-old Asah1+/+ and Asah1P361R/P361R mice (C,D). The retinal layers of Asah1P361R/P361R mice are less organized. The GCL contains many vacuoles and granular cytoplasmic staining (E). Numerous inflammatory cells (white arrowheads) are seen along the ONH in 8-9-week-old Asah1P361R/P361R mice (F). In Asah1P361R/P361R mice that exhibit significant retinal dysplasia, loss/sloughing of the GCL is observed (white arrows), as well as retinal detachment leading to presence of hemorrhaging (black arrowhead; G). Immunohistochemistry (IHC) staining for Mac-2 in Asah1+/+ and Asah1P361R/P361R mice (H-J). Many of the invading inflammatory cells that near the optic nerve vessels are macrophages (white arrow head; O). Invading cells can also be detected in the choroid layers where there are retinal folds (P). Scale bars are 50 µm.

Ultrastructural analysis of the retina in 8-9-week-old Asah1P361R/P361R mice revealed extensive storage pathology that is absent in 8-9-week-old control mice (Figure 20A,B,F-H).

Ganglion cells in 8-9-week-old Asah1P361R/P361R mice contained excess of zebra-like storage bodies (Figure 20C,D). Accumulation of storage bodies was also present in endothelial cells of the inner plexiform layer (IPL) (Figure 20D). In regions near the IPL, 8-9-week-old

Asah1P361R/P361R mice displayed abnormal macrophage-like cells with excessive curved semi linear tubular bodies (CTB) that were within the vicinity of capillaries (Figure 20E). CTB-like storage vacuoles were also observed in the endothelial cells of the choroid capillary, and melanosome containing cells of the outer choroid (Figure 20I). Lastly, storage bodies were also present within scleral fibroblasts (Figure 20J,K).

Figure 20. Ultrastructure pathology in ganglion cells and optic nerve disc in Asah1P361R/P361R mice.

Electron micrographs highlighting the GCL (nuclei-black asterisks), inner plexiform layer (IPL) of 8-9-week-old Asah1+/+ mouse (A,B). Electron micrograph of 8-9-week-old Asah1P361R/P361R mice highlighting the GCL and IPL. GCL (nuclei-black asterisks) contain numerous zebra-like storage vacuoles (white arrow heads) in their cytoplasm (C,D). Zebra-like storage vacuoles (black arrow) are present in endothelial cells near the capillary lumina (white asterisks; D). Adjacent to the IPL, zebra-like storage (white arrow) can be seen near astrocytic-like cells (E). Electrons micrographs of the choroid in Asah1+/+ mouse (choroid capillary lumen: black asterisks; F,G). Scleral fibroblasts in Asah1+/+ mouse (nuclei: white asterisks) are spindle shaped and surrounded by collagen- rich extracellular matrix (H). Electron micrograph of Asah1P361R/P361R mouse retina where the retinal pigmented epithelium is detached from Bruch’s membrane (black arrow head; potential processing artifact) endothelial cells of the choroid capillaries (lumen: black asterisk) contains CTB-like storage vacuoles (small black arrow, and higher magnification inset) (I). CTB-like storage bodies are found in melanosome containing cells in the outer choroid (small black arrows) and scleral fibroblasts (large black arrows) (J,K). All scale bars are 5 µm.

3.4.6 Optic neuropathy and storage pathology

H&E stained optic nerves from 8-9-week-old Asah1+/+ and Asah1P361R/P361R mice

(Figure 21,B). Asah1P361R/P361R mice displayed a thickening of the ON pial sheath (Figure

21B). ONs also contained an accumulation of mononuclear cellular infiltrate with abundant vacuolated cytoplasm and/or cytoplasm distended with granular material that preferentially accumulated in the areas proximal to the optic vein (Figure 21B). IHC stain for Mac-2 revealed many infiltrated macrophages present along the ON track in 8-9-week-old

Asah1P361R/P361R but not 8-9-week-old Asah1+/+ mice (Figure 21C,D).

Ultrastructure analysis of 8-9-week-old Asah1+/+ and Asah1P361R/P361R ONs revealed abnormal storage in Asah1P361R/P361R macrophages accumulated in the intraneuronal perivascular spaces (Figure 21E-H). These macrophage and fibroblast-like cells contained

CTB storage in vacuoles. Abnormal storage vacuoles were also present in the endothelial cells (both pial and intraneuronal) of vessels, and capillaries (Figure 22B,C,E,F).

Figure 21. Optic nerve pathology in Asah1P361R/P361R mice.

Light micrograph of H&E and IHC stained optic nerve (ON) sections for Mac-2 from 8-9-week-old Asah1+/+ and Asah1P361R/P361R mice (A-D). Scale bars are 100 µm. Electron micrographs of a normal ON from 8-9- week-old Asah1+/+ mice, with astrocytic processes (black arrows) (E,F). Electron micrograph of ON from 8-9- week-old Asah1P361R/P361R mice with macrophage-like cells filled with curved semi linear tubular body (CTB) like storage bodies (with white arrows), and astrocytic process that contained zebra-like storage bodies (black arrow) (G,H). Scale bars are 5 µm.

Figure 22. Ultrastructure pathology of the ON of Asah1P361R/P361R mice.

Representative electron micrograph featuring vascular cells in the ON of 8-9-week-old Asah1+/+ and Asah1P361R/P361R mice. Normal structured pial vessel in a Asah1+/+ mouse (A). Pial vessel in a Asah1P361R/P361R mice that is thickened by accumulation of vacuolated cells with macrophage and fibroblast morphology (B,C). Storage pathology is also present in the endothelial cells that line the optic vein (B,C). Normal intraneuronal capillary in the Asah1+/+ mouse (D). Endothelial cells and peri-capillary macrophages display storage vacuoles in Asah1P361R/P361R mice (E,F). Abbreviations and symbols: EC ncl – endothelial cell nucleus, AST ncl – astrocyte nucleus, OLG ncl – oligodendrocyte nucleus, MCP ncl – macrophage nucleus, SMC ncl – smooth muscle cell nucleus, EG ncl – eosinophil granulocyte nucleus, OV – optic vein, black arrows – zebra-like storage bodies, white arrows – partly cleared storage vacuoles with curvilinear tubular (Farber) bodies, white arrowheads – partly cleared MCP storage vacuoles with curvilinear tubular (Farber) bodies, black arrowheads – EG granules. Scale bars = 2 µm.

3.4.7 Presence of Astrogliosis

To further characterize neuroinflammation within the visual system, we assessed retinal and optic nerve tissue from 3-4 and 8-9-week-old mice by immunofluorescence (IF) staining (Figure 23). We stained for glial fibrillary acidic protein (GFAP), a label for glial cells such as astrocytes and Müller cells, and for ionized calcium-binding adaptor molecule 1

(Iba-1), a marker for microglia and macrophages. Increased GFAP staining was found on the NFL/GCL of retinal sections from 3-4-week-old Asah1P361R/P361R mice (Figure 23B). On occasion, we also noticed GFAP positive staining on the INL and outer plexiform layer (OPL) of retinal sections from 3-4-week-old mutant mice (Figure 23B). GFAP staining in the retina was significantly more pronounced and affected multiple layers in 8-9-week-old

Asah1P361R/P361R mice (Figure 23D). A similar trend was observed in ON tissue from

Asah1P361R/P361R mice, which displayed increased GFAP staining at both 3-4 and 8-9 weeks of age in comparison to controls (Figure 23E-H). Together these observations suggest progressive astrocytosis and Müller cell activation in the ocular system of ACDase deficient mice.

The presence of activated microglia is a severe phenotype that affects the brains of

Asah1P361R/P361R mice [Sikora et al. 2017]. To assess whether activation of microglia was present in the visual system, retinal and ON tissue were stained for Iba-1. The retina of

Asah1P361R/P361R mice did not contain many cells that were positive for Iba-1. On occasion we observed some microglia-like cells that had a branching phenotype in 8-9-week-old

Asah1P361R/P361R mice, however it was infrequent, and those cells were not specific to one particular region (Figure 23B,D). Within the optic nerve, cells positive for Iba-1 did not have the activated phenotype.

Figure 23. Presence of astrogliosis and activated Müller cells

Immunofluorescence (IF) staining for glial fibrillary acidic protein (GFAP), Ionized calcium binding adapter molecule 1 (iba-1), and 4',6-diamidino-2-phenylindole (DAPI) on retina (A-D) and optic nerves (E-H) of 3-4 and 8- 9-week-old Asah1+/+ and Asah1P361R/P361R mice. GFAP is a marker for astrocytes and Müller cells, and Iba-1 is a marker for microglia and macrophages (white arrow). Scale bars are 50 µm.

100

3.4.8 Neuroaxonal dystrophy and reduced axonal density

Staining for H&E on ON tissue from Asah1+/+ and Asah1P361R/P361R mice revealed less pronounced cytoplasmic staining in the Asah1P361R/P361R mice. The decreased cytoplasmic staining was present by 3-4-week-old and decreased became fainter by 8-9 weeks of age in

Asah1P361R/P361R mice (Figure 24A). When we stained corresponding ON tissue from Asah1+/+ and Asah1P361R/P361R mice with Luxol fast blue (LFB) we observed a similar trend. ON samples from Asah1P361R/P361R mice displayed a reduced LFB staining intensity. Together this suggests perturbations to the ON architecture in the Asah1P361R/P361R mice (Figure 24B).

Furthermore, amongst Asah1P361R/P361R animals, the LFB stained ONs appeared more disorganized and faint in 8-9-week-old samples versus 3-4-week-old samples suggesting demyelination (Figure 24B). Ultrastructure analysis on ON tissue (inside and periphery) from

8-9-week-old animals, revealed that the ON of mutant mice had a qualitative reduction in axon density (Figure 24C,D,E). ON from Asah1P361R/P361R mice also showed discernable storage vacuoles in astrocyte-like cells, and zebra-like storage vacuoles within the cytoplasm of oligodendrocytes (Figure 24G,H,J,K). Signs of neuroaxonal dystrophy in some axons via the formation of axonal spheroids were also observed in the ON of

Asah1P361R/P361R mice (Figure 24D,E).

To assess myelin, we measured the G-ratio (axon diameter/fiber diameter, where fiber diameter represents the sum of axon diameter and myelin sheath thickness) on the ON from 8-9-week-old mice. Our analysis revealed a significant reduction in G-ratio in

Asah1P361R/P361R in comparison to control Asah1+/+ mice (Figure 24L). No differences were detected in ON axon diameter between 8-9-week-old Asah1+/+ and Asah1P361R/P361R mice

(Figure 24M). Further evaluation revealed a reduction in the density of myelinated axons in

Asah1P361R/P361R ONs, however, no changes in the percentage of myelinated axons were observed, in the axons that were assessed (Figure 24N,O).

101

Figure 24. Optic nerve dystrophy and storage pathology

Representative light micrographs of longitudinal ON sections stained for H&E and luxol fast blue (LFB) from 3-4 and 8-9-week-old Asah1+/+ and Asah1P361R/P361R mice (A,B). Electron micrographs of ON axons in 8-9-week-old Asah1+/+ and Asah1P361R/P361R mice. Representative illustration of the ON axon diameters and axonal density in Asah1+/+ mice (C). Asah1P361R/P361R mice appear to show less density in ON axons (D,E). Presence of neuronal axonal dystrophy (axonal spheroid) containing accumulated axonal cargo (highlighted by black rectangle in panel D and enlarged in panel E. An astrocyte from an Asah1+/+ mouse with finely structured cytoplasmic extensions (F). Hyperplastic astrocytes from Asah1P361R/P361R mice with zebra-like and CTB containing storage vacuoles (G,H). Normal structured oligodendrocyte from an Asah1+/+ mouse (I). Oligodendrocytes containing zebra-like storage bodies in cytoplasm (J,K). Abbreviations and symbols: AST – astrocyte, AST ncl – astrocyte nucleus, OLG ncl – oligodendrocyte nucleus, black arrows – zebra-like storage bodies, white arrow – storage vacuole with curvilinear tubular (Farber) bodies. Scale bars are 2 µm. The G-ratio (axon diameter/fiber diameter) was plotted against axon diameter. In 8-9-week-old animals, Asah1P361R/P361R mice (1350 axons; n=3) show a difference in myelination compared to Asah1+/+ mice (1333 axons; n=3), (p <0.001 from clustered rank sum test) (L). Comparison of axon diameters (M). Quantification of myelinated axon density and percentage of myelinated axons (N,O).

102

3.4.9 Altered sphingolipid profile in the retina

Sphingolipids were measured via liquid chromatography-mass spectrometry in retinal lipid extracts from 8-9-week-old Asah1+/+ mice and Asah1P361R/P361R mice. Sphingolipids measured included ceramide, ceramide-1-phosphate (C1P), sphingomyelin (SM), and monohexosylceramide (MHC). In 8-9-week-old control Asah1+/+ mice, SM was the most abundant of the measured sphingolipids, followed by ceramide, MHC and C1P (Figure 25A).

In 8-9-week-old Asah1P361R/P361R mice the most abundant sphingolipid measured was ceramide, followed by SM, MHC, and C1P (Figure 25A). As ceramide was dramatically elevated Asah1P361R/P361R it was not a surprise to find a significant reduction in the abundance of SM (Figure 25A).

Analysis of the individual species for each class of sphingolipid between Asah1+/+ and Asah1P361R/P361R mice revealed a pan increase in all evaluated ceramide species in

Asah1P361R/P361R mice (Figure 25B). While the fold change of total SM was unchanged in

Asah1P361R/P361R mice, SM 18:0 was found slightly elevated, and SM 24:1 and 24:0 were slightly depressed (Figure 25C). For MHC, the C18:0 species had the highest fold change in the 8-9-week-old Asah1P361R/P361R mice (Figure 25D). This significant increase was associated with a total increase in MHC (Figure 25D). Lastly, no significant differences in

C1P were noted between 8-9-week-old Asah1+/+ and Asah1P361R/P361R mice (Figure 25E).

Alterations to the relative abundance of each class of sphingolipids was also detected between 8-9-week-old Asah1+/+ and Asah1P361R/P361R mice. A decrease in the percentage of C16:0 ceramide was associated with an increase in all other ceramide species measured (Figure 25F). An increase in SM18:0 and a slight decrease in the other

SM species was observed (Figure 25G). The greatest change in MHC abundance was an increase in the C18:0 species which coincided with decreased C16:0, C20:0, C24:1 and

103

C24:0 species (Figure 25H). Lastly, for C1P a decrease in C24:1 resulted in an increase in

C24:0 (Figure 25).

Figure 25. Altered sphingolipid species in the retina of Asah1P361R/P361R mice.

Sphingolipids were quantified from 8-9-week-old control Asah1+/+ and Asah1P361R/P361R mice by liquid chromatography-mass spectrometry. Percentage composition of the different classes of sphingolipids (A). The fold change over Asah1+/+ of individual isoform within each sphingolipid class (B-E). The percent composition of the various sphingolipid classes measured (F-I). n = 5 animals for each genotype. ∗p< 0.05, ∗∗p< 0.01 ∗∗∗p < 0.001.

104

3.5 Discussion and conclusion

ACDase deficiency manifests in a wide spectrum of clinical signs. The cardinal phenotypes include painful joints, formation of nodules, and aphonia [Levade et al. 2014]. In classical and more severe variants, neurological involvement is common and is often associated with functional and behavioral deficits [Eviatar et al. 1986; Ehlert et al. 2007].

Ocular manifestations are commonly seen in patients that also display neurological dysfunction, with the most common phenotype being the formation of a cherry red spot

[Cogan et al. 1966; Moser et al. 1969; Zarbin et al. 1988; Cvitanovic-Sojat et al. 2011;

Zielonka et al. 2017]. Herein, we have provided an analysis of the ocular pathology present in the ACDase deficient mouse. We demonstrate that a deficiency in ACDase leads to a broad range of ophthalmic abnormalities. We highlight findings from non-invasive ocular imaging, characterize the affected cell types, and demonstrate visual impairment.

While ocular disease is often present in FD, reports on eye phenotype have thus far been restricted to superficial/observational descriptions. In the literature, there are only two studies on the visual aspects of FD. The first was published in 1966 and applied light microscopy and histological techniques to reveal the presence of lipid granules within the ganglion cells of a FD patient who died at 11-months of age [Cogan et al. 1966]. The second report, published in 1985, described the retinal ultrastructure of a FD patient who died before the age of 3 [Zarbin et al. 1985]. Five types of cytoplasmic inclusions were identified in the patient, with the most abundant being described as “flattened stacks of osmophilic lamellae”

[Zarbin et al. 1985]. The retinal ganglion cells, glia and phagocytic cells all showed significant inclusions [Zarbin et al. 1985]. The findings from these reports are recapitulated in our murine model. Our H&E retinal tissue staining revealed a disorganized and vacuolated

GCL in Asah1P361R/P361R mice. Ultrastructure analyses of the retina also revealed significant storage pathology in various cell type much like those reported by Zarbin and colleagues

105

[Zarbin et al. 1985]. Retinal pathology such as this has been noted in retinal ganglion cells in other LSDs such as Niemann-Pick disease Type A, Pompe disease, and Tay-Sachs disease

[Cogan and Kuwabara 1968; Brownstein et al. 1980; Yanovitch et al. 2010]. In this study, we observed significant storage pathology in endothelial cells, astrocytes, oligodendrocytes, and macrophage-like cells in the retina and or optic nerve in Asah1P361R/P361R mice.

A heightened state of inflammation is characteristic of FD. Increased presence of inflammatory cells has been previously reported in the lungs, brain, and hematopoietic organs in our mouse model of FD [Sikora et al. 2017; Yu et al. 2017; Dworski et al. 2015].

Thus, finding that a deficiency in ACDase activity also leads to inflammation in the eye and optic nerve was unsurprising. Through non-invasive ocular imaging, we noted the formations of granulomas and nodules in the anterior chamber. Additionally, deposits proximal to the

ONH of the fundus were seen with cSLO, and hyperreflectivity was noted along the ONH and vitreous body by OCT. These observations can be explained by the presence of inflammatory cells in our histopathological analyses of retinal and optic nerve tissue. IHC staining for Mac-2 and ultrastructure analyses demonstrated a significant presence of abnormal looking macrophages primarily along the choroid and other vessels along the retina and optic nerve. We previously reported increased vascular permeability in the

Asah1P361R/P361R mice [Yu et al. 2017]. While that study was focused on lung pathology, one experiment involved a tail vein injection of Evans blue dye (EBD), and measured EBD accumulation in various organs in Asah1P361R/P361R mice [Yu et al. 2017]. While not reported in that study, we noted qualitatively that Asah1P361R/P361R mouse eyes were more intensely stained than controls. We have also previously demonstrated elevations in inflammatory cytokines such as monocyte chemoattractant protein (MCP)-1, MCP-3 and MCP-5 [Dworski et al. 2017; Yu et al. 2017; Yu et al. 2018] in Asah1P361R/P361R serum and organs. In the case of MCP-1 we also verified its increase in FD patient plasma [Dworski et al. 2017]. While cytokines were not measured in the retina it is possible the combination of vascular leakage

106 and altered cytokine production previously reported could contribute to the macrophage and neutrophil recruitment described herein.

We further characterized neuroinflammation via IF staining for GFAP and iba-1 in retinal and optic nerve tissue. Asah1P361R/P361R mice showed a progressive accumulation of

GFAP in the retina and optic nerve suggesting activation of astrocytes and Müller cells.

While we noted significant macrophage infiltration, iba-1 staining was largely unchanged, suggesting limited microglia involvement and that the majority of infiltrated macrophages did not express iba-1. This data contrasts with our previous findings where we showed significant microglia and astrocyte activation in the brains of Asah1P361R/P361R mice [Sikora et al. 2017; Yu et al. 2018]. Even within the brain, microglia were preferentially found in the subcortical white matter and the CA1 region of the hippocampus of Asah1P361R/P361R mice

[Sikora et al. 2017]. It is possible that the eye like other regions of the brain are less affected by activated microglia.

Retinal dysplasia was also a prevalent feature in Asah1P361R/P361R globes. Over 75% of globes analyzed had intermediate to severe retinal folding. Within the retinal folds inflammatory cells were often discernable. Retinal folding was most prevalent in the regions near the ONH. This may be due to the proximity of the central optic vein and artery, the region where we also noted increased macrophage infiltration. The phenotype seen in eyes of Asah1P361R/P361R mice share some similarities with the rodent models of experimental autoimmune uveoretinitis (EAU) [Gasparin et al. 2012]. During the early phase after EAU induction, cellular infiltrates can be observed in the vitreous and optic nerve head of mice eyes [Chen et al. 2013]. As mice transition into the acute phase of EAU, inflammation is more severe, and retinal folding and edema will occur [Commodaro et al. 2010; Chan et al.

1990; Caspi et al. 2008].

Furthermore, the association between inflammation and retinal dysplasia has been reported in other models such as cats infected with feline leukemia virus, dogs that were

107 inoculated with canine herpes, and mice that were infected with Sindbis virus. [Albert et al.

1976; Albert et al. 1977; Carreras et al. 1982]. The range of retinal dysplasia seen in the

Asah1P361R/P361R mice is likely due to varying rates of inflammation. Other disease phenotypes that we have previously demonstrated to exhibit variability in the

Asah1P361R/P361R mice include presence of hydrocephalus, and the varying intensities of inflammatory cytokine expression in plasma [Sikora et al. 2017; Dworski et al. 2017].

Sphingolipid metabolites have been shown to both regulate the function and trafficking of inflammatory cells, and conversely there is some evidence demonstrating that sphingolipid metabolism can be altered due to inflammation [Rivera et al. 2008; Cyster and

Schwab 2012; Maceyka and Spiegel 2014]. This has been shown in the eye where cases of acute uveitis can change both lipid and sphingolipid profiles [Wang 2017]. As well direct intraocular injection of bioactive sphingolipids such as C8:0 in mice can induce a dose dependent effect in recruiting inflammatory cells [Chan et al. 2012]. In fact, due to these and other observations, targeting the sphingolipid pathway has been suggested as an approach to treat uveitis [Copland et al. 2012]. One of the most promising agents is FTY720 or fingolimod, an approved drug for the treatment of multiple sclerosis (MS). FTY720 has shown significant anti-inflammatory effects in EAU rodent models [Kurose et al. 2000;

Commodaro et al. 2010; Raveney et al. 2008]. FTY720 is an analogue of Sph that is derived from myriocin, a potent inhibitor of the de novo ceramide synthesis pathway [Chueh and

Kahan 2003]. FTY720 is phosphorylated by sphingosine-kinase-2 in a similar manner to the modification of Sph to sphingosine-1-phosphate (S1P) [Chueh and Kahan 2003]. In addition to anti-inflammatory properties via S1P binding, FTY720 has also been shown to inhibit ceramide synthase (CerS), the class of enzyme that regulates ceramide acyl chain length during synthesis [Berdyshev et al. 2009]. Due to its anti-inflammatory effects, and potential decrease ceramides, it is possible that treatment with FTY720 in conjunction with other anti- inflammatory compounds may impede pathogenesis of FD.

108

Maintaining sphingolipid equilibrium is essential for proper health, and vision. Ocular manifestations are commonly reported in LSDs and animal models have been instrumental in understanding disease progression. Studies on the acid sphingomyelinase (ASMase) deficient mouse, the model for Niemann-Pick disease, has revealed progressive retinal degeneration, and inflammation [Wu et al. 2015; Dannhausen et al. 2015]. In the context of

Sandhoff disease, work on the β-hexosaminidase (Hexβ) mutant mouse showed robust storage pathology preferentially in the retinal ganglion cells, and also impaired neurite outgrowth from cultured retinal explants [Sango et al. 2005; Denny et al. 2007]. Decreased

ERG amplitude has also been documented in mice deficient in ASMase, HexB, ceramide synthase CerS1, CerS2 and CerS4 [Dannhausen et al. 2015; Wu et al. 2015; Brüggen et al.

2016]. The similarity in disease manifestation seen in these reports and our current study suggests that insults to sphingolipid metabolism enzymes lead to comparable pathology as homeostatic control of sphingolipids is tightly regulated.

Ceramide is the central signaling lipid in the sphingolipid pathway. In more common retinal diseases such as retinitis pigmentosa, increases in ceramide can lead to downstream effects such as photoreceptor cell death [Sanvicens and Cotter 2006]. One study in the retinal degeneration 10 (rd10) mouse model showed that ceramides are increased with progressive photoreceptor degeneration and that treatment with the inhibitor myriocin could reduce ceramide and protect photoreceptors from apoptosis [Strettoi et al. 2010]. Another study showed a neuroprotective effect of FTY720 that acted independent of its immunosuppressive action in a rat model of light induced retinal degeneration [Chen et al.

2013]. Together these reports demonstrate the deleterious effects of ceramide in the retina.

In the case of chronic ceramide accumulation such as in FD, alternative pathways may be switched on in response to sphingolipid buildup. Further work will be required to fully understand the role of ceramide signaling in FD and visual biology.

109

Our study has provided the first in-depth analysis of the ocular manifestations and optic nerve pathology in the ACDase deficient mouse. We have highlighted parallels between human cases and our model particularly in terms of inflammation and storage pathology, and central nervous system decline. Given the lack of patient cases and available human tissue to study, our mouse model has aimed to fill this gap. Our results demonstrate a progressive decline in retinal function coinciding with severe retinal and optic pathology and sphingolipid accumulation in the Asah1P361R/P361R mice. While these insights are exploratory, they may provide insight into the possible ocular pathology present in classical and severe variants of human FD.

Currently, there is an ongoing study on clinicaltrials.gov (ID NCT03233841) to understand the natural history of FD. While our data provides insight on ocular pathology, most importantly we have demonstrated the feasibility of non-invasive ocular imaging to assess and track ocular pathology. Use of non-invasive ocular imaging in FD patients may not only aid in diagnosis but serve to monitor the inflammatory and neurological consequences of FD. Furthermore, recombinant enzyme therapy is currently being developed for the treatment of FD as well as interest in gene therapy [He et al. 2017;

Alayoubi et al. 2013; Ramsubir et al. 2008]. Noninvasive monitoring of the ocular conditions may also be a modality to assess the efficacy of these and future therapies.

110

Chapter 4

Chronic Lung Injury and Impaired Pulmonary Function in a

Mouse Model of Acid Ceramidase Deficiency

This work is adapted from Yu et al., 2017 with permission from the American Journal of Physiology- Lung Cellular and Molecular Physiology.

4.1 Abstract

Farber disease (FD) is a debilitating lysosomal storage disorder (LSD) caused by a deficiency of acid ceramidase (ACDase) activity due to mutations in the gene ASAH1.

Patients with ACDase deficiency may develop a spectrum of clinical phenotypes. Severe cases of FD are frequently associated with neurological involvement, failure to thrive, and respiratory complications. Mice homozygous (Asah1P361R/P361R) for an orthologous patient mutation in Asah1 recapitulate human FD. In this study, we show significant impairment in lung function, including low compliance and increased airway resistance in a mouse model of ACDase deficiency. Impaired lung mechanics in Farber mice resulted in decreased blood oxygenation and increased red blood cell production. Inflammatory cells were recruited to both perivascular and peribronchial areas of the lung. We observed large vacuolated foamy histiocytes that were full of storage material. An increase in vascular permeability led to protein leakage, edema, and impacted surfactant homeostasis in the lungs of Asah1P361R/P361R mice. Bronchial alveolar lavage fluid (BALF) extraction and analysis revealed accumulation of a highly turbid lipoprotein-like substance that was composed in part of surfactants, phospholipids, and ceramides. The phospholipid composition of BALF

111 from Asah1P361R/P361R mice was severely altered, with an increase in both phosphatidylethanolamine (PE) and sphingomyelin (SM). Ceramides were also found at significantly higher levels in both BALF and lung tissue from Asah1P361R/P361R mice when compared with levels from wild-type animals. We demonstrate that a deficiency in ACDase leads to sphingolipid and phospholipid imbalance, chronic lung injury caused by significant inflammation, and increased vascular permeability, leading to impaired lung function.

112

4.2 Introduction

Farber disease (FD) (OMIM #228000) is a rare Lysosomal Storage Disorder (LSD) caused by an inherited deficiency in acid ceramidase (ACDase) activity; the enzyme that hydrolyzes the bioactive sphingolipid ceramide into sphingosine and a fatty acid. Case reports catalogue ACDase deficiency as a spectrum disorder with varying symptoms.

Patients with the classical form of FD die in early childhood [Devi et al. 2006; Ahmad et al.

2009; Levade et al. 2014].

Manifestations of ACDase deficiency include growth of subcutaneous nodules that cause pain in the joints, development of a hoarse cry that progresses to aphonia, and severe neurological deficits. Other characteristics of this debilitating disorder may include hepatosplenomegaly and respiratory difficulties [Levade et al. 2014]. Decreased lung function and increased respiratory infections are commonly described in patients with FD.

While there are no detailed studies addressing lung involvement per se, case reports have mentioned impairments such as atelectasis [Samuelsson et al. 1972], lung consolidation, and calcification of nodules in lungs [Toppet et al. 1978]. The most reported cause of death is respiratory complications from bronchial pneumonia [Zetterström 1958; Sugita et al. 1972;

Barriere and Gillot 1973]. Despite this, aside from superficial descriptions of pulmonary manifestations within these occasional case reports, there has not been a systemic analysis of the pathophysiology of lungs in FD.

Lipoproteinosis causes severe lung pathology in several other LSDs such as

Gaucher disease and Niemann-Pick disease types B, C1, and C2 [Schneider et al. 1977;

Griese et al. 2010; Sideris and Josephson 2016; Sheth et al. 2017; Sikora et al. 2017].

Corresponding mouse models have been instrumental in characterizing disease progression

[Dhami et al. 2001; Ikegami et al. 2003; Buccoliero et al. 2004]. We previously reported the first viable murine model of ACDase deficiency, wherein a human ASAH1 mutation (proline

113

(P) 362 to arginine (R)) was 'knocked-in' to a conserved region of the murine Asah1 locus

(P361R) [Li et al. 1999]. Mice homozygous (Asah1P361R/P361R) for this mutation develop a phenotype that mirrors symptoms reported in FD patients including: systemic accumulation of ceramides, shortened life spans, decreased weight, hepatosplenomegaly, accumulation of foamy macrophages in organs, elevation of inflammatory cytokines, and failure to thrive

[Alayoubi et al. 2013].

Ceramide accumulation is a hallmark of FD. Ceramides and their metabolites are vital constituents of membranes that regulate a diverse array of cellular responses and functions - ranging from inflammation to apoptosis [Hannun and Obeid 2008]. Recently, several important studies have linked sphingolipid signaling with lung health and pathologies

[Uhlig and Gulbins 2008; Yang and Uhlig 2011; Tibboel et al. 2014].

Here, we present a systematic examination of the lung pathology in the

Asah1P361R/P361R mouse with the objective of obtaining better insights into the consequences of ACDase deficiency. Specifically, we present the signature of the various phospholipids and sphingolipids that accumulate in the lung due to ACDase deficiency. We also performed a combination of functional, biochemical, histological, ultrastructural and lipidomic approaches within the lungs and bronchial alveolar lavage fluid (BALF). We report that

ACDase deficiency leads to chronic lung injury characterized by inflammation, increased vascular permeability, and perturbed surfactant and phospholipid homeostasis, all culminating in impaired lung function.

114

4.3 Materials and methods

4.3.1 Animal breeding

To generate homozygous Asah1P361R/P361R mice, we crossed heterozygotes [Alayoubi et al. 2013]. All animal procedures were approved and carried out in strict accordance with the policies of the University Health Network (UHN) Animal Care Committee and the Medical

College of Wisconsin (MCW) Biomedical Resource Center Advisory Committee.

4.3.2 Pulmonary function test

Mice were anesthetized with an intraperitoneal injection of ketamine-xylazine (50 mg/kg and 10 mg/kg) and underwent tracheostomy followed by intubation with an 18-gauge cannula (BD, Biosciences Canada, Mississauga, Canada), which was then connected to the

FlexiVent system (Scireq Inc., Montreal, Canada) for lung function measurements. Prior to measurement, the FlexiVent system was calibrated for mouse weight and gender. Mice were ventilated at 150 breaths/min with a tidal volume of 10 ml/kg. The total lung capacity

(TLC) and Snapshop (single frequency forced oscillation) maneuvers were applied. For the

Snapshot maneuver, measurements for resistance, elastance, and compliance were recorded with the forced oscillation technique that applied a step-wise incremental pressure from 2 cmH2O to 30 cmH2O, and an oscillatory waveform set at 150 breaths/min.

4.3.3 Blood gas, oximetry, and blood counts

For blood gas measurements, mice were anesthetized with an intraperitoneal injection of ketamine-xylazine (50 mg/kg and 10 mg/kg, respectively). A 20-gauge catheter was inserted into the trachea. Each mouse was then connected to the Servo-I mechanical

115 ventilator (Maquet, Germany) for respiratory support with a peak inspiratory pressure (PIP) of 10 cmH2O, positive end expiratory pressure (PEEP) of 2 cmH2O, and a respiratory rate

(RR) of 120 breaths/min. Following a 5-minute acclimatization on the ventilator, the chest cavity was exposed, and the left ventricle of the heart identified. Oxygenated blood was collected using a 25-gauge heparinized needle/syringe and immediately transferred to a

Radio Medical ABL800 Flex (Radio Medical, Copenhagen, Denmark) blood gas analyzer.

Partial pressure of O2 (PaO2), partial pressure of CO2 (PaCO2), and blood pH were measured. Fraction of inspired O2 (FiO2) of 21% was used to calculate the PaO2/ FiO2 ratio.

The MouseOX Plus apparatus (Star Life Sciences Corp, Oakmont, PA) was used to measure pulse, respiration rate, and arterial oxygen saturation (SaO2). Blood samples were taken from the saphenous vein of animals into EDTA-coated tubes. Complete blood counts were performed on a Hemavet 950 FS machine (Drew Scientific Group, Miami, FL).

4.3.4 Lung histopathology and immunohistochemistry

Harvested mouse lungs were inflated with 10% phosphate-buffered formalin and then immersed in the same fixative solution for 24-48h. The lungs were trimmed; the lobes were separated and embedded in paraffin and then sectioned at 4 μm. Sections were stained with hematoxylin and eosin (H&E) and Masson’s trichrome. For immunohistochemistry, the following primary and secondary antibodies and reagents were used: rat anti-mouse neutrophil (clone 7/4; Cedarlane, Burlington, Canada); rat anti-mouse

Mac-2 (Galectin-3 clone M3/38; Cedarlane, Burlington, Canada); rabbit anti-Pro-SP-C

(Sevenhills Bioreagents, Cincinnati, OH); biotinylated rabbit anti-rat (Vector Laboratories,

Burlingame, CA); biotinylated goat anti-rabbit IgG (Vector Laboratories); avidin-biotin/HRP

(Vector Laboratories); DAB kit (Vector Laboratories) and Vectastain ABC Elite Standard kit

(Vector Laboratories). For the primary antibodies, we relied on the testing of the specificity of

116 staining as performed by the respective companies from whom those reagents were purchased. Our own assays included a control wherein secondary antibody was added alone; this measured background signal from staining of nonspecific epitopes. Slides were assessed by a pathologist and no signs of atelectasis or tissue deformation were observed.

All histology slides were scanned on the Aperio AT2 apparatus (Leica Biosystems, Buffalo

Grove, IL) and analyzed using Aperio ImageScope analysis software (Leica Biosystems,

Grove, IL).

4.3.5 Electron Microscopy

Post-euthanasia and lung inflation with 4% paraformaldehyde, lungs were immersed in the same fixative solution for 24-48h. Lungs of each animal were sampled (4-6 individual samples) from the central (hilar) and peripheral (subpleural) areas. Each sample was post- fixed in 3% glutaraldehyde. Samples were further washed and post-fixed with 2% OsO4 in phosphate buffer. After dehydration, samples were then embedded in Durcupan Epon

(Fluka, Hatfield, PA). Ultrathin sections (60 nm) were cut from selected tissue blocks using an Ultracut ultramicrotome (Reichert-Jung, Munich, Germany) and placed on copper grids

(regular Cu 200 Mesh, SPI, West Chester, PA, USA). Ultrathin sections were stained with uranyl acetate and lead citrate and examined with a JEOL 1400+ (JEOL, Tokyo, Japan) transmission electron microscope (TEM) equipped with Olympus Veleta CCD camera and

Radius software.

4.3.6 Alveolar type II cell quantification

Morphometric analyses were performed on the right upper lobes. Lungs were prepared and inflated as previously mentioned; no signs of macroscopic atelectasis were

117 observed. Lung sections were taken from randomly oriented blocks and then stained with an anti-Pro-SP-C antibody for immunohistochemistry detection. Scoring was performed by two independent observers who were blinded from group allocation. The dimension of each field scored was 250µm x 450µm. Quantification was performed on three tissue sections per animal; 10 random non-overlapping images were scored per section. The individual section average was calculated from the cell counts of the 10 images, and then an average of the three sections was used to produce a count for each individual animal.

4.3.7 Western Blots

Lung samples were collected, lysed in RIPA lysis buffer (Thermo Scientific, Waltham,

MA), and homogenized with a Dounce tissue homogenizer. Forty µg of protein was separated by 12% SDS-PAGE, transferred to PVDF membranes by wet transfer, and blocked overnight at 4°C with 5% nonfat dry milk. The following primary and secondary antibodies were used: anti-β-actin (Millipore, Billerica, MA; 1:1000 dilution); anti-Surfactant

Protein (SP) -A (Santa Cruz Biotechnology Inc., Dallas, TX; 1:300 dilution); anti-SP-B

(Millipore; 1:500 dilution); anti-SP-C (Santa Cruz; 1:500); anti-SP-D (Abcam, Cambridge,

MA; 1:1000 dilution); anti-rabbit IgG-HRP (Thermo Scientific); anti-mouse IgG-HRP (BD,

Biosciences Canada, Mississauga, Canada); anti-goat-IgG-HRP (Santa Cruz); and anti- mouse IgG-HRP (GE Healthcare, Waukesha, WI). The secondary antibody was detected with ECL (Thermo Scientific Pierce, Waltham, MA) and autoradiography film.

4.3.8 BALF cytospin and differentials

Mice were euthanized by CO2 inhalation. An 18-gauge catheter (BD, Biosciences

Canada) was inserted and tied into the trachea. The lungs were washed three times by gentle flushing with 0.9 ml of cold phosphate-buffered saline (PBS). BALF was collected into

118

1.5 ml tubes and centrifuged at 300 x g for 10 min. The supernatant was removed and stored at -80°C. The cell pellet was resuspended in 300 to 500 µl of PBS and used for total cell counts, which were performed on trypan blue-stained samples on a Countess II FL (Life

Technologies, Carlsbad, CA). The remainder of the resuspended cells were diluted to a total volume of 150 µl with PBS and a cytospin was performed with the Shandon CytoSpin III

Cytocentrifuge (Thermo Shandon, Waltham, MA) for 5 min at 800 rpm. For cell type quantification, cytospins of 2 x 105 cells per slide were prepared and stained with the Kwik-

Diff kit (Thermo Scientific Pierce, Waltham, MA). Differential cell counts were calculated by averaging the cell counts in 10 non-overlapping zones, where a total of 200 cells were scored in each zone. Cell counts were scored by a 'blinded' observer.

4.3.9 BALF turbidity, cytokine analysis and ELISA

Turbidity was determined by measuring BALF supernatant on the NanoDrop One

(Thermo Scientific, Wilmington, DE) apparatus at 600 nm. Cytokine levels in mouse BALF were analyzed with the Cytokine 20-Plex Mouse Panel (Thermo Scientific Pierce) as per the manufacturer’s instructions. Luminescence was quantified on the Luminex 100 instrument

(Luminex, Austin, TX). Samples with a low bead count (<45 beads) were omitted. IgG, IgM, albumin (Bethyl Laboratories, Montgomery, TX), and surfactant proteins A, B, C, and D were measured in BALF using commercially-available ELISA kits for mouse samples (Cloud-

Clone Corp., Wuhan, China) as per the manufacturer’s instructions.

4.3.10 Vascular permeability and wet-to-dry ratio

To test for vascular permeability, 150 µl of Evans Blue dye (EBD) solution (Thermo

Scientific Pierce) in PBS (50 mg/kg) was injected into the tail vein of mice. Thirty minutes after injection, animals were euthanized by CO2 inhalation and perfused with PBS. Organs

119 were extracted, weighed, and dried overnight at 55oC. Dried organs were incubated in 2 ml of formamide (Thermo Scientific Pierce) at 60°C for 24 hours, followed by centrifugation at

5000 x g for 30 min. The supernatant was collected, and the absorbance was measured at

620 nm and 740 nm (Biotek ELx800, Winooski, VT). The following formula was used to adjust for the contaminating hemoglobin pigment: E620(EBD) = E620 − (1.426 × E740 + 0.030)

[Standiford et al. 1995]. To calculate the wet-to-dry lung weight ratio, age-matched mice were used. Post-euthanization by CO2 inhalation, lungs were rinsed in PBS, weighed immediately, and then transferred to an oven at 60°C. Lung samples were reweighed after four days to obtain dry weights.

4.3.11 Phospholipid and sphingolipid analyses

Lipids from BALF and post-lavage lungs were extracted in CH2Cl2 / MeOH (2% AA) / water (2.5:2.5:2, v/v/v) in the presence of internal standards: ceramide d18:1/15:0 (16 ng);

Sphingomyelin (SM) d18:1/12:0 (16 ng); phosphatidylethanolamine (PE) 12:0/12:0 (180ng); phosphatidylcholine (PC) 13:0/13:0 (16 ng); phosphatidylinositol (PI) 14:1/17:0 (30 ng); and phosphatidylserine (PS) 12:0/12:0 (156.25 ng) (Avanti Polar Lipids, Inc, Alabaster, AL) according to a procedure adapted from Bligh and Dyer [Bligh and Dyer 1959]. The relative quantifications of the phospholipid (PC, PE, PI, PS) and sphingolipid (ceramide and SM) species were obtained in the multiple reaction monitoring mode on a triple quadrupole liquid chromatrography mass spectrometer (Agilent 6460, Santa Clara, CA) equipped with a

Kinetex HILIC column (Phenomenex, 50 x 4,6 mm, 2,6 μm Torrance, CA) as previously described [Sommer et al. 2006; Loiseau et al. 2015].

120

4.3.12 Statistical analyses

Data are expressed as means ± SEM. Statistical significance between Asah1+/+ and

Asah1P361R/P361R groups was determined by Student’s t-test. Significant differences are expressed in the figures as *P < 0.05, **P < 0.01, and ***P < 0.001. No corrections for multiple comparisons were performed for this exploratory study.

121

4.4 Results

4.4.1 Impaired lung mechanics and decreased blood oxygenation

To assess respiratory function in Asah1P361R/P361R mice, lung mechanics were measured with the FlexiVent system. Lung volume measurement revealed a decrease in total lung capacity (Figure 26A) in 8- to 9-week-old Asah1P361R/P361R mice compared to age- matched Asah1+/+ mice, which is in line with the decreased animal size we previously reported [Alayoubi et al. 2013]. Within the single frequency force oscillation (FOT) program, lungs from Asah1P361R/P361R mice displayed increased airway resistance and increased elastance, as well as decreased compliance (Figure 26B-D).

Figure 26. Impaired lung mechanics in the Asah1P361R/P361R mice.

Total inspiratory capacity (A), dynamic elastance (B), dynamic compliance (C), and resistance (D) were measured as described in the Materials and Methods. All comparisons were made between Asah1+/+ and Asah1P361R/P361R mice (n=5 mice for each genotype) at 5 and 8-9 weeks of age. *p<0.05, ***p<0.001.

122

Lung mechanics were also investigated in younger ACDase-deficient and control mice at 5 weeks of age. No difference was seen in lung volume - possibly because the mice are relatively similar in size at this age - but an analogous impairment was observed in airway resistance, elastance, and compliance (Figure 26A-D).

We then assessed whether the impairment in lung mechanics had any effect on blood oxygenation. Arterial blood gas measurements from 8- to 9-week-old Asah1P361R/P361R mice revealed a lower PaO2, an increased PaCO2, and noticeable decreases in the

+/+ pO2/FIO2 ratio and compared to Asah1 mice (Figure 27A,B,D). Arterial blood analysis also revealed a decrease in blood pH suggesting possible respiratory or combined respiratory/metabolic acidosis (Figure 27F). Mouse pulse oximetry measurements confirmed

P361R/P361R that Asah1 mice had significantly lower SaO2 levels (Figure 27C). In addition,

Asah1P361R/P361R mice displayed both increased respiration and increased heart rates (Figure

27E,G). Previously, we reported that Asah1P361R/P361R mice had increased levels of red blood cells (RBCs) from complete blood counts (CBCs) [Alayoubi et al. 2013]. Here, we further characterized the RBCs and show that, in addition to developing an elevated cell count,

RBCs from Asah1P361R/P361Rmice also demonstrated an increase in both hemoglobin and hematocrit values, a lower mean corpuscular volume (MCV), and no change in mean corpuscular hemoglobin concentration (MCHC) (Figure 27H-L). Taken together, these vital readings suggest that Asah1P361R/P361R mice demonstrate signs of chronic hypoxemia/hypercapnia and polycythemia.

123

Figure 27. Impaired blood oxygenation and polycythemia in the Asah1P361R/P361R mice.

Arterial blood gas measurements: arterial blood oxygen partial pressure (PaO2; A), arterial partial pressure of carbon dioxide (PaCO O2; B), the ratio of arterial oxygen partial pressure to fractional inspired oxygen (D), and blood pH measurements (F); n = 5 mice for each genotype. Pulse oximeter measurements: arterial oxygen saturation (SaO2) percentage (C), respiratory rate (E), and heartrate (G); n = 5 each. Erythrocyte indices from hematology analysis: red blood cell (RBC) counts (H), hemoglobin (Hb) levels (I), hematocrit (J), mean corpuscular volume (MCV; K), and mean corpuscular hemoglobin concentration (MCHC; L); n = 10 mice for each genotype. All comparisons were made between 8- and 9-wk-old Asah1+/+ and Asah1P361R/P361R mice. *p < 0.05; **p < 0.01; ***p < 0.001.

124

Hematoxylin and eosin (H&E) staining of 8-9-week-old Asah1P361R/P361R inflated mouse lungs revealed signs of inflammation. There was an increased presence of both alveolar (Figure 28B) and interstitially localized (Figure 28C) macrophages with foamy cytoplasm. Interstitial macrophages were abundant in perivascular and peribronchial tissue, as well as in the terminal alveolar septa (Figure 28B,C). Importantly, these cells were found both singly and in aggregated clusters. Lungs from 8-9-week-old Asah1P361R/P361R mice also contained substantial amounts of eosinophilic cellular debris scattered throughout the alveolar space. To further characterize the extent of inflammation, immunohistochemistry

(IHC) with antibodies specific for Mac-2 (galectin-3) and a mouse neutrophil marker was performed. The accumulations of foamy macrophages was greatly highlighted (Figure

28E,G and H). Macrophages were variable in size and on occasion we could detect large foamy multinucleated cells. Similarly, IHC staining for a neutrophil marker, showed that recruitment of neutrophils is widely abundant through the lung parenchyma of

Asah1P361R/P361R mice (Figure 28J).

125

Figure 28. Immune infiltration in Asah1P361R/P361R mice.

Light micrographs of 8-9-week-old Asah1+/+ and Asah1P361R/P361R mice lung sections stained with hematoxylin and eosin (H&E) (A-C), large number of large foamy storage laden macrophages (alveolar and interstitial) are found in the Asah1P361R/P361R mouse. Immunohistochemistry (IHC) for MAC-2 (D-H) and a neutrophil marker (I, J) show significant infiltration by the foamy macrophages and increased presence of neutrophils in the mutants. The interstitial foamy macrophages can be found individually or clustered in the alveolar septa (C, E, G, H, and J) and peribronchial/perivascular tissue (E). Symbols: black arrowheads- foamy interstitial macrophages; empty arrowheads- foamy alveolar macrophages; black arrows- cells with segmented nuclei/neutrophil granulocytes; empty arrows-abnormal alveolar debris; * – bronchial lumen; ** - lumen of the hilar vessel. Scale bars are 100m (A, D, E); 25m (B, C, G, and H); and 50m (F, I, and J).

126

Ultrastructure analyses were performed on both 8-9-weeks-old Asah1+/+ and

Asah1P361R/P361R mice lungs (Figure 30). Extensive storage products were found in interstitial as well as within resident alveolar macrophages in 8-9-weeks-old Asah1P361R/P361R mice lungs. Ultrastructural pathology of these cells was hallmarked by extensive cytoplasmic vacuolization (Figure 30B). The limiting membranes of many of these vacuoles were fusing.

Importantly, curved-linear tubular profiles (Farber bodies, Figure 30C) represented the almost exclusive contents of these storage vacuoles. On the contrary, storage vacuoles observed in alveolar macrophages were structurally more pleiomorphic and many of them adopted the lamellar “zebra-body” appearance (Figure 30E). On occasion, the alveolar storage-laden macrophages displayed degenerative cytoplasmic changes and loss of cytoplasmic membrane integrity (Figure 30F). Extracellular alveolar space contained excessive amounts of biological debris that was formed by membranous lamellar whorls and condensed Farber body-like/fingerprint-like profiles (Figure 30H-I). In addition to macrophages, subcellular storage pathology could be found in pulmonary endothelial cells

(Figure 31B,C) and ciliary cells of the respiratory epithelium (Figure 31 E,F). Lastly, we observed subtle storage abnormalities in the microvillous epithelial and Clara cells, (data not shown).

127

Figure 29. Ultrastructural pathology in macrophages and contents of alveolar space in Asah1P361R/P361R mice.

Ultrastructural pathology in macrophages and contents of alveolar space in Asah1P361R/P361R mice. Electron micrographs of 8- to 9-wk-old Asah1+/+ (A,D, and G) and Asah1P361R/P361R mouse lungs. B: Features include macrophages in the alveolar septa (nucleus marked by a white asterisk). The macrophages that contain excess cytoplasmic vacuolar storage bodies are labeled with open arrowhead. Black arrows demark alveolar type II cells. C: Storage bodies of the interstitial macrophages contain curved linear tubular profiles (Farber bodies in C; detail corresponding to the box in B). E: An alveolar macrophage laden with pleomorphic storage vacuoles [white asterisk, nucleus; black arrowheads, Zebra bodies; empty arrow heads, vacuoles containing Farber bodies; compare with Asah1+/+ alveolar macrophages (white asterisks) in A and D]. White arrows highlight erythrocytes in the alveolar septal capillary. F: Discontinuity and rupture of the cytoplasmic membrane of the alveolar macrophage (is marked by black arrows in F; detail corresponding to the box in E). H: the abnormal extracellular alveolar content that is formed partly by concentric membranous whorls (black arrowheads) and Farber/finger print-like profiles (open arrowheads). The latter pattern is shown in detail in I (and corresponds to the box in H). Compare with limited amounts of concentric lamellar structures (black arrowhead in D) or completely clear alveolar spaces in Asah1+/+ mice (G). Scale bars, 5 (A, B, D, and E) and 3 m (G and H).

128

Figure 30. Ultrastructural pathology of pulmonary endothelial cells and respiratory epithelial cells Electron micrographs of 8-9-week-old Asah1P361R/P361R mice lungs featuring endothelial cells in pulmonary capillaries containing storage vacuoles (empty arrowheads) with curved-linear tubular profiles (Farber bodies) (B, C - detail corresponding to the rectangle in B). Lamellar “Zebra” storage bodies can also be identified within ciliated cells of the respiratory epithelium (white arrowheads) (E, F - detail corresponding to the rectangle in E). Compare the findings in the Asah1P361R/P361R mice to Asah1+/+ mice (A and D) Scale bars are 1.2m ;(A and B) and 2m (D and E).

4.4.2 Absence of lung fibrosis but presence of lung edema

Slides of H&E stained lung tissue from Asah1P361R/P361R mice did not show significant perturbation (e.g., large areas of atelectasis, excessive emphysema, and/or widespread alveolar edema) in tissue architecture. (Figure 28B,C). Trichrome staining was performed on lung sections to rule our fibrosis. While Asah1P361R/P361R mice lungs did not appear to show fibrosis and collagen accumulation, we did however notice that the cellular debris within the

129 parenchyma stained positive for collagen (Figure 31B). This non-specific staining may be caused by the released products of the large macrophages. Since the lung contains increased cell infiltration and cellular debris, lung tissue weights were performed.

Measurement of both the normalized wet and dry lung weights from Asah1P361R/P361R mice revealed an increased tissue weight. This may indicate pulmonary remodeling or increased cellularity (Figure 31C,D). Interestingly the wet-to-dry ratio as normalized to body weight was higher in lungs from Asah1P361R/P361R mice, indicating the likely presence of lung edema

(Figure

31E).

Figure 31. Absence of fibrosis but presence of interstitial edema

Light micrographs of 8-9-week-old Asah1+/+ and Asah1P361R/P361R mice lung sections were stained with Trichrome (A and B – alveolar storage macrophages (open arrowheads), interstitial storage macrophages (black arrowheads), alveolar debris (black arrows)), scale bars represent 100m. Lung wet and dry weights were normalized to body weight. Wet lung/body weight percentage (C), dry lung/body weight percentage (D), and wet- to-dry lung ratio were normalized to body weight (E); n=10 mice at 8-9 weeks of age for each genotype. ***p<0.001.

130

4.4.3 Elevation of leukocytes in lung

To quantitate and characterize the immune cell infiltration in the Asah1P361R/P361R mouse lungs, cells collected from BALF were stained with the Kwik-Diff reagent for cytology.

Analysis of the slides revealed significantly more inflammatory cells in the lungs of

Asah1P361R/P361R mice (Figure 32A,B). Proteinaceous material was also widely present throughout the background of the slide. Evaluation of blood cell types revealed that

Asah1P361R/P361R mice had significantly higher quantities of all infiltrating immune cells in

BALF samples (Figure 32C-F). In addition, the percentage of multinucleated macrophages within the total macrophage pool in BALF from Asah1P361R/P361R mice was several-fold higher than that in their Asah1+/+ counterparts (Figure 32G).

Figure 32. Immune cell infiltration in bronchoalveolar lavage fluid of Asah1P361R/P361R mice

Light micrograph of BALF cells, scale bar represents 100µm at equal 1:2 dilutions (A, B). Scale bars are 100 m. Cell counts and differentials from Kwik-Diff stained cytospin slides in BALF samples show BALF cell count (C), macrophages (D), neutrophils (E), lymphocytes (F) and percentage of multinucleated macrophages of total macrophages count (G). All comparisons were made between 8-9-week-old Asah1+/+ and Asah1P361R/P361R mice. n=5 mice for each genotype. ***p<0.001.

131

4.4.4 Vascular permeability in multiple organs.

The consistency of the BALF collected from the 8-9-week-old Asah1P361R/P361R mice lungs was milky and opaque (Figure 33A). Analysis of that BALF revealed markedly increased turbidity (Figure 33B), accumulation of albumin, and the presence of IgG and IgM antibodies (Figure 33C-E), which are indicative of protein leakage. Compared to age- matched Asah1+/+ mice, the lungs, heart, thymus, liver, and spleen of Asah1P361R/P361R mice demonstrated increased vascular permeability as measured by parenchymal dye accumulation following Evans Blue injections (Figure 33F,G).

Figure 33. Vascular permeability leads to protein leakage in Asah1P361R/P361R mice.

BALF was characterized showing opaque samples from Asah1P361R/P361R vs Asah1+/+ (A) and turbidity as measured by absorbance at 600 nm (B); n=8 from 8-9-week-old mice for each genotype. Levels of albumin (C), IgG antibodies (D), and IgM antibodies (E) collected from BALF were determined by ELISA from 8-9-week-old mice; n=5 mice for each genotype. Evans Blue (EB) dye injected via tail vein revealed dye accumulation (F-G) in the lungs and other organs in 8-9-week-old Asah1+/+ and Asah1P361R/P361R mice; n=5 mice for each genotype. *p<0.05 **p<0.01, ***p<0.001.

132

4.4.5 Increased cytokine production in BALF

Systemic inflammation is a characteristic of FD. A multiplex cytokine panel assay was performed on BALF from 8-9-week-old mice. An increase in MCP-1 concentration was seen in BALF from Asah1P361R/P361R mouse lungs (Figure 34A), similar to the MCP-1 elevation we previously reported in the plasma of Asah1P361R/P361R mice and FD patients

[Dworski et al. 2017]. IL-4, VEGF, and TNFα concentrations were also increased (Figure

34B-D). No significant change was found in MIP-1α or GM-CSF levels (Figure 34E-F). No changes were observed for bFGF, IL-1β, IL-10, IL-13, IL-6, IL-17, IL-5, IGNγ, IL-2, IP-10,

MIG, or KC (Figure 35). No data was obtained for the IL-20 and IL-1α samples due to low

bead counts (data not shown).

Figure 34. Elevated cytokines in BALF of Asah1P361R/P361R mice.

133

Cytokines were measured in BALF samples of 8-9-week-old animals using a multiplex assay. Levels of CCL2 (MCP-1) (A), interleukin 4 (IL-4) (B), vascular endothelial growth factor (VEGF) (C), tumor necrosis factor alpha (TNFα) (D), CCL3 (MIP-1α) (E), and granulocyte macrophage colony-stimulating factor (GM-CSF) (F) were measured; n=6 mice for each genotype. *p<0.05 **p<0.01, **p<0.001.

Figure 35. Unchanged cytokines in BALF of Asah1P361R/P361R mice.

134

Cytokines were measured in BALF samples of 8-9-week-old animals using a multiplex assay. Levels of basic fibroblast growth factor (bFGF) (A), IL-1b (B), IL-10 (C), IL-13 (D), IL-6 (E), IL-17 (F), IL-5 (G), Interferon gamma (IFN-y) (H), IL-2 (I), Interferon gamma-induced protein 10 (IP-10) (J), Monocyte induced by gamma interferon (MIG (K), Keratinocyte chemoattractant (KC) (L) were measured; n=6 mice for each genotype.

4.4.6 Abnormal lamellar body formation and increased expression of

surfactant proteins

In contrast to cell types listed earlier, ultrastructural analyses of alveolar type II cells in Asah1P361R/P361R mice did not show any apparent storage bodies within the cytoplasm

(Figure 36B). Unexpectedly, we observed that the surfactant-containing lamellar bodies that were readily detected in Asah1+/+ (Figure 36A) mice showed partial condensed contents in

Asah1P361R/P361R mice (Figure 36B). To access expression of surfactants, we performed IHC staining for pro-surfactant C (pro-SP-C) on lung sections from 8-9-week-old Asah1P361R/P361R mice compared to Asah1+/+ mice (Figure 36D,E). This staining, which marks alveolar type II cells, also showed positive staining in patches of alveolar cellular debris and within the foamy alveolar macrophages (Figure 36E). Quantitation of alveolar type II cells from 8-9- week-old pro-SP-C stained slides of Asah1+/+ and Asah1P361R/P361R mice showed no differences (Figure 36C). Western blot analyses of lung tissue lysates confirmed the increase in SP-A levels; of note a second unknown band of lower molecular weight was also demarcated (Figure 36F). SP-C and SP-D levels were increased; while SP-B was present at normal levels when compared to samples from Asah1+/+ lungs (Figure 36F). Since surfactants are also secreted into the lung airspaces, we further measured surfactant proteins in BALF via ELISA. In BALF, the content of all surfactant proteins was higher in samples from Asah1P361R/P361R mice than from Asah1+/+ controls (Figure 36G-J). This observed increase in surfactant protein maybe a compensatory mechanism caused by

135 increased production and/or decreased degradation in the Asah1P361R/P361R mice as a result of chronic lung injury.

Figure 36. Lamellar body formation and accumulation of surfactant proteins in lungs of Asah1P361R/P361R mice.

Electron micrograph of an alveolar type II cell in Asah1+/+ lungs containing normal formed lamellar bodies (black arrowheads) (A). Electron micrograph of an alveolar type II cell in Asah1P361R/P361R lung suggesting an altered phenotype of the surfactant containing granules (B). Contents are largely condensed into electron-dense clumps (black arrowheads). Quantitation of alveolar type II cells from IHC staining for pro-surfactant protein-C in Asah1+/+ and Asah1P361R/P361R mice (pro-SP-C); n=6 mice for each genotype (C). Micrographs of IHC for pro-SP- C in Asah1+/+ (D, alveolar type II cells (black arrowheads)) and Asah1P361R/P361R mice (E, alveolar type II cells

136

(black arrowheads), alveolar storage macrophages (open arrowheads)) at 8-9 weeks of age. Western blots from lung tissue lysate show levels of SP-A, SP-B, SP-C, SP-D, and beta-Actin in samples from Asah1+/+ and Asah1P361R/P361R mice (F). ELISA was performed on BALF samples from 8-9-week-old mice for SP-A (G), SP-B (H), SP-C (I) and SP-D (J) from Asah1+/+ and Asah1P361R/P361R samples; n=6-8 mice for each genotype. *p<0.05, ***p<0.001. Scale bars are 2m (A), 1.2m (B) and 50m (D, E).

4.4.7 Accumulation and disruption of phospholipid composition

Total phospholipids were analyzed by mass spectrometry in BALF samples obtained from 8- to 9-week-old mice. There were shifts in the phospholipid composition of BALF from

Asah1P361R/P361R mice compared to Asah1+/+ mice. SM and PE were increased in abundance in BALF obtained from Asah1P361R/P361R mice, while PC levels were decreased (Figure 36A).

Although PC decreased proportionally as a percent of total phospholipids, the total amount of measurable PC was still significantly increased (Figure 37B, C). This includes an increase in Sat-PC, the major phospholipid components in pulmonary surfactant (Figure 37C).

Significant increases were also detected in all lipid species of PI, PE, and PS in

Asah1P361R/P361R BALF samples (Figure 37E-G). When we analyzed the post-lavage lungs, only the PC was increased; there was no change in PE, PS, or PI (Figure 37H-K). This observation suggests that much of the accumulation in BALF is from secreted lipids that are released into the air space.

137

Figure 37. Impaired phospholipid homeostasis in BALF of Asah1P361R/P361R mice.

BALF phospholipid composition (A), total fold-change of various phospholipids in BALF (B), total saturated-PC in BALF (C), phosphatidylcholine (PC) species in BALF (D), phosphatidylinositol (PI) species in in BALF species (E), phosphatidylethanolamine (PE) species in BALF (F), and phosphatidylserine (PS) species in BALF (G). Post-lavage lung tissue phospholipid profiles as measured by mass spectrometry. Molecular species of PC (H), PI (I), PE (J), and PS (K) in lung lysates. Lipid concentrations were calculated as ng/mg protein of BALF or lung tissue. All samples were obtained from 8-9-week-old mice; n=6 mice for each genotype. *p<0.05 **p<0.01, ***p<0.001.

138

4.4.8 Ceramide and sphingomyelin accumulation in BALF and lungs

We have previously reported that ceramides are increased in various organs in

Asah1P361R/P361R mice [Alayoubi et al. 2013]. Here we quantified ceramides in both BALF and the post-lavage lung tissue. In BALF, all ceramide species were significantly increased in

Asah1P361R/P361R mice compared to Asah1+/+. Total ceramides were increased ~300-fold and the most prominently elevated species being C24:1 followed by C16:0 (Figure 38A). In addition, the relative abundance of some individual ceramide species in BALF was altered, in that: the C24:1 increased whereas C22:0 and C24:0 decreased (Figure 38B). The finding of increased ceramide species was the same in lung tissue proper. There, however, total ceramides were only increased 12-fold and no changes were seen in the relative abundances. The fact that we can detect a significantly higher-fold difference of ceramides within the BALF may suggest that the Asah1P361R/P361R lungs are leaking out ceramides due to increased vascular permeability (Figure 38C, D).

When we analyzed the SM content, all species accumulated in the Asah1P361R/P361R mouse BALF. We detected a 25-fold increase of total SM species in Asah1P361R/P361R mice

BALF when compared to Asah1+/+ (Figure 38E). This contrasted with the results from the post-lavage tissue where only one species, SM 16:1, was significantly increased (Figure

38G). Changes in the relative SM pattern were similar to those seen in the ceramide pattern, wherein BALF from Asah1P361R/P361R mice had significant changes in composition abundance but post-lavage tissue samples did not (Figure 38F, H).

139

Figure 38. Significant accumulation of both Ceramide and SM in BALF and lungs of Asah1P361R/P361R mice.

Ceramide species in BALF (A), relative abundance of ceramide species in BALF (B), ceramide species in post- lavage lung tissue, (C) relative abundance of ceramide species in post-lavage lung tissue (D), SM species in BALF (E), relative abundance of SM species in BALF (F), SM species in post-lavage lung tissue (G) and (H) relative abundance of SM species in post-lavage lung tissue as measured by mass spectrometry (H). Lipid concentrations were calculated as ng/mg protein of BALF or lung tissue. All samples obtained from 8-9-week-old mice; n=6 mice for each genotype. *p<0.05 **p<0.01, ***p<0.001.

140

4.5 Discussion and conclusion

Deficiencies in ACDase result in systemic accumulation of ceramide. Patient symptoms may manifest along a spectrum with the cardinal phenotypes including presence of subcutaneous nodules, joint pain, and a hoarse voice [Levade et al. 2014]. Respiratory complications have been documented in several case reports of Farber disease: nodular opacities in x-rays have been seen and the terminal respiratory distress has resulted in early morbidity [Pellissier et al. 1986; Fiumara et al. 1993]. Here, we provide a comprehensive analysis of the lung pathology present in the ACDase-deficient mouse. We demonstrate that a deficiency in ACDase leads to chronic lung injury and impaired pulmonary function.

Importantly, we also provide insights that could underlie such modifications. These alterations are likely a result of a several factors. A heightened inflammatory state which led to significant pulmonary inflammation resulted in an increase in vascular permeability, a buildup of storage material, and impaired surfactant homeostasis. Lung mechanic tests on

Asah1P361R/P361R mice showed significantly decreased values in lung compliance and an increase in airway resistance.

Of the few Farber case reports that mention lung involvement, most only describe the presence of infiltrates and the development of infections at the end of life [Ehlert et al.

2007]. Here, we provide the first evidence that ACDase deficiency leads to a wider scope of lung defects thus expanding the role of respiratory involvement in FD.

Light and electron microscopy analyses showed an inflammatory phenotype similar to other tissues present in Asah1P361R/P361R mice [Alayoubi et al. 2013; Sikora et al. 2017].

The pulmonary pathology was dominated by the presence of the abnormal storage-laden macrophages. These cells populated individually and in clustered aggregates in the interstitial peribronchial and perivascular tissue as well as in the alveolar septa. This observation mirrors those found in early cases of FD, where patients were reported to have

141 pervasive infiltrating foam cells in the lung and lipid laden macrophages from BALF samples.

[Farber et al. 1957; Antonarakis et al. 1984; Yeager et al. 2000]. Alveolar macrophages, which are important surfactant scavengers, were also severely affected by the storage pathology. The histopathological, immunohistochemical, and ultrastructural findings we observed suggest that some of these abnormal alveolar macrophages degenerate and their released contents contribute to the substrate burden for other surrounding cells involved in phagocytosis. This released content may in part contribute to the proteinaceous and highly turbid material that was present throughout the air spaces and in BALF samples. Increased immune cell recruitment and lipoprotein accumulation is a recurring observation in animal models of sphingolipid enzyme deficiencies and LSDs. For example, the Cers2-deficient mouse, the S1P lyase knock-out mouse, and the acid sphingomyelinase-deficient mouse all have increased alveolar macrophages and/or the presence of protein-like material within the alveolar space or in BALF samples [Dhami et al. 2001; Vogel et al. 2009; Petrache et al.

2013]. This similarity in phenotype suggests that impairment in sphingolipid-metabolizing enzymes may lead to comparable pathology because their metabolism is tightly interconnected. It also demonstrates that homeostasis of these bioactive lipids is vital for proper respiratory function.

The pro-inflammatory state in Asah1P361R/P361R mouse lungs results in significant immune cell infiltration; especially macrophages and neutrophils as shown by histology and

BALF analysis. Cellular infiltration, formation of large foamy macrophages, and the increased formation and release of surfactants together may be contributing factors to the decreased lung compliance observed in both 5 and 8-to 9-week-old Asah1P361R/P361R mice.

Lung mechanics revealed an increase in airway resistance, which was detectable by 5- weeks of age. While we do not fully understand the cause of this, we speculate one potential factor may be the overall inflammatory response characteristic of FD as patients have been reported to have formation of granulomas and edema in both the larynx and vocal chords

142

[Farber et al. 1957]. These phenotypes, detected as early as 5-weeks of age in

Asah1P361R/P361R mice, demonstrate that chronic lung impairment begins early in this model.

The inflammation and protein accumulation in the lungs of Asah1P361R/P361R mice may be explained in part by increases in the levels of cytokines such as MCP-1. We previously demonstrated that this cytokine is elevated in the plasma of Farber patients and treatment by hematopoietic stem cell transplantation normalizes MCP-1 levels [Dworski et al. 2017]. In addition to its role in the recruitment of monocytes, there is evidence that during lung injury

MCP-1 also mediates the recruitment of neutrophils into the lung [Balamayooran et al. 2011].

TNFα and VEGF, which are both elevated in BALF from Asah1P361R/P361R mice, are also implicated in lung inflammation, edema, and promotion of vascular permeability [Mura et al.

2010; Yang et al. 2010; Patel et al. 2013]. While the cytokines increased in BALF from

Asah1P361R/P361R mice may cause lung complications, this may actually be a secondary consequence of the accumulation of ceramide.

Accumulation of albumin, IgG, and IgM in Asah1P361R/P361R BALF demonstrates impairment of lung vasculature barriers. Sphingolipids have been shown to alter pulmonary vascular permeability [Kuebler et al. 2010]. For example, S1P enhances barrier function by its interaction with the G-protein coupled S1P1 receptor and subsequent downstream activation of a small Rho GTPase [Wang and Dudek 2009]. Ceramide, however, has been shown to function oppositely, by increasing vascular permeability via the regulation of endothelial Ca2+ signaling and the formation of nitric oxide (NO) [Yang et al. 2010]. The relationship between ceramide and edema formation has been demonstrated in an in vivo experiment wherein ceramide was directly instilled in rat lungs [Ryan et al. 2003]. Another study demonstrated that lung edema caused by platelet-activating factor (PAF) is mediated by ASMase and ceramide [Göggel et al. 2004]. Compared to control mice, lung edema was attenuated in ASMase-deficient mice when they were treated with PAF [Göggel et al. 2004].

In the same study, direct perfusion of C2:0 ceramide into isolated Asah1+/+ rat lungs resulted

143 in interstitial edema formation [Göggel et al. 2004]. Further, when control mice were treated with PAF and ceramide-specific antibodies, the resulting lung edema was less severe than when the animals were only treated with PAF [Göggel et al. 2004]. These studies allude to sphingomyelin-derived ceramide as the key source of ceramides that affect barrier function.

Our results coincide with these observations, as we also report edema from our wet-to-dry ratio measurements along with an increase in all ceramide species from our MS analyses.

Taken together with the EB dye accumulation we measured in the lungs, vascular permeability and edema are contributing factors to the impaired lung function we reported.

Another distinguishing phenotype in our Asah1P361R/P361R mice is the drastic accumulation of lipoproteinaceous material within the lung. The milky consistency of mutant

BALF and increased recruitment of inflammatory cells suggests that Asah1P361R/P361R mice may be more prone to lung infections. However, when we collected BALF samples from

Asah1P361R/P361R and Asah1+/+ mice for bacteria cultures, we discovered there was no significant difference between the quantity and species of colonies grown on BALF-cultured plates (data not shown). This finding could be more related to the highly-controlled environment within which the animals were maintained, however, rather than to the biology of the system.

The turbid BALF observed in Asah1P361R/P361R mice is likely a result of massive protein extravasation into the air space resulting in a pulmonary alveolar proteinosis (PAP)-like phenotype. PAP is characterized by compromised macrophage function due to a deficiency in GM-CSF that leads to impaired surfactant catabolism [Seymour and Presneill 2002]. Here, both Asah1P361R/P361R and Asah1+/+ mice had similar levels of GM-CSF in their BALF, and macrophages from Asah1P361R/P361R mouse lungs stained positive for pro-surfactant protein-C suggesting active surfactant catabolism. Interestingly, BALF from Asah1P361R/P361R mice were seen to have elevations in IL-4 and MCP-1, which has been noted in rare cases of

144

PAP [Iyonaga et al. 1999; Ikegami et al. 2000]. Together this suggests that Asah1P361R/P361R mice have an atypical presentation of PAP.

When we characterized the BALF and tissue lysates in Asah1P361R/P361R mice, we detected increases in all surfactant proteins and saturated-PC. In addition, though IHC staining for pro-SP-C was significantly increased in the Asah1P361R/P361R mice, there was no increase in the number of marked alveolar type II cells. Rather, the increased pro-SP-C signal was due to staining within the foamy macrophages and extracellular debris. Together this suggests that Asah1P361R/P361R mice have increased surfactant production and release.

Unexpectedly, from the TEM observations, the alveolar type II cells in the Asah1P361R/P361R mice exhibited condensed (non-lamellar) contents in the secretory bodies. While further work is required, this could suggest an abnormality in the formation of surfactant and/or differences in the lipid composition of the lamellar bodies. The secretory lamellar bodies within alveolar type II cells are a subclass of lysosome-related organelles [Weaver et al.

2002]. Due to related functions between lysosomes and secretory bodies, it is possible that

ACDase deficiency may contribute to the abnormal secretory function of alveolar type II cells resulting in the irregular lamellar body phenotype. Furthermore, there have been reports demonstrating that surfactant surface tension can be perturbed in the presence of hemoglobin, albumin, and other plasma proteins [Seeger et al. 1985; Holm et al. 1985; Holm and Notter 1987]. In addition, ceramides themselves may be able to impede surfactant function [Ryan et al. 2003]. This relationship in our Asah1P361R/P361R mice has not been determined and will require future study.

From our lipidomic analyses of sphingolipids and phospholipids we found that BALF sample contained significantly increased levels of the measured compounds compared to the tissue proper. This finding, added with the observation of degeneration of some alveolar macrophages, suggests that there may be inefficient macrophage scavenging activity further aggravating the imbalance in surfactant homeostasis in Asah1P361R/P361R mice.

145

The phospholipid composition analyses of tissues from the Asah1P361R/P361R mice demonstrated a markedly altered relative abundance of each phospholipid type.

Asah1P361R/P361R mice were observed to have a reduction in the percentage of PC coinciding with an increase in SM and PE phospholipids. Interestingly, other LSDs such as Niemann-

Pick types A, B, C1, and C2, as well as Sandhoff disease display a combination of either dysfunction in surfactant homeostasis and/or changes to phospholipid abundance indicating the importance of tight sphingolipid regulation in this context [Ikegami et al. 2003; Buccoliero et al. 2004; Bjurulf et al. 2008].

The change in phospholipid profiles and impaired surfactant homeostasis appears to be a generalized characteristic of lung injury. One such example is in patients who develop acute respiratory distress syndrome (ARDS). BALF from patients with this critical illness have been shown to have a decrease in PC and an associated increase in PI and SM

[Gunther et al. 1996]. Furthermore, the BALF from ARDS patients has been shown to acquire an elevated SM-to-PC ratio [Hallman et al. 1982], and glycolipids, such as lactosyl- ceramides, accumulate in the airspaces [Rauvala and Hallman 1984].

Ceramide, a central signaling sphingolipid, is essential for maintenance of lung health. Various respiratory diseases have been shown to manifest when the sphingolipid balance is perturbed [Ghidoni et al. 2015; Petrache and Berdyshev 2016]. Increases in ceramide due to various stimuli can lead to downstream effect such as lung cell death, in the case of exposure to cigarette smoke and COPD. Also, the use of sphingosine analogs to reduce ceramide levels may provide an antimicrobial effect in models of cystic fibrosis

[Pewzner-Jung et al. 2014]. These reports demonstrate that ceramide production in the lung can be triggered acutely by a variety of stimuli. In the case of chronic ceramide accumulation such as in FD, alternative pathways may be activated in response to the sphingolipid buildup.

It is clear that more work will be required to fully understand the role of ceramide signaling in

FD and lung pathobiology.

146

In the Asah1P361R/P361R mouse, ceramide accumulation is chronic and likely progressive. The mouse has a drastic reduction in lifespan and recapitulates many features of human FD [Alayoubi et al. 2013]. The lung involvement brought on by significant inflammation and accumulation of lipoprotein and surfactant within the respiratory system suggests that impaired oxygenation and chronic lung injury is a potential contributor to mortality. Asah1P361R/P361R mice typically die between 8-10 weeks of age, so our analyses represent phenotypes seen near end-of-life [Alayoubi et al. 2013]. The precise mechanisms of how the various species of ceramide act at the molecular level within the lungs to cause the spectrum of phenotypes observed remain to be uncovered. Within the human context, respiratory involvement is most prevalent in the severe forms of FD [Farber et al. 1957;

Abul-Haj et al. 1962; Samuelsson et al. 1972]. The main phenotype documented in patients are the infiltration of foamy cells and proteinaceous material in the lungs [Farber et al. 1957;

Moser 1989; Tanaka et al. 1979]. Some cases have also noted presence of respiratory stridor, autopsy descriptions of poorly expanded lungs, and atelectasis from radiology

[Farber et al. 1957; Bierman et al. 1966; Samuelsson and Zetterström 1971; Pellissier et al.

1986]. While we don’t report the presence of lung atelectasis in the Asah1P361R/P361R mice, the increased airway resistance and low compliance may be consequences of the aforementioned human observations. From our study, several parallels may be drawn from the mouse and human. However, due to the wide clinical spectrum of FD these results should be taken with caution as there may also be important species differences. For example, one case revealed that out of 6 patients who had attenuated forms of FD, only 3 were reported to have lung involvement [Fiumara et al. 1993]. Nonetheless, our results do offer clinical insight since respiratory failure and pneumonia are contributors to mortality in classical FD [Farber et al. 1957; Fiumara et al. 1993; Kim et al. 1998]. This susceptibility is likely caused by a few reasons; immune dysfunction, poor clinical state, and possibility impaired bronchial clearance. These reasons along with the expanded lung phenotype seen

147 in the ACDase mouse warrants the need to perform pulmonary tests in FD patients as part of a natural history study. Furthermore, from a therapeutic perspective, our study indicates there might be merit in evaluating whether treatment with partial lung lavage, commonly performed on patients with PAP, may help manage the pulmonary aspects of this disorder

[Seymour and Presneill 2002]. In addition, with the advent of recombinant enzyme therapy, use of aerosolized ACDase in conjunction with other treatments may help to manage respiratory symptoms and assist in treatment of patients with FD [Schuchman 2016; He et al.

2017].

148

Chapter 5

Liver Injury and Altered Gene Expression Changes in Acid

Ceramidase Deficiency

5.1 Abstract

Farber Disease (FD) is a rare Lysosomal Storage Disorder (LSD) characterized by systemic ceramide accumulation caused by a deficiency in acid ceramidase (ACDase). In its classic form, FD manifests with painful lipogranulomatous nodules in extremities and joints, respiratory complications, and neurological involvement. Hepatosplenomegaly is commonly reported, and severe cases have documented significant liver failure as a cause of early death. Mice homozygous for an orthologous patient mutation in the ACDase gene

(Asah1P361R/P361R) recapitulate the classical form of human FD. In this study, we demonstrate impaired liver function and elevation of various liver injury markers in Asah1P361R/P361R mice as early as 5 weeks of age. Histopathology analyses demonstrated significant recruitment and formation of foamy macrophages, invasion of neutrophils, progressive tissue fibrosis, increased cell proliferation and death, and significant storage pathology within various liver cell types. Lipidomic analyses revealed significant reduction of various lipids in both serum and liver tissue. A significant accumulation of ceramide and other sphingolipids both in liver and hepatocytes was noted. Sphingolipid acyl chains were altered, with an increase in long acyl chain sphingolipids coinciding with a decrease in ultra-long acyl chains. Lastly, transcriptome analyses on hepatocytes revealed a significant alteration in gene expression.

149

Pathway analyses revealed activation of genes related to inflammation, and a deactivation of lipid metabolism pathways. Lastly, misregulated genes within the sphingolipid pathway are documented.

150

5.2 Introduction

Farber disease (FD) (OMIM #228000) also known as Farber’s lipogranulomatosis, or acid ceramidase (ACDase) deficiency is an ultra-rare Lysosomal Storage Disorder (LSD) caused by an inherited mutation in the ASAH1 gene [Levade et al. 2014]. ACDase is a hydrolase that breaks down the bioactive sphingolipid ceramide into sphingosine (Sph) and a free fatty acid. To date, there have been 152 cases of FD reported in the literature [Yu et al. 2018]. While our understanding of the natural course of FD is incomplete, case reports suggest that FD can manifest along a wide clinical spectrum [Zielonka et al. 2017]. Patient with attenuated FD may live to early adulthood, while patients with the classic or severe variants will die during infancy [Levade et al. 2014]. The most common clinical manifestations include: formation of subcutaneous nodules, joint contractures, and voice hoarseness [Farber et al. 1957; Ehlert et al. 2007; Levade et al. 2014]. Patients who develop more severe FD may also acquire nervous system involvement, respiratory disease, and hepatosplenomegaly [Kim et al. 2016; Zetterström 1958; Salo et al. 2003; Ehlert et al. 2007].

As of date there is no known cure for FD. Hematopoietic stem cell transplantation

(HSCT) may be an effective option for patients with little to no neurological involvement

[Ehlert et al. 2006]. Beyond that, most patients receive symptomatic treatment [Levade et al.

2014].

Several case reports on FD patients have documented hepatomegaly during examination, and presence of foamy histiocytes and ceramide accumulation in post-mortem liver tissue [Farber et al. 1957; Abul-Haj et al. 1962; Samuelsson and Zetterström 1971;

Tanaka et al. 1979; Antonarakis et al. 1983]. Additionally, hepatic dysfunction is present in the severe visceral variant of FD (Type IV) [Levade et al. 2014; Willis et al. 2008; Nowaczyk et al. 1996]. In one such case, an infant was misdiagnosed with neonatal hepatitis and

151 treated with a liver transplant before being properly diagnosed with FD [Salo et al. 2003].

However, aside from superficial descriptions in infrequent case reports, there has not been a detailed study performed on the liver pathophysiology in ACDase deficiency.

Proper sphingolipid homeostasis is vital for health, and insult to this tightly regulated balance results in disease [Coant et al. 2017; Hannun and Obeid 2017]. This is apparent in

FD and other related LSDs where inherited lysosomal enzyme deficiencies leads to sphingolipid accumulation and pathogenesis. LSDs such as Niemann-Pick types A,B,

Gaucher, and GM1 gangliosidosis all exhibit forms of liver disease and corresponding mouse models have been important in highlighting disease progression and the extent of liver involvement [Matsuda et al. 1997; García-Ruiz et al. 2003; Mistry et al. 2010;

Jennemann et al. 2010].

In this study, we utilize the Asah1P361R/P361R mouse model for ACDase deficiency that we previously generated [Alayoubi et al. 2013]. Asah1P361R/P361R mice have a reported FD patient mutation, proline (P) 362 to arginine (R), “knocked-in” to the corresponding murine

ASAH1 locus (P361R) [Alayoubi et al. 2013]. Homozygous Asah1P361R/P361R mice develop and exhibit phenotypes that recapitulate clinical signs and symptoms seen in FD patients including; extensive inflammation, a reduction in life span, ceramide accumulation, pulmonary deficits, neurological pathology, and failure to thrive [Alayoubi et al. 2013;

Dworski et al. 2015; Dworski et al. 2017; Yu et al. 2017; Yu et al. 2018].

Ceramide build-up is an essential feature of FD. Ceramides and their metabolites are key constituents of membranes that regulate a multitude of cellular functions including inflammation, cell proliferation, and apoptosis [Futerman and Hannun 2004; Hannun and

Obeid 2008; Mullen et al. 2012]. Studies have shown that impairment to enzymes within the ceramide pathway can lead to significant downstream effects to the liver in more common

152 diseases [Adams et al. 2004; Yang et al. 2009]. Due to the role of ceramides and associated sphingolipids in hepatic health we performed a comprehensive evaluation of liver pathology and highlight the gene expression profile in Asah1P361R/P361R mouse hepatocytes. In this study, we report that ACDase deficiency leads to perturbed liver function, characterized by significant inflammation, liver injury, perturbed lipid profiles, and sphingolipid accumulation, which consequentially leads to alteration in genes that affect inflammation as well as lipid and sphingolipid homeostasis.

153

5.3 Material and Methods

5.3.1 Animal use, breeding and genotyping

All animal procedures and experiments were approved and carried out in strict adherence to the policies set forth by the University Health Network Animal Care Committee and the Medical College of Wisconsin Institutional Animal Care and Use Committee. To generate homozygous Asah1P361R/P361R mice, we crossed Asah1+/P361R heterozygotes as previously reported [Alayoubi et al. 2013]. Genotyping was conducted by PCR performed on

DNA from mouse ear notches as previously reported [Alayoubi et al. 2013].

5.3.2 Animal liver weights and collagen assay

Mice were weighed on a small animal scale and then euthanized with CO2 gas. The entire livers were immediately dissected out, weighed, and collected on dry ice for further analyses. To access collagen levels in liver tissue a Sircol Collagen Assay (Biocolor Ltd,

United Kingdom) was performed following manufacturer instructions.

5.3.3 Serum biochemistry and ELISAs

For serum separation, mouse blood was collected via cardiac puncture into serum separator SST microtainers (BD Biosciences, San Jose, CA). Sample tubes were inverted 8-

10 times immediately after blood collection. Samples were set at room temperature for 30 minutes to allow for clotting. Tubes were then centrifuged at 1200 x g for 10 minutes. The top serum portion was collected and immediately stored at -80 °C until use. All samples that contained significant hemolysis were excluded from the study. Serum metabolites and enzymes were measured using both the VetScan Comprehensive Diagnostic Profile and

Mammalian Liver Profile reagent rotors (Abaxis, Union City, CA) following manufacturer’s instructions on the Abaxis VetScan VS2 chemistry analyzer (Abaxis, Union City, CA). Amino

154 aspartate transferase (AST) levels in serum were measured with the AST ELISA kit (Cloud-

Clone Corp., Wuhan China) following manufacturer’s instructions.

5.3.4 Histopathology and immunohistochemistry

As mentioned above, mice were euthanized by CO2 inhalation. Cardiac perfusion with ice-cold PBS was performed with a 24 G needle and liver specimens were immediately fixed in 10% phosphate-buffered formalin for 24-48 hours. Fixed tissues were consequently set, embedded in paraffin, and sectioned at 4 μm. Liver sections were stained with hematoxylin and eosin (H&E), Masson’s trichrome, and reticulin. For liver immunohistochemistry (IHC) the following primary antibodies were used: anti-mouse neutrophil, clone 7/4 (Cedarlane, Burlington, Canada) and rat anti-mouse Mac-2 (Galectin-3) clone M3/38 (Cedarlane), rabbit anti-cleaved caspase-3 (Cell Signaling Technology,

Danvers, MA), rabbit anti-Ki-67 (Thermo Scientific). In situ detection of cell death was performed via a Terminal deoxynucleotidyl transferase (TdT) dUTP Nick-End Labeling

(TUNEL) assay (Promega, Madison, WI). The following secondary antibodies and reagents were then used to detect primary antibodies and TUNEL: rabbit anti-rat IgG, biotinylated

(Vector Laboratories, Burlingame, CA); goat anti-rabbit IgG, biotinylated (Vector

Laboratories); goat anti-rat IgG, biotinylated (Vector Laboratories); donkey anti-chicken IgG

(Jackson ImmunoResearch USA, West Grove, PA); donkey anti-rabbit IgG, biotinylated

(ImmunoResearch); avidin-biotin/HRP (Vector Laboratories); DAB kit (Vector Laboratories) and Vectastain Elite ABC kit (Vector Laboratories). Histology slides were scanned on the

Aperio AT2 histology slide scanner (Leica Biosystems, Buffalo Grove, IL) or NanoZoomer

2.0-HT histology slide scanner (Hamamatsu Photonics, Ichinocho, Japan). Scanned micrographs were analyzed with Aperio ImageScope analysis software (Leica Biosystems,

Buffalo Grove, IL).

155

5.3.5 Pathologic Evaluation

H&E stained liver slides from Asah1+/+ and Asah1P361R/P361R mice at 3,5,7 and 9 weeks of age were analyzed for liver pathology. We developed an injury score with four main parameters. 1) Histiocytic inflammation [0 – normal/none, 1 – minimal (<5%), 2 – mild

(5-10%), 3 – moderate (>10-25%, scattered multinucleated giant cells), 4 – marked (>25%, positive for histiocytic rafts within vessels, endothelial cells with lipid inclusions, and mononuclear giant cells)]. 2) Neutrophilic inflammation [(0 – normal/none, 1 – minimal (<5% section affected), 2 – mild (5-10%), 3 – moderate (>10-25%), and 4 – marked (>25%)]. 3)

Fibrosis [(0 – normal/none, 1 – minimal (focal), 2 – mild (>5-10%), 3 – moderate (>10-25%) and 4 – marked (>25%, dissecting fibrosis portal triads, periportal fibrosis)]. 4) Change in hepatocyte morphology [(0 – normal/none, 1 – mild microvesiculation (<10%) or sporadic single cell necrosis, 2 – moderate microvesiculation (10-25%), 3 – generalized microvesiculation (>25-50%) and 4 – diffuse microvesiculation]. Injury score was evaluated by a pathologist who was masked/blinded to the experimental groups. A total of three animals per genotype and age group were scored. The sum of the four parameters were calculated as an individual liver score, and consequently the age group injury score was calculated by averaging the 3 individual liver scores for each genotype.

5.3.6 Quantitation of cell death and proliferation

Micrographs from Ki67 and cleaved caspase-3 stained slides were further analyzed using the Aperio IHC Nuclear Image Analysis algorithm (Leica Biosystems). In brief 3-4 individual liver samples were collected from Asah1+/+ and Asah1P361R/P361R mice at both 5 and

9 weeks of age. Tissues were prepared as described above. Three consecutive sections per animal were obtained and individually stained for Ki67 and cleaved caspase-3. Image analysis was performed on 10 random non-overlapping fields (250 µm x 450 µm) per

156 section. Section average was obtained by averaging the 10 individual fields, and a total animal value was obtained by averaging the section averages. Data from the algorithm was obtained and expressed as a percentage of positive nuclear staining.

5.3.7 Transmission electron microscopy

Post-euthanasia, 8-9-week-old mice were opened and transcardially perfused with

4% paraformaldehyde (PFA). Liver tissue was collected and stored in the same fixative for

48h. Multiple liver samples (~1-1.5 mm3) were cut from the right lateral, median, and left lateral lobes of each animal and were processed for transmission electron microscopy

(TEM). Each sample was post-fixed in 3% glutaraldehyde for four hours at room temperature, washed with phosphate buffered saline solution (PBS) and contrasted overnight at 8°C with 2% OsO4 in PBS. After treatment with a dehydration series, samples were then sequentially placed in polymerization mixtures (1:1 for 1.5 hour and 3:1 overnight at room temperature) of Durcupan Epon (Fluka, Hatfield, PA) and propylene oxide, respectively. After polymerization ultrathin sections (60 nm) were cut from prepared tissue blocks using an Ultracut ultramicrotome (Reichert-Jung, Munich, Germany) and placed on copper grids. The samples on the grids were contrasted with uranyl acetate and incubated in lead citrate. Tissue sections were observed, and images were taken using a JEOL 1400+

(JEOL, Tokyo, Japan) TEM equipped with Olympus Veleta CCD camera and Radius software.

5.3.8 Murine liver perfusion and hepatocyte isolation

Liver perfusion and hepatocyte isolation was adapted from Azuma and colleagues

[Azuma et al. 2007]. In brief, five-week-old Asah1+/+ and Asah1P361R/P361R mice were anesthetized with inhaled 5% isoflurane and opened to expose the liver and portal vein.

157

Prior to perfusion, the portal vein was cannulated with a 26 G ¾ inch peripheral IV catheter

(Zoetis, Inc., Parsippany-Troy, NJ) connected by tubing to the Watson-Marlow 323S/D peristaltic pump (Watson-Marlow Fluid Technology Group., Wilmington, MA). Once the catheter was inserted the vena cava was cut and perfused. The liver was first flushed with

Solution 1: 40-50 ml of Earle’s balanced salt solution (EBSS) (Thermo Scientific) supplemented with 1x EDTA (Thermo Scientific) and 1X HEPES (Thermo Scientific) at 40oC for 10 minutes. The liver was next flushed with Solution 2: 20-30 ml of EBSS containing calcium and magnesium (Thermo Scientific) with 1x HEPES (Solution 2) at 40oC for 3-5 minutes. Lastly, the liver was flushed with 40-50 ml of Solution 2 supplemented with 1 mL of

Liberase (Roche Diagnostics., Indianapolis, IN) at 26 Wunsch units/mL (Solution 3) at 40oC for 10 minutes. Once digested the liver was excised and transferred into a Petri dish with

Solution 2 where connective and unwanted tissue was removed manually. The digested liver was then transferred to a new Petri dish containing 10-15 mL of cold hepatocyte culture medium (DMEM, 10% FCS, 1% Pen-Strep, 1% Glutamine, and 0.1% 1M HEPES) where the liver was minced. The tissue and cell suspension was transferred into a 50 mL conical tube and allowed to settle. After settling, the supernatant was filtered through a 100 μm cell strainer. The suspension was further washed and passed through a 100 μm cell strainer with 10-15 mL of hepatocyte culture media and filtered. A total of 50 mL of filtered cells were collected. The collected cells were then passed once more through a 70 μm cell strainer and spun at 300 g for 3 mins at 4oC. The cell pellet was resuspended in 5 mL of hepatocyte culture medium and counted using Trypan Blue exclusion with the Countess II FL automated cell counter (Life Technologies, Carlsbad, CA). 2.5 million cells were seeded in 25 sq cm

o tissue culture flasks and incubated at 37 C with 5% CO2. After overnight incubation, cells were washed 2 times with PBS to eliminate cellular debris before further analyses.

158

5.3.9 Flow cytometry analyses of macrophages

Flow cytometry was performed on liver and hepatocyte-enriched cultures samples.

To obtain a single cell suspension of the liver, cells were forced through a 40 µm nylon cell strainer in PBS with 2% fetal calf serum (FCS). Cells were collected, washed, and resuspended in PBS with 2% FCS and counted on a hemocytometer. Liver cell and hepatocyte-enriched cultures were stained for 30 mins at 4°C with the CD11b (Mac-1) FITC

(Biolegend; M1/70) antibodies to assess for abundance of monocytes/macrophages. Flow cytometry was performed using FACSDiva software (BD Biosciences) on either Fortessa or

ARIA II (BD Biosciences) cytometers. Data was analyzed using FlowJo software (Tree Star

Inc, Ashland, OR).

5.3.10 RNA isolation, library preparation for RNASeq and sequencing

To analyze the transcriptome of 5-week-old Asah1P361R/P361R mouse hepatocytes- enriched cultures we applied an RNAseq approach. First, total RNA was isolated from approximately two million cells using TRIzol (Life Technologies). RNA was purified using

RNeasy spin columns (Qiagen, Crawley, UK) according to the manufacturer’s instructions.

Total RNA integrity was then confirmed using the Agilent 2100 Bioanalyzer (Agilent

Technologies, Santa Clara, CA). RNA-seq library preparation and sequencing was performed by the Sequencing Core, Medical College of Wisconsin, as follows: libraries were prepared from 100 ng of total RNA using the Truseq Stranded mRNA Kit on the Illumina

NeoPreo automated library preparation instrument (Illumina Inc., San Diego, CA). The quality of the prepared libraries were analyzed using the DNA1000 assay on the Agilent

Bioanalyzer 2100 (Agilent Technologies) and quantified via qRT-PCR using the Kapa

159

Biosystems Library Quantification Kit (Kapa Biosystems, Inc., Wilmington, MA) and Bio-Rad

CFX384 real-time PCR instrument (Bio-Rad Laboratories, Inc, Des Plaines, IL). Libraries were diluted and pooled at equimolar ratios and sequenced on a 1 x 50 single-read Miseq run to assess for cluster density and index distribution. The library pool concentration was then normalized then ran on a HiSeq 2500 Flowcell (Illumina technology) and 2 x 125 bp paired-end sequenced across 2 lanes at a sample density of 8 samples per lane. Upon completion of the sequencing run, the data were demultiplexed using Illumina’s Bcl2Fastq v1.8.4 conversion software (Illumina Inc).

5.3.11 Quantitative real time PCR

For control purposes, we assessed RNA expression of high-scoring genes from

RNA-seq using quantitative real time PCR. Total RNA was isolated with TRIzol reagent.

After retro-transcription, we performed Real Time PCR with Applied Biosystems™

PowerUp™ SYBR™ Green Master Mix (Thermo Scientific), following the manufacturer’s instructions, on a ViiA7 System (Applied Biosystems, Life Technologies Cooperation,

Carlsbad, CA, USA). The primer sequences used are in Table 9.

160

Genes NCBI reference Forward Reverse Size

ApoM NM_018816.1 TCAACTAACTGCGCTGGGAA TGTAGAGGAGAAAGCGCTGG 368 CerK NM_145475.4 GCTACATGATACCCACGCCC ATGAGGATGAGGGGAGGCCAT 658

Cers2 NM_001320492.1 GTCAGTGCCTAGTGCCTCT CGGACGTAATTGGCAAACCA 792

Cers4 NM_026058.4 AGACTCAACGCTGGTTCAGG AACTGGCTCGTCATCACTGG 728

Dgat2 NM_026384.3 TGCTTCGCGAGTACCTGATG AAAGGATCTGCAGTCGTCCC 608

ApoA1 NM_009692.4 CAGCGGCAGAGACTATGTGT GGAATTCGTCCAGGTAGGGC 242

Psap NM_001146120.1 GTTGATGCACATGCAACCCA GCACTGCTTCTGGTAAGGGT 682

Vcam1 NM_011693.3 GCTGCGAGTCACCATTGTTC TGTTCATGAGCTGGTCACCC 330

Cxcr4 NM_001356509.1 AAACCTCTGAGGCGTTTGGT CTATCGGGGTAAAGGCGGTC 651

Gapdh NM_001289726.1 AACTCAGGAGAGTGTTTCCTCG TCGTGGTTCACACCCATCAC 442

Ugcg NM_011673.3 ACCGAGGTTGGAGGTTTTCA CACAGCGTAACCTGTAGCGA 420

Table 9 Yield and quality of RNA for RNA-seq analysis

Each sample represented an individual mouse. WT: Asah1+/+. Hom: Asah1P361R/P361R.

5.3.12 Sphingolipids in liver tissue and hepatocytes

Liver tissue (~300 mg) was homogenized in ~600 µL of 2% CHAPS with the Bullet blender (Next Advance, Inc., Troy, NY) using 0.5 mm zirconium oxide beads. Lipids were extracted from 50 µL of tissue lysate with 200 µL isopropanol. The following internal standards were used for sphingolipid measurements: ceramide 100 ng (d18:1/ 22:0) d4

(Medical University of South Carolina (MUSC) Lipidomics Core, Charleston, SC); monohexosylceramide (MHC) 100 ng (d18:1/17:0) (Avanti Polar Lipids Inc., Alabaster, AL); sphingomyelin (SM), 1000 ng (d18:1/17:0) (Avanti Polar Lipids Inc.); and ceramide-1- phosphate (C1P) (d18:1/16:0), (d18:1/24:0) and (d18:1/24:1) (Matreya Inc., Pleasant Gap,

PA) (Avanti Polar Lipids Inc.); and 100 ng sphingosine (Sph) d7 (Avanti Polar Lipids Inc.)

[Bielawski et al. 2010].

For enriched hepatocytes ~2.0 x 106 - 3.0 x 106 were lysed in 200 µL of PBS by passing through a 28 G needle 15 times. Lipids were extracted from 100 µL of each cell

161 homogenate with 400 µL of isopropanol. The supernatant was reconstituted with 250 µL of water for LC-MS/MS analysis. The following internal standards were used for sphingolipid measurements: ceramide 100 ng (d18:1/ 17:0) d4 (MUSC Lipidomics Core); MHC 10 ng

(d18:1/17:0) (Avanti Polar Lipids Inc.); SM, 300 ng (d18:1/17:0) (Avanti Polar Lipids Inc.); and 100 ng C1P (d18:1/12:0) (Avanti Polar Lipids Inc.). Sample analysis for sphingolipids from liver and hepatocytes were both performed using the Shimadzu 20AD HPLC system and a Leap PAL autosampler coupled to a triple quadrupole mass spectrometer (API-4000:

Applied Biosystems) operated in the Multiple Reaction Monitoring (MRM) mode. The positive ion electrospray ionization (ESI) mode was used for detection of sphingolipids.

Samples were injected in duplicate for data averaging. Data processing was conducted with

Analyst 1.5.1 (Applied Biosystems). The relative quantification of sphingolipids was obtained and expressed as the peak area ratios of the analytes to the corresponding internal standards. The data are presented as fold change relative to the averaged sphingolipid values detected in control Asah1+/+ mice.

5.3.13 Measurement of lipids in liver and serum

Serum and liver tissue were collected from mice that were fasted overnight (~12 hours) post CO2 euthanization. Liver samples were homogenized as mentioned above.

Lipids were extracted as previously mentioned from 50 µL of serum and liver lysates. The following internal standards were used to measure lipids: Phosphatidylcholine (PC) (14:1-

14:1) (Avanti Polar Lipids Inc,); Triglyceride (TG) (17:0-17:0-17:0) (Avanti Polar Lipids Inc,);

Free Fatty Acid (FFA) d4 (16:0) (Avanti Polar Lipids Inc,). Internal standards were added to the samples before extraction. Extracted FFA samples were further derivatized by aminomethyl phenyl pyridium (AMPP) into FA-AMPP derivatives to obtain increased sensitivity in MS.

162

Free cholesterol (FC) and cholesteryl esters were extracted with 200 µL isopropanol.

After vortexing and centrifugation, 10 µL of the supernatant was used for total cholesterol

(TC) and the remaining supernatant was used for FC. Five µg of deuterated cholesterol-d7

(Avanti Polar Lipids Inc,) was added as internal standard for FC and TC samples prior to lipid extraction. Measurement of lipids was performed using a Shimadzu 10A HPLC system and a Shimadzu SIL-20AC HT autosampler coupled to a Thermo Scientific TSQ Quantum

Ultra triple quadrupole mass spectrometer (API-4000: Applied Biosystems) operated in

MRM mode for PC, FFA, TC and SRM mode under ESI (+) for TC and FC samples. Data processing was conducted with Analyst 1.5.1 (Applied Biosystems).

5.3.14 Statistical analyses

Data are expressed as means ± standard error. Unless otherwise stated data was analyzed with a student t-test. Serum biochemistry, ELISA, and FACS analyses were analyzed with a 1-way ANOVA followed by a Tukey post-test. All statistics were analyzed using GraphPad Prism 6.0 (GraphPad Software Inc, La Jolla, CA). Significant differences are expressed in the figures as *p < 0.05, **p < 0.01, and ***p < 0.001. For RNAseq, the

Benjamini–Hochberg method was used to control for false discovery rates. False discovery rates of <0.05 were considered significant. The 95% confidence interval was considered significant.

163

5.4 Results

5.4.1 Hepatomegaly and liver injury

Asah1P361R/P361R mice demonstrate impaired growth and decreased body weight post

3 weeks of age (Figure 39A) [Alayoubi et al. 2013]. Liver weight from Asah1+/+ mice were significantly heavier than Asah1P361R/P361R mice from 7 weeks onward (Figure 39B).

Normalizing liver tissue weight for body weight revealed significant hepatomegaly in

Asah1P361R/P361R mice as early as 3 weeks of age (Figure 39C). We measured various biochemical analytes in serum of 5 and 9-week-old Asah1+/+ and Asah1P361R/P361R mice to monitor organ function and disease. Analyses of common liver injury markers in serum of

Asah1P361R/P361R mice revealed significantly higher expression of aspartate aminotransferase, alanine aminotransferase, and alkaline phosphatase by 5 weeks of age (Figure 39D-F).

Gamma-glutamyl transferase, an enzyme which is often analyzed in conjunction with alkaline phosphatase to distinguish bone versus liver injury, was also significantly increased in Asah1P361R/P361R mice at 9 weeks of age (Figure 39G) further confirming liver involvement.

No changes were detected in total bilirubin between Asah1+/+ and Asah1P361R/P361R mice, however, total bilirubin did trend upward between 5 and 9 weeks of age (Figure 39H). Both blood urea nitrogen and bile acids were also found to be significantly increased in

Asah1P361R/P361R mice at 9 weeks of age (Figure 39I,J). A reduction in serum albumin and total cholesterol levels were only detected between Asah1+/+ and Asah1P361R/P361R mice at 9 weeks of age (Figure 39K,L). Glucose levels in serum were significantly decreased in both 5 and 9-week-old Asah1P361R/P361R mice compared to controls. Lastly creatinine, a commonly used indicator of renal function, and troponin, a protein marker for cardiac health, were found unchanged in Asah1+/+ and Asah1P361R/P361R mice (Figure 39N,O).

164

Figure 39. Hepatomegaly, liver injury and perturbed metabolites in Asah1P361R/P361R mice

Growth curve measuring Asah1+/+ and Asah1P361R/P361R mouse weight over time (n=8 for both genotypes) (A). Liver tissue weight recorded over time (n=8 per genotype) (B). Liver weight measured as a percentage of body weight over time (n=8 per genotype) (C). ELISA for aspartate aminotransferase (AST) measured in serum from 5 and 9-week-old mice (n=5 per genotype) (D). Liver specific and general metabolites and enzymes were analyzed in serum from 5 and 9-week-old mice for each genotype. Analytes from the biochemistry panels include alanine aminotransferase (ALT) (E), alkaline phosphatase (ALP) (F), gamma-glutamyl transferase (GGT) (G), total bilirubin (TB) (H), blood urea nitrogen (BUN) (I), bile acids (BA) (J), albumin (K), total cholesterol (TC) (L), glucose (M), creatinine (N), and troponin (O). All comparisons were made between 5 and 9-week-old mice (n=6 per genotype). ns (not significant), *p<0.05, **p<0.01, ***p<0.001.

165

5.4.2 Liver pathology and cellular infiltration

Hematoxylin and eosin (H&E) staining of 9-week-old Asah1P361R/P361R liver tissue revealed significant signs of inflammation in comparison to control Asah1+/+ livers (Figure

40A,B). Macrophages in Asah1P361R/P361R livers displayed a pale eosinophilic and foamy appearance and contained vesiculated cytoplasmic material (Figure 40B). Multifocal coalescing accumulations of foamy macrophages were also seen in the parenchyma and surrounding vessels (Figure 40B). On occasion, rafts of cells formed by 3-4 foamy macrophages were found perivascularly, and liver sinusoid macrophages (Kupffer cells) also showed a foamy appearance that occasionally formed granulomas that disrupted the hepatocellular trabeculae (Figure 40B). In some areas neutrophil-like cells were noted towards the periphery of histiocytic aggregates. Within these aggregates, single cell necrosis was observed in hepatocytes (Figure 40B).

To further characterize the extent of inflammation, immunohistochemistry (IHC) with antibodies specific to Mac-2 (galectin-3) and a mouse neutrophil marker was performed on

3,5,7 and 9-week-old liver sections from both Asah1+/+ and Asah1P361R/P361R mice (Figure

40C-J). Increased inflammation was visible by 3 weeks of age in Asah1P361R/P361R livers

(Figure 40G). Between 5 and 7 weeks of age, Asah1P361R/P361R livers experienced an accelerated rate of neutrophil and macrophage recruitment (Figure 40H,I). Macrophage appeared to increase in both size and foaminess as the mice aged. By 9 weeks-of-age

Asah1P361R/P361R mice displayed extensive neutrophil and macrophage infiltration (Figure

40J).

166

Figure 40. Significant inflammation and pathology in Asah1P361R/P361R mice

Light micrographs of liver sections stained for hematoxylin and eosin (H&E) in 9-week-old Asah1+/+ and Asah1P361R/P361R mice (A-B). H&E micrographs and immunohistochemistry (IHC) for Anti-Neu and Mac-2 in 3,5,7, and 9-week-old Asah1+/+ mice (C-F) and Asah1P361R/P361R mice (G-J). The drawn rectangles correspond to the adjacent magnified panel. Scale bars for all micrographs indicate 100 µm

167

5.4.3 Liver storage pathology

TEM was performed on liver sections from Asah1+/+ control and Asah1P361R/P361R mice

(Figure 41). Ultrastructure analyses within the portal field of Asah1+/+ control mice revealed normal cells (Figure 41A), however, the portal field of Asah1P361R/P361R mice displayed cells that contained excessive collagen fibers in 9-week-old Asah1P361R/P361R livers (Figure 41B).

Many atypical macrophages were observed with abnormal storage vacuoles containing fine granular-fibrillary content and curvilinear tubular bodies (CTB), also known as Farber bodies, in the cytoplasm (Figure 41B). Analyses of the Kupffer cells in the liver sinusoid revealed a similar pattern of storage pathology where the cytoplasm of macrophages was laden with abnormal storage vacuoles (Figure 41D). Endothelial cells were also found to accumulate

CTB-like storage material in Asah1P361R/P361R livers (Figure 41D).

Hepatocytes in the livers of Asah1+/+ control mice had a normal cellular phenotype

(Figure 42A), however hepatocytes from Asah1P361R/P361R mice also displayed storage vacuoles in the cytoplasm (Figure 42B). However phenotypically, the size and abundance of the individual vacuoles appeared to be reduced in comparison to those found in portal field macrophages and Kupffer cells. We observed bi-nucleated hepatocytes, as well as individual hepatocytes with degenerative vacuolar changes within their cytoplasm that appeared phenotypically different from the vacuolar storage pathology (Figure 42B). Further analyses of the liver sinusoids revealed abnormal storage vacuoles in hepatic stellate (Ito) cells in Asah1P361R/P361R mice (Figure 42D). Finally, within the bile ducts of Asah1P361R/P361R mice we observed Zebra-like storage material in the epithelial cells (Figure 43B,C). The storage material present these bile duct cells were phenotypically different than the CTB-like material present in other liver cells of Asah1P361R/P361R mice (Figure 43B,C).

168

Figure 41. Storage pathology within the portal fields and liver sinusoid in Asah1P361R/P361R mice

Electron micrograph of representative portal fields in a 9-week-old Asah1+/+ mouse (A) (rectangle corresponds to the adjacent magnified panel). Electron micrograph of a portal field in a 9-week-old Asah1P361R/P361R liver with excessive collagen fibers and infiltrating macrophage displaying cytoplasm containing excessive storage vacuoles with curved-linear tubular profiles (CTB aka ‘Farber bodies’) (B) (rectangle corresponds to the adjacent magnified panel). Representative electron micrograph of liver sinusoid in a 9-week-old Asah1+/+ mouse (C). Electron micrograph of liver sinusoid in liver of a 9-week-old Asah1P361R/P361R mouse featuring excessive storage vacuoles with CTB-like content (black arrows) (D). Higher magnification of storage material in the cytoplasm of Kupffer and endothelial cells (Annotated rectangles 1,2 correspond to the two adjacent panels). Annotations in figures include BD- bile duct lumen, EC – endothelial cell, EC ncl – nucleus of an endothelial cell, ery – erythrocytes, glc – glycogen, HPC ncl – nucleus of a hepatocyte, KC – Kupffer cell, KC ncl – nucleus of a Kupffer cell, LS – liver sinusoid lumen, MCP-macrophage, MCP ncl- nucleus of a macrophage, PV- portal vein lumen. Scale bars on micrographs indicate 2µm.

169

Figure 42 Hepatocyte and hepatic stellate cell storage pathology in Asah1P361R/P361R mice

Electron micrographs of representative hepatocytes in 9-week-old Asah1+/+ mice (A). Electron micrograph of a representative 9-week-old Asah1P361R/P361R mouse featuring a bi-nucleated hepatocyte with membrane bound storage vacuoles containing CTBs in the cytoplasm (black arrows) (B) (rectangle corresponds to the adjacent magnified panel). Electron micrograph of an endothelial and hepatic stellate (Ito) cell in a representative 9-week- old Asah1+/+ mouse located in the perisinusoidal space with cytoplasmic lipid droplet (C) (rectangle corresponds to the adjacent magnified panel). Electron micrograph of a representative 9-week-old Asah1P361R/P361R mouse featuring hepatic stellate (Ito) cell with membrane bound storage vacuoles (black arrows) (D) (rectangle corresponds to the adjacent magnified panel). Annotations include: Black arrowhead- endothelial fenestrations in liver sinusoid, BC Nuc – bile canaliculus, EC ncl – nucleus of an endothelial cell, ery – erythrocytes, glc – glycogen, HPC ncl – nucleus of a hepatocyte, ld - lipid droplet, LS – liver sinusoid lumen. Scale bars on micrographs indicate 2 µm.

170

Figure 43 Storage pathology in bile duct epithelia of Asah1P361R/P361R mice

Electron micrograph of a representative bile duct in cross section from a 9-week-old Asah1+/+ mouse (A) (rectangle corresponds to the adjacent magnified panel). Bile duct epithelia in 9-week-old Asah1P361R/P361R mouse containing storage vacuoles (black arrows) with “zebra-like” content (B) (rectangle corresponds to the adjacent magnified panel). Microvilli at the apical surface of the bile duct epithelia are highlighted by white arrows, and bile ducts epithelial tight junctions are highlighted by white arrowheads (C) (rectangle corresponds to the adjacent magnified panel). Annotations; BD- bile duct lumen, B-nucleus of bile duct epithelial cell. Scale bar on micrographs indicate 2 µm.

171

5.4.4 Progressive liver fibrosis

Formation of granulomas was the most obvious phenotype in the liver of

Asah1P361R/P361R mice. However, signs of fibrosis were also apparent as pale pink H&E stained regions. We confirmed this with Masson’s trichrome and reticulin staining, and soluble collagen measurements, on liver tissue from Asah1+/+ and Asah1P361R/P361R mice

(Figure 44A-F). Masson’s trichrome stains were performed on liver sections from 3,5,7, and

9-week-old mice revealing a progressive fibrotic phenotype in the livers of Asah1P361R/P361R mice (Figure 44A-D). In the livers of 3 and 5-week-old Asah1P361R/P361R mice, the fibrosis appeared primarily localized near the central veins (Figure 44A,B). However, by 7 and 9 weeks of age, distension of fibrosis was most apparent in the regions where unstained foamy macrophages were localized (Figure 44C,D). The livers of 9-week-old Asah1P361R/P361R mice revealed the presence of collagen and reticular fibers in the portal fields and along the central vein (Figure 44D,E). To further quantify the fibrotic phenotype, soluble collagen was measured in liver tissue from Asah1+/+ and Asah1P361R/P361R mice. Asah1P361R/P361R mouse livers displayed significantly increased collagen by 5 weeks of age in comparison to age- matched controls (Figure 44F). While Asah1+/+ control mice displayed no further statistical increase in collagen staining by 9 weeks of age, Asah1P361R/P361R mouse livers displayed a significant increase in collagen between 5 and 9-weeks of age (Figure 44F).

172

Figure 44 Progressive liver fibrosis in Asah1P361R/P361R mice

Light micrographs of liver sections stained with Masson’s trichrome from 3,5, and 7-week-old Asah1+/+ and Asah1P361R/P361R mice (A-C). Masson’s trichrome (D) and reticulin (E) stained liver sections from 9-week-old Asah1+/+ and Asah1P361R/P361R mice. Scale bars for all light micrographs indicate 100 µm. Soluble collagen content measured from the livers of 5 and 9-week-old Asah1+/+ and Asah1P361R/P361R mice (F). n=6 mice for each genotype. ***p<0.001.

173

5.4.5 Increased cell proliferation and cell death in the liver

Histopathology revealed mild necrosis and the presence of hepatocyte dropout

(Table 10). To assess whether there was also increased cell proliferation and also apoptotic cell death, we performed IHC for Ki67 and cleaved caspase-3 in 3,5,7, and -9-week-old liver sections from both Asah1+/+ and Asah1P361R/P361R mice (Figure 45A-D). Image analysis of IHC stained liver sections revealed an increase in Ki67 and cleaved caspase-3 nuclear staining by 7 weeks of age in Asah1P361R/P361R mice compared to age matched controls (Figure

45E,F). Also, liver sections from 7 and 9-week-old Asah1P361R/P361R mice stained positive for terminal deoxynucleotidyl transferase dUTP nick end labeling (TUNEL) in areas that were positive for cleaved caspase-3 (Figure 45C,D).

Sample Histiocytic Inflammation Fibrosis Hepatic Total size Infiltration (neutrophilic) (0-4) Change Score (0-4) (0-4) (0-4) 9-week-old 5 0 0 0 0 0 Asah1+/+ 5-week-old 5 2.6 ±0.6 1.6 ±0.2 1.8 ±0.4 1.0 ±0.3 7.0 ±1.2 Asah1P361R/P361R 9-week-old 5 3.6 ±0.2 2.4 ±0.2 3.0 ±0.3 1.4 ±0.2 10.4 ±0.5 Asah1P361R/P361R

Table 10. Liver Injury Score

Average liver injury score ± SEM. 9-week-old Asah1+/+ mice revealed no liver injury. 5 and 9-week-old Asah1P361R/P361R mice show progressive liver injury

174

Figure 45 Increased cell death and proliferation in liver of Asah1P361R/P361R mice

Light micrographs of liver sections stained for Ki67 and cleaved caspase-3 in 3,5,7, and 9-week-old Asah1+/+ and Asah1P361R/P361R mice (A-D). Light micrographs of liver sections stained for terminal deoxynucleotidyl transferase (TdT) dUTP Nick-End Labeling (TUNEL) in 7 and 9-week-old Asah1+/+ and Asah1P361R/P361R mice (C-D). Scale bars for all micrographs indicate 100 µm. Percentage of Ki67 positive cells from 3,5,7 and 9-week-old Asah1+/+ and Asah1P361R/P361R mice (E). Percentage of cleaved capsase-3 positive cells from 3,5,7 and 9-week-old Asah1+/+ and Asah1P361R/P361R mice (F).

175

5.4.6 Liver injury score

To quantify the extent of liver injury in Asah1P361R/P361R mice we created a liver injury score that assessed histiocytic infiltration, neutrophil mediated inflammation, fibrosis, and hepatocyte morphology on liver tissue stained with H&E. Evaluation by a pathologist blinded to the genotypes of tissues revealed no injury in livers of 9-week-old Asah1+/+ mice.

Significant liver injury was present by 5 weeks of age and worsened in 9-week-old

Asah1P361R/P361R mice (Table 10). Based on our liver injury score, the most affected condition in Asah1P361R/P361R mice were formation of histiocytes, followed by fibrosis, neutrophil mediated inflammation and lastly cellular changes to hepatocyte morphology (Table 10).

5.4.7 Alteration to lipids in plasma and liver tissue

Asah1P361R/P361R mice lose weight as disease manifests (Figure 39A) and display reduced fat pad size over time (data not shown). Our serum biochemistry panel also showed a reduction in total cholesterol suggesting that Asah1P361R/P361R mice may experience hypocholesterolemia (Figure 39L). However, those analyses were not performed on fasted mice. Thus, we quantified free fatty acids (FFA), triglycerides (TG), phosphatidylcholine

(PC), free (FC) and total cholesterol (TC) both in serum and liver tissue from 8-9-week-old

Asah1+/+ and Asah1P361R/P361R mice that were fasted overnight with LC-MS. Asah1P361R/P361R mice displayed significantly decreased FFA, and TG in both the liver tissue and plasma

(Figure 46A,B,G,H). PC, FC, and TC were found to be significantly decreased in the plasma, but no changes in PC, FC and TC were detected in the liver lysates from Asah1P361R/P361R mice (Figure 46C-F, I-K).

176

Figure 46 Decreased lipids in serum and liver of Asah1P361R/P361R mice

Free fatty acid (FFA) (A,F), triglyceride (TG) (B,G), phosphatidylcholine (PC) (C,H), free cholesterol (FC) (D,I), and total cholesterol (TC) (E,J) measured in liver and plasma samples from 8-9-week-old Asah1+/+ and Asah1P361R/P361R mice; n=5 mice for each genotype. *p<0.05 **p<0.01.

177

5.4.8 Altered sphingolipid profile in the liver

Liver lysate sphingolipids from 8-9-week-old from Asah1+/+ and Asah1P361R/P361R mice were quantified with LC-MS. Liver tissue from Asah1P361R/P361R mice showed a significant increase in all species of ceramides, monohexosylceramides (MHC), and ceramide-1- phosphate (C1P) measured (Figure 47A,D,E). While total sphingomyelin (SM) was significantly increased in the liver lysates from Asah1P361R/P361R mice, the greatest increase was seen in the SM16:0 and SM18:0 species, whereas no changes were seen in SM22:0 and SM24:0 (Figure 47B). Notably, the C16:0 and C18:0 species were also the most elevated in the aforementioned ceramide and MHC analysis (Figure 47A,D). Sphingosine

(Sph) on the other hand was significantly decreased in liver lysates from Asah1P361R/P361R mice (Figure 47C). When we calculated the relative abundance of each class of sphingolipid, we found that there was an increase in the C16:0/SM16:0 and C18:0/SM18:0 species and a decrease in C22:0/SM22:0, C24:1/SM24:1 and C24:0/SM24:0 in species ceramide, SM, and MHC in the liver lysates of Asah1P361R/P361R animals (Figure 47F-H). Of the C1P species we measured, the relative abundance of C1P was unaffected (Figure 47I).

178

Figure 47 Sphingolipid accumulation in liver of Asah1P361R/P361R mice

Ceramide species in liver lysates from Asah1+/+ and Asah1P361R/P361R mice (A). Sphingomyelin (SM) species in liver lysates from Asah1+/+ and Asah1P361R/P361R mice (B). Sphingosine (Sph) in liver lysates from Asah1+/+ and Asah1P361R/P361R mice (C). Monohexosylceramide (MHC) species in liver lysates from Asah1+/+ and Asah1P361R/P361R mice (D). Ceramide-1-phosphate (C1P) species in liver lysates from Asah1+/+ and Asah1P361R/P361R mice (E). Relative abundance of ceramide species in liver lysates from Asah1+/+ and Asah1P361R/P361R mice (F). Relative abundance of SM species in liver lysates from Asah1+/+ and Asah1P361R/P361R mice (G). Relative abundance of MHC species in liver lysates from Asah1+/+ and Asah1P361R/P361R mice (H). Relative abundance of C1P species in liver lysates from Asah1+/+ and Asah1P361R/P361R mice (I) All samples were obtained from 8-9-week-old Asah1+/+ and Asah1P361R/P361R mice; n=4 mice for each genotype. *p<0.05 **p<0.01, ***p<0.001.

179

5.4.9 Hepatocyte isolation and sphingolipid profile

Lipidomic and genetic analyses were conducted on hepatocytes-enriched cultures.

We performed/adapted the commonly used ‘two-step’ collagenase liver perfusion method to isolate primary hepatocytes and reduce macrophage contamination from Asah1+/+ and

Asah1P361R/P361R livers [Lee et al. 2013; Azuma et al. 2007]. Additionally, we selected the 5- week time point since it represents mid-point in the lifespan of Asah1P361R/P361R mice and presents a milder inflammatory phenotype. Flow cytometry for CD11b (Macrophage-1 antigen) demonstrated a significant reduction in macrophages compared to non-perfused liver tissue from 5-week-old Asah1P361R/P361R mice (Figure 48). All analyses on primary hepatocyte-enriched cells were conducted following removal of debris and non-adherent cells and overnight culture.

Figure 48 Reduction of CD11b cells in hepatocyte-enriched cultures from Asah1P361R/P361R mice

Representative flow cytometry histograms displaying macrophage-1 antigen (CD11b) positive cells in hepatocyte-enriched cultures from 5-week-old Asah1+/+ mouse (A), dissociated cells from 5-week-old Asah1P361R/P361R mouse liver tissue (B), and hepatocyte-enriched cultures from 5-week-old Asah1P361R/P361R (C). Percentage of CD11b positive cells from liver and hepatocytes-enriched cultures measured by flow cytometry conducted on 5-week-old Asah1+/+ and Asah1P361R/P361R mice (D). n=4 ***p<0.001, ns- nonsignificant.

180

When we measured sphingolipids from hepatocyte-enriched cultures we found similar alterations to those seen in liver tissue. Hepatocyte-enriched cultures from

Asah1P361R/P361R mice had significantly increased levels of total ceramide, SM and MHC

(Figure 49A,C,H). The C16:0/SM16:0 and C18:0/SM18:0 species showed the largest fold changes for ceramide, SM and MHC (Figure 49A,C,E). Measurements for C1P and Sph were below the limit of detection (data no shown). Analyses of the relative abundance of each sphingolipid class in hepatocyte-enriched cultures revealed again that there was a significant increase in the percentage of C16:0/SM16:0 and C18:0/SM18:0 and reduction in percentage of C22:0/SM22:0, C24:1/SM24:1 and C24:0/SM24:0 in Asah1P361R/P361R samples

(Figure 49B,D,F).

181

Figure 49 Sphingolipid accumulation in hepatocyte-enriched cultures from Asah1P361R/P361R mice

Quantification of ceramide species in hepatocyte-enriched cultures from Asah1P361R/P361R mice in comparison to Asah1+/+ mice (A). Relative abundance of ceramide species in hepatocyte-enriched cultures from Asah1+/+ and Asah1P361R/P361R mice (B). Quantification of SM species in hepatocyte-enriched cultures from Asah1P361R/P361R mice in comparison to Asah1+/+ mice (C). Relative abundance of SM species in hepatocyte-enriched cultures from Asah1+/+ and Asah1P361R/P361R mice (D). Quantification of MHC species in hepatocyte-enriched cultures from Asah1P361R/P361R mice in comparison to Asah1+/+ mice (E). Relative abundance of MHC species in hepatocyte- enriched cultures from Asah1+/+ and Asah1P361R/P361R mice (F). Hepatocytes-enriched cultures were isolated from 5-week-old Asah1+/+ and Asah1P361R/P361R mice as previously described; n=4 samples for each genotype. *p<0.05 **p<0.01, ***p<0.001.

182

5.4.10 Differential gene expression in hepatocytes

We performed RNAseq analysis on hepatocytes-enriched cultures derived from both

5-week-old Asah1+/+ and Asah1P361R/P361R mice. The yield of the RNAseq analysis, quality of the sequencing reads, and mapping rates are summarized in Table 11. Substantial differences in the transcriptomes were detected between samples from Asah1+/+ and

Asah1P361R/P361R mice. Multiple genes showed altered expression (false discovery rate <0.05) between the samples of 3 Asah1+/+ control and 3 Asah1P361R/P361R mice; with more than 70 transcripts showing a >5-fold change compared to control.

Sample Group Yield # Reads % of >= Mean % of ID (Mbases) Q30 Quality mapping Bases Score (PF) rate (PF) 983 WT_5_week 2,957 24,215,148 92.3 34.9 83.3 923 WT_5_week 2,827 22,877,600 97.6 36.1 95.6 928 WT_5_week 3,066 24,841,108 97.5 36.0 94.0 982 HOM_5_week 2,722 22,048,596 94.0 35.2 88.6 922 HOM_5_week 2,813 22,734,564 97.6 36.1 95.2 985 HOM_5_week 2,976 24,215,666 93.4 35.1 87.5

Table 11. Yield and quality of RNA for RNAseq analysis

Each sample represented an individual mouse. WT: Asah1+/+. Hom: Asah1P361R/P361R.

5.4.11 Pathway clustering for inflammatory response and lipid homeostasis

To gain insight into pathway changes in ACDase deficiency, we analyzed all genes differentially expressed in hepatocyte-enriched cultures from Asah1+/+ and Asah1P361R/P361R mice using Ingenuity Pathway Analysis (IPA) software. The top 2000 genes up or down regulated are presented as a heat map in (Figure 50A). Pathway analyses conducted in the

183

IPA software identified 168 molecules (p-value 8.22 E-7 to 8.24 E-26) organized into activated inflammatory response pathways (Figure 50B). Additionally, pathway analyses identified 146 molecules (p-value 8.54 E-7 to 7.49 E-39) categorized into deactivated lipid metabolism and injury response pathways (Figure 50C).

Figure 50 Differential gene expression and altered pathways in Asah1P361R/P361R mice

Heatmap analysis of the 2000 genes that displayed the most significant variation in relative expression (Z-score) between Asah1+/+ (left) and Asah1P361R/P361R (right) mice (A). Genes induced are depicted in red, genes repressed are depicted in blue (at least 2-fold change in expression, false discovery rate < 1E-5). IPA analysis identified significantly activated (red) and deactivated (blue) pathways plotted according to -log(p-value) (B,C).

One of the activated inflammatory response pathways from the IPA analyses was the leukocyte extravasation pathway (z score = 3.61; 15 of 211 genes associated, 7.1%, p-value

= 3.91E-6) (Figure 51A). Cluster analysis of genes within this pathway revealed significant differences between the transcriptomes of hepatocyte-enriched cultures from 5-week-old

Asah1+/+ and Asah1P361R/P361R mice (Figure 51A). Activation of this pathway is likely a key contributor to the massive and progressive inflammatory phenotype seen in the

Asah1P361R/P361R mouse. Within this pathway, vascular cell adhesion molecule 1 (Vcam1) and

184 chemokine (C-X-C motif) receptor 4 (Cxcr4) were significantly upregulated in the hepatocyte-enriched cultures of Asah1P361R/P361R mice (Vcam1= 3.04-fold change, p-value=

1.5E-4) (Cxcr4= 4.87-fold change, p-value 1.5 E-4). We reconfirmed the overexpression of

Vcam1 and Cxcr4 by Real-time PCR in a separate cohort of Asah1P361R/P361R mice (Figure

51C, D).

One of the top deactivated lipid metabolism and injury response pathways from the

IPA analysis was the liver X receptor/retinoid x receptor (LXR/RXR) pathway (z score -3; 18 of 121 genes, 14.9% p-value= 2.75 E-12) (Figure 51B). Genes within this pathway play a role in the homeostasis of cholesterol, and triglyceride metabolism. A number of the apolipoproteins were downregulated. Two of which were Apolipoprotein M (ApoM) (-4.37- fold change, p-value 5.0 E-5) and Apolipoprotein A2 (ApoA2) (-5.78-fold change, p-value

2.5E-4) and. Impaired apolipoprotein activity may contribute to the reduced free fatty acid and triglyceride levels in liver and serum in the Asah1P361R/P361R mouse. Both ApoM and

ApoA2 were validated in a different cohort of samples by Real-time PCR (Figure 51E,F).

185

Figure 51 Misregulated genes in leukocyte extravasation signaling and LXR/RXR activation pathways in Asah1P361R/P361R mice

Hierarchical clustering of significantly altered genes involved in the leukocyte extravasation signaling (A) and LXR/RXR activation (B) pathways were analyzed in 3 Asah1+/+ (left) and 3 Asah1P361R/P361R (right) mice. Genes induced are depicted in red, genes repressed are depicted in blue. Real-time PCR analyses confirm leukocyte extravasation signaling pathway member vascular cellular adhesion molecule 1 (Vcam1) (C), and chemokine (C- X-C motif) receptor 4 (Cxcr4) (D) modulation in a different cohort of 4 Asah1+/+ and 4 Asah1P361R/P361R mice. Real-time PCR analyses confirmed LXR/RXR activation pathway member Apolipoprotein M (ApoM) (E), and Apolipoprotein A2 (ApoA2) (F) and modulation in a different cohort of 4 Asah1+/+ and 4 Asah1P361R/P361R mice. *p<0.05 **p<0.01.

5.4.12 Perturbed sphingolipid homeostasis

As ACDase deficiency directly impacts the sphingolipid pathway we analyzed the genes involved in sphingolipid metabolism in our RNAseq analyses on the hepatocyte- enriched cultures from Asah1P361R/P361R mice. We revealed 146 molecules confirmed to be misregulated (p-value= 8.54 E-7 to 7.42 E-39). Cluster analysis revealed that many of the most significant genes involved in this pathway were dysregulated in the hepatocyte- enriched cultures from Asah1P361R/P361R mice (Figure 52A).

186

Ceramide kinase (CerK) a gene that encodes for the enzyme that converts ceramide to C1P was found to be significantly upregulated by RNAseq analysis and real-time PCR

(4.25-fold change, p-value= 5.00 E-5) (Figure 52A,B). Also upregulated was Psap which encodes for a glycoprotein that generates four different polypeptides known as saposin A, B,

C, and D (Figure 52A,C). The four saposins act as activators of lysosomal hydrolase for sphingolipid catabolism. Additionally, we observed a significant upregulation (1.69-fold change, p-value 1.77 E-05) in ceramide glucosyltransferase (Ugcg) a gene that is involved in the production of glucocerebrosides (Figure 52A,D). In the case of ceramide, saposin D activates ACDase. Upregulation of Psap may occur as a compensation to increase activation of the defective ACDase. Upregulation of these three genes appear to be triggered in response to ceramide buildup.

From our RNAseq analysis, we also identified downregulated genes that encode both ceramide synthase 2 (Cers2) (-0.30-fold change, p-value 4.60 E-12) and Cers4 (-1.47- fold change, p-value 4.0E-5) (Figure 52A,E,F). Ceramide synthases are proteins that catalyze the synthesis of ceramide. Cers2 is the enzyme that primarily synthesizes very long acyl chain (C22-24) ceramides, and Cers4 primarily synthesizes (C18-C22). Both Cers2 and

Cers4 were significantly downregulated in our real-time PCR. Additionally, diacylglycerol acyltransferase 2 (Dgat2) (-3.5-fold change, p-value 5.0-E-5), a gene that encodes for an enzyme that catalyzes triglyceride and acyl ceramide generation was downregulated in both our RNAseq and Real-time PCR reactions (Figure 52A,G). Collectively it is possible that downregulation of these genes may in part explain the alterations in abundance of ceramides within the liver and hepatocytes-enriched cultures from Asah1P361R/P361R mice.

187

Figure 52 Altered genes in sphingolipid metabolism in Asah1P361R/P361R mice

Hierarchical clustering of significantly altered genes involved in the sphingolipid metabolism pathway analyzed in 3 Asah1+/+ (left) and 3 Asah1P361R/P361R (right) mice (A). Genes induced are depicted in red, genes repressed are depicted in blue. Real-time PCR analyses to confirm misregulation of sphingolipid metabolism pathway members ceramide kinase (Cerk) (B), prosaposin (Psap) (C), ceramide glucosyltransferase (Ugcg) (D), ceramide synthase 2 (Cers2) (E), ceramide synthase 4 (Cers4) (F), diacylglycerol o-acyltransferase 2 (Dgat2) (G), in a different cohort of 4 Asah1+/+ and 4 Asah1P361R/P361R mice. *p<0.05 **p<0.01, ***p<0.001.

188

5.5 Discussion and conclusion

The clinical spectrum of ACDase is broad. The most common clinical manifestations include subcutaneous nodules, joint contractures, and a hoarse voice. Beyond that patients also develop neurological, respiratory, and other visceral disease. ACDase deficient mice also develop significant pathology to the hematopoietic, central nervous, and pulmonary systems [Dworski et al. 2015; Yu et al. 2017; Sikora et al. 2017]. In this report, we add to this body of work by documenting the liver pathology of the ACDase deficient mouse. We further catalogued the lipid and sphingolipid profiles and for the first time highlight change in gene expression in the ACDase deficient mouse.

Within the limited literature on FD, the most typical hepatic sign is hepatomegaly where the liver is often palpable during a clinical examination [Farber 1952; Abul-Haj et al.

1962; Samuelsson and Zetterström 1971; Tanaka et al. 1979; Antonarakis et al. 1984]. In one of Dr. Sidney Farber’s original case of three patients, he noted presence of hepatic necrosis, faint cytoplasmic staining in both hepatocytes and Kupffer cells and dilation to the liver sinusoid [Farber et al. 1957]. Other post-mortem reports have also demonstrated significant presence of atypical histiocytes that are highly vacuolated [Antonarakis et al.

1984; Levade et al. 2014; Bao et al. 2017]. Additionally, TEM analyses of liver biopsies have revealed cells containing large intracytoplasmic inclusions many of which contained low electron dense amorphous material [Hoof and Hers 1968; Tanaka et al. 1979; Abenoza and

Sibley 1987]. Our analyses of the ACDase deficient mice have revealed a high degree of similarity. Light microscopy analyses reveal significant inflammation as early as 3 weeks of age. By 9 weeks of age, the liver of Asah1P361R/P361R mice are filled with large foamy macrophages and show presence of tissue fibrosis. Both observations were also present in the spleen and thymus of Asah1P361R/P361R mice [Dworski et al. 2015]. Similarly, our TEM

189 analyses also revealed compelling storage pathology within the recruited macrophages, resident Kupffer cells, and other hepatic cells within both the liver sinusoid and portal fields.

Liver enlargement and histiocytosis are most common in the classical type 1 variant of FD, however the most severe liver pathology documented is in type 4 FD. This variant is the “visceral-neonate” form [Levade et al. 2014]. Patient will exhibit significant organomegaly, visceral histiocytosis and usually do not live past 1 year of age [Levade et al.

2014]. Of the handful of cases available, presence of cholestatic jaundice, liver fibrosis, ascites, and liver injury have also been documented [Salo et al. 2003; Willis et al. 2008;

Nowaczyk et al. 1996]. In one of these atypical cases, a 6-month old infant that showed severe liver pathology was misdiagnosed with neonatal hepatitis. The patient later underwent a liver transplantation which partially normalized liver function [Salo et al. 2003].

Proper diagnosis was only achieved after the appearance of the more common arthritic-like symptoms [Salo et al. 2003]. Cases such as these have expanded the clinical spectrum of

FD, where liver involvement may precede the appearance of nodules. In our study, we also see signs that are present in type 4 FD. Namely, an elevation of liver injury markers, progression of fibrosis, and cell death. While the Asah1P361R/P361R mouse contains a mutation that is orthologous for known patient mutation with classical FD, our data suggests that the pathology seen in the Asah1P361R/P361R mouse can also extend our understanding of the severe variants of FD.

A heighted state of inflammation is characteristic of FD. This phenotype is well replicated in the Asah1P361R/P361R mouse, and severely affects the liver. Ultrastructure analyses revealed presence of storage pathology in various hepatic cell types, however macrophages and Kupffer cells appeared the most affected. Phenotypically these cells contained both larger and a higher quantity of storage vacuoles. We have previously demonstrated that MCP-1 and other inflammatory cytokines are elevated in serum and tissues including the liver [Alayoubi et al. 2013]. Furthermore, we recently reported that

190 genetic ablation of MCP-1 was able to reduce circulating monocytes, decrease ceramides and partially extend the life of the Asah1P361R/P361R mouse [Yu et al. 2018]. This study demonstrated that inhibiting the inflammatory cascade may be a potential therapeutic target.

While identifying therapeutic targets is important, we performed transcriptome analyses to uncover altered genes and pathways to better understand the pathology in FD.

Analyses on the most significantly affected genes from our RNAseq experiment revealed an upregulation to pathways related to inflammation. One of the upregulated genes in the leukocyte extravasation pathway that we validated with Real-time PCR was Vcam1, a gene that functions in leukocyte-endothelial cell adhesion. We recently reported presence of chronic lung injury in the lungs of the Asah1P361R/P361R mice. One of the findings from this study was significant impairment of to the vascular permeability resulting is vascular leakage in various organs including the liver [Yu et al. 2017]. Upregulation of Vcam1 in the liver may contribute to this observation and partly explain for the progressive infiltration of immune cells.

Interestingly, a number of pathways relating to lipid homeostasis were significantly deactivated, which may in part explain for the reduction of TG and FFA in both liver and serum samples. Asah1P361R/P361R mice start decreasing their total body weight starting around

5-weeks of age and accompanied by a reduction in fat [Alayoubi et al. 2013]. While we did not assess adipose tissue content, during necropsy we often saw an absence of fat pads in the visceral, anterior subcutaneous and posterior subcutaneous regions (data not shown).

Additionally, a number of case reports have reported progressive weight loss [Ozaki et al.

1978; Amirhakimi et al. 1976; Fiumara et al. 1993]. In fact, one of these patients with the mild variant of FD was described lacking cutaneous fat [Fiumara et al. 1993]. One of significantly downregulated pathways was the LXR/RXR activation, which plays a role in cholesterol absorption and excretion. Within this pathway, a number of genes for apolipoproteins were downregulated. ApoM and ApoA2 encode for lipoproteins found in

191 high-density lipoprotein (HDL) particles. Deactivation of these pathways would explain the change in lipid content in the Asah1P361R/P361R mice.

Disturbance in lipid homeostasis has also been demonstrated in other mutant models of sphingolipid metabolism. One study on the S1P lyase deficient mouse demonstrated elevated lipid storage in both serum and liver, but decreased adiposity

[Bektas et al. 2010]. Another report on the acid sphingomyelinase and LDL receptor double deficient mouse demonstrated reduced hepatic TG accumulation and less fat accumulation compared to controls when placed on a high fat diet [Deevska et al. 2009]. Furthermore, a series of studies have demonstrated the relationship of adiponectin in regulating ceramidase activity in the maintenance of insulin resistance [Summers and Nelson 2005; Holland et al.

2007; Holland et al. 2011]. One recent study from that same group demonstrated that overexpressing ACDase in the liver led to improved insulin sensitivity and prevented hepatic steatosis in mice on a high fat diet [Xia et al. 2015]. While more work will be required to understand this mechanism, collectively these studies and ours, demonstrates that insults to the sphingolipid pathway can lead to functional consequences that cross over to other lipid metabolic pathways.

Lipidomic analyses on our liver lysates and hepatocytes revealed significant accumulation of total ceramide, SM and MHC. Assessment of the abundance of the ceramides revealed a proportional increase in the C16:0/SM16:0 and C18:0/SM18:0 acyl chain sphingolipids and a reduction in the C22:0/SM22:0 and C24:0/SM24:0 acyl chain sphingolipids in the liver of Asah1P361R/P361R mice. This alteration in long chain versus very long chain species is specific to the liver. Previous analyses of ceramides in the lung did not reveal alterations in acyl chain composition, and another study on the brain of

Asah1P361R/P361R mice revealed the opposite where there was an increase in the C24:0 species and a decrease in the C18:0 species [Sikora et al. 2017; Yu et al. 2017]. These differences in ceramide acyl chain distribution are partly due to the expression of the 6 CerS

192

[Mullen et al. 2012]. Our data revealed a significant reduction in Cers2 and Cers4 expression. In normal animals, both Cers2 and Cers4 have high expression in the mammalian liver [Levy and Futerman 2010]. Cers2 is important for the synthesis of the VLC ceramides which include C20-C26 whereas Cers4 are important for the formation of C18-

C22 ceramides [Levy and Futerman 2010]. Decreased expression of Cers2 and Cers4 may partly explain the reduced abundance of C22:0 and C24:0 species. Work on the Cers2 deficient mouse revealed a similar pattern where the C16:0 species were significantly elevated to account for the reduction in C22:0 and C24:0 ceramides. Analyses of the liver of these mice also revealed some similarities in liver pathology such as increased cell death, increase in injury markers, inflammation and fibrosis [Pewzner-Jung et al. 2010].

Ceramide has often been attributed to being pro apoptotic while S1P as pro survival

[Maceyka et al. 2002]. Particular to this rheostat is the role of C16:0 ceramide in the promotion of apoptosis [Thomas et al. 1999]. One study demonstrated that overexpression of neutral ceramidase (NCDase) led to a protective effect from TNFα induced apoptosis in rat hepatocytes. Not only was C16:0 reduced, but an increase in S1P levels was detected

[Osawa et al. 2005b]. Another report showed that knockdown of Cers2 led to increased cell death in HeLa cells treated with cisplatin, and a shift in sphingolipid composition from C24 to

C16 [Sassa et al. 2012]. With respect to our study, the increased cell death noted maybe in part be due to the increased abundance of C16:0 ceramide. Furthermore, while S1P was not detected, we did see a significant decrease of SM in liver lysates from Asah1P361R/P361R mice.

The decrease in SM from defective ceramide breakdown would presumably lead to less S1P and consequently increasing susceptibility for cell death.

Lastly, we demonstrated a significant upregulation in Cerk and Ugcg. Cerk converts ceramide to C1P, and Ugcg is an important enzyme in the biosynthesis of glucosylceramides. This upregulation may explain for the increased level of C1P and MHC in liver lysates. Additionally, Psap which encodes for a precursor glycoprotein that cleaves

193 into 4 sphingolipid activator proteins collectively called saposin A-D [Bradova et al. 1993].

Since saposin D acts as a stimulant for ACDase, it is possible that an increase in Psap is stimulated in order to boost enzyme activity [Azuma et al. 1994]. Additionally, since a pan increase in sphingolipids was detected, upregulation of Psap may impact the activity of other sphingolipid metabolizing enzymes. Taken together, these results demonstrate that mutations to Asah1 not only lead to sphingolipid buildup, but also significant alteration in genes that control sphingolipid homeostasis.

Our study has provided a comprehensive analysis of hepatic pathology in the

ACDase deficient mouse. We have highlighted parallels between human cases and our model particularly in terms of liver injury, inflammation, and ultrastructure pathology. Due to the lack of available patient tissues, our mouse model has aimed to fill this gap. Our data demonstrates a progressive inflammation, fibrosis leading to increased liver injury, and cell death. Our transcriptome analyses provide insights into key pathway changes in the

Asah1P361R/P361R mice. These insights although exploratory provide an understanding on the pathogenesis and the perturbed metabolic regulation present in the Asah1P361R/P361R mice.

194

Chapter 6

MCP-1 Impedes Pathogenesis of Acid Ceramidase Deficiency

This work is adapted from Yu et al., 2018 with permission of Scientific Reports

6.1 Abstract

Farber Disease (FD) is an ultra-rare Lysosomal Storage Disorder caused by deficient acid ceramidase (ACDase) activity. Patients with ACDase deficiency manifest a spectrum of symptoms including formation of nodules, painful joints, and a hoarse voice.

Classic FD patients will develop histiocytes in organs and die in childhood. Monocyte chemotactic protein (MCP-1; CCL2) is significantly elevated in both FD patients and a mouse model we previously generated. Here, to further study MCP-1 in FD, we created an ACDase;MCP-1 double mutant mouse. We show that deletion of MCP-1 reduced leukocytosis, delayed weight loss, and improved lifespan. Reduced inflammation and fibrosis were observed in livers from double mutant animals. Bronchial alveolar lavage fluid analyses revealed a reduction in cellular infiltrates and protein accumulation.

Furthermore, reduced sphingolipid accumulation was observed in the lung and liver but not in the brain. The neurological and hematopoietic defects observed in FD mice were maintained. A compensatory cytokine response was found in the double mutants, however, that may contribute to continued signs of inflammation and injury. Taken together, targeting a reduction of MCP-1 opens the door to a better understanding of the mechanistic consequences of ceramide accumulation and may even delay the progression of FD in some organ systems.

195

6.2 Introduction

The gene ASAH1 encodes acid ceramidase (ACDase, EC 3.5.1.23), a hydrolase that degrades the bioactive lipid ceramide into sphingosine and a free fatty acid [Levade et al.

2014]. Mutations in ASAH1 can result in acid ceramidase deficiency, which may manifest as

Farber Disease (FD, Farber Lipogranulomatosis, OMIM #22000), a rare inherited Lysosomal

Storage Disorder (LSD). While patients who are diagnosed with FD can display a range of symptoms, the cardinal signs are painful joints, formation of subcutaneous nodules, and the development of a hoarse voice that may lead to aphonia. In severe cases, patients may have neurologic involvement, develop hepatosplenomegaly, and fail to thrive, leading to childhood lethality [Ehlert et al. 2007; Ahmad et al. 2009]. We previously generated a mouse model of ACDase-deficiency that has high fidelity with the human disorder [Alayoubi et al.

2013]. In this model, the P361R mouse mutation (which is orthologous to the human P362R mutation) was knocked in [Li et al. 1999]. Mice homozygous for the mutation

(Asah1P361R/P361R) recapitulate the classical form of human FD with a phenotype that includes systemic ceramide accumulation, increased leukocytosis, and enlargement of visceral organs [Alayoubi et al. 2013]. The mice are much smaller than littermates and progressive macrophage infiltration is present in the pulmonary, hematopoietic, hepatic, and central nervous systems [Alayoubi et al. 2013]. Large foamy macrophages and associated tissue injury is also commonly observed.

The mouse model of FD that we developed also demonstrates a unique cytokine profile that includes dramatically higher levels of MCP-1 in plasma [Dworski et al. 2017].

Further, we have also recently shown that FD patients have increased levels of MCP-1, keratinocyte chemoattractant (KC), macrophage inflammatory protein-1α (MIP-1α), and the inflammatory cytokine interferon gamma-induced protein-10 (IP-10) in their plasma [Dworski

196 et al. 2017]. In the mouse, MCP-1 is the most elevated and its highest concentration is detected at the end of the animal's life. MCP-1 is a potent chemoattractant that is responsible for the recruitment of monocytes. It is a member of the C-C chemokine subfamily and functions by binding to its G-protein coupled receptor: chemokine receptor 2

(CCR2). Both MCP-1 and CCR2 have been widely studied for their role in monocyte recruitment during infection and for their other roles during inflammation [Deshmane et al.

2009]. Expression of MCP-1 can be detected in a variety of cell types, including endothelial, smooth muscle, astrocytic, and microglial cells [Cushing et al. 1990; Standiford et al. 1995].

However, the major sources of MCP-1 are monocytes and macrophages [Yoshimura et al.

1989]. Numerous FD case reports have demonstrated that histiocytic infiltration is present in many tissues and that plasma chitotriosidase is significantly elevated [Ehlert et al. 2007;

Saygi et al. 2015]. This inflammatory phenotype is reflected in our Asah1P361R/P361R mouse, which displays macrophage and neutrophil infiltration in various tissues and organs

[Alayoubi et al. 2013; Dworski et al. 2015]. In clinical practice, the management strategy for the treatment of FD focuses largely on managing patient symptoms to date. Hematopoietic stem cell transplantation (HSCT) may be an option for certain Farber patients as well as enzyme replacement therapy (ERT) [He et al. 2017]. Several case reports have demonstrated that BMT reduced the number and size of subcutaneous nodules and decreased joint pain in patients with FD [Yeager et al. 2000; Vormoor et al. 2004]. Along these lines, we previously demonstrated that MCP-1 levels are reduced in FD patient plasma post-BMT [Dworski et al. 2017]. This further suggests that there may be a relationship between MCP-1 and FD symptoms, and supports MCP-1 as a prospective biomarker for FD.

Here, we generated and characterized an Asah1 and MCP-1/CCL2 double deficient mouse (herein called Asah1P361R/P361R;MCP-1-/-) to further understand the role of MCP-1 in

197

FD development and progression. In this study, we show that ablation of MCP-1 can impede symptoms of FD. This is an important insight into the pathogenesis of this disorder and opens the door to a whole new line of therapy.

198

6.3 Material and Methods

6.3.1 Animal use, breeding and genotyping

To generate homozygous Asah1P361R/P361R mice, we crossed Asah1+/P361R heterozygotes as previously reported [Alayoubi et al. 2013]. CCL2tm1Rol (MCP-1-/-) knockout mice [Lu et al. 1998] were purchased from Jackson Laboratory (Bar Harbor, MA). Male

MCP-1-/- mice were crossed to female Asah1+/P361R mice to generate doubly heterozygous animals (Asah1+/P361R;MCP-1+/-). These doubly heterozygous mice were mated with MCP-1-/- mice to generate Asah1+/P361R;MCP-1-/- mice. Subsequently Asah1+/P361R;MCP-1-/- mice were mated with each other to generate Asah1P361R/P361R and MCP-1-/- homozygous double mutants (Asah1P361R/P361R ;MCP-1-/-). Genotypes were confirmed via PCR using genomic

DNA from ear notches. For MCP-1 genotyping, a set of three primers were used, 5-GCC

AGA GGC CAC TTG TGT AG-3, 5-TGA CAG TCC CCA GAG TCA CA-3, and 5-TCA TTG

GGA TCA TCT TGC TG-3, yielding a 287 bp product for the wild-type gene and a 179 bp product for the knocked-out gene. To detect the wild-type allele of Asah1, we used the primers 5-CAG AAG GTA TGC GGC ATC GTC ATA C-3 and 5-AGG GCC ATA CAG AGA

AAC CCT GTC TC-3, which yielded a 379 bp product. For the Asah1 knock-in allele, we used the primers 5-TCA AGG CTT GAC TTT GGG GCA C-3 and 5-GCT GGA CGT AAA

CTC CTC TTC AGA CC-3, which amplifies a 469 bp product from the neomycin resistance cassette. All animal procedures were approved and carried out in strict adherence to the policies of the University Health Network (UHN) Animal Care Committee and the Medical

College of Wisconsin (MCW) Institutional Animal Care and Use Committee (IACUC).

199

6.3.2 Animal data and organ weights

Mice survival was recorded for Kaplan-Meier curves. For organ weight, mice were euthanized with CO2 gas and immediately weighed. Organs were collected immediately and weighed. Brain edema was determined by brain wet-to-dry weight ratio. Whole brains were weighed to determine wet weight and then were placed in a laboratory oven (Qualtech

Industries Denver, CO) to evaporate for 48 hours. Dried samples were weighed to obtain dry weight. Skin stretch was quantified with a digital Vernier caliper (Fisher Scientific, Waltham,

MA) by measuring the height of the tent formed when scruffing the mouse.

6.3.3 Peripheral blood and serum biochemistry, ELISA, and cytokine analyses

Peripheral blood was collected by cardiac puncture into EDTA-coated microtainers

(BD, Biosciences Canada, Mississauga, Canada). Blood samples were inserted into a

Hemavet (Drew Scientific Group, Waterbury CT) for complete blood count (CBC) analyses.

For serum separation, mouse blood was collected via cardiac puncture into serum separator

SST microtainers (BD Biosciences). Sample tubes were gently inverted, and the blood was allowed to clot at room temperature for 30 minutes. Samples were then centrifuged at 1200 x g for 10 minutes. The serum portion was immediately collected and stored in -80 °C until use. Mouse cytokine levels were measured on collected serum with the Cytokine-20-Plex mouse panel (Thermo Scientific Pierce, Waltham, MA) following the manufacturer's instructions. Luminescence was measured on the Luminex 100 instrument (Luminex, Austin,

TX). Data with a low bead count (<45 beads) was omitted. Additionally, we measured MCP-

3 and MCP-5 in mouse serum with MCP-3 and MCP-5 ELISA kits (Thermo Scientific,

Waltham, MA). Serum metabolites and liver enzymes in serum were measured using the

200

VetScan Comprehensive Diagnostic Profile and Mammalian Liver Profile (Abaxis Union City,

CA) on the Abaxis VetScan VS2 (Abaxis Union City, CA). An AST ELISA kit (Cloud-Clone

Corp. Wuhan China) was used to assay for serum amino aspartate transferase (AST) levels.

6.3.4 BALF turbidity, cytospin & differential

A 20-gauge catheter (BD, Biosciences) was inserted into the trachea of CO2- euthanized mice and secured. The lungs were flushed three times by gentle washing with

1.0 ml of ice-cold phosphate buffered saline (PBS). BALF was collected in 1.5 ml tubes and centrifuged at 300 x g at 4 °C for 10 minutes. The supernatant was removed and stored at -

80 °C. BALF supernatant was measured for turbidity on the NanoDrop One (Thermo

Scientific Wilmington DE) at OD600. Protein concentration of BALF supernatant was measured using a commercially available bicinchoninic acid (BCA) assay kit (Thermo

Scientific Pierce, Waltham, MA). Surfactant proteins A, B, C, and D in BALF supernatant were measured using ELISA kits for mouse material (Cloud-Clone Corp. Wuhan, China) as per the manufacturer’s instructions.

For cell analyses, BALF cell pellets were resuspended in 500 µl of PBS and varying dilutions were used to measure total cell counts, which were performed on trypan blue- stained samples on the Countess II FL automated cell counter (Life Technologies, Carlsbad,

CA). The remainder of the resuspended cells were diluted in various concentrations to a total volume of 200 µl with PBS and a cytospin was performed with the Shandon CytoSpin

III Cytocentrifuge (Thermo Shandon, Waltham, MA) for 5 mins at 300 g. For cell type analyses, cytospins that contained 200,000 cells per slide were stained with the Kwik-Diff kit

(Thermo Scientific Pierce, Waltham, MA). Differential cell counts were obtained by averaging counts from 10 non-overlapping zones with a total of 200 cells scored in each zone.

201

6.3.5 Histopathology and immunohistochemistry

As mentioned above, mice were euthanized by CO2 inhalation. Cardiac perfusion with ice- cold PBS was performed with a 24 G needle and, following necropsy, organs were immediately fixed in 10% phosphate-buffered formalin for 24-48 hours. Organs were subsequently embedded in paraffin and sectioned at 4 μm. Dissected tibias were decalcified with Cal-Ex II fixative/decalcifier per the manufacturer's instructions (Fisher Scientific.) prior to embedding and sectioning. For analysis of the hematopoietic system, paraffin-embedded thymus, tibia, and spleen samples were stained with hematoxylin and eosin (H&E).

Immunohistochemistry (IHC) was then performed with the following primary antibodies:

B220 (BD Biosciences #550286; 1:100), CD3 (Sigma #C7930; 1:1000), and Mac-2 for myeloid cells (Cedarlane #CL8942AP; 1:2000). For liver analyses, tissue sections were stained for H&E and Masson’s trichrome. Liver IHC was performed using the following primary antibodies: anti-mouse neutrophil, clone 7/4 (Cedarlane, Burlington, Canada) and rat anti-mouse Mac-2 (Galectin-3) clone M3/38 (Cedarlane). For brain analyses, tissue sections were stained with H&E and IHC was performed using the following primary antibodies: rabbit anti-ionized, calcium-binding adaptor molecule 1 (Iba-1) (Wako Chemicals

USA, Cambridge, MA) and chicken anti-glial fibrillary acidic protein (GFAP) (Aves Lab Inc,

Tigard, OR). The following secondary antibodies and kits were then used to detect primary antibodies: rabbit anti-rat IgG, biotinylated (Vector Laboratories, Burlingame, CA); goat anti- rabbit IgG, biotinylated (Vector Laboratories); goat anti-rat IgG, biotinylated (Vector

Laboratories); donkey anti-chicken IgG (Jackson ImmunoResearch USA, West Grove PA); donkey anti-rabbit IgG, biotinylated (ImmunoResearch); avidin-biotin/HRP (Vector

Laboratories ); DAB kit (Vector Laboratories) and Vectastain Elite ABC kit (Vector

Laboratories). Histology slides were scanned on the Aperio AT2 histology slide scanner

(Leica Biosystems, Buffalo Grove, IL) or NanoZoomer 2.0-HT histology slide scanner

202

(Hamamatsu Photonics, Ichinocho, Japan). Scanned micrographs were analyzed with

Aperio ImageScope analysis software (Leica Biosystems, Grove, IL).

6.3.6 Flow cytometry of hematopoietic cells

Flow cytometry was performed on bone marrow, spleen, and thymus samples. For the latter two, cells were forced through a 40 µm nylon cell strainer in PBS with 2% fetal calf serum (FCS). For the bone marrows, cells were collected by flushing femurs and tibias with

PBS containing 2% FCS. After RBC lysis, collected cells were washed and resuspended in

PBS with 2% FCS and counted on a hemocytometer. For fluorescence-activated cell sorting

(FACS) analyses, cells were stained for 30 mins at 4° C with the following antibodies: CD3

FITC (BioLegend San Diego, CA; 17A2), CD4 PE (BD Biosciences; GK1.5), CD8 APC (BD

Biosciences; 53-6.7), CD19 BV605 (Biolegend; 6D5), CD43 APC (Biolegend; S11), IgD e450 (eBiosciences; 11-26 Waltham, MA), IgM PE-Cy-7 (Biolegend; RMM-1), B220 PE

(eBiosciences RA3-6B2), Gr-1 FITC (Biolegend RB-6-8C5), Mac1 PE (BD Biosciences;

M1/70), and CD11c APC (eBiosciences; N418). FACS analysis was performed on a LSR

Fortessa flow cytometer (BD Biosciences) with FACSDiva software (BD Biosciences). Data was analyzed with FlowJo software (Tree Star Inc, Ashland, OR).

6.3.7 Mass spectrometry for sphingolipids

Brain, liver, and lung tissue samples were homogenized in 500 µl PBS with the Omni

Bead Raptor 24 tissue homogenizer (Omni International, Inc., Kennesaw, GA) using 2.8 mm ceramic beads. Lipids were extracted from 50 µl of tissue lysate with 200 µl isopropanol.

Two sets of lipid mass spectrometry analyses were performed. In the first experiment; the following internal standards were used for sphingolipid measurements: ceramide 100 ng

203

(d18:1/ 22:0) d4 (Medical University of South Carolina (MUSC) Lipidomics Core, Charleston,

SC); monohexosylceramide 100 ng (d18:1/17:0) (Avanti Polar Lipids Inc., Alabaster, AL); sphingomyelin, 1000 ng (d18:1/17:0) (Avanti Polar Lipids Inc.); and ceramide-1-phosphate

(C1P) (d18:1/16:0), (d18:1/24:0) and (d18:1/24:1) (Matreya Inc., Pleasant Gap, PA); (Avanti

Polar Lipids Inc.) [Bielawski et al. 2010]. For the second experiment the following internal standards were used: 100 ng of d7-Sphingosine (Sph) (Avanti Polar Lipids Inc.); and 100 ng of d7-sphingosine-1-phosphate (S1P) (Avanti Polar Lipids Inc.). Relative quantification (i.e. peak: area ratio) was obtained for the analytes in comparison to those from their corresponding internal standards.

Samples were analyzed on the Shimadzu 20AD HPLC system using reverse-phase

C18 HPLC columns (Agilent Co., Santa Clara, CA) and a Leap PAL autosampler coupled to a triple quadrupole mass spectrometer (API-4000: Applied Biosystems, Carlsbad, CA) operated in MRM (multiple reaction mode) at the MUSC Lipidomics Core. Positive-ion ESI mode was used to detect all sphingolipids. Tissue extraction samples were injected in duplicate for data averaging. The Analyst 1.5.1 software was used for data analysis (Applied

Biosystems). Data are presented as fold change relative to the averaged sphingolipid values detected in samples from 8-9-week-old Asah1+/+;MCP-1+/+ mice.

6.3.8 Quantification and image analyses

For the quantitative evaluation of GFAP and Iba-1 staining in brain samples, five non-overlapping images (20x magnification, 485 µm x 285 µm) for each anatomical region of interest were obtained from scanned micrographs for each slide. Three consecutive brain sections per mouse were stained, scanned, and analyzed. All quantitation was performed with Fiji-ImageJ analysis software (National Institutes of Health, Bethesda, MD).

204

6.3.9 Behavioral tests

Motor coordination and muscle strength were tested with an accelerating rotating rod

(rotarod) and grip strength meter. Circadian rhythms were taken into account; all tests were performed between 8 am and 12 pm. To reduce non-study related variability, all tests were performed with the same apparatus in the same room. For the IITC Rotarod (IITC Inc. Life

Science, Woodland Hills CA), mice were placed on the device and subjected to acceleration from 4 rpm to 40 rpm for a total of 5 minutes. The time when the mice fell off - measured from start - was then recorded. Animals were subjected to three trials per day for two consecutive days; with the same individual performing all tests. For the grip strength test, each animal was weighed prior to the trial. Both total-limb grip strength and fore-limb grip strength were measured with a digital grip strength meter (Columbus Instruments,

Columbus, OH). For this assay, mice were placed on the metal grid and allowed to grip with either all total limbs or just forelimbs. After confirmation of grip, the mouse tail was pulled posteriorly parallel to the grid at a constant rate until release. The grip strength, measured by Gram-force, was recorded for three consecutive trials and then averaged. The grip strength was expressed as mean grip strength relative to weight Gram-force/gram body weight for both total limbs and forelimbs.

6.3.10 Statistical analyses

Data are expressed as means ± standard error. Survival significance was tested with the log-rank (Mantal-Cox) test comparing pairs of curves. All subsequent data was analyzed with a 1-way ANOVA followed by a Tukey post-test. All statistics were analyzed using

GraphPad Prism 5.0 (GraphPad Software Inc, La Jolla, CA). Significant differences are expressed in the figures as *p < 0.05, **p < 0.01, and ***p < 0.001.

205

6.4 Results

6.4.1 Deletion of MCP-1 improves the course of FD

Asah1P361R/P361R;MCP-1+/+ mice demonstrated an expected life span of 7-9 weeks

[Alayoubi et al. 2013]. Asah1P361R/P361R;MCP-1-/- mice displayed a significantly increased lifespan to a median age of 13 weeks (Figure 53A). Asah1P361R/P361R;MCP-1-/- mice also weighed more between 5 and 6 weeks of age than Asah1P361R/P361R;MCP-1+/+ mice (Figure

53B-C). No differences in body weight were found after 7 weeks of age, however, between

Asah1P361R/P361R;MCP-1+/+ and Asah1P361R/P361R;MCP-1-/- mice (Figure 53B and D).

Asah1P361R/P361R;MCP-1+/+ mice also develop a tight skin phenotype that renders them difficult to restrain [Lopez-Vasquez et al. 2016]. Asah1P361R/P361R;MCP-1+/+ and

Asah1P361R/P361R;MCP-1-/- mice skin stretch length was found to be significantly different at 5 weeks of age (Figure 53E). No differences were detected in lifespan, weight, or skin stretch length between Asah1P361R/P361R;MCP-1+/- and Asah1P361R/P361R;MCP-1+/+ mice (Figure 54), therefore the remainder of this study focused exclusively on homozygous

Asah1P361R/P361R;MCP-1-/- mice.

Organomegaly is a common phenotype present in classical FD and the

Asah1P361R/P361R;MCP-1+/+ mouse [Alayoubi et al. 2013]. Deletion of MCP-1 resulted in a delay in organ weight accumulation when normalized to body weight (Figure 53). In terms of absolute organ weights, only the thymus showed a significant change in 8-9-week-old

Asah1P361R/P361R;MCP-1-/- mice when compared to age-matched controls (Figure 55). The weight differences were only revealed when the organ weight was normalized with body weight. The kidney and liver weight as a percentage of total body weight in the

Asah1P361R/P361R;MCP-1-/- mice were within the normal ranges at 8-9-weeks of age (Figure

53,L).

206

There was a decrease in brain, thymus, lung, and spleen weights as a percentage of total body weight when compared to Asah1P361R/P361R;MCP-1+/+ mice (Figure 53F,G,I,K). Other than an increase in brain weight as a percentage of body weight (Figure 53F), there were no other weight increases between 8-9 and 11-12 week-old Asah1P361R/P361R;MCP-1-/- mice

(Figure 53G-L).

Figure 53. Increased survival and weight gain in Asah1P361R/P361R ;MCP-1-/- mice.

(A) Kaplan-Meier survival plot (n=10-15 per genotype), (B) Growth curve measured in weight versus age (n=10 per genotype), (C) 5-week-old weight by genotype, (D) 7-week-old mice weights by genotype, (E) Skin stretch curves measured as length versus age (n=10 per genotype). Organ weight measurements as a percentage of body weight for brain (F), thymus (G), heart (H), lung (I), kidney (J), spleen (K), and liver (L). n=10 samples were analyzed from 8-9-weeks-old mice for each genotype and from 11-12-week-old Asah1P361R/P361R;MCP-1-/- mice. All comparisons were made between 8-9-week-old mice for each genotype and samples from 11-12-week-old Asah1P361R/P361R;MCP-1-/- mice. All comparisons were made between 8-9-week-old Asah1+/+;MCP-1+/+, ns (not significant), *p<0.05, **p<0.01, ***p<0.001.

207

Figure 54. No improvement of disease course in Asah1P361R/P361R ;MCP-1+/- heterozygous mice

Kaplan-Meier survival plot (n=10 per genotype) (A), growth curve measured in weights versus age (n=10 per genotype) (B), Skin stretch curves measured as length versus age (n=10 per genotype) (C).

Figure 55. Absolute organ weights of mice are largely unchanged in Asah1P361R/P361R ;MCP-1- /- mice

Absolute organ weights for brain (A), thymus (B), heart (C), lung (D), kidney (E), spleen (F), and liver (G). n=10 samples were analyzed from 8-9-weeks-old mice for each genotype and from 11-12-week-old Asah1P361R/P361R;MCP-1-/- mice. All comparisons were made between 8-9-week-old mice for each genotype and samples from 11-12-week-old Asah1P361R/P361R;MCP-1-/- mice. All comparisons were made between 8-9-week-old Asah1+/+;MCP-1+/+, ns (not significant), *p<0.05, **p<0.01, ***p<0.001.

208

6.4.2 Peripheral blood cell counts

Complete blood counts (CBCs) from peripheral blood draws of 8-9-week-old

Asah1P361R/P361R;MCP-1-/- mice showed a normalization in circulating red blood cells (RBC) and white blood cells (WBC) compared to 8-9-week-old Asah1P361R/P361R;MCP-1+/+ mice

(Figure 56A-B). However, by 11-12 weeks of age the numbers of WBCs in

Asah1P361R/P361R;MCP-1-/- mice were similar to that of 8-9-week-old Asah1P361R/P361R;MCP-1+/+ mice (Figure 56B). Absolute lymphocyte cell counts were unchanged between genotypes; however, based on the differential cell counts of WBCs, lymphocyte percentages were significantly decreased between controls, Asah1P361R/P361R;MCP-1+/+, and

Asah1P361R/P361R;MCP-1-/- mice (Figure 56C-D). Increased monocytes, neutrophils, and eosinophils are common in 8-9-week-old Asah1P361R/P361R;MCP-1+/+ mice [Alayoubi et al.

2013]. A reduction in monocytes was found in both the absolute and differential cell counts between the 8-9-week-old Asah1P361R/P361R; MCP-1+/+ mice and the 8-9 and 11-12-week-old

Asah1P361R/P361R;MCP-1-/- mice (Figure 56E-F). Neutrophils cell counts were not statistically significant between 8-9-week-old Asah1P361R/P361R;MCP-1+/+ and 8-9 and 11-12

Asah1P361R/P361R; MCP-1+/+ mice (Figure 56G-H). Basophil cell counts did not change significantly either (Figure 56I-J). Absolute eosinophil cell counts were similar to control mice in 8-9-week-old Asah1P361R/P361R;MCP-1-/- mice, but were elevated in 11-12-week-old

Asah1P361R/P361R;MCP-1-/- mice (Figure 56K). However, based on the differential cell counts, the percentage of eosinophils did not differ significantly between any of the genotypes

(Figure 56L).

209

Figure 56. MCP-1 deletion impedes leukocytosis in ACDase-deficiency.

Complete blood count (CBC) measurements showing total red blood cell (RBC) count (A), total white blood cell (WBC) count (B), lymphocyte count (C), lymphocyte differential (D), monocyte count (E), monocyte differential (F), neutrophil count (G), neutrophil differential (H), basophil count (I), basophil differential (J), eosinophil count (K), and eosinophil differential (L). n=10 samples were analyzed from 8-9-week-old mice for each genotype and 11-12-week-old Asah1P361R/P361R;MCP-1-/- double mutant mice. All comparisons were made between 8-9-week- old Asah1+/+;MCP-1+/+, Asah1P361R/P361R;MCP-1+/+, Asah1P361R/P361R;MCP-1-/- and 11-12-week-old Asah1P361R/P361R;MCP-1-/- mice. ns (not significant), *p<0.05, **p<0.01, ***p<0.001.

210

6.4.3 MCP-1 deletion does not normalize hematopoiesis

We previously reported that the bone marrow of Asah1P361R/P361R;MCP-1+/+ mice displayed a dramatic reduction in the pre-B, pro-B, immature-B, and transitional-B cells

[Dworski et al. 2015]. FACS analysis of cells from the bone marrow of 9-week-old

Asah1P361R/P361R;MCP-1-/- mice also showed a reduction in B-cell progenitors (Figure 57A).

There were no differences in the various B-cell progenitor populations between 9-week-old

Asah1P361R/P361R;MCP-1+/+ and Asah1P361R/P361R;MCP-1-/- mice (Figure 57C). H&E staining of

9-week-old Asah1P361R/P361R;MCP-1-/- and both 9- and 12-week-old Asah1P361R/P361R;MCP-1-/- tibias revealed a pale architecture compared to both Asah+/+;MCP-1+/+ and Asah1+/+;MCP-1-/- samples. These light H&E areas were stained positive for Mac-2 but not for B220 and CD-3, which demonstrate myeloid lineage-mediated infiltration (Figure 57B). Similarly, FACS analysis of T-cells from thymic tissue of 9-week-old Asah1P361R/P361R;MCP-1-/- mice demonstrated a significant reduction in CD4+;CD8+ T-cell numbers compared to

Asah1P361R/P361R;MCP-1+/+ mice with no differences seen between both genotypes in the various T-cell subsets (Figure 57D and Figure 58). Histological analysis performed on 9- week-old thymic tissue obtained from Asah1P361R/P361R;MCP-1-/- mice showed pale H&E staining that was additionally characterized by the presence of Mac-2 staining. Tissue injury and macrophage infiltration appear progressive as the staining of tissue from 12 weeks of age Asah1P361R/P361R;MCP-1-/- mice became more pale for H&E and more intense for Mac-2

(Figure 58). FACS analysis of cells from 9-week-old spleen tissue of Asah1P361R/P361R;MCP-1-

/- mice showed a significant reduction in CD11b+ and Gr1+ granulocytes compared to cells from 9-week-old Asah1P361R/P361R;MCP-1+/+ mice (Figure 57E and Figure 59).

211

The spleen parenchyma in the 9-week-old Asah1P361R/P361R;MCP-1+/+ mice is also highly disrupted by the presence of foamy macrophages and demonstrates a reduction in

B220+ and CD3+ cells (Figure 59). This phenotype is present in the 9-week-old

Asah1P361R/P361R;MCP-1-/- mice but to a lesser extent (Figure 59). However, by 12 weeks of age the phenotype becomes exacerbated and mirrors that typically seen in the spleen of 9- week-old Asah1P361R/P361R;MCP-1+/+ mice (Figure 59). MCP-1 deletion also appears to decrease the rate of tissue destruction in the spleen.

Figure 57. Perturbed hematopoiesis is retained in Asah1P361R/P361R ;MCP-1-/- mice

(A) FACS plots showing B-cell lineage staining in BM from 9-week-old mice of all genotypes. (B) Histology staining of BM with H&E along with anti-B220, anti-CD-3, and anti-Mac-2 antibodies. (C) The absolute number of Pro-B, Pre-B, immature B, transitional B, and mature B cells in mice bone marrow (BM). (D) The absolute number of CD11b+ and Gr-1+ cells in mouse BM, thymus, and spleens. (E) The absolute number of CD4+ CD8-, CD4- CD8+, and CD4+ CD8+ cells in mouse thymus. 9- and 12-week-old mice were used for the histopathology analyses. Original magnification of BM micrographs at 10x where scale bar represents 300µm. Spleen and thymus micrographs are at 4x magnification and the scale bar represents 100µm. n=3-4 mice at 9 weeks of age were used for FACS analysis and cell counts. ns (not significant), *p<0.05 **p<0.01, ***p<0.001.

212

Figure 58. Absence of T-cell population in Asah1P361R/P361R ;MCP-1-/- mice

Results of flow cytometry analyses performed on thymus tissue from 9-week-old mice of all genotypes following staining for CD4 and CD8 cells (A). Thymus histology from 9-week-old mice of all genotypes and from 12-week- old Asah1P361R/P361R ;MCP-1-/- mice. Tissues were stained for H&E, B220, CD-3, and Mac-2 (B). n=3-4 mice at 9 weeks of age were used for FACS analysis and cell counts.

213

Figure 59. Mild impedance of granulocyte infiltration in Asah1P361R/P361R ;MCP-1-/- mice

Results of flow cytometry analyses performed on spleen tissue from 9-week-old mice of all genotypes following staining for CD11b and GR-1 cells (A). Spleen histology from 9-week-old mice of all genotypes and from 12- week-old Asah1P361R/P361R ;MCP-1-/- mice. Tissues were stained for H&E, B220, CD-3, and Mac-2 (B). n=3-4 mice at 9 weeks of age were used for FACS analysis and cell counts.

214

6.4.4 Liver inflammation and injury makers are reduced

Histological analyses (via H&E, Trichrome and IHC staining for Mac-2) demonstrated significant tissue disruption, fibrosis, and macrophage infiltration in liver tissue from 8-9- week-old Asah1P361R/P361R;MCP-1+/+ mice compared to controls (Figure 60A-C). The infiltration and fibrosis present in 8-9-week-old Asah1P361R/P361R;MCP-1+/+ mouse livers was more pervasive compared to samples from 8-9-week-old Asah1P361R/P361R;MCP-1-/- animals

(Figure 60C & D). The formation of large, foamy histiocytes and fibrotic regions increased by

11-12 weeks of age in Asah1P361R/P361R;MCP-1-/- mice (Figure 60E) but not to the extent seen in Asah1P361R/P361R;MCP-1+/+ mice. Albumin levels in serum showed a significant decrease in samples from Asah1P361R/P361R;MCP-1+/+ mice when compared to those from aged-matched controls (Figure 60F). Ablation of MCP-1 normalized albumin levels in 8-9-week-old

Asah1P361R/P361R;MCP-1-/- mice, which was later lost in 11-12-week-old animals when compared to 8-9-week-old Asah1+/+;MCP-1+/+ mice (Figure 60F). Although a decrease was detected in 11-12-week-old double mutants, there was no significant change when compared to 8-9-week-old Asah1P361R/P361R;MCP-1-/- mice (Figure 60F). We also measured various enzymes that are associated with liver injury. At 8-9 weeks of age, alkaline phosphatase (ALP) levels were increased in Asah1P361R/P361R;MCP-1+/+ mice compared to age-matched controls (Figure 60G). ALP from 8-9-week-old Asah1P361R/P361R;MCP-1-/- mice was not different from age-matched controls, but 11-12-week-old Asah1P361R/P361R;MCP-1-/- mice did display elevated ALP, similar to the levels seen in 8-9-week-old

Asah1P361R/P361R;MCP-1+/+ mice (Figure 60G). In comparison to control mice, alanine aminotransferase (ALT) was elevated similarly in 8-9-week-old Asah1P361R/P361R;MCP-1+/+ mice and in both 8-9 and 11-12-week-old Asah1P361R/P361R;MCP-1-/- mice (Figure 60H). In comparison to control mice, aspartate aminotransferase (AST) levels were elevated in both

215

8-9 and 11-12-week-old Asah1P361R/P361R;MCP-1-/- mouse samples but not to the same extent as that seen in samples from 8-9-week-old Asah1P361R/P361R;MCP-1+/+ mice (Figure 60I).

Figure 60. Delayed signs of liver injury and fibrosis in Asah1P361R/P361R;MCP-1-/- mice

Light micrographs of liver sections stained for hematoxylin and eosin (H&E), Masson’s trichrome, and immunohistochemistry (IHC) for Mac-2 in 8-9-week-old Asah1+/+;MCP-1+/+ mice (A), 8-9-week-old Asah1+/+;MCP- 1-/- mice (B), 8-9-week-old Asah1P361R/P361R;MCP-1+/+ mice (C), 8-9-week-old Asah1P361R/P361R;MCP-1-/- mice (D), and 11-12-week-old Asah1P361R/P361R;MCP-1-/- mice (E). The original magnification of the top panel is 20x where the scale bar represents 100µm and the original magnification of the bottom panel is 40x where the scale bar represents 50µm. Liver function metabolites and enzymes were analyzed in serum from 8-9-week-old mice for each genotype, and samples from 11-12-week-old Asah1P361R/P361R;MCP-1-/- mice. Analytes from the biochemistry panel include; Albumin (F), alkaline phosphatase (ALP) (G), alanine aminotransferase (ALT) (H),

216 n=7-9 mice per group. ELISA for aspartate aminotransferase (AST) (I). All comparisons were made between 8- 9-week-old Asah1P361R/P361R;MCP-1+/+, 8-9-week-old Asah1P361R/P361R;MCP-1-/- and 11-12-week-old Asah1P361R/P361R;MCP-1-/- mice. n=8 mice per group. ns (not significant), *p<0.05, **p<0.01,***p<0.001. 6.4.5 Reduction of pulmonary infiltrates and protein accumulation

Recently, we demonstrated that Asah1P361R/P361R;MCP-1+/+ mice develop chronic lung injury and inflammation [Yu et al. 2017]. Here we examined the effects of deletion of MCP-1 on those parameters. Micrographs of representative BALF Cytospin slides showed significant cellularity and proteinaceous material in samples from both 8-9-week-old

Asah1P361R/P361R;MCP-1+/+ and 8-9-week-old Asah1P361R/P361R;MCP-1-/- mice in comparison to controls (Figure 61A-D). That said, samples from 8-9-week-old Asah1P361R/P361R;MCP-1-/- mice demonstrated a reduction in infiltrating cells and a lesser abundance of extracellular protein-stained debris than samples from the 8-9-week-old Asah1P361R/P361R;MCP-1+/+ mice

(Figure 61D). That latter phenotype became even more pronounced in the 11-12-week-old

Asah1P361R/P361R;MCP-1-/- samples (Figure 61E). This observation was also reflected in BALF supernatant turbidity and protein concentration analyses. At 8-9 weeks of age, BALF turbidity and protein levels from Asah1P361R/P361R;MCP-1-/- were significantly lower than in 8-9- week-old Asah1P361R/P361R;MCP-1+/+ mice (Figure 61F-G). However, by 11-12 weeks of age,

BALF turbidity and protein levels from Asah1P361R/P361R;MCP-1-/- mice had increased to levels similar to 8-9-week-old Asah1P361R/P361R;MCP-1+/+ mice (Figure 61F-G). Quantitation of BALF cells was performed on Kwik-Diff stained Cytospin slides. Total cell counts in samples from

8-9-week-old Asah1P361R/P361R;MCP-1-/- mice were not significantly different from samples from age-matched Asah1P361R/P361R;MCP-1+/+ mice, but the Asah1P361R/P361R;MCP-1+/+ mice did display a significant increase in cell recruitment between 8-9 and 11-12 weeks of age

(Figure 61H). BALF cell differentials from 8-9-week-old Asah1P361R/P361R;MCP-1-/- mice revealed an increase in macrophage abundance and reductions in neutrophil, lymphocyte, and percentage of multinucleated macrophages when compared to age-matched

Asah1P361R/P361R;MCP-1+/+ mice (Figure 61I-L). The percentage of total macrophage,

217 neutrophil, and percentage of multinucleated macrophages in the BALF of 11-12-week-old

Asah1P361R/P361R;MCP-1-/- mice had normalized to the levels found in 8-9-week-old

Asah1P361R/P361R;MCP-1+/+ mice (Figure 61I-L). Thus, it appears that deletion of MCP-1 may impede cellular recruitment to the lungs. This trend was also observed when an ELISA was performed on BALF supernatant to detect surfactant proteins (SP). There we observed a reduction of detectable SP-B and SP-C in samples from 8-9-week-old Asah1P361R/P361R;MCP-

1-/- mice. By 11-12 weeks of age SP-B and SP-C in the BALF of Asah1P361R/P361R;MCP-1-/- mice were detected at levels similar to 8-9-week-old Asah1P361R/P361R;MCP-1+/+ mice (Figure

61N and O). Interestingly, though elevated in BALF supernatants from all Asah1P361R/P361R mice, no significant changes were seen in Surfactant Protein (SP) A or D with altered MCP-

1 expression (Figure 61M and P).

218

Figure 61. Bronchial alveolar lavage fluid (BALF) from Asah1P361R/P361R;MCP-1-/- mice lungs displays lessened signs of pulmonary inflammation and surfactant protein accumulation.

Representative light micrographs of Cytospin BALF slides from 8-9-week-old mice in all genotypes and 11-12- week-old from Asah1P361R/P361R;MCP-1-/- mice. The original magnification was set at 20x and the scale bar represents 100µm at equal 1:2 dilutions (A-E). BALF supernatant from 8-9-week-old mice in all genotypes and 11-12-week-old from Asah1P361R/P361R;MCP-1-/- mice measuring turbidity by absorbance at 600nm (F) and protein concentration from Bicinchoninic acid (BCA) assay (G) n=5-6. Cell counts, and differentials were performed on BALF cell pellets from 8-9-week-old mice in all genotypes and from 11-12-week-old Asah1P361R/P361R;MCP-1-/- mice. Kwik-Diff stained slides show BALF cell count (H), macrophages (I), neutrophils (J), lymphocytes (K), and percentage of multinucleated macrophages of total macrophages count (L) n=5. ELISAs were performed on BALF supernatant from 8-9-week-old mice in all genotypes and 11-12-week-old from Asah1P361R/P361R;MCP-1-/- mice for SP-A (M), SP-B (N), SP-C (O) and SP-D (P) n=5-6. All comparisons were made between 8-9-week-old Asah1+/+;MCP-1+/+, Asah1P361R/P361R;MCP-1+/+, Asah1P361R/P361R;MCP-1-/- and 11-12-week-old Asah1P361R/P361R;MCP-1-/- mice. ns (not significant), *p<0.05, **p<0.01, ***p<0.001.

219

6.4.6 Brain and behavioral phenotypes persist in MCP-1 deficient mice

Asah1P361R/P361R;MCP-1+/+ mice develop significant central nervous system defects including neuroinflammation, demyelination, neurodegeneration, and hydrocephaly. These become progressively more severe as the disease manifests fully [Sikora et al. 2017]. To assess the state of neuroinflammation, brain tissues were stained with GFAP and Iba-1 to highlight activated astrocytes and microglia. Brain samples from 8-9-week-old

Asah1P361R/P361R;MCP-1+/+ and both 8-9 and 11-12-week-old Asah1P361R/P361R;MCP-1-/- mice have similar Iba-1 and GFAP staining patterns in the cerebellum, thalamus, and cerebral cortex (Figure 62A-G). This may indicate that there is no additive effect of inflammation between 8-9 and 11-12 weeks of age in these animals. From brain wet-to-dry weight ratios, no changes in brain edema were found between 8-9-week-old Asah1P361R/P361R;MCP-1+/+ and both 8-9 and 11-12-week-old Asah1P361R/P361R;MCP-1-/- mice (Figure 62H).

6.4.7 Behavioral phenotypes unchanged in MCP-1 deficient mice

The Asah1P361R/P361R;MCP-1+/+ mice replicate behavioral features that are typically present in FD patients [Sikora et al. 2017]. Amongst these the Asah1P361R/P361R;MCP-1+/+ mice have shown decreased locomotion, decreased exploratory behavior, increased thigmotaxis, and weakness in grip strength and rotarod testing [Sikora et al. 2017]. To assess whether MCP-1 ablation may improve behavior parameters, we repeated the grip strength and Rotarod tests. In the grip strength test, both 8-9 and 11-12-week-old

Asah1P361R/P361R;MCP-1-/- mice displayed significantly decreased ability on both total limb and forelimb strength tests when compared to age-matched control mice (Figure 62I and J).

However, when we normalize for body weight, 8-9-week-old Asah1P361R/P361R;MCP-1-/- mice showed no differences in total limb or forelimb strength compared to age-matched controls

220

(Figure 62K-L). At 11-12 weeks, Asah1P361R/P361R;MCP-1-/- mice lost their total limb strength but maintained a hind-limb strength advantage (Figure 62K-L). This may suggest that hind- limb weakness occurs prior to forelimb weakness in our model. However, due to their tight skin and our subsequent difficulty in scruffing Asah1P361R/P361R;MCP-1+/+ and

Asah1P361R/P361R;MCP-1-/- mice, we could not reliably obtain hind-limb data. Lastly, the

Rotarod test demonstrated that 8-9 and 11-12 week-old Asah1P361R/P361R;MCP-1-/- animals displayed a motor and endurance phenotype that largely mirrors what we reported in 8-9- week-old Asah1P361R/P361R;MCP-1+/+ mice [Sikora et al. 2017] (Figure 62M). Taken together, there appears to be no improvement in neuroinflammation or brain edema, and little improvement in age-matched motor strength, in the test animals.

221

Figure 62. Brain and behavioral deficits persist in Asah1P361R/P361R;MCP-1-/- mice

Representative micrographs from 8-9-week-old and 11-12-week-old mice from all genotypes featuring cerebellum, thalamus, and cortex regions stained for GFAP and Iba-1. Original magnification of cerebellum samples is 10x. The scale bar represents 100 µm and the original magnifications of the thalamus and cortex samples were 20x wherein the scale bar represents 50µm (A-E). The percentage of GFAP and Iba-1 positive staining from the cerebellum, thalamus, and cortex regions. Quantitation was performed on micrographs from 8-9 and 11-12-week-old mice. n= 3-4 mice per genotype (F and G). Brain wet-to-dry weight ratios were performed on 8-9 and 11-12-week-old mice. n= 3-4 mice per genotype (H). Behavioral tests were performed on 8-9 and 11-12- week-old mice for grip strength (I-L) and Rotarod proficiency (M). n= 8-10 per genotype. ns (not significant), *p<0.05 **p<0.01, ***p<0.001.

222

6.4.8 Tissue-specific changes in sphingolipid profiles due to MCP-1 deletion

Sphingolipid profiles were measured via liquid chromatography-mass spectrometry

(LC-MS) in lipid extracts from liver, lung, and brain tissue lysates from control and test animals. Sphingolipids assayed included ceramide, C1P, SM, MHC, Sph, and S1P. The most abundant sphingolipid in all three tissues in 8-9-week-old Asah1P361R/P361R;MCP-1+/+ mice was ceramide (Figure 63A-C). The greatest changes in abundance were in the lung, where 8-9-week-old Asah1P361R/P361R;MCP-1-/- mice displayed increased ceramide and SM levels compared to control mice, but had a significant reduction in ceramide and increased

SM when compared to 8-9-week-old Asah1P361R/P361R;MCP-1+/+ mice (Figure 63B).

Interestingly, this shift was no longer present in tissues from 11-12-week-old

Asah1P361R/P361R;MCP-1-/- mice. When we examined liver tissues, 8-9-week-old

Asah1P361R/P361R;MCP-1-/- mice had a reduction in total and C24:1 ceramide compared to samples from 8-9-week-old Asah1P361R/P361R;MCP-1+/+ mice. By 12 weeks of age, the C24:1 and total ceramide content in liver were increased to the same level as that seen in 8-9- week-old Asah1P361R/P361R;MCP-1+/+ mice. There were no differences between

Asah1P361R/P361R;MCP-1+/+ and Asah1P361R/P361R;MCP-1-/- mice for SM, MHCs, C1P, and S1P for the liver (Figure 63J, M, Figure 64 and Figure 65). Sph in liver was unchanged between

Asah1P361R/P361R;MCP-1-/- mice and age-matched controls, however a decrease was found when compared to the 11-12-week-old group (Figure 65A).

The greatest changes in ceramide levels were seen in the lungs. There we observed a reduction in C24:1, C24:0, and total ceramide content in samples from 8-9-week-old

Asah1P361R/P361R;MCP-1-/- animals compared to those from age-matched

Asah1P361R/P361R;MCP-1+/+ mice (Figure 63E). This relationship is reflected in a significant increase in SM24:1, SM24:0, and total SMs between samples from 8-9-week-old

223

Asah1P361R/P361R;MCP-1-/- animals compared to those from age-matched

Asah1P361R/P361R;MCP-1+/+ mice (Figure 63K). In the lungs, we observed a significant reduction in C24:0 MHCs, total MHCs, C24:0 C1P, and total C1P species but no changes in

Sph and S1P between samples from 8-9 and 11-12-week-old Asah1P361R/P361R;MCP-1-/- animals compared to those from 9-week-old Asah1P361R/P361R;MCP-1+/+ mice (Figure 64 and

Figure 65). In brain lysates, there were no observed changes in the sphingolipid species we queried between samples from 8-9 and 11-12-week-old Asah1P361R/P361R;MCP-1-/- animals compared to those from 9-week-old Asah1P361R/P361R;MCP-1+/+ mice.

224

Figure 63. Sphingolipid accumulation and abundance changes are organ specific in the Asah1P361R/P361R;MCP-1-/- mice

Percentage distribution of all total sphingolipids measured in liver, lung, and brain tissue lysates (A-C). Ceramide species in liver, lung, and brain (D-F). Relative abundance of ceramide species in liver, lung, and brain (G-I). Sphingomyelin (SM) species in liver, lung, and brain (J-L) and relative abundance of SM species in liver, lung, and brain (M-O). n=4-6 per genotype. All comparisons were made between samples from 8-9-week-old Asah1P361R/P361R;MCP-1+/+, 8-9-week-old Asah1P361R/P361R;MCP-1-/- and 11-12-week-old Asah1P361R/P361R;MCP-1-/- mice. ns (not significant), * represents post-hoc test compared to Asah1+/+;MCP-1+/+ *p<0.05, **p<0.01,***p<0.001. # represents post-hoc test compared to Asah1P361R/P361R;MCP-1+/+ #p<0.05

225

Figure 64. Monohexosylceramide (MHC) and Ceramide-1-phosphate (C1P) quantification

MHC species in liver (A), relative abundance of MHC species in liver (B), MHC species in lung (C), relative abundance of MHC species in lung (D), MHC species in brain (E), relative abundance of MHC species in brain (F), C1P species in liver (G), relative abundance of C1P species in liver (H), C1P species in lung (I), relative abundance of C1P species in lung (J), C1P species in brain (K), and relative abundance of C1P species in brain (L) were determined. n= 4-6 per genotype. All comparisons were made between 8-9-week-old Asah1P361R/P361R;MCP-1+/+, 8-9-week-old Asah1P361R/P361R;MCP-1-/- and 11-12-week-old Asah1P361R/P361R;MCP-1-/- mice. ns (not significant), * represent post-hoc test compared Asah1+/+;MCP-1+/+ to *p<0.05, **p<0.01,***p<0.001. # represent post-hoc test compared to Asah1P361R/P361R;MCP-1+/+ #p<0.05.

226

Figure 65. Sphingosine (Sph) and sphingosine-1-phosphate (S1P) evaluation

Sph fold change measurements in liver, lung and brain lipid extracts (A). and S1P fold change in liver, lung and brain lipid extracts (B). n= 4-6 per genotype. All comparisons were made between 8-9-week-old Asah1P361R/P361R;MCP-1+/+, 8-9-week-old Asah1P361R/P361R;MCP-1-/- and 11-12-week-old Asah1P361R/P361R;MCP-1-/- mice. ns (not significant), * represent post-hoc test compared Asah1+/+;MCP-1+/+ to *p<0.05, **p<0.01,***p<0.001.

6.4.9 Unique cytokine signature double mutants

To access whether the absence of MCP-1 may affect cytokine profiles, a multiplex mouse cytokine panel was queried using serum samples. As we have seen before, MCP-1 levels were found to be dramatically increased in 8-9-week-old Asah1P361R/P361R;MCP-1+/+ mice [Alayoubi et al. 2013]. As expected, no MCP-1 was found in serum samples from 8-9 and 11-12-week-old Asah1P361R/P361R;MCP-1-/- mice (Figure 66A). Interestingly, the levels of monocyte inflammatory protein-1 alpha (MIP-1α) in samples from 8-9-week-old

227

Asah1P361R/P361R;MCP-1-/- mice was significantly higher compared to those obtained from age-matched Asah1P361R/P361R;MCP-1+/+ mice; these values were even further increased in samples from such mice at 11-12 weeks of age (Figure 66B). Keratinocyte chemoattractant

(KC) concentrations peaked in samples from both 8-9-week-old Asah1P361R/P361R;MCP-1+/+ and Asah1P361R/P361R;MCP-1-/- mice. While KC levels were detected at lower levels in 11-12- week-old samples compared to 8-9-week-old samples in Asah1P361R/P361R;MCP-1+/+, they were nonetheless statistically elevated when compared to controls (Figure 66C). Interferon gamma-induced protein 10 (IP-10) concentrations in serum from 8-9 and 11-12-week-old

Asah1P361R/P361R;MCP-1-/- mice were elevated to similar levels as those seen in samples from

8-9-week-old Asah1P361R/P361R;MCP-1+/+ mice (Figure 66E). Interestingly, monocyte induced by gamma interferon (MIG), Interleukin (IL) IL-12, and IL1α were not significantly increased in samples from 8-9-week-old Asah1P361R/P361R;MCP-1+/+ mice when compared to control mice but were indeed significantly increased in samples from 8-9-week-old Asah1P361R/P361R;

MCP-1-/- mice (Figure 66D,F,G). Although not statistically significant, IL1β, and IL13 concentrations displayed a trend towards an increase in 11-12-week-old

Asah1P361R/P361R;MCP-1-/- samples (Figure 67). No differences were seen for basic fibroblast growth factor (bFGF), IL-4, IL-5, IL-6, IL-10, IL-17, granulocyte macrophage colony- stimulating factor (GM-CSF), interferon gamma (IFNγ), or vascular endothelial growth factor

(VEGF) (Figure 67). We further measured MCP-3 and MCP-5 levels by ELISA in mouse serum. Both MCP-3 and MCP-5 levels were significantly elevated in 8-9-week-old

Asah1P361R/P361R;MCP-1+/+ when compared to age-matched controls (Figure 66H, I). MCP-3 and MCP-5 levels were also elevated in both 8-9 and 11-12-week-old Asah1P361R/P361R;MCP-

1-/-. No differences in MCP-3 and MCP-5 levels between samples from 8-9-week-old

Asah1P361R/P361R;MCP-1+/+, 8-9 and 11-12-week-old Asah1P361R/P361R; MCP-1-/- mice were observed.

228

Figure 66. Altered cytokines in the serum of Asah1P361R/P361R;MCP-1-/- mice.

Cytokines were measured in serum of 8-9-week-old mice for each genotype and 11-12-week-old Asah1P361R/P361R;MCP-1-/- mice. Levels of CCL2 (MCP-1) (A), CCL3 (MIP-1α) (B), keratinocyte chemoattractant (KC) (C), CXCL9 (MIG) (D), CXCL10 (IP-10) (E), interleukin-12 (IL-12) (F), and interleukin-1-alpha (IL-1α) (G) were measured. n = 8. Levels of CCL7 (MCP-3) (H) and CCL12 (MCP-5) (I) measured with ELISA. n= 4-5. All comparisons were made between 8-9-week-old Asah1+/+;MCP-1+/+, Asah1P361R/P361R;MCP-1+/+, Asah1P361R/P361R;MCP-1-/- and 11-12-week-old Asah1P361R/P361R;MCP-1-/- mice. ns (not significant), *p<0.05, **p<0.01, ***p<0.001.

229

Figure 67. Cytokines that were unchanged in serum analyses.

Serum cytokine levels were measured in samples from 8-9-week-old mice for each genotype, and from 11-12- week-old Asah1P361R/P361R;MCP-1-/- mice. Levels of FGF-β (A), IL-1β (B), IL-4 (C), IL-5 (D), IL-6 (E), IL-10 (F), IL- 13 (G), IL-17 (H), GM-CSF (I), IFNγ (J), VEGF (K) were determined. n = 8. All comparisons were made between 8-9-week-old Asah1+/+;MCP-1+/+, Asah1P361R/P361R;MCP-1+/+, Asah1P361R/P361R; MCP-1-/- and 11-12-week-old Asah1P361R/P361R;MCP-1-/- mice. ns (not significant), *p<0.05, ***p<0.001.

230

6.5 Discussion and conclusion

The signs and symptoms of FD can manifest along a spectrum. In its moderate form,

FD patients will develop arthritis-like symptoms such as joint pain and nodule formation

[Ehlert et al. 2007]. In its severe classical form, however, FD patients will often have significant inflammation, neurological complications, and respiratory failure [Levade et al.

2014]. In our FD mouse, widespread inflammation has been reported to affect many organs in the Asah1P361R/P361R;MCP-1+/+ mouse [Alayoubi et al. 2013; Dworski et al. 2015; Yu et al.

2017]. Here, we examine multiple parameters of FD manifestations in the context of deletion of a key chemokine (MCP-1) that we identified previously as a biomarker for this disorder

[Dworski et al. 2017]. Our analyses of the hematopoietic, hepatic, pulmonary, and neurological systems demonstrate that the severity of MCP-1-mediated inflammation is organ-dependent. Our results here further demonstrate that attenuation of MCP-1 in FD mice can affect ceramide accumulation, improve survival, and delay pathology.

The Asah1P361R/P361;MCP-1+/+ mouse has impaired hematopoiesis characterized by a decline in pre- and pro-B cell numbers in the bone marrow of aged mice and a reduction of

CD4+ CD8+ T-cells within the thymus of 5-weeks-of-age animals that progressively worsens12. Despite increased longevity and a delay in weight loss and skin elasticity at 8-9- weeks-of-age, the Asah1P361R/P361R;MCP-1-/- mice share a similar hematological profile to the

Asah1P361R/P361R;MCP-1+/+ mice. Histological analyses revealed tissue damage to the architecture of the bone marrow, thymus, and spleen with 11-12 weeks of age

Asah1P361R/P361R;MCP-1-/- mice, sharing a profile that is very similar to 8-9-week-old

Asah1P361R/P361R;MCP-1+/+ mice. Since age-matched Asah1P361R/P361R;MCP-1-/- mice show reduced leukocytosis and a reduction in granulocytes in the spleen, it suggests that absence of MCP-1 can slow the rate of macrophage infiltration; this also suggests, however, that the bone marrow and thymus may be more sensitive to inflammation and injury when the

231 sphingolipid balance is perturbed. This phenotype has also been noted in studies on the

S1P lyase deficient mouse (SGPL-/-), where investigators showed impaired B-cell development in the bone marrow, thymic atrophy associated with a reduction in CD4+ and

CD8+ thymocytes, and an increase in thymic ceramide levels [Vogel et al. 2009; Weber et al.

2009].

Deletion of MCP-1 also leads to improvement in the hepatic system. In comparison to age-matched control animals, there was a reduction in liver injury enzymes, decreased formation of characteristic large foamy macrophages, and less tissue fibrosis in the double mutants. MCP-1 expression within the liver has been shown to play a role in hepatic inflammation and fibrosis in patients with hepatitis C [Mühlbauer et al. 2003]. Similarly, inhibition of MCP-1 with the structured L-enantiomeric RNA oligonucleotide mNOX-E36 (aka

Spiegelmer) has shown promise in reducing fibrosis in murine models of chronic liver disease [Baeck et al. 2012]. Since we also see a reduction in accumulated ceramides, it is possible that, similar to the lungs, an absence of MCP-1 is reducing the rate of recruitment of cells which is reducing the total amounts of ceramides as well as the inflammation- induced injury that we normally observe in Asah1P361R/P361R;MCP-1+/+ mice.

Infiltration of immune cells into the lungs and respiratory failure can be a significant problem in FD [Levade et al. 2014]. Recently, we have shown that Asah1P361R/P361R;MCP-1+/+ mice develop chronic lung injury caused by a combination of inflammation and vascular leakage [Yu et al. 2017]. Here, we find that MCP-1 deletion results in significant improvement of lung phenotypes. BALF from Asah1P361R/P361R;MCP-1-/- mice revealed both lower protein content and lower cell counts than that from Asah1P361R/P361R;MCP-1+/+ mice. In addition, there was a mild reduction in lipo-protein-like material in the BALF, which also correlated with a reduction in various surfactant proteins. Surprisingly, in addition to reduced inflammation in the lungs, we also observed a reduction in all measured sphingolipids in lung tissue lysates (Figure 64B). One explanation for this may be the reduced recruitment of

232 cells within the lung parenchyma. Another explanation may be the reduced inflammatory cues produced by MCP-1 localized within the lungs; the role of MCP-1 as an instigator of inflammation and as a potential biomarker has been widely reported [Chiao et al. 2011;

Ruhwald et al. 2009]. Increased MCP-1 can also be detected in rare diseases like pulmonary alveolar proteinosis, chronic obstructive pulmonary disease, as well as in severe cases of acute lung injury [Iyonaga et al. 1999; Suga et al. 1999; Bhatia et al. 2012]. In addition, use of anti-rat MCP-1 antibodies has shown efficacy in reducing not only the size, but the quantity of glucan-induced pulmonary granulomas vasculitis in rats [Flory et al.

1993]. Due to the widespread role of MCP-1 in lung pathology, it is thus possible that in the context of FD, MCP-1-mediated recruitment of inflammatory cells may be more penetrant and pernicious in the lung than other organs in our mouse model.

The brains of Asah1P361R/P361R;MCP-1+/+ mice show an accumulation and a shift in the relative abundance of various sphingolipids, the presence of cellular storage bodies in various neuronal cells, signs of inflammation, and elevated MCP-1 levels [Alayoubi et al.

2013; Sikora et al. 2017]. Histological analyses on age-matched brains from

Asah1P361R/P361R;MCP-1-/- mice also demonstrated significant astrocytosis and large activated microglia (Figure 62D-E). Though not statistically significant, there also appears to be a trend towards an increase in percentage staining of Iba-1 from 8-9-week-old to 11-12-week- old brains of Asah1P361R/P361R; MCP-1-/- mice. Behavioral assays on the Asah1P361R/P361R;

MCP-1-/- mice demonstrated little change from controls, however, we do show a mild improvement in forelimb grip strength force when normalized for body weight (Figure 62J,

L). Together, this demonstrates that deletion in MCP-1 may have a minor effect in delaying the signs of neural inflammation in our Farber mice.

Inhibition of inflammatory cytokines has shown variable results in LSD brain models.

For example, in the Hexβ-/-;Mip-1α-/- mouse, deletion of Mip-1α led to an improvement in

233 brain pathology [Wu and Proia 2004]. The opposite outcome was observed when the same cross was performed with the Niemann-Pick Type C (NPC-/-) mouse model [Lopez et al.

2012]. Both of those LSD models demonstrate some shared phenotypes; however, in the latter cross the mice experienced a decreased lifespan [Wu and Proia 2004; Lopez et al.

2012]. The group that characterized this outcome rationalized that in the context of NPC, it is possible that inflammation in the brain may serve a protective role [Lopez et al. 2012]. This theme is also shared in the case of globoid cell leukodystrophy (GLD; Krabbe’s disease), which shares some neurological phenotypes with the Farber mouse such as activated microglia/macrophages and demyelination [Sikora et al. 2017]. To genetically remove macrophages, one group crossed the GLD Twitcher (twi-/-) mouse to the macrophage- deficient osteoporotic (CSF-1op/op) mouse [Kondo et al. 2011]. Double mutants there not only had a shorter lifespan but also manifested an exacerbated neurological phenotype compared to control twi-/- mice [Kondo et al. 2011].

Brain sphingolipid profiles of our Asah1P361R/P361R;MCP-1-/- mice were largely unchanged compared to Asah1P361R/P361R;MCP-1+/+ mice. In our analyses of Sph there was a minor but significant reduction in Sph for both the Asah1P361R/P361R;MCP-1+/+, and

Asah1P361R/P361R;MCP-1-/- groups. This observation deviates somewhat from our previously published results wherein we detected a modest increase in Sph in brain lysates from

ACDase deficient mice [Sikora et al. 2017]. The variability could be a combination of the different tissue homogenization protocols, lipid extraction methods, internal standards used, instrumentation and LC/MS protocols as these were processed at different sites.

Interestingly this contrasting result may be exclusive to Sph, however, as all other sphingolipids measured in both studies displayed the same trend [Sikora et al. 2017].

Nonetheless, elevated ceramide levels were noted in the brain. Increases in ceramides are also detected in various CNS conditions, such as in multiple sclerosis (MS)

234

[van Doorn et al. 2012]. This study showed that activated astrocytes that were preferably localized to MS-related brain lesions were the main contributors of the increased levels of ceramides detected [van Doorn et al. 2012]. Furthermore, inhibition of the de novo pathway of ceramide synthesis with the S1P analogue FTY720 was able to attenuate reactive phenotype of astrocytes [van Doorn et al. 2012]. Therefore, it is possible that the inflammatory consequences of increased MCP-1 in the brain are masked by the deleterious effects of significant ceramide and sphingolipid accumulation

Our study demonstrated that genetic ablation of MCP-1 allowed Asah1P361R/P361R mice to survive up to 14 weeks of age. This increased longevity coincided with reduced pathology in the lungs and liver of 8-9-week-old Asah1P361R/P361R; MCP-1-/- and age matched controls. By 11-12-weeks of age the disease phenotypes in Asah1P361R/P361R; MCP-1-/- mice largely mirrors that of 8-9-week-old Asah1 P361R/P361R; MCP-1+/+ mice. While little to no changes were seen in the hematopoietic and neurological systems, the absence of MCP-1 slowed the deleterious effects of ceramide-induced inflammation. In addition, we found new changes in the cytokine profiles that may, in part, explain the impeded pathology in the

Asah1P361R/P361R; MCP-1-/- mice.

KC expression was previously found to peak at 7-8 weeks and decline by 9 weeks in

Asah1P361R/P361R;MCP-1+/+ mice [Dworski et al. 2017]. In this study, deletion of MCP-1 caused a shift where KC was seen to peak at 8-9-weeks and later decrease in 11-12-week samples

(Figure 66C). Further, our data also highlights a possible compensatory accumulation of the cytokines MIP-1α, MIG, IL-12, and IL-1α in the double mutant animals. Along these lines, we previously showed that MIP-1α levels increase by 5 weeks and peaked at 9 weeks of age in

Asah1P361R/P361R;MCP-1+/+ mice [Dworski et al. 2017]. In this study, we found no statistical significance for MIP-1α in 8-9-week-old Asah1P361R/P361R;MCP-1+/+ mice samples. This may in part be due to sample variability in the Asah1P361R/P361R;MCP-1+/+ samples. However, MIP-

235

1α was elevated and significant in both 8-9 and 11-12-week-old Asah1P361R/P361;MCP-1-/- mice samples. MIP-1α, a well-studied chemokine, is known for its roles in macrophage recruitment and inflammatory responses. In fact, other studies of LSDs such as Gaucher disease and Sandhoff’s disease have reported elevations of MIP-1α as part of disease progression in patient and mouse models [Wu and Proia 2004; Van Breemen et al. 2007]. In

Gaucher disease, where glucosylceramide accumulates, infiltration of tissues and formation of large foamy macrophages is also widely abundant [Jmoudiak and Futerman 2005; Mistry et al. 2010]. Both MCP-1 and MIP-1α have been found to be elevated in the serum samples of Gaucher patients [Pavlova et al. 2011]. Large foamy macrophages are a shared phenotype present in both Farber and Gaucher disease. Due to this similarity and the tight balance of sphingolipid metabolism, the compensatory accumulation of MIP-1α may yet be another shared sign. While more work is required, it is possible that the inflammation observed in both Farber disease and Gaucher disease manifest similarly.

Here, we also report elevations in MIG in the serum of our Asah1P361R/P361;MCP-1-/- mice. In contrary, MIG from Asah1P361R/P361;MCP-1+/+ mice samples were not found to be statistically significant in this study but appear to show a trend towards an increase, which is also mirrored in FD patient plasma [Dworski et al. 2017]. Similar findings were noted for IL-

12. Lastly, IL-1α was elevated in the double mutants. This was unexpected as IL-1α in our previous data was variably expressed in mouse plasma [Dworski et al. 2017]. There is evidence that activation of IL-1 may lead to the accumulation of ceramide and expression of the inflammatory enzyme COX-2 along with prostaglandin E2 production [Homaidan et al.

2002]. More work will be required to elucidate the roles of these cytokines in FD, however, these data together demonstrate the complexity and nuanced cytokine/chemokine regulation that might be involved in the inflammatory response seen in FD.

236

Analyses of MCP-3 and MCP-5 by ELISA in mouse serum revealed significantly increased levels of these factors in samples from 8-9-week-old Asah1P361R/P361;MCP-1+/+ mice. This observation demonstrates that other agonists of CCR2 are also contributors to the inflammatory phenotype seen in ACDase deficiency [Franci et al. 1995; Sarafi et al.

1997]. While the MCP-3 and MCP-5 levels were not statistically significant between 8-9 and

11-12-weeks in the Asah1P361R/P361;MCP-1-/- mice, there appeared to be a trend towards an increase. The high levels of MCP-3 and MCP-5 would partly explain the continued presence of foamy macrophages in Asah1P361R/P361;MCP-1-/- mice. A previous study has shown that

MCP-3-/- mice that were exposed to Listeria monocytogenes infection showed a similar inflammatory phenotype as those of MCP-1-/- mice [Jia et al. 2008]. That study demonstrated that both MCP-1 and MCP-3 act in parallel on CCR2 in the context of innate immune defense [Jia et al. 2008]. This phenomenon may apply in our model as the ablation of MCP-

1 alone provided a modest lifespan improvement to Asah1P361R/P361 mice. Furthermore, another chemokine, MCP-4, which was not measured in this study, has also been shown to be an agonist for CCR2 [Berkhout et al. 1997]. Out of all three chemokines measured, MCP-

1 was detected at the highest level in serum samples of Asah1P361R/P361;MCP-1+/+ mice, suggesting a dominant role for that factor in macrophage recruitment in ACDase deficiency.

While additional work will be required, the discovery of elevated MCP-3 and MCP-5 further underscores the importance of CCR2 signaling in ACDase biology.

MCP-1 is a potent chemo-attractant that has been implicated in exacerbating the pathology of an array of diseases. Genetic crosses between the MCP-1-/- mouse and other disease models ranging from the Charcot-Marie-Tooth mouse model to the more common atherosclerosis mice models have been previously performed [Dawson et al. 1999; Groh et al. 2010; Kohl et al. 2010]. Results showed that the absence of MCP-1 led to attenuation of

237 neuroinflammation in the case of the Charcot-Marie-Tooth model, and reduction in aortic lesions the atherosclerotic ApoE deficient mouse [Dawson et al. 1999; Groh et al. 2010].

MCP-1 has also been identified as a therapeutic target for other conditions including kidney and heart disease [Tesch 2008; Niu and Kolattukudy 2009]. Use of the structured L- enantiomeric RNA oligonucleotide mNOX-E36 (aka Spiegelmer) to inhibit MCP-1 has shown promise in murine models of chronic liver disease [Baeck et al. 2012]. Similarly, another group has shown that use of the MCP-1 inhibitor 2-Methyl-2-[[1-(phenylmethyl)-1H-indazol-

3yl]methoxy]propanoic acid (aka Bindarit) can protect mice when induced with acute pancreatitis, and limit proteinuria in rat models of renal disease [Bhatia et al. 2005; Zoja et al. 2015].

A heightened inflammatory response is common in many LSDs that involve sphingolipid storage as well as other more common diseases [El Alwani et al. 2006;

Maceyka and Spiegel 2014]. In the case of MCP-1 release, one study has showed that C1P can promote MCP-1 release in various cell types via the downstream PI3K/Akt, MEK/ ERK, and p38 pathways [Arana et al. 2013]. Another study has demonstrated that inhibition of de novo synthesized ceramide can lead to IL-6 and MCP-1 secretion from adipocytes in vitro

[Hamada et al. 2014]. This observation was also replicated in an experiment wherein obese mice, that were treated with myriocin - a potent inhibitor of the de novo pathway of ceramide synthesis showed a reduction in ceramide levels and MCP-1 mRNA expression [Yang et al.

2009].

Currently the treatment options for FD are limited to anti-inflammatory therapy and, in some patients, BMT. Human recombinant acid ceramidase (rhACDase) is currently being developed and has shown great promise as a potential therapy for FD [He et al. 2017]. A recent study showed that treating the ACDase-deficient mice with rhACDase significantly improved survival, and reduced MCP-1 with repeated infusions [He et al. 2017]. Still, human

238 studies and clinical approval of this treatment modality are still a ways off. Our study has demonstrated that targeting MCP-1 may delay the course of FD, and contribute to reduced inflammation within the respiratory and visceral systems. While more work will be required, the use of targeted anti-inflammatory treatments in conjunction with substrate reduction therapy, enzyme replacement therapy, or gene therapy may also create a synergistic benefit for FD management [Medin et al. 1999; Ramsubir et al. 2008; Walia et al. 2011; Schuchman

2016].

239

Chapter 7 General Discussion

7.1 Evaluating our hypotheses, and the success of our aims

7.1.1 Aim 1

Hypothesis: ACDase deficiency results in perturbed ocular and respiratory function.

Through this aim we successfully identified significant impairment to both the respiratory and ocular systems in Asah1P361R/P361R mice. Common findings that were present in the eye and lung of the Asah1P361R/P361R mouse included the accumulation of sphingolipids and recruitment of inflammatory cells. Large foamy macrophages were highly prominent in the lung parenchyma, as well as in the optic nerve and eye. Ultrastructure analyses revealed significant storage pathology in various cell types. Resident and recruited macrophages appeared most affected in both the eye and lung.

Infiltration of inflammatory cells into the eye and lung is likely related to perturbed vascular permeability. Analyses performed on BALF revealed significant albumin and immunoglobulin expression. In the eye, we observed infiltration of cells within the vascular layer of the GCL, as well as cells entering from the choroid layer.

ERG revealed a decreased retinal response in 3-4-week-old Asah1P361R/P361R mice compared to controls. Interestingly when ERG was repeated on 8-9-week-old

Asah1P361R/P361R mice, their retinal responses were found to be worse and, in some conditions, undetectable. Physiological tests were performed to assess lung and eye function. Decreased lung compliance and an increased in airway resistance was detected in

240

5-week-old Asah1P361R/P361R mice. Lung impairment was similar in 9-week-old

Asah1P361R/P361R mice.

A limitation in these studies was the phenotypic variability of Asah1P361R/P361R mice.

The mice displayed a range of retinal dysplasia, and significant variability in lung mechanics, and cytokine levels in BALF.

In the lung analyses, we proposed that ACDase deficiency leads to impaired surfactant homeostasis, and decreased blood oxygenation [Yu et al. 2017]. One limitation is that we did not characterize whether surfactant activity was altered. This could be addressed by assessing surfactant surface tension purified from BALF with a surfactometer.

Nonetheless, our data presents a detailed understanding of pulmonary involvement and highlights potential treatment avenues such as whole lung lavage to address respiratory complications in FD patients.

There are two further limitations present in our eye study. The first is the inability to conclusively assess the degree of visual impairment. Previous studies have demonstrated significant behavioral deficits in these mice. While we demonstrate decreased visual acuity in the “fake cliff test” the reduced movement might be confounded by other neurological deficits [Sikora et al. 2017]. The second limitation is that ocular manifestations may not be present in all cases of FD. From case reports eye pathology is most common in the classic and severe neurological variant of FD [Zielonka et al. 2017]. Based on this pattern our findings are likely most relevant to severe cases of FD. Importantly, our work demonstrates the application of noninvasive imaging in FD and LSD diseases. Use of these techniques may help to screen patients and monitor the effects of future therapeutics.

241

7.1.2 Aim 2

Hypothesis: ACDase deficiency results in liver injury and impaired expression of genes that impact sphingolipid metabolism.

In this aim, we were able to investigate the progression of liver pathology and report an altered gene expression profile in hepatocyte-enriched cultures from Asah1P361R/P361R mice. Similar to our findings in Aim 1, inflammation was significant and progressive in

Asah1P361R/P361R mice. Increases in select liver injury markers were detectable by 5 weeks of age. Injury became more severe and liver function deteriorated by 9 weeks of age.

Analyses of the transcriptome in the Asah1P361R/P361R mice revealed a significant number of genes that showed differential gene expression. Analyzing the most significantly altered genes revealed inflammation and leukocyte recruitment pathway activation. While this finding was not unexpected, we had not previously performed any such analyses. These findings will help to inform future studies exploring the source and mechanism of inflammation in the Asah1P361R/P361R mice. One of the unexpected changes observed was the deactivation of pathways related to lipid homeostasis. While phenotypically we observed a decrease in adipose tissue, we had no expectation of finding altered genes in the maintenance of lipid metabolism. This finding is substantial as it indicates that ACDase deficiency can have an impact on lipids outside of the sphingolipid pathway.

Important observations were made on the expression of genes in the sphingolipid pathway. Decreases in the mRNA expression of Cers2 and Cers4 may partlyexplain for the increase of C16:0 ceramide and a decrease in the C24:0 ceramide species. Increases to

Cerk and Ugcg may be equally important as their upregulation informs which pathways might compensate for ceramide backlog.

One potential limitation in our study is the purity of our hepatocyte-enriched cultures.

The 5-week time point was chosen in Asah1P361R/P361R mice as it represents the midpoint of

242 the animal’s life, and because the inflammatory phenotype is not as severe as in end stage animals. We reasoned that perfusion of the liver would reduce circulating cells such as macrophages. While our cell isolation was able to reduce CD11b positive macrophages, we cannot rule out the presence of other cell types.

A second drawback to this study was that our data only provides transcriptional insight, further protein analyses will be required to form definitive conclusions. Also, candidate genes and pathways identified as perturbed likely only represent the biology of

ACDase deficient hepatocytes. Confirmation in other tissues will also be required in future studies. Nonetheless, findings from this study have demonstrated that the liver pathology in

Asah1P361R/P361R mice recapitulates the phenotypes seen in the both the classic type 1 FD, and the neonatal visceral type 4 FD.

7.1.3 Aim 3

Hypothesis: Ablation of MCP-1 will decrease inflammation and hamper pathogenesis.

In this aim we, provided evidence that the absence of MCP-1 in the Asah1P361R/P361R mice impedes inflammation, decreased ceramide accumulation in specific organs, and provides a modest extension in lifespan. Deletion of MCP-1 had the greatest effect in the lung with significant reduction in inflammation, expression of surfactant proteins and turbidity in BALF samples from Asah1P361R/P361R;MCP-1-/- mice when compared to

Asah1P361R/P361R;MCP-1+/+ mice. The liver from Asah1P361R/P361R;MCP-1-/- mice also showed less fibrosis and cellular infiltration compared to Asah1P361R/P361R mice. Lipidomic analyses revealed a significant decrease in ceramides in the lungs and the liver of

Asah1P361R/P361R;MCP-1-/- mice compared to Asah1P361R/P361R;MCP-1+/+ mice. No changes were seen in the brain and hematopoietic organs. This suggests that MCP-1 may play a

243 greater role in disease manifestation in the lung and liver compared to other organs.

Importantly, we also revealed changes in the cytokine profile, and identified novel elevations in MCP-3 and MCP-5 in serum from both Asah1P361R/P361R and Asah1P361R/P361R;MCP-1-/- mice.

Since the aim of the study was to impede monocyte migration from the affected tissues, we were surprised that ablation of MCP-1 only impeded inflammation in certain tissues. One possible explanation is that the brain and hematopoietic organs are more sensitive to full MCP-1 ablation than the liver and lung, and therefore prone to produce compensating cytokines like Mip-1α, MIG, IL-1α, and IL-12. To answer this question, we could test semi-ablation with siRNA or inhibitor specific to MCP-1 and remeasure cytokines.

One further follow-up would measure the cytokine profiles specific to each organ. This would be valuable for the brain since the cortex and thalamus were more affected with activated microglia and glial cells than the cerebellum in both Asah1P361R/P361R;MCP-1+/+ and

Asah1P361R/P361R;MCP-1-/- mice. Therefore, cytokine analysis of the different brain regions might be more informative.

As little neurologic or hematopoietic benefit was found by deletion of MCP-1, our data suggests that targeting MCP-1 alone may provide better outcome in some certain organ systems such as the lung and liver, rather than others. Thus, therapeutic strategies targeting MCP-1 might be effective in treating specifically mild or attenuated variants of FD where neurological decline is absent. Until a cure is available, a combination therapy approach to target multiple inflammatory pathways will likely be the best option for managing

FD.

244

7.2 Pathogenesis in the ACDase deficient mouse

7.2.1 Phenotypes manifest post weening

We observed that mutant Asah1P361R/P361R mice appear physically normal until 3-4 weeks of age, a time point that coincides with mouse weening. Around this age,

Asah1P361R/P361R mice begin to differentiate from wild-type and heterozygous littermates and display skin feeling more rigid and tough. This feature continues to progress as the mice age

[Lopez-Vasquez et al. 2016].

One potential explanation is that prior to weaning, Asah1P361R/P361R mice are receiving functional ACDase from the milk of the nursing mother. It is possible that sufficient levels of

ACDase are transferred to curb substrate buildup during early development. While no study has demonstrated this hypothesis, a few studies support this possibility. One study has demonstrated that human milk contains a high abundance of prosaposin [Kondoh et al.

1991]. Prosaposin is the precursor protein that activates several different lysosomal hydrolases. In this study, addition of purified prosaposin showed a slight activation of β- glucosidase activity [Kondoh et al. 1991]. Another study has demonstrated that ASMase can be transferred from mother to progeny in tsetse flies [Benoit et al. 2012]. Finally, one early study of ERT showed that administration of human α-glucosidase purified from the milk of transgenic rabbits could improve the clinical condition of infantile Pompe disease patients

[Van den Hout et al. 2001].

Another potential hypothesis for the delay in phenotype is that a certain threshold of ceramide is required before manifestations of pathology. While a longitudinal analysis of ceramide expression has not been performed on Asah1P361R/P361R mice, results discussed in

Section 6.3.8 suggest this possibility. In this ceramide analyses while we did not test for

245 statistical significance between the 8-9 and 11-12-week time points, ceramide and SM in the lung and liver appeared have an increasing trend [Yu et al. 2018].

Additionally, based on ultrastructure analyses of the brain, presence of zebra or CTB storage bodies are more prevalent in older animals and particularly in macrophage [Sikora et al. 2017]. While there is no direct evidence that storage bodies in vacuoles are comprised of ceramide, one study did show that exogenous treatment of fibroblasts with ceramide led to the development of lysosomal storage bodies [Rutsaert et al. 1977].

7.2.2 Pathology is organ-specific

ACDase deficiency leads to pathology in multiple organs and tissue, however, not all organs are equally affected. The brain and the hematopoietic organs including the bone marrow, spleen, and thymus appear to be the most severely affected [Dworski et al. 2015;

Sikora et al. 2017]. The next most affected tissues are the liver, lung, and eyes [Yu et al.

2017]. In contrast, heart and kidney do not show presence of cellular infiltration [Alayoubi et al. 2013]. While no formal physiological tests have been performed, these organs appear phenotypically the least affected.

One of our theories is that excess ceramide triggers activation of inflammatory pathways and leukocyte recruitment as evidenced by high levels of inflammatory cytokines such as MCP-1 in the Asah1P361R/P361R mouse [Alayoubi et al. 2013]. Presumably these recruited cells are also burdened with ceramide accumulation that may trigger further cytokine release resulting in a positive feedback response.

One hypothesis as to why the heart and kidney appear less affected may be that the need for ACDase is different in each organ/tissue. In an early study characterizing the human ASAH1 gene, it was revealed that the organs with the greatest mRNA expression were the heart and kidney [Li et al. 1999]. Diagnosis of FD is based on enzyme activity from

246 cultured cells [Levade et al. 2014]. Fibroblasts from classical FD patients express <10% normal enzyme activity. As different tissues express different amounts of ASAH1 mRNA, it is probable that different organs also express different ASAH1 levels. One study attempting to address this performed acid and neutral ceramidase (NCDase) activity assays on 7 different rat tissues. The kidney followed by the brain showed the greatest enzyme activity, whereas the heart had the third lowest level of enzyme activity [Spence et al. 1986]. We have measured enzyme activity in the Asah1P361R/P361R mouse and revealed a greater reduction of enzyme activity in the liver, spleen, and thymus compared to the heart and brain (ACDase activity in the kidney was not performed) [Alayoubi et al. 2013].

Furthermore, measurement of ceramide in the Asah1P361R/P361R mice revealed the greatest accumulation in the liver, lung and spleen and least accumulation in the brain, heart and kidney [Alayoubi et al. 2013]. Taken together these studies provide at least a partial molecular explanation as to why the kidney and heart from Asah1P361R/P361R mice may be more protected versus the other organs. While the exact amount of enzyme activity required to elicit a phenotype is unknown, we suggest that it must be below 50% of normal Asah1+/+ levels as no apparent pathology is seen in Asah1+/P361R heterozygous mice, or human carriers of FD.

7.2.3 Potential role of resident macrophages in FD pathology

A heightened state of inflammation is characteristic of Asah1P361R/P361R mice, and FD patients. Our analyses of the eye (Chapter 3), lung, (Chapter 4), and liver (Chapter 5) have demonstrated that inflammation is progressive over time. Histological analyses of end stage animals revealed significant cellular infiltration consisting of many neutrophils and macrophages. From ultrastructure analyses, many of these macrophages were filled with storage vacuoles containing zebra and or CTB bodies.

247

In Section 7.2.2 we postulated that the heart and kidney showed reduced signs of disease versus other organs due to higher ACDase activity. Another hypothesis might be that tissues with a higher abundance of resident macrophages, or macrophage-like cells, will be more inclined to elicit a pro-inflammatory effect leading to recruitment factor expression and further inflammation. Phenotypically the bone marrow, spleen and thymus appear to be the most affected with inflammatory cells [Dworski et al. 2015]. Disruption to the tissue architecture has been reported in end-stage Asah1P361R/P361R mice [Dworski et al. 2015]. It is possible that these organs are more affected due to their hematopoietic function and origin.

The organs that are also severely affected with inflammation are the brain, liver, and lungs.

All of these organs contain tissue resident macrophages. One study found that normal adult mouse brains contained ~10-15% microglia [Lawson et al. 1992]. In rat livers, another study found that Kupffer cells account for ~15% of the total liver cell population [Bouwens et al.

1986]. In the lungs, one report demonstrated that ~3% of all cells in the murine lung were alveolar macrophages and another ~2% were composed of interstitial macrophages

[Lehnert et al. 1985]. Based on our data, the brain and liver from end stage mice showed signs of tissue damage whereas the lung was unaffected. This could be due to the increased inflammatory contribution from these resident cells.

Resident macrophages are also present in the heart and kidney though these two organs do not display a significantly increased inflammatory phenotype in Asah1P361R/P361R mice. A mouse study showed that ~4.7% of all the cells in the heart were leukocytes with the majority of them being monocytes, and cardiac macrophages [Pinto et al. 2015]. The kidney is also known to contain several distinct populations of macrophage-like cells [Hume and Gordon 1983; Schlondorff 1987].

Both the brain and liver contained significantly increased amounts of endogenous resident macrophage-like cells in Asah1P361R/P361R mice. After the hematopoietic organs the brain and liver display the most tissue destruction in the Asah1P361R/P361R mice [Dworski et al.

248

2015]. As mentioned above, the lung has been shown to contain ~5% resident macrophages, but we did not find significant tissue disruption in our study. Rather one of our main observations was that inflammation was present in the airway and alveolar space [Yu et al. 2017]. Ceramides and phospholipids were significantly elevated in the BALF component in comparison to the tissue suggesting that lipids may be secreted due to vascular leakage or from infiltrated and resident macrophages.

The heart was found to contain resident macrophages at a lower abundance than the brain and liver. Taken together it is possible that the penetrance of pathology in the different organs of the Asah1P361R/P361R mouse is dependent on both the endogenous enzyme level and the tissue specific response of resident macrophages. Further work will be required to investigate the biology of ACDase deficient macrophages.

7.2.4 Limitations of the ACDase deficient murine model

Asah1P361R/P361R mice recapitulate many of the clinical phenotypes reported in FD patient case reports, however, there are several disease features that are limited in the mouse model. Asah1P361R/P361R mice lack subcutaneous nodules along their joints. There is no clear current explanation for this phenomenon. Many patients have been identified through the presence of the cardinal triad symptoms: the development of painful nodules, joint contracture, and aphonia. An absence of these signs in Asah1P361R/P361R mice limits our understanding of the biology of nodules in FD and how to potentially treat them. The presence of nodules are an important indicator of FD progression, in fact one outcome measure in an active clinical study assessing the natural history of FD is data from a nodule impact questionnaire [Solyom et al. 2018]. Nodule reduction has also been used as an indicator of the efficacy of BMT [Ehlert et al. 2006]. The main clinical manifestation in

249 patients with mild FD is typically nodule formation and joint pain. Due to its similarity to JIA, misdiagnosis is common [Levade et al. 2014].

Polyarticular nodules are a common feature in arthritis [Sayah and English 2005].

Interestingly, the formation of nodules is also not widely apparent in some arthritis mouse models. In the Freund’s complete adjuvant (CFA) induction model of JIA, mice that were injected with CFA displayed significant inflammation but no nodules [Avau et al. 2014].

However, the mouse digits developed a swollen phenotype [Avau et al. 2014]. In the TNFα overexpressing mouse commonly used as a model for rheumatoid arthritis, significant joint inflammation has been reported [Li and Schwarz 2003]. Additionally, these mice develop shorter metacarpals, clubbed paws and occasional joint dislocation but no nodules [Li and

Schwarz 2003]. Crossing a T-cell receptor transgenic mouse that recognizes an epitope from bovine RNase with a NOD mouse resulted in a new mouse line that spontaneously develops inflammation and mimics human rheumatoid arthritis [Kouskoff et al. 1996]. This mouse develops severe joint destruction, swollen joints, and paw deformities [Kouskoff et al.

1996]. Collectively these studies indicate that the formation of nodules may also be absent from arthritic mouse models. However, it is also possible that the manifestation of nodules is different between humans and mice. Beyond nodule formation, other commonly reported arthritic-like signs are painful joints, and the degradation of cartilage and bone.

Asah1P361R/P361R mice have disorganized bone growth plates and defects in cartilage formation suggesting that they do manifest signs of polyarticular disease [Schuchman et al.

2016].

Another difference between the Asah1P361R/P361R mice and FD in humans is the absence of a cherry-red spot and corneal clouding. These two ocular phenotypes are also common in other LSDs [Hayasaka 1983]. In the human retina, there is an oval shaped structure called the macula which contains two or more layers of ganglion cells. Within the macula is a small depression called the fovea, a region of the eye where visual acuity is

250 highest due to a large concentration of photoreceptors and no ganglion cells. The ganglion cells of various LSD patients are filled with storage products. Phenotypically cellular storage in the ganglion cells in the macula results in the appearance of a red fovea or cherry-red- spot [Chen et al. 2014]. This phenomenon is absent in all rodents as their eyes contain a different composition of photoreceptors and no macula or fovea. Corneal clouding is attributed to substrate buildup within both the epithelial or stroma layer of the cornea [Chen et al. 2014]. In Asah1P361R/P361R mice we did not observe a clouding phenotype, but we did see presence of uveitis in the anterior chamber. Additionally, ultrastructure analyses in

Asah1P361R/P361R mice confirmed presence of storage pathology in various cells, and while we did not examine the epithelial and stromal layers of the cornea it is possible that storage pathology is also present in these regions.

The last limitation of Asah1P361R/P361R mice is the restricted modeling of FD due to the specific mutation these mice have. Asah1P361R/P361R mice are orthologous for the human

P362R mutation, which replicates the classical variant of FD [Li et al. 1999]. Our data in

Chapters 5 and 6 indicate that Asah1P361R/P361R mice develop phenotypes that are also seen in the more severe variants of FD. A recent cross-sectional study demonstrated that patients from reports published after the year 2000 lived longer than patients from studies published prior [Zielonka et al. 2017]. This study attributes the increased longevity to better treatments, improved diagnosis and increased identification of patients with mild variants of FD [Zielonka et al. 2017]. Patients with mild FD variants have been reported to live well into adulthood [Yu et al. 2018].

Since the Asah1P361R/P361R mice die at 9 weeks of age it is possible that some phenotypes most relevant to the attenuated forms of FD are masked or manifest at a later time.

251

7.3 Characterization of sphingolipids in Asah1P361R/P361R mice

7.3.1 Sphingolipid accumulation in Asah1P361R/P361R mice

In this thesis, we have identified the sphingolipids that accumulate in the lung, ocular, and hepatic system in the Asah1P361R/P361R mouse model of FD. This data builds on our previous sphingolipid measurements performed on the brain of Asah1P361R/P361R mice [Sikora et al. 2017]. Collectively these data highlight the differential changes in measured sphingolipids with respect to each tissue. In the case of ceramide, our data shows a pan increase in all ceramides measured. However, the relative abundance of the individual ceramide species shows a differential alteration depending on the organ or tissue type in

Asah1P361R/P361R mice.

7.3.2 Sphingolipids in the respiratory system

In Chapter 4, we quantified sphingolipids in BALF and lung tissue from

Asah1P361R/P361R and Asah1+/+ mice. In BALF from Asah1P361R/P361R mice we observed a reduction in C22:0 and C24:0 ceramides that coincided with an increase in the C24:1 ceramide. This contrasts with measurements from lung tissue where no alterations were found in the percent composition of ceramide. Another contrasting finding was the difference in ceramides accumulation. BALF samples from Asah1P361R/P361R mice had a ~300-fold increase versus ~12-fold increase in the lung tissue (Section 3.3.8) Additionally, our ultrastructure data showed the presence of macrophages full of storage vacuoles. On occasion, abnormal macrophages were found with a ruptured phenotype. Additionally, we could observe CTB-like material within the lung parenchyma. This data suggests that ceramide and other sphingolipids may be leaching out into the lung space, or that as the

252 foamy macrophages rupture, some of the engulfed contents such as ceramide are released into the lung.

Additionally, we noticed an increase in C24:1 ceramide both in total amount and percent composition in BALF samples from Asah1P361R/P361R mice. This observation also extended to SM where we saw increased accumulation and percent composition of SM24:0 and SM24:1. This suggests that a backlog of C24:1 may be metabolized to SM in an effort to curb ceramide accumulation.

7.3.3 Sphingolipids in the ocular system

Measurement of sphingolipids in retinal tissue revealed a different storage pattern from the lung and liver. Compared to aged matched controls the retina in Asah1P361R/P361R mice showed an ~8-fold increase in ceramide, and a ~2-fold increase in MHC (Section

4.3.9). SM was unchanged. It is important to note that in Asah1+/+ mice SM accounted for

>70% of total sphingolipids, and although there was no major change in SM levels between

Asah1+/+ and Asah1P361R/P361R mice, due to the significant elevation of ceramide and MHC in

Asah1P361R/P361R mice SM accounted for only ~30% of all sphingolipids measured. This clearly demonstrates the importance of ACDase in maintaining lipid balance in the retina.

Analysis of the relative abundance of ceramides in the retina of Asah1P361R/P361R mice revealed a decrease in the composition of the C16:0 and an increase in the C18:0. This pattern was similar in MHC where a decrease in the percentage of C16:0 led to an increase in C18:0.

In the lung, we did not measure MHC, so we cannot speculate to which pathway ceramides are preferentially metabolized. In retinal tissue, we did not detect any major changes in SM or C1P, suggesting that the accumulated ceramide appears to be primarily metabolized to glucosylceramide and or galactosylceramide. Glycosphingolipids are in high

253 abundance and are essential for nervous system development, so a backlog of ceramide may just preferentially trigger production of MHC due to enzyme substrate availability

[Robert et al. 2009]. Support for this hypothesis comes from LSDs where glycosphingolipids accumulate [Cogan and Kuwabara 1968]. A number of these diseases share similar ocular and CNS phenotypes with FD. Two examples are Gaucher disease, which accumulates galactosylceramide and Krabbe disease, which stores glucosylceramide. Both of these

LSDs display a variety of ocular manifestations including the presence of storage pathology, optic neuropathy and retinal disease [Emery et al. 1972; Brownstein et al. 1978; Seidova et al. 2009; McNeill et al. 2013; Chen et al. 2014]. While we cannot rule out the contribution of other sphingolipids in pathology, accumulation of glycosphingolipids certainly results in nervous system pathology [Hannun and Obeid 2017].

One caveat to our sphingolipid analyses is that we only assessed lipids in the retina, and therefore our data is not representative of the whole eye. This is important as studies investigating the retina, trabecular mesh and aqueous humor have demonstrated that different ocular structures contain their own respective lipid profiles [Brush et al. 2010;

Aljohani et al. 2013; Aljohani et al. 2014]. Another example is in a study where retinal function and sphingolipids were measured in the Cers1, Cers2, Cers4 mutant mice [Brüggen et al. 2016]. Beyond the changes in ERG response in these mutant mice, this study performed sphingolipid measurements specifically in the cornea and retina. Retinas from wild-type control mice contained a higher abundance of C16:0 and C18:0 ceramide whereas the cornea contained a higher abundance of C22:0, C24:0 and C24:1 ceramide [Brüggen et al. 2016].

One explanation for the difference in ceramide preference between the cornea and retina may be that the cornea acts as an important barrier in the eye, whereas the retina contains cells that link to the central nervous system. Therefore, an accumulation of VLC

254 ceramide may serve a structural role. Importantly, these studies demonstrate the complexity and diversity of lipid composition even within a single organ.

7.3.4 Sphingolipids in the hepatic system

In our liver analyses ceramides were measured in both liver tissue and hepatocyte- enriched cultures. Similar to our findings in the lung and retina, all ceramide species measured were significantly elevated. Compared to aged-matched controls Asah1P361R/P361R mice developed a ~25-fold increase in liver tissue and a ~17-fold increase in hepatocyte- enriched culture volume (Section 5.3.8 and Section 5.3.9). The C16:0 ceramide was the most elevated species in both the liver and hepatocytes from Asah1P361R/P361R mice. In fact, both C16:0 MHC and SM16:0 both were the highest species measured. Severe alteration was also present in relative abundance of ceramide. The percentage of C16:0 ceramide showed the greatest increase and C24:0 ceramide the largest decrease in percentage composition in both the liver and hepatocytes from Asah1P361R/P361R mice. Both SM and MHC showed an increase in the abundance of the C16:0/SM16:0 species.

From our transcriptome analyses we found a number of sphingolipid metabolizing enzymes that were upregulated including CerK and Ugcg (Section 5.3.12). CerK converts ceramide to C1P, and Ugcg is an enzyme that converts ceramide to glucosylceramide.

Upregulation of these two genes is reflected in our sphingolipid measurements as both C1P and MHC were both elevated in the liver. Unfortunately, C1P was not detectable in our hepatocytes but MHC was increased. Furthermore, we also reported downregulation of

Cers2 and Cers4 the enzymes that regulate acyl chain length of ceramide in both the de novo and salvage pathways [Mullen et al. 2012]. Cers4 is highly expressed in heart and liver and is important in the formation of C18:0-22:0 ceramides. Downregulation of this gene may explain the reduction of C22:0 species. Cers2 which has high expression in the liver is

255 important in forming C22:0-C26:0 ceramides. Downregulation of Cers2 could also explain the significant reduction in the C24:0 and C22:0 ceramides.

7.3.5 Complexity and limitations in understanding ceramide metabolism

In this thesis, we have catalogued the sphingolipids that accumulate in the lung, retina, and liver of Asah1P361R/P361R mice. Our data shows a pan increase in ceramides and other sphingolipid classes. Importantly, we have highlighted alterations in the relative abundance of various sphingolipid classes. The organs we measured also confirm that sphingolipid composition differs for each organ and tissue. Furthermore, the sphingolipid data from BALF may have potential value as a biomarker in assessing FD involvement in the lung. One example that is under development is a mass spectrometry approach to screen FD by measuring C26:0 ceramides in dry blood spots [Cozma et al. 2017].

While interest in sphingolipid biology has increased over the last decade, there is still a gap in our understanding of the function of different species within a sphingolipid class.

Our sphingolipid data is largely descriptive and functional interpretation should be taken with caution. One example that demonstrates the complexity of sphingolipids in tissue distribution is from a collaboration between our lab and Dr. Richard Drake’s laboratory at MUSC.

Kidneys from Asah1P361R/P361R mice were used to develop a matrix-assisted laser desorption ionization (MALDI) imaging technique for sphingolipids [Jones et al. 2014]. MALDI imaging provides on-tissue spatial localization of lipids, similar to what you might expect from IHC.

MALDI imaging of kidneys from Asah1P361R/P361R and Asah1+/+ mice revealed higher intensities in the mutants, but similarities in terms of sphingolipid localization [Jones et al.

2014]. S1P had a higher abundance in the medulla versus the cortex [Jones et al. 2014]. In ceramides, the C24:1 species was highly localized to the medulla, whereas the C16:0 species was localized to the cortex [Jones et al. 2014]. When MALDI imaging was repeated

256 on brain samples, changes in the spatial localization of sphingolipids in Asah1P361R/P361R mice were observed [Sikora et al. 2017]. For ceramides, the two most abundant species were

C16:0 and C18:0, and as expected Asah1P361R/P361R mice displayed higher intensities than

Asah1+/+ mice [Sikora et al. 2017]. In terms of localization, brains from both Asah1+/+ and

Asah1P361R/P361R mice showed higher abundance of C16:0 ceramide in the cerebellar granule cell layer and cerebral cortex, but the C16:0 ceramide intensity was lower in the cerebellar white matter [Sikora et al. 2017]. For C18:0 ceramide, equal distribution across the brain was found in Asah1+/+ mice, but in Asah1P361R/P361R mice C18:0 ceramide was primarily localized to the in corpus callosum and septal nuclei [Sikora et al. 2017]. These two studies demonstrate that different anatomical regions will contain different sphingolipid expression patterns. Secondly, they reveal that a deficiency in ACDase not only caused sphingolipids accumulation, but also modifications in sphingolipid localization and abundance.

The textbook view of sphingolipid metabolism consists of a basic schema where sphingolipids are metabolized by a few key enzymes (Figure 5). Unfortunately, the reality is much more complicated. Just for ceramide alone, it is estimated that over 28 unique enzymes can interact with it as either a substrate or a byproduct [Hannun and Obeid 2011].

Ceramide lie at the epicenter of sphingolipid metabolism, they are precursors to sphingomyelin, C1P, and the entire glycosphingolipid pathway. Furthermore, degradation of ceramide leads to Sph and S1P. Our data suggests that a backlog of sphingolipids may only occurs up to the ceramidase (CDase) step as we found a decrease in Sph in the liver, lung and brain [Yu et al. 2018]. There are five CDases that degrade ceramide. It is possible that

NCDase and the 3 alkaline CDases contribute to degrading some ceramides leading to the detection of Sph (Section 1.4). One hypothesis for this might be that ceramides that are normally targeted to the lysosome could be re-trafficked to organelles where the other 4

CDase are catalytically active.

257

Formation of ceramide in mammals is regulated in part by 6 ceramide synthases

(Section 1.3.5). They catalyze the acetylation of sphinganine with specific fatty acid chains lengths. Our data demonstrates that ACDase deficiency alters ceramide synthase expression. One major caveat to our data is the incomplete profiling of ceramide. This is largely a technical issue as there are limited ceramide standards, and sensitive MS techniques are restricted to specialized centers. One example that demonstrates this is a study where a specialized mass spectrometry technique identified 342 unique species of ceramide in the stratum corneum layer of human skin [Masukawa et al. 2008]. In addition to the large number of lipids, they also discovered a new class of ceramide that consisted of α- hydroxy fatty acid and dihydrosphingosine moieties [Masukawa et al. 2008]. Furthermore, a recent study using an updated lipidomic approach identified just under 1000 species of ceramides in the stratum corneum layer of the skin [Tsugawa et al. 2017]. Even though these two studies were performed on skin, the large number of species detected suggests that many more species are also likely present in other organs.

7.3.6 Technical limitations with our lipidomic analysis

The term lipidomic was first coined by Spener and colleagues in 2003 and refers to the quantitative measurement of lipids within cells, tissues, or an organism [Spener et al.

2003]. Advances to MS ionization modules and technological improvements to mass accuracy and resolution have greatly expanded our tools to study complex lipids [Gross

2017]. Furthermore, with increased interest in lipid biology many open resources have been established to facilities data analysis, protocol exchange, and lipid characterization such as

LIPID MAPS, Lipid Library and LipidBank [Fahy et al. 2009; Yasugi and Watanabe 2002;

Christie 2012]. Nonetheless, there are still several inherent challenges present in the study of lipids and sphingolipids. One of the first limitations that affects lipid researchers is

258 considerable variation in techniques used to extract lipids. Mammalian sphingolipids contain both extreme hydrophobic (1-deoxyceramides) and hydrophilic (some gangliosides) properties [Merrill and Sullards 2017]. Due to this caveat, there is no one protocol that can extract all sphingolipid classes with uniform yields. Another difficulty is the ongoing effort needed for method development when attempting to incorporate and characterize complex sphingolipids. For example, to measure some complex glycosphingolipids, samples may require multiple rounds of fragmentation to separate the glycosidic bonds and the ceramide backbone [Sullards et al. 2011]. However, the main fundamental problem for sphingolipidomic analysis is the broad number of compounds and their large number of isomers. A prime example are the glycosphingolipids where the same chemical formula is used (i.e. glucosylceramide and galactosylceramide). With regards to our study the last example is a clear limitation. A class of sphingolipids measured in the retina, liver, lung, and brain was MHC. MHC represents both glucosylceramide and glucosylceramide, and due to protocol and technical limitations we are unable to separate the two classes of glycosphingolipids. Nonetheless, MS technologies are improving and becoming instrumental for the study of lipid biology. As better methods are developed, a set of standardized protocols must be freely accessible to the community to ensure data is consistent and reproducible.

259

7.4 Improving understanding of FD

7.4.1 Data from FD mouse extends our understanding of human FD

To date, 152 cases of FD have been reported. Due to the scarcity of cases, and access to patient tissues for studies, experiments performed on the Asah1P361R/P361R mice are valuable for improving the understanding of ACDase deficiency. In this thesis, I have highlighted how ACDase deficiency can affect the respiratory, ocular and hepatic systems.

Furthermore, we created the Asah1P361R/P361R; MCP-1-/- mouse line to elucidate the role of

MCP-1. This later study demonstrates the value of Asah1P361R/P361R mouse for genetic and outcome studies. Without a doubt the Asah1P361R/P361R mouse line will continue to be an important tool for more in-depth mechanistic studies and importantly for evaluating efficacy of future therapies.

7.4.2 FD Diagnosis and screening technology

Currently definitive diagnosis of FD is achieved through biochemical testing for deficient ACDase enzyme activity. Enzyme activity tests are traditionally performed on cultured fibroblasts or leukocytes and are limited to specialized testing sites. Within the last few years reducing costs in sequencing technology has led to an increased use of whole exome sequencing (WES) as part of the diagnostic and screening process. Recently, several patients of FD and SMA-PME were identified this way [Teoh et al. 2016; Bao et al.

2017]. Furthermore, genetic sequencing has been important in the discovery of new variants of ACDase deficiency. For SMA-PME the connection with ASAH1 was made only after ruling out mutations in survival of motor neuron 1 (SMN1), the gene associated with classical SMA

[Zhou et al. 2012]. Additionally, one case that performed WES on a family of three long lived

FD patients identified a new osteoarticular phenotype of FD. In this later case mutations in

260

ASAH1 were identified after sequencing for metallopeptidase 2 and 4 (MMP2 and MMP4), genes associated with recessively inherited osteolysis yielded no pathological mutations

[Bonafé et al. 2016]. Taken together these cases demonstrate the value of performing genetic sequencing in patients with atypical symptoms.

7.4.3 Mutations causing ACDase deficiency

In Section 1.5 we highlighted the broad clinical spectrum of ACDase deficiency.

Mutations to ASAH1 can lead to a lethal childhood disorder, and on the other side of the spectrum a progressive neuromuscular disease. One hypothesis in our laboratory is that there may be a cluster of mutations specific for the different forms of ACDase deficiency. A patient in a recently published case report presented with muscle weakness common in

SMA-PME, but had also formed nodules as in classical FD [Teoh et al. 2016]. Sequencing results revealed a compound mutation that had not previously been reported [Teoh et al.

2016]. To investigate this possibility, we can curate the known ACDase mutations and map them on the predicted 3D structure of the protein.

From our literature search we identified 57 unique pathological mutations (Table 12) for FD, SMA-PME and the one case of Polyarticular Arthritis and SMA [Yu et al. 2018].

These mutations were curated from case reports, and the Human Gene Mutation Database

[Stenson et al. 2009]. The predicted 3D structure for the human and mouse ACDase were the same as those previously published [Alayoubi et al. 2013]. Mapping of the mutations revealed that the 27 mutations for FD were located on the beta chain, and 17 on the alpha chain (Figure 68). For SMA-PME, out a total of 11 unique mutations, 6 were on the alpha chain and 5 were on the beta chain. In total, the majority of mutations were found on the beta subunit. ACDase is first formed as a precursor protein predicted to be cleaved in an autoproteolytic manner through Cys143, Arg159 and Asp162, all of which are located on the

261 beta subunit [Shtraizent et al. 2008]. ASAH1 contains 6 glycosylation sites on the beta subunit, 3 of which are essential for the formation of the heteromeric enzyme [Ferlinz et al.

2001]. Based on these studies, mutations that are located near these sites would likely result in a deleterious effect on the enzyme. Currently, there is still limited data to demonstrate specific mutation hot spots, however as more FD patients are identified we will gain a better understanding of the genotype to phenotype relationship.

c.677G>C Missense Exon 9 p.R226P FD Heteroallelic NI 1 [Bashyam et al.262 2014]

Reported Mutation Amino Acid Allelic ACDase DNA Change Locus Phenotyp Cases Reference type Change Status Activity e

c.66G>C Missense Exon 1 p.Q22H FD NI NI 1 [Zhang et al. 2000]

c.67C>G Missense Exon 1 p.H23D FD NI NI 1 [Zhang et al. 2000]

c.77C>G Missense Exon 1 p.P26R SMA-PME Heteroallelic NI 1 [Behin et al. 2015]

c.92G>T Missense Exon 2 p.C31F FD Homoallelic NI 2 [Saygi et al. 2015; Cozma et al. 2017] Homoallelic & c.107A>G Missense Exon 2 p.Y36C FD NI 4 [Bär et al. 2001; Cozma et al. 2017] Heteroallelic Homoallelic & [Filosto et al. 2016; Sathe and Pearson c.124A>G Missense Exon 2 p.T42A SMA-PME <10% 4 Heteroallelic 2014; Cozma et al. 2017] [Zhou et al. 2012; Rubboli et al. 2015; Giráldez et al. 2015; Oguz Akarsu et al. c.125C>T Missense Exon 2 p.T42M SMA-PME Homoallelic 32% 12 2016; Johannsen et al. 2015; Yildiz et al. 2018] c.125+1G>A Insertion Intron 2 -- SMA-PME Heteroallelic NI 2 [Behin et al. 2015; Cozma et al. 2017] c.126- undetect Deletion Exon 3-5 p.Y42Rfs*10 FD Heteroallelic 1 [Alves et al. 2013] 3941_382+1358del able c.177C>G Nonsense Exon 3 p.Y59* SMA-PME Heteroallelic NI 1 [Rubboli et al. 2015]

c.223_224insC Insertion Exon 3 pV75Afs*6 SMA-PME Heteroallelic NI 1 [Rubboli et al. 2015]

c.212C>A Missense Exon 3 p.P71Q FD Heteroallelic NI 1 [Chikova et al. 2014]

c.256_257insA Insertion Exon 4 p.T86Nfs*13 FD Heteroallelic NI 1 [Bao et al. 2017]

c.290_292delTGG Deletion Exon 4 p.V96del FD Homoallelic 37% 1 [Muramatsu et al. 2002]

c.290T>A Missense Exon 4 p.V97E FD Heteroallelic 35% 1 [Muramatsu et al. 2002]

c.290T>G Missense Exon 4 p.V97G FD Homoallelic NI 2 [Chedrawi et al. 2012]

c.314T>C Missense Exon 4 p.L105P FD Heteroallelic NI 1 [Bao et al. 2017] c.383-16_383- Deletion Intron 5 -- FD Heteroallelic NI 1 [Bashyam et al. 2014] 12delTTTTC c.372T>A Missense Exon 6 p.D124E FD Heteroallelic NI 1 [Chikova et al. 2014]

c.408T>A Missense Exon 6 p.F136L FD Heteroallelic NI 1 [Bashyam et al. 2014]

c.412G>T Missense Exon 6 p.E139* FD Heteroallelic NI 1 [Bär et al. 2001] FD & Homoallelic & c.410A>G Missense Exon 6 p.Y137C NI 2 [Kernohan et al. 2017; Cozma et al. 2017] SMA-PME Heteroallelic c.410_411delAT Deletion Exon 6 p.Y137* FD Heteroallelic NI 1 [Torcoletti et al. 2014] Homoallelic & [Li et al. 1999; Bär et al. 2001; Zhang et al. c.413A>T Missense Exon 6 p.E138V FD <5% 4 Heteroallelic 2012] [Bär et al. 2001; Rubboli et al. 2015; c.456A>C Missense Exon 6 p.K152N SMA-PME Heteroallelic <20% 5 Kernohan et al. 2017; Cozma et al. 2017] [Muranjan et al. 2012; Bashyam et al. c.457+4A>G Splicing Intron 6 -- FD Homoallelic NI 2 2014] undetect c.502G>T Missense Exon 7 p.G168W FD Homoallelic 1 [Cvitanovic-Sojat et al. 2011] able Homoallelic & [Bashyam et al. 2014; Bonafé et al. 2016; c.505T>C Missense Exon 8 p.W169R FD <10% 7 Heteroallelic Cozma et al. 2017; Al Jasmi 2012] Polyarticul c.518A>T Missense Exon 8 p.N173I ar Arthritis Heteroallelic <10% 1 [Teoh et al. 2016] and SMA [Cozma et al. 2017; Sathe and Pearson c.536C>T Missense Exon 8 p.T179I SMA-PME Heteroallelic NI 3 2014] c.538G>A Missense Exon 8 p.E180K FD Heteroallelic NI 1 [Bashyam et al. 2014]

c.544C>G Missense Exon 8 p.L182V FD Homoallelic NI 4 [Devi et al. 2006; Bashyam et al. 2014]

c.593T>C Missense Exon 8 p.V198A FD Heteroallelic NI 1 [Bashyam et al. 2014] Polyarticul c.594_599dupCTT Duplicatio Exon 8 F199_K200dup ar Arthritis Heteroallelic <10% 1 [Teoh et al. 2016] CAA n and SMA c.665C>A Missense Exon 9 p.T222K FD Homoallelic <5% 1 [Koch et al. 1996; Bär et al. 2001]

263

Homoallelic & [Bashyam et al. 2014; Muramatsu et al. c.703G>C Missense Exon 9 p.G235A FD 2% 3 Heteroallelic 2002] c.704G>A Missense Exon 9 p.G235D FD Heteroallelic NI 1 [Torcoletti et al. 2014]

c.704-2A>G Splicing Exon 9 -- FD Homoallelic NI 1 [Cozma et al. 2017]

Homoallelic & [Kostik et al. 2013; Bonafé et al. 2016;Li et c.760A>G Missense Exon 10 p.R254G FD <10% 3 Heteroallelic al. 1999; Chikova et al. 2014]

c.770T>C Missense Exon 10 p.L257P FD Homoallelic NI 1 [Hugle et al. 2014]

c.833C>T Missense Exon 11 p.P278L FD Homoallelic NI 1 [Levade et al. 2014]

c.850G>T Nonsense Exon 11 p.G284X SMA-PME Heteroallelic <10% 1 [Dyment et al. 2014]

c.886C>T Missense Exon 11 p.R296X SMA-PME Heteroallelic <20% 1 [Gan et al. 2015]

c.917+4A>G Splicing Intron 11 -- FD Heteroallelic NI 1 [Alves et al. 2013]

c.958A>G Missense Exon 12 p.N320D FD Homoallelic <15% 1 [Bär et al. 2001]

c.959A>G Missense Exon 12 p.N320S FD Homoallelic NI 1 [Bashyam et al. 2014]

c.991G>A Missense Exon 12 p.D331N FD Heteroallelic NI 1 [Bär et al. 2001] Homoallelic & c.997C>T Missense Exon 12 p.P333C FD NI 3 [Cozma et al. 2017; Kim et al. 2016] Heteroallelic c.997C>G Missense Exon 12 p.P333G FD Heteroallelic NI 4 [Bashyam et al. 2014; Cozma et al. 2017]

c.998G>A Missense Exon 12 p.P333H FD Homoallelic NI 1 [Bashyam et al. 2014]

c.1085C>G Missense Exon 13 p.P362R FD Homoallelic <5% 2 [Li et al. 1999]

c.1084C>A Missense Exon 13 p.P362T FD Heteroallelic NI 1 [Bashyam et al. 2014]

c.1096A>C Missense Exon 13 p.K366Q FD Heteroallelic NI 2 [Al Jasmi 2012; Cozma et al. 2017]

c.1105G>A Missense Exon 13 p.V369I FD Heteroallelic NI 1 [Muramatsu et al. 2002]

c.1098+1G>T Splicing Intron 13 p.N348_K366del FD Heteroallelic NI 1 [Bär et al. 2001] c.1186_1187insT Insertion Exon 14 p.*396L FD NI NI 1 [Zhang et al. 2000]

Table 12. Pathological mutations in ACDase deficiency

List of the known published mutations that result in ACDase deficiency. The list details the type of mutation, the change in amino acid, allelic status, enzyme activity level, number of reported patients with mutation and the reference cited. NI- no information. Adapted from [Yu et al. 2018]

264

Figure 68. Mutations mapped to predicted ACDase structure

Mutations curated in Table 12 are annotated in the predicted mouse (A) and human (B) ACDase. Red amino acids are mutations that represent FD, lavender amino acids represent cases of SMA-PME, and orange amino acids are for the single case of FD and SMA-PME and polyarticular arthritis. The alpha subunit (left) is in labelled in green, and the beta subunit (right) is in blue. Adapted from [Alayoubi et al. 2013].

265

7.4.4 Improved FD diagnosis and treatment

For patients who manifest attenuated FD, delayed diagnosis or misdiagnosis is unfortunately a key concern [Solyom et al. 2015; Zielonka et al. 2017]. This thesis has demonstrated the role of MCP-1 in the pathogenesis of FD (Chapter 6) [Yu et al. 2018].

MCP-1 has previously been shown to be dramatically elevated in both plasma samples from

Asah1P361R/P361R mice and FD patients [Dworski et al. 2017]. Furthermore, MCP-1 is absent in JIA plasma samples demonstrating its potential as a biomarker for distinguishing FD from

JIA [Dworski et al. 2017]. Work on the Asah1P361R/P361R;MCP-1-/- mice has revealed that they produce high levels of MCP-3, and MCP-5 [Yu et al. 2018]. While we will need to validate

MCP-3 and MCP-5 levels in patients, it is possible that these and other chemokines in our analyses are also applicable as biomarkers for FD. Additionally, combining multiple cytokines in a single test may increase confidence when screening patients for FD.

Current treatment for FD is based on addressing patient symptoms. Beyond this,

HSCT has been performed in select patients [Ehlert et al. 2006]. Early attempts of HSCT in patients with classical FD showed modest efficacy in reducing nodule size, however, due to neurological involvement, these patients continued to deteriorate and eventually died

[Souillet et al. 1989; Yeager et al. 2000]. Work in the Asah1P361R/P361R mice has demonstrated that neurological pathology begins as early as 3 weeks of age [Sikora et al.

2017]. This pathology is also present in the liver and hematopoietic organs. Treatment with

HSCT would reduce inflammation in the peripheral and visceral organs, however, if tissue damage has already occurred HSCT would only halt further damage. The difficulty with treatments only effective with early diagnoses can be seen in another LSD, Fabry disease, where a study found that out of 45 patients, the average time to obtain a correct diagnosis was 19.7 years [Marchesoni et al. 2010]. Patients who are diagnosed later in life would likely

266 not respond as well to treatment because of irreversible tissue damage [Rombach et al.

2013].

In the context of FD, early diagnosis and intervention is even more imperative as symptoms can manifest during infancy. One approach would be to perform HSCT shortly after birth before manifestation of symptoms. This strategy has been demonstrated to be effective in globoid cell leukodystrophy where patients were treated with umbilical-cord blood either as asymptomatic newborns between 12 to 44 days old or symptomatic infants between 142 to 352 days old [Escolar et al. 2005]. A three-year follow-up showed normal enzyme levels and neurodevelopment in most of the patients that were treated when asymptomatic [Escolar et al. 2005]. Whereas infants who underwent treatment after onset of symptoms showed minimal neurological improvements [Escolar et al. 2005]. Another potential avenue is to intervene prior to birth in utero. One study showed that administration of human mononuclear blood-cord cells into a heterozygous mouse model for the LSD

Sanfilippo type B 5 days into pregnancy allowed injected cells to transmigrate and diffuse into the embryo [Garbuzova-Davis et al. 2006]. Tests performed on E12.5 embryos showed presence of the transplanted cells throughout the brain and liver, and correction of enzyme deficiency up to heterozygous levels [Garbuzova-Davis et al. 2006]. While this study was only proof of concept, it nonetheless demonstrated that early intervention is possible.

However, further advances in FD screening will be required in order for such a therapeutic approach to be a possibility.

7.4.5 Current and emerging research in ACDase deficiency

It is an exciting time in ACDase deficiency research. Comprehensive studies have been completed on Asah1P361R/P361R mice. These studies help to increase our understanding of the pathological role of ACDase deficiency in various organs. Particularly we have been

267 able to highlight the specific sphingolipids classes that accumulate and identify cytokines that drive inflammation.

Other developments in FD research are occurring outside our group. One study used a mass spectrometry approach to show a proof of concept that C26:0 ceramide was a potential biomarker for identifying FD patients [Cozma et al. 2017]. This group not only showed the possibility of differentiating classical versus mild FD patients but also demonstrated that their methods were applicable on dry blood spots [Cozma et al. 2017].

These new biomarker targets are important findings as they may mean that newborn screening could become available soon for FD.

Another study of interest is a clinical observational protocol aiming to document the natural history of Farber disease in 40 patients over a period of 21 months. The study will include a mixture of patient who have and have not undergone HSCT. As of the date of writing this thesis, the trial is still recruiting patients and aims to be complete in late 2019

(clinicaltrials.gov ID NCT03233841) [Solyom et al. 2018].

Finally, within the last few years there has been ongoing effort in developing human recombinant ACDase (rhACDase) for FD. A proof of concept study has already been completed using the Asah1P361R/P361R mice as a model. This study showed that administration of rhACDase as early as 3 days of age could modestly increase lifespan, normalize spleen size, reduced MCP-1 levels and significantly reduce in ceramide within tissue [He et al. 2017]. Pharmacokinetic studies have demonstrated that the maximally

P361R/P361R effective dose of rhACDase in Asah1 mice is 10 mg/kg, which led to a Cmax value of

1.25 µg/ml [Gaukel et al. 2018a]. Distribution studies revealed that rhACDase was taken up in a number of tissues, but the highest concentration was found in the liver and spleen at more than 25 times that of serum [Gaukel et al. 2018a]. Work on predicting the human equivalent dosage of rhACDase has revealed that a 10mg/kg dose in a mouse is roughly equivalent to 3-5mg/kg dose in humans or a 4-5mg/kg dose in a 15kg child [Gaukel et al.

268

2018b]. Orphan designation for rhACDase has already been granted by the US and

European regulatory agencies and plans for a clinical trial are underway. rhACDase is expected to be an effective treatment for patients with mild and intermediate forms of FD.

Based on the work in this thesis, we would anticipate that rhACDase would be most effective in the liver, followed by the lung, and least effective in the ocular and neurological system due to the inability of the enzyme to penetrate the blood brain barrier. Our group has previously seen that administration of human ACDase-encoded lentiviral vectors into

Asah1P361R/P361R neonate mice could impede inflammation and increase life span [Alayoubi et al. 2013]. Future and ongoing work in our lab will be focused on ex vivo and other novel gene therapy approaches.

269

Chapter 8 Conclusions

ACDase deficiency is a complicated disease that manifests as either a lethal childhood LSD or a debilitating neuromuscular disorder. Regardless of the variant, ACDase deficiency leads to significant disease and early death. In this thesis, I have demonstrated the presence of chronic lung injury, impaired ocular function, and progressive liver disease in Asah1P361R/P361R mice, our mouse model of ACDase deficiency.

In these mice, a common theme shared amongst these organ systems is inflammatory involvement. Recruitment of neutrophils and macrophages is progressive and a main contributor to organ dysfunction. Lipidomic analyses have revealed sphingolipid accumulation and alterations to the composition of different sphingolipid classes.

Our work has also highlighted significant changes to the liver transcriptome.

Upregulation of inflammatory pathways increase our understanding of genes that contribute to the inflammatory phenotype. The downregulated pathways associated with lipid hemostasis demonstrate that insults to the sphingolipid pathway can have significant consequences to other lipid pathways. Furthermore, within the sphingolipid metabolic pathway we have highlighted the roles of Cers2 and Cers4 in contributing to the transformed sphingolipid abundance in the liver of Asah1P361R/P361R mice. Additionally, upregulation of

CerK and Ugcg, two genes that encode enzymes that metabolize ceramide into other sphingolipids, indicates two pathways that regulate the backlog of ceramide.

MCP-1 is significantly upregulated in both Asah1P361R/P361R mice and FD patients. It has been proposed as a potential biomarker to differentiate FD patients and those with JIA.

In Chapter 6, we demonstrated that ablation of MCP-1 can delay pathology and provide a modest increase in lifespan in Asah1P361R/P361R;MCP-1-/- double mutant mice. Interestingly, deletion of MCP-1 provided the most improvement in the lung, followed by the liver and no

270 correction to the brain and hematopoietic organs. Decreased inflammation was noted in the lung and liver as well as decreased ceramide accumulation, suggesting that inflammation may directly impact ceramide production.

Cytokine analyses of double mutants revealed a change in cytokine profile.

Furthermore, in this project we also discovered that MCP-3 and MCP-5 are significantly elevated in the Asah1P361R/P361R mice which may explain the continued inflammation seen in

Asah1P361R/P361R;MCP-1-/- mice. Our work has demonstrated that targeting MCP-1 in conjunction with other therapies may improve some clinical manifestations of FD.

ACDase deficiency is a complex disorder. This thesis has provided insights into the role of ACDase deficiency in various organs, the sphingolipids that are most affected, and gene candidates for future targeted projects. While FD and SMA-PME are both ultra rare disorders, there is a broad interest in ACDase and ceramides as they are relevant in other more common diseases such as diabetes, and cancer.

Recent developments in ACDase deficiency include establishing the link between

ASAH1 mutations and SMA-PME, newly proposed biomarkers of disease and diagnostic tools, increased effort in improving the awareness of FD, and development of rhACDase for

ERT. Collectively, this body of work validates Sir Archibald Garrod’s wish when he stated

“Let us hope there will always be some who will seek to guess the riddles and to learn the lessons of rare maladies” during his address to the Medical Society of London in 1928.

271

Chapter 9 Future Directions

The work described in this thesis provides an increased understanding of the pathology of FD but also proposes many questions for consideration. Additionally, the results from these studies will be applicable to assessing the efficacy of future therapeutics for ACDase deficiency.

9.1 Pathobiology of the Asah1P361R/P361R mouse

One of the main phenotypes that has to date not been thoroughly characterized is the skin of Asah1P361R/P361R mice. By 3-4 weeks of age Asah1P361R/P361R mice may be distinguished from heterozygous and homozygous littermates by the rigidity of their skin.

This phenotype has been touched upon in Section 6.3.1 where we showed that the scruff

(back skin stretch) length in Asah1P361R/P361R mice was significantly shorter than aged matched controls. As the Asah1P361R/P361R mice age, their skin develops a tighter phenotype.

In the literature one case described a patient with a phenotype similar to tight skin syndrome

[El-Kamah et al. 2009]. Other cases have reported the presence of fibrosis, granulomas and storage pathology in skin biopsies [Schmoeckel 1980; Burck et al. 1985; Zappatini-Tommasi et al. 1992]. More common dermatologic diseases such as psoriasis, atopic dermatitis, and ichthyosis have been associated with altered sphingolipid composition and metabolism

[Moskot et al. 2018]. Studying the skin phenotype will further improve the understanding FD.

Additionally, analyses of the skin would offer insight into ACDase in skin biology that might be translated to the aforementioned skin disorders. In fact, a study focused on the skin of alkaline ceramidase 1 (Acer1) deficient mice demonstrated accumulation of ceramides in skin, altered sebaceous gland architecture and increased transepidermal water loss

272

[Liakath‐Ali et al. 2016]. Due to the robust skin phenotype, and related function of ACDase it is possible that similar phenotypes are present in our mice. To this end, characterization of the skin pathology and sphingolipid accumulation in the various skin layers should be performed. Additionally, targeted tests to assess water-retention and permeability-barrier function as ceramide and Sph would be important to address any defects to the lipid barrier.

Significant motor and behavioral impairments have also been demonstrated in the

Asah1P361R/P361R mice [Sikora et al. 2017]. FD Patients that display neurological involvement develop ambulatory issues and may be wheelchair bound later in life [Eviatar et al. 1986].

Thus, one area that should be explored in future Asah1P361R/P361R mouse studies is peripheral nervous system pathology, specifically an analysis of neurons that innervate muscles, and musculature itself. Observationally some Asah1P361R/P361R mice appear to develop a mild hind limb paralysis, perform poorly on strength tests, and show signs of wasting. As SMA-PME is a neuromuscular disease, pathology present in Asah1P361R/P361R mice may help in understanding SMA-PME. In addition to pathological analyses, physiological tests should be performed to assess the rate of nerve conduction velocity as some cases of FD have reported the presence of nerve compression due to excessive storage in Schwann cells [Zappatini-Tommasi et al. 1992]. Lastly in terms of muscle, one attenuated FD patient with no neurological deficits developed severe muscle atrophy and weakness [Fiumara et al. 1993]. Therefore, a combination of biochemical tests to assess for myositis and physiological tests such as muscle force measurements and electromyography should be considered.

273

9.2 Studies on ACDase protein trafficking and folding

We have documented 57 pathological mutations in ASAH1 (Table 12). While enzyme activity studies have been completed on various mutants, there have not been detailed studies conducted on enzyme trafficking for any of the known FD mutations. ACDase is a lysosomal enzyme. ACDase is synthesized as a precursor protein in the ER, after addition of

N-linked glycosylation signals, the precursor is subjected to ER quality control. Following folding, the enzyme undergoes further processing in the Golgi, and finally traffics to the lysosome where it is cleaved in an autoproteolytic manner [Ferlinz et al. 2001]. As FD displays a broad clinical spectrum, different mutations could impact enzyme trafficking along any of the above steps. In the case of Gaucher disease, there have been over 300 mutations identified, and many mutant variants are recognized as misfolded proteins [Maor et al. 2013]. In a study of Sanfilippo syndrome type C, 17 of 21 mutations led to misfolding of the enzyme heparan sulfate acetyl-CoA [Feldhammer et al. 2009].

A future project that would greatly benefit the FD field is to assess ACDase misfolding and impaired trafficking. A potential experiment would use classic, mild, and

SMA-PME patient fibroblasts and perform confocal immunofluorescence microscopy. Cells could be stained for the lysosomal-late endosomal compartment using LysoTracker Red, ER with Calnexin, the lysosome with Lamp-1 or Lamp-2, and lastly an antibody for ACDase.

This experiment might give us some insight into whether mutant forms of ACDase get stuck in the ER or another compartment due to misfolding. Accumulation of misfolded proteins in the ER has been shown to trigger a signaling cascade known as unfolded protein response

(UPR) [Cao and Kaufman 2012]. If the preliminary experiment suggests impaired enzyme trafficking, it would also be important to explore this mechanism further. Misfolded proteins are translocated to the cytoplasm where they consequently undergo degradation in a process known as ER associated degradation (ERAD). In Gaucher disease, the level of

274

ERAD has been correlated with disease severity [Maor et al. 2013]. It would be interesting to see if this relationship is also present in ACDase deficiency and whether the degree of

ERAD correlates with enzyme activity in Asah1P361R/P361R mice.

275

9.3 Patient specific induced pluripotent stem cells

Identification and quantification of sphingolipids in the Asah1P361R/P361R mouse has been a focus in this thesis. While it is informative at a systemic level, the data generated represents the sphingolipid from a mixture of cell types in each organ studied. This rationale was one of the reasons for isolating primary hepatocytes in the liver. Unfortunately, culturing primary cells from different organs and obtaining a pure population is technically challenging and not feasible in terms of human patients. A potential follow-up project would be to generate induced pluripotent stem cells (iPSC) from patient fibroblasts. These patient derived iPSCs could be obtained from patients with different variants of FD and SMA-PME.

If it proves difficult to obtain these samples, an alternative strategy would be to take existing fibroblasts or iPSC cell lines and perform genome editing with an engineered nuclease like the CRISPR-Cas9 system to induce the known mutations (Table 12). The value in these iPSC cell lines would be the ability to study the role of ASAH1 mutations in various differentiated cells. This project could focus on differentiating cells that are most severely affected such as neurons, Schwann cells, chondrocytes and monocytes/macrophages.

Using LS-MS we could further assess the sphingolipid profiles of the different cell types. We could also use real-time PCR to assess changes in sphingolipid metabolizing enzymes.

Importantly acquiring iPSC may allow us to test and screen small molecules potentially useful in novel chaperon or substrate reduction therapies.

276

9.4 Ex vivo gene therapy for FD

In our original description of the Asah1P361R/P361R mouse, we demonstrated that direct administration of lentivirus expressing ACDase in Asah1P361R/P361R neonate mice provided a modest benefit by decreasing both ceramide accumulation and inflammation however it did not normalize the lifespan of the mouse [Alayoubi et al. 2013]. An explanation for the limited success could be due to an inability of the enzyme to penetrate the blood brain barrier and reach regions beyond the peripheral organs.

In this future project, HSCs would be isolated from wild-type mice and transduced with lentivirus expressing ACDase. After Asah1P361R/P361R mice have had their bone marrow conditioned, transduced cells would be transplanted to allow for BM engraftment. The anticipated results would be production of supraphysiologial levels of enzyme by the daughter cells [Rastall and Amalfitano 2015]. This treatment may also provide some benefit to the CNS. One recent clinical protocol using this approach to treat metachromatic leukodystrophy showed that patients treated with LV transduced HSCs had a clear therapeutic benefit [Biffi et al. 2013]. In this study, brain MRI scans performed on treated patients after a two-year follow-up were largely normal compared to untreated patient, where demyelination and diffuse atrophy is common [Biffi et al. 2013].

A large proportion of the pathology that has been characterized in this thesis will be highly applicable as we can repeat many of these physiologic tests to assess how various organs responded to treatment. The pathology in Asah1P361R/P361R mice is severe, it is my opinion that even if a treatment shows a modest improvement in this model it should not be discarded as it may in fact provide a more potent therapeutic benefit in the many less severe forms of ACDase seen in the clinic.

277

Finally, our research group has established expertise in ex vivo gene therapy for another treating LSDs [Huang et al. 2017]. Most relevant is an ongoing ex vivo gene therapy tial for the treatment of Fabry disease (clinicaltrials.gov ID NCT02800070). Our ultimate goal would be to apply this platform technology first in our murine model, and then transition this treat option to ACDase deficient patients.

278

References

Abenoza, P., and R. K. Sibley. 1987. Farber's disease: a fine structural study. Ultrastructural pathology 11(4): 397-403.

Abul-Haj, S. K., D. G. Martz, W. F. Douglas, and L. J. Geppert. 1962. Farber's disease: Report of a case with observations on its histogenesis and notes on the nature of the stored material. The Journal of pediatrics 61(2): 221-232.

Adams, J. M.,2nd, T. Pratipanawatr, R. Berria, E. Wang, R. A. DeFronzo, M. C. Sullards, and L. J. Mandarino. 2004. Ceramide content is increased in skeletal muscle from obese insulin-resistant humans. Diabetes 53(1): 25-31.

Aerts, J. M., C. E. Hollak, R. G. Boot, J. E. Groener, and M. Maas. 2006. Substrate reduction therapy of glycosphingolipid storage disorders. Journal of inherited metabolic disease 29(2): 449-456.

Aerts, J. M., W. W. Kallemeijn, W. Wegdam, M. J. Ferraz, M. J. van Breemen, N. Dekker, G. Kramer, B. J. Poorthuis, J. E. Groener, and J. Cox-Brinkman. 2011. Biomarkers in the diagnosis of lysosomal storage disorders: proteins, lipids, and inhibodies. Journal of inherited metabolic disease 34(3): 605- 619.

Ahmad, A., A. U. Mazhar, and M. Anwar. 2009. Farber disease: a rare neurodegenerative disorder. J Coll Physicians Surg Pak 19(1): 67-68.

Aimo, L., R. Liechti, N. Hyka-Nouspikel, A. Niknejad, A. Gleizes, L. Göttz, D. Kuznetsov, F. P. David, van der Goot, F Gisou, and H. Riezman. 2015. The SwissLipids knowledgebase for lipid biology. Bioinformatics 31(17): 2860-2866.

Al Jasmi, F. 2012. A novel mutation in an atypical presentation of the rare infantile Farber disease. Brain and Development 34(6): 533-535.

Alayoubi, A. M., J. C. Wang, B. C. Au, S. Carpentier, V. Garcia, S. Dworski, S. El-Ghamrasni, et al. 2013. Systemic ceramide accumulation leads to severe and varied pathological consequences. EMBO molecular medicine 5(6): 827-842.

Albert, D. M., M. Lahav, L. E. Carmichael, and D. H. Percy. 1976. Canine herpes-induced retinal dysplasia and associated ocular anomalies. Investigative ophthalmology & visual science 15(4): 267- 278.

Albert, D. M., M. Lahav, E. D. Colby, J. A. Shadduck, and D. N. Sang. 1977. Retinal neoplasia and dysplasia. I. Induction by feline leukemia virus. Investigative ophthalmology & visual science 16(4): 325-337.

Aldenhoven, M., J. Boelens, and T. J. de Koning. 2008. The clinical outcome of Hurler syndrome after stem cell transplantation. Biology of Blood and Marrow Transplantation 14(5): 485-498.

Aldenhoven, M., P. Orchard, J. Kurtzberg, R. Wynn, A. O'Meara, P. Veys, A. Fischer, V. Valayannopoulos, B. Neven, and A. Rovelli. 2014. Predictors of Long-Term Clinical Outcome in Hurler Syndrome Patients after Successful Hematopoietic Cell Transplantation: An International Study. Biology of Blood and Marrow Transplantation 20(2): S79.

279

Aljohani, A. J., G. Edwards, Y. Guerra, S. Dubovy, D. Miller, R. K. Lee, and S. K. Bhattacharya. 2014. Human trabecular meshwork sphingolipid and ceramide profiles and potential latent fungal commensalism. Investigative ophthalmology & visual science 55(6): 3413-3422.

Aljohani, A. J., G. C. Munguba, Y. Guerra, R. K. Lee, and S. K. Bhattacharya. 2013. Sphingolipids and ceramides in human aqueous humor. Molecular vision 19: 1966.

Alves, M. Q., E. Le Trionnaire, I. Ribeiro, S. Carpentier, K. Harzer, T. Levade, and M. G. Ribeiro. 2013. Molecular basis of acid ceramidase deficiency in a neonatal form of Farber disease: Identification of the first large deletion in ASAH1 gene. Molecular genetics and metabolism 109(3): 276-281.

Amirhakimi, G. H., P. Haghighi, M. A. Ghalambor, and S. Honari. 1976. Familial lipogranulomatosis (Farber's disease). Clinical genetics 9(6): 625-630.

Antonarakis, S., D. Valle, H. W. Moser, A. Moser, S. J. Qualman, and W. H. Zinkham. 1984. Phenotypic variability in siblings with Farber disease. The Journal of pediatrics 104(3): 406-409.

Antonarakis, S., D. Valle, H. Moser, W. Zinkham, and S. Qualman. 1983. Farber’s lipogranulomatosis: variability of expression and clinical overlap with histiocytosis. Pediatr.Res 17: 206A.

Arana, L., M. Ordonez, A. Ouro, I. G. Rivera, P. Gangoiti, M. Trueba, and A. Gomez-Munoz. 2013. Ceramide 1-phosphate induces macrophage chemoattractant protein-1 release: involvement in ceramide 1-phosphate-stimulated cell migration. American journal of physiology.Endocrinology and metabolism 304(11): 1213.

Arana, L., P. Gangoiti, A. Ouro, M. Trueba, and A. Gómez-Muñoz. 2010. Ceramide and ceramide 1- phosphate in health and disease. Lipids in health and disease 9(1): 15.

Avau, A., T. Mitera, S. Put, K. Put, E. Brisse, J. Filtjens, C. Uyttenhove, J. Snick, A. Liston, and G. Leclercq. 2014. Systemic Juvenile Idiopathic Arthritis–like Syndrome in Mice Following Stimulation of the Immune System With Freund's Complete Adjuvant: Regulation by Interferon‐γ. Arthritis & rheumatology 66(5): 1340-1351.

Azuma, H., N. Paulk, A. Ranade, C. Dorrell, M. Al-Dhalimy, E. Ellis, S. Strom, M. A. Kay, M. Finegold, and M. Grompe. 2007. Robust expansion of human hepatocytes in Fah−/−/Rag2−/−/Il2rg−/− mice. Nature biotechnology 25(8): 903.

Azuma, N., J. S. Obrien, H. W. Moser, and Y. Kishimoto. 1994. Stimulation of acid ceramidase activity by saposin D. Archives of Biochemistry and Biophysics 311(2): 354-357.

Baeck, C., A. Wehr, K. R. Karlmark, F. Heymann, M. Vucur, N. Gassler, S. Huss, et al. 2012. Pharmacological inhibition of the chemokine CCL2 (MCP-1) diminishes liver macrophage infiltration and steatohepatitis in chronic hepatic injury. Gut 61(3): 416-426.

Balamayooran, G., S. Batra, T. Balamayooran, S. Cai, and S. Jeyaseelan. 2011. Monocyte chemoattractant protein 1 regulates pulmonary host defense via neutrophil recruitment during Escherichia coli infection. Infection and immunity 79(7): 2567-2577.

Banugaria, S. G., S. N. Prater, Y. Ng, J. A. Kobori, R. S. Finkel, R. L. Ladda, Y. Chen, A. S. Rosenberg, and P. S. Kishnani. 2011. The impact of antibodies on clinical outcomes in diseases treated with therapeutic protein: lessons learned from infantile Pompe disease. Genetics in Medicine 13(8): 729-736.

280

Banugaria, S. G., S. N. Prater, T. T. Patel, S. M. DeArmey, C. Milleson, K. B. Sheets, D. S. Bali, C. W. Rehder, J. A. Raiman, and R. A. Wang. 2013. Algorithm for the early diagnosis and treatment of patients with cross reactive immunologic material-negative classic infantile pompe disease: a step towards improving the efficacy of ERT. PLoS One 8(6): e67052.

Bao, X. H., J. M. Tian, T. Y. Ji, and X. Z. Chang. 2017. A case report of childhood Farber's disease and literature review. Zhonghua er ke za zhi = Chinese journal of pediatrics 55(1): 54-58.

Bär, J., T. Linke, K. Ferlinz, U. Neumann, E. H. Schuchman, and K. Sandhoff. 2001. Molecular analysis of acid ceramidase deficiency in patients with Farber disease. Human mutation 17(3): 199- 209.

Barriere, H., and F. Gillot. 1973. La lipogranulomatose de Farber. Nouv Presse Med 2: 767.

Barton, N. W., R. O. Brady, J. M. Dambrosia, A. M. Di Bisceglie, S. H. Doppelt, S. C. Hill, H. J. Mankin, et al. 1991. Replacement therapy for inherited enzyme deficiency—macrophage-targeted glucocerebrosidase for Gaucher's disease. New England Journal of Medicine 324(21): 1464-1470.

Bashyam, M. D., A. K. Chaudhary, M. Kiran, V. Reddy, H. A. Nagarajaram, A. Dalal, L. Bashyam, D. Suri, A. Gupta, and N. Gupta. 2014. Molecular analyses of novel ASAH1 mutations causing Farber lipogranulomatosis: analyses of exonic splicing enhancer inactivating mutation. Clinical genetics 86(6): 530-538.

Beck, M. 2007. New therapeutic options for lysosomal storage disorders: enzyme replacement, small molecules and gene therapy. Human genetics 121(1): 1-22.

Beck, M., H. Moser, and K. sandhoff. 2014. Acid Ceramidase Deficiency: Farber Lipogranulomatosis and Spinal Muscular Atrophy Associated with Progressive Myoclonic Epilepsy (35): 395-402.

Bedia, C., L. Camacho, J. L. Abad, G. Fabriàs, and T. Levade. 2010. A simple fluorogenic method for determination of acid ceramidase activity and diagnosis of Farber disease. Journal of lipid research 51(12): 3542-3547.

Bedia, C., J. Casas, N. Andrieu-Abadie, G. Fabriàs, and T. Levade. 2011. Acid ceramidase expression modulates the sensitivity of A375 melanoma cells to dacarbazine. Journal of Biological Chemistry 286(32): 28200-28209.

Bedia, C., J. Casas, V. Garcia, T. Levade, and G. Fabriàs. 2007. Synthesis of a Novel Ceramide Analogue and its Use in a High‐Throughput Fluorogenic Assay for Ceramidases. Chembiochem 8(6): 642-648.

Behin, A., I. Nelson, G. Bonne, N. Romero, and R. Froissart. 2015. Mutations in ASAH1 may cause spinal muscular atrophy. Neuromuscular Disorders (25): S225.

Bektas, M., M. L. Allende, B. G. Lee, W. Chen, M. J. Amar, A. T. Remaley, J. D. Saba, and R. L. Proia. 2010. Sphingosine 1-phosphate lyase deficiency disrupts lipid homeostasis in liver. Journal of Biological Chemistry 285(14): 10880-10889.

Ben‐Yoseph, Y., R. Gagne, M. R. Parvathy, D. A. Mitchell, and T. Momoi. 1989. Leukocyte and plasma N‐laurylsphingosine deacylase (ceramidase) in Farber disease. Clinical genetics 36(1): 38-42.

Berdyshev, E. V., I. Gorshkova, A. Skobeleva, R. Bittman, X. Lu, S. M. Dudek, T. Mirzapoiazova, J. G. Garcia, and V. Natarajan. 2009. FTY720 inhibits ceramide synthases and up-regulates dihydrosphingosine 1-phosphate formation in human lung endothelial cells. Journal of Biological Chemistry 284(9): 5467-5477.

281

Berkhout, T. A., H. M. Sarau, K. Moores, J. R. White, N. Elshourbagy, E. Appelbaum, T. J. Brawner, M. Reape, J. Makwana, and J. J. Foley. 1997. Cloning, in vitro expression, and functional characterization of a novel human CC chemokine of the monocyte chemotactic protein (MCP) family (MCP-4) that binds and signals through the CC chemokine receptor 2B. Journal of Biological Chemistry 272(26): 16404-16413.

Bernardo, K., R. Hurwitz, T. Zenk, R. J. Desnick, K. Ferlinz, E. H. Schuchman, and K. Sandhoff. 1995. Purification, characterization, and biosynthesis of human acid ceramidase. Journal of Biological Chemistry 270(19): 11098-11102.

Beutler, E., N. J. Nguyen, M. W. Henneberger, J. M. Smolec, R. A. McPherson, C. West, and T. Gelbart. 1993. Gaucher disease: gene frequencies in the Ashkenazi Jewish population. American Journal of Human Genetics 52(1): 85-88.

Bhatia, M., R. D. Ramnath, L. Chevali, and A. Guglielmotti. 2005. Treatment with bindarit, a blocker of MCP-1 synthesis, protects mice against acute pancreatitis. American journal of physiology.Gastrointestinal and liver physiology 288(6): 1259.

Bhatia, M., R. L. Zemans, and S. Jeyaseelan. 2012. Role of chemokines in the pathogenesis of acute lung injury. American journal of respiratory cell and molecular biology 46(5): 566-572.

Bielawski, J., J. S. Pierce, J. Snider, B. Rembiesa, Z. M. Szulc, and A. Bielawska. 2010. Sphingolipid analysis by high performance liquid chromatography-tandem mass spectrometry (HPLC-MS/MS). Sphingolipids as Signaling and Regulatory Molecules: 46-59.

Bierman, S. M., T. Edgington, V. D. Newcomer, and C. M. Pearson. 1966. A disorder of mucopolysaccharide metabolism with articular, respiratory, and neurologic manifestations. Arthritis & Rheumatology 9(4): 620-630.

Biffi, A., E. Montini, L. Lorioli, M. Cesani, F. Fumagalli, T. Plati, C. Baldoli, et al. 2013. Lentiviral hematopoietic stem cell gene therapy benefits metachromatic leukodystrophy. Science (New York, N.Y.) 341(6148): 1233158.

Biffi, A., A. Capotondo, S. Fasano, U. Del Carro, S. Marchesini, H. Azuma, M. C. Malaguti, S. Amadio, R. Brambilla, and M. Grompe. 2006. Gene therapy of metachromatic leukodystrophy reverses neurological damage and deficits in mice. Journal of Clinical Investigation 116(11): 3070.

Bitsikas, V., I. R. C. Jnr, and B. J. Nichols. 2014. Correction: Clathrin-independent pathways do not contribute significantly to endocytic flux. eLife 3: e05256.

Bjurulf, B., S. Spetalen, A. Erichsen, M. T. Vanier, E. H. Strom, and P. Stromme. 2008. Niemann-Pick disease type C2 presenting as fatal pulmonary alveolar lipoproteinosis: morphological findings in lung and nervous tissue. Medical science monitor : international medical journal of experimental and clinical research 14(8): 5.

Bligh, E. G., and W. J. Dyer. 1959. A rapid method of total lipid extraction and purification. Canadian journal of biochemistry and physiology 37(8): 911-917.

Blitzer, M. G., and G. A. McDowell. 1992. Tay-Sachs disease as a model for screening inborn errors. Clinics in laboratory medicine 12(3): 463-480.

Blott, E. J., and G. M. Griffiths. 2002. Secretory lysosomes. Nature reviews Molecular cell biology 3(2): 122-131.

282

Boado, R. J., J. Z. Lu, E. K. Hui, H. Lin, and W. M. Pardridge. 2016. Insulin Receptor Antibody− α-N- Acetylglucosaminidase Fusion Protein Penetrates the Primate Blood–Brain Barrier and Reduces Glycosoaminoglycans in Sanfilippo Type B Fibroblasts. Molecular pharmaceutics 13(4): 1385-1392.

Boelens, J. J., M. Aldenhoven, D. Purtill, A. Ruggeri, T. Defor, R. Wynn, E. Wraith, M. Cavazzana- Calvo, A. Rovelli, and A. Fischer. 2013. Eurocord; Inborn Errors Working Party of European Blood and Marrow Transplant group; Duke University Blood and Marrow Transplantation Program; Centre for International Blood and Marrow Research. Outcomes of transplantation using various hematopoietic cell sources in children with Hurler syndrome after myeloablative conditioning. Blood 121(19): 3981-3987.

Boelens, J. J., R. F. Wynn, A. O'meara, P. Veys, Y. Bertrand, G. Souillet, J. E. Wraith, A. Fischer, M. Cavazzana-Calvo, and K. W. Sykora. 2007. Outcomes of hematopoietic stem cell transplantation for Hurler's syndrome in Europe: a risk factor analysis for graft failure. Bone marrow transplantation 40(3): 225-233.

Boelens, J. J., P. J. Orchard, and R. F. Wynn. 2014. Transplantation in inborn errors of metabolism: current considerations and future perspectives. British journal of haematology 167(3): 293-303.

Bonafé, L., A. Kariminejad, J. Li, B. Royer‐Bertrand, V. Garcia, S. Mahdavi, B. Bozorgmehr, R. L. Lachman, L. Mittaz‐Crettol, and B. Campos‐Xavier. 2016. Brief Report: Peripheral Osteolysis in Adults Linked to ASAH1 (Acid Ceramidase) Mutations: A New Presentation of Farber's Disease. Arthritis & Rheumatology 68(9): 2323-2327.

Bouwens, L., M. Baekeland, R. de Zanger, and E. Wisse. 1986. Quantitation, tissue distribution and proliferation kinetics of Kupffer cells in normal rat liver. Hepatology 6(4): 718-722.

Bradova, V., F. Šmíd, B. Ulrich-Bott, W. Roggendorf, B. C. Paton, and K. Harzer. 1993. Prosaposin deficiency: further characterization of the sphingolipid activator protein-deficient sibs. Human genetics 92(2): 143-152.

Brady, R. O., P. G. Pentchev, A. E. Gal, S. R. Hibbert, and A. S. Dekaban. 1974. Replacement therapy for inherited enzyme deficiency: use of purified glucocerebrosidase in Gaucher's disease. New England Journal of Medicine 291(19): 989-993.

Breen, C., R. F. Wynn, A. O'Meara, E. O'Mahony, S. Rust, J. Imrie, and J. E. Wraith. 2013. Developmental outcome post allogenic bone marrow transplant for Niemann Pick Type C2. Molecular genetics and metabolism 108(1): 82-84.

Brownstein, S., S. Carpenter, R. C. Polomeno, and J. M. Little. 1980. Sandhoff's Disease (Gm2 Gangliosidosis Type 2): Histopathology and Ultrastructure of the Eye. Archives of Ophthalmology 98(6): 1089-1097.

Brownstein, S., K. Meagher-Villemure, R. C. Polomeno, and J. M. Little. 1978. Optic nerve in globoid leukodystrophy (Krabbe's disease): Ultrastructural changes. Archives of Ophthalmology 96(5): 864- 870.

Brüggen, B., C. Kremser, A. Bickert, P. Ebel, K. Dorp, K. Schultz, P. Dörmann, K. Willecke, and K. Dedek. 2016. Defective ceramide synthases in mice cause reduced amplitudes in electroretinograms and altered sphingolipid composition in retina and cornea. European Journal of Neuroscience 44(1): 1700-1713.

Brush, R. S., J. A. Tran, K. R. Henry, M. E. McClellan, M. H. Elliott, and M. N. A. Mandal. 2010. Retinal Sphingolipids and Their Very-Long-Chain Fatty Acid–Containing Species. Investigative ophthalmology & visual science 51(9): 4422-4431.

283

Buccoliero, R., L. Ginzburg, and A. H. Futerman. 2004. Elevation of lung surfactant phosphatidylcholine in mouse models of Sandhoff and of Niemann–Pick A disease. Journal of inherited metabolic disease 27(5): 641-648.

Burck, U., H. W. Moser, H. H. Goebel, R. Grüttner, and K. R. Held. 1985. A case of lipogranulomatosis Farber: some clinical and ultrastructural aspects. European journal of pediatrics 143(3): 203-208.

Burke, W., B. Tarini, N. A. Press, and J. P. Evans. 2011. Genetic screening. Epidemiologic reviews 33(1): 148-164.

Burton, B. K., J. Charrow, G. E. Hoganson, D. Waggoner, B. Tinkle, S. R. Braddock, M. Schneider, et al. 2017. Newborn Screening for Lysosomal Storage Disorders in Illinois: The Initial 15-Month Experience. The Journal of pediatrics 190: 130-135.

Cao, S. S., and R. J. Kaufman. 2012. Unfolded protein response. Current biology 22(16): R626.

Carreras, B., D. E. Griffin, and A. M. Silverstein. 1982. Sindbis virus-induced ocular immunopathology. Investigative ophthalmology & visual science 22(5): 571-578.

Cartigny, B., J. Libert, A. H. Fensom, J. J. Martin, J. L. Dhondt, D. Wyart, G. Fontaine, and J. P. Farriaux. 1985. Clinical diagnosis of a new case of ceramidase deficiency (Farber's disease). Journal of inherited metabolic disease 8(1): 8.

Caspi, R. R., P. B. Silver, D. Luger, J. Tang, L. M. Cortes, G. Pennesi, M. J. Mattapallil, and C. Chan. 2008. Mouse models of experimental autoimmune uveitis. Ophthalmic research 40(3-4): 169-174.

Chan, A. Y., R. Chen, D. S. Stone, A. Eckerd, and N. Mandal. 2012. Bioactive Sphingolipids As Mediators Of Ocular Inflammation: Observations In An Animal Model. Investigative ophthalmology & visual science 53(14): 1239.

Chan, C., R. R. Caspi, M. Ni, W. C. Leake, B. Wiggert, G. J. Chader, and R. B. Nussenblatt. 1990. Pathology of experimental autoimmune uveoretinitis in mice. Journal of Autoimmunity 3(3): 247-255.

Chandwani, R., and A. S. Kuwar. 2002. Farber's disease. Indian pediatrics 39(5): 502.

Chanoki, M., M. Ishii, K. Fukai, H. Kobayashi, T. Hamada, K. Murakami, and A. Tanaka. 1989. Farber's lipogranulomatosis in siblings: Light and electron microscopic studies. British Journal of Dermatology 121(6): 779-785.

Chatelut, M., J. Feunteun, K. Harzer, A. H. Fensom, J. Basile, R. Salvayre, and T. Levade. 1996. A simple method for screening for Farber disease on cultured skin fibroblasts. Clinica chimica acta 245(1): 61-71.

Chaurasia, B., and S. A. Summers. 2015. Ceramides–lipotoxic inducers of metabolic disorders. Trends in Endocrinology & Metabolism 26(10): 538-550.

Chedrawi, A. K., Z. N. Al-Hassnan, M. Al-Muhaizea, D. Colak, B. Al-Younes, A. Albakheet, S. Tulba, and N. Kaya. 2012. Novel V97G ASAH1 mutation found in Farber disease patients: unique appearance of the disease with an intermediate severity, and marked early involvement of central and peripheral nervous system. Brain and Development 34(5): 400-404.

Chen, H., J. T. Tran, A. Eckerd, T. P. Huynh, M. H. Elliott, R. S. Brush, and N. A. Mandal. 2013. Inhibition of de novo ceramide biosynthesis by FTY720 protects rat retina from light-induced degeneration. Journal of lipid research 54(6): 1616-1629.

284

Chen, H., A. Y. Chan, D. U. Stone, and N. A. Mandal. 2014. Beyond the cherry-red spot: Ocular manifestations of sphingolipid-mediated neurodegenerative and inflammatory disorders. Survey of ophthalmology 59(1): 64-76.

Chen, H., J. A. Tran, R. S. Brush, A. Saadi, A. K. Rahman, M. Yu, D. Yasumura, M. T. Matthes, K. Ahern, and H. Yang. 2012. Ceramide signaling in retinal degeneration: 553-558.

Chen, J., H. Qian, R. Horai, C. Chan, and R. R. Caspi. 2013. Use of optical coherence tomography and electroretinography to evaluate retinal pathology in a mouse model of autoimmune uveitis. PloS one 8(5): e63904.

Chiao, Y. A., Q. Dai, J. Zhang, J. Lin, E. F. Lopez, S. S. Ahuja, Y. M. Chou, M. L. Lindsey, and Y. F. Jin. 2011. Multi-analyte profiling reveals matrix metalloproteinase-9 and monocyte chemotactic protein-1 as plasma biomarkers of cardiac aging. Circulation.Cardiovascular genetics 4(4): 455-462.

Chikova, I. A., N. V. Buchinskaya, М М Kostik, V. V. Avramenko, O. L. Krasnogorskaya, R. A. Nasirov, T. Levade, and V. G. Chasnyk. 2014. Farber Disease- Disease Description with Case Reports. Voprosy Sovremennoj Pediatrii 13(6): 78-84.

Chiu, S. J., X. T. Li, P. Nicholas, C. A. Toth, J. A. Izatt, and S. Farsiu. 2010. Automatic segmentation of seven retinal layers in SDOCT images congruent with expert manual segmentation. Optics express 18(18): 19413-19428.

Christie, W. W. 2012. Welcome to the AOCS Lipid Library. The lipid library.

Chueh, S. J., and B. D. Kahan. 2003. Update on FTY720: review of mechanisms and clinical results. Current Opinion in Organ Transplantation 8(4): 288-298.

Clarke, J. T. 2007. Narrative Review: Fabry DiseaseFabry Disease. Annals of Internal Medicine 146(6): 425-433.

Coant, N., W. Sakamoto, C. Mao, and Y. A. Hannun. 2017. Ceramidases, roles in sphingolipid metabolism and in health and disease. Advances in biological regulation 63: 122-131.

Cogan, D. G., and T. Kuwabara. 1968. The sphingolipidoses and the eye. Archives of Ophthalmology 79(4): 437-452.

Cogan, D. G., T. Kuwabara, H. Moser, and G. W. Hazard. 1966. Retinopathy in a case of Farber's lipogranulomatosis. Archives of Ophthalmology 75(6): 752-757.

Cohen, I. J., P. Mistry, H. Baris, and M. Sands. 2016. Therapy for Lysosomal Storage Disorders: How Therapies are being Developed for Orphan Diseases. Pediatric Endocrinology Reviews (PER) 13: 1.

Colamaria, V., L. Giardina, M. Simeone, A. Salviati, A. H. Fensom, and B. Dalla Bernardina. 1992. Neurologic progressive form (type 5) of Farber's lipogranulomatosis (ceramidase deficiency) in monozygotic twins.

Collins, M., and A. Thrasher. 2015. Gene therapy: progress and predictions 282(1821): 20143003.

Commodaro, A. G., J. P. S. Peron, C. T. Lopes, C. Arslanian, R. Belfort, L. V. Rizzo, and V. Bueno. 2010. Evaluation of experimental autoimmune uveitis in mice treated with FTY720. Investigative ophthalmology & visual science 51(5): 2568-2574.

Copland, D. A., J. Liu, L. P. Schewitz-Bowers, V. Brinkmann, K. Anderson, L. B. Nicholson, and A. D. Dick. 2012. Therapeutic dosing of fingolimod (FTY720) prevents cell infiltration, rapidly suppresses

285 ocular inflammation, and maintains the blood-ocular barrier. The American journal of pathology 180(2): 672-681.

Coutinho, M. F., and S. Alves. 2016. From rare to common and back again: 60years of lysosomal dysfunction. Molecular genetics and metabolism 117(2): 53-65.

Coutinho, M. F., M. J. Prata, and S. Alves. 2012. Mannose-6-phosphate pathway: a review on its role in lysosomal function and dysfunction. Molecular genetics and metabolism 105(4): 542-550.

Cowart, L. A., Z. Szulc, A. Bielawska, and Y. A. Hannun. 2002. Structural determinants of sphingolipid recognition by commercially available anti-ceramide antibodies. Journal of lipid research 43(12): 2042-2048.

Cox, T., R. Lachmann, C. Hollak, J. Aerts, S. van Weely, M. Hrebícek, F. Platt, T. Butters, R. Dwek, and C. Moyses. 2000. Novel oral treatment of Gaucher's disease with N-butyldeoxynojirimycin (OGT 918) to decrease substrate biosynthesis. The Lancet 355(9214): 1481-1485.

Cozma, C., M. I. Iurascu, S. Eichler, M. Hovakimyan, O. Brandau, S. Zielke, T. Bottcher, A. K. Giese, J. Lukas, and A. Rolfs. 2017. C26-Ceramide as highly sensitive biomarker for the diagnosis of Farber Disease. Scientific reports 7(1): 2.

Cremesti, A. E., and A. S. Fischl. 2000. Current methods for the identification and quantitation of ceramides: an overview. Lipids 35(9): 937-945.

Crocker, A. C., J. Cohen, and S. Farber. 1967. The “lipogranulomatosis” syndrome; Review, with report of patient showing milder involvement. Inborn disorders of sphingolipid metabolism.Pergamon Press, New York: 485-503.

Cushing, S. D., J. A. Berliner, A. J. Valente, M. C. Territo, M. Navab, F. Parhami, R. Gerrity, C. J. Schwartz, and A. M. Fogelman. 1990. Minimally modified low density lipoprotein induces monocyte chemotactic protein 1 in human endothelial cells and smooth muscle cells. Proceedings of the National Academy of Sciences of the United States of America 87(13): 5134-5138.

Cvitanovic-Sojat, L., R. G. Juraski, F. Sabourdy, A. H. Fensom, K. Fumic, E. Paschke, and T. Levade. 2011. Farber lipogranulomatosis type 1–late presentation and early death in a Croatian boy with a novel homozygous ASAH1 mutation. european journal of paediatric neurology 15(2): 171-173.

Cyster, J. G., and S. R. Schwab. 2012. Sphingosine-1-phosphate and lymphocyte egress from lymphoid organs. Annual Review of Immunology 30: 69-94.

Dagan, A., V. Agmon, S. Gatt, and T. Dinur. 2000. Section II. Methods for Analyzing Aspects of Sphingolipid Metabolism in Intact Cells-23 Synthesis of Fluorescent Substrates and Their Application to Study of Sphingolipid Metabolism in Vitro and in. Methods in enzymology 312: 293-303.

Dannhausen, K., M. Karlstetter, A. Caramoy, C. Volz, H. Jägle, G. Liebisch, O. Utermöhlen, and T. Langmann. 2015. Acid sphingomyelinase (aSMase) deficiency leads to abnormal microglia behavior and disturbed retinal function. Biochemical and biophysical research communications 464(2): 434- 440.

Das, A. M., F. Lagler, M. Beck, M. Scarpa, and C. Lampe. 2017. Lysosomal Storage Diseases: Challenges in Multiprofessional Patient Care with Enzyme Replacement Therapy. Klinische Padiatrie 229(3): 168-174.

Dawson, T. C., W. A. Kuziel, T. A. Osahar, and N. Maeda. 1999. Absence of CC chemokine receptor- 2 reduces atherosclerosis in apolipoprotein E-deficient mice. Atherosclerosis 143(1): 205-211.

286

De Duve, C. 1959. The function of intracellular hydrolases. Experimental cell research 7: 169-182.

De Duve, C. 1975. Exploring cells with a centrifuge. Science 189(4198): 186-194.

De Duve, C., and R. Wattiaux. 1966. Functions of lysosomes. Annual Review of Physiology 28(1): 435-492.

D'Ecclesia, G., V. Scorrano, C. Bernardini, and V. Poli. 1985. Unusual familial association of epilepsy, myoclonus and muscular atrophy. Case report. Rivista di neurologia 55(5): 313-316.

Deevska, G. M., and M. N. Nikolova-Karakashian. 2017. The expanding role of sphingolipids in lipid droplet biogenesis. Biochimica et Biophysica Acta (BBA)-Molecular and Cell Biology of Lipids 1862(10): 1155-1165.

Deevska, G. M., K. A. Rozenova, N. V. Giltiay, M. A. Chambers, J. White, B. B. Boyanovsky, J. Wei, A. Daugherty, E. J. Smart, and M. B. Reid. 2009. Acid sphingomyelinase deficiency prevents diet- induced hepatic triacylglycerol accumulation and hyperglycemia in mice. Journal of Biological Chemistry 284(13): 8359-8368.

Denny, C. A., J. Alroy, B. S. Pawlyk, M. A. Sandberg, A. D’Azzo, and T. N. Seyfried. 2007. Neurochemical, morphological, and neurophysiological abnormalities in retinas of Sandhoff and GM1 gangliosidosis mice. Journal of neurochemistry 101(5): 1294-1302.

Deshmane, S. L., S. Kremlev, S. Amini, and B. E. Sawaya. 2009. Monocyte chemoattractant protein- 1 (MCP-1): an overview. Journal of interferon & cytokine research 29(6): 313-326.

Devi, A. R. R., M. Gopikrishna, R. Ratheesh, G. Savithri, G. Swarnalata, and M. Bashyam. 2006. Farber lipogranulomatosis: clinical and molecular genetic analysis reveals a novel mutation in an Indian family. Journal of human genetics 51(9): 811-814.

Dhami, R., X. He, R. E. Gordon, and E. H. Schuchman. 2001. Analysis of the lung pathology and alveolar macrophage function in the acid sphingomyelinase–deficient mouse model of Niemann-Pick disease. Laboratory investigation 81(7): 987-999.

Dubra, A., and Z. Harvey. 2010. Registration of 2D images from fast scanning ophthalmic instruments. Biomedical Image Registration: 60-71.

Dulaney, J. T., A. Milunsky, J. B. Sidbury, N. Hobolth, and H. W. Moser. 1976. Diagnosis of lipogranulomatosis (Farber disease) by use of cultured fibroblasts. The Journal of pediatrics 89(1): 59-61.

Dustin, P., M. Tondeur, G. Jonniaux, E. Vamos-Hurwitz, and S. Pelc. 1973. La maladie de Farber. Etude anatomo-clinique et ultrastructurale. Bull Acad R Med Belg 128: 733-762.

Dworski, S., A. Berger, C. Furlonger, J. M. Moreau, M. Yoshimitsu, J. Trentadue, B. C. Au, C. J. Paige, and J. A. Medin. 2015. Markedly perturbed hematopoiesis in acid ceramidase deficient mice. Haematologica 100(5): 162.

Dworski, S., P. Lu, A. Khan, B. Maranda, J. J. Mitchell, R. Parini, M. Di Rocco, et al. 2017. Acid Ceramidase Deficiency is characterized by a unique plasma cytokine and ceramide profile that is altered by therapy. Biochimica et Biophysica Acta (BBA)-Molecular Basis of Disease 1863(2): 386- 394.

287

Dyment, D. A., E. Sell, M. R. Vanstone, A. C. Smith, D. Garandeau, V. Garcia, S. Carpentier, et al. 2014. Evidence for clinical, genetic and biochemical variability in spinal muscular atrophy with progressive myoclonic epilepsy. Clinical genetics 86(6): 558-563.

Ehlert, K., J. Roth, M. Frosch, N. Fehse, N. Zander, and J. Vormoor. 2006. Farber's disease without central nervous system involvement: bone-marrow transplantation provides a promising new approach. Annals of the Rheumatic Diseases 65(12): 1665-1666.

Ehlert, K., M. Frosch, N. Fehse, A. Zander, J. Roth, and J. Vormoor. 2007. Farber disease: clinical presentation, pathogenesis and a new approach to treatment. Pediatric Rheumatology 5(1): 15.

Ekici, B., D. Kürkçü, and M. Çaliskan. 2012. Farber disease: A clinical diagnosis. Journal of pediatric neurosciences.

El Alwani, M., B. X. Wu, L. M. Obeid, and Y. A. Hannun. 2006. Bioactive sphingolipids in the modulation of the inflammatory response. Pharmacology & therapeutics 112(1): 171-183.

El Bawab, S., A. Bielawska, and Y. A. Hannun. 1999. Purification and characterization of a membrane-bound nonlysosomal ceramidase from rat brain. Journal of Biological Chemistry 274(39): 27948-27955.

El Bawab, S., P. Roddy, T. Qian, A. Bielawska, J. J. Lemasters, and Y. A. Hannun. 2000. Molecular cloning and characterization of a human mitochondrial ceramidase. Journal of Biological Chemistry 275(28): 21508-21513.

El-Darouti, M. A. 2013. A Child with Hoarse Cry, Subcutaneous Nodules and Joint Contracture: 107- 112.

Eliyahu, E., J. H. Park, N. Shtraizent, X. He, and E. H. Schuchman. 2007. Acid ceramidase is a novel factor required for early embryo survival. FASEB journal : official publication of the Federation of American Societies for Experimental Biology 21(7): 1403-1409.

Eliyahu, E., N. Shtraizent, R. Shalgi, and E. H. Schuchman. 2012. Construction of conditional acid ceramidase knockout mice and in vivo effects on oocyte development and fertility. Cellular physiology and biochemistry : international journal of experimental cellular physiology, biochemistry, and pharmacology 30(3): 735-748.

El-Kamah, G. Y., M. A. El-darouti, A. I. Kotoury, and M. I. Mostafa. 2009. Farber disease overlapping with stiff skin syndrome: Expanding the spectrum. Egyptian Journal of Medical Human Genetics 10(1).

Emery, J. M., W. R. Green, and D. S. Huff. 1972. Krabbe's disease: Histopathology and ultrastructure of the eye. American Journal of Ophthalmology 74(3): 400-406.

Erfan, M., A. U. Haque, and S. A. Ahmed. 2015. Farber's disease: a case report. Int.J.Pathol. 13: 115-119.

Escolar, M. L., M. D. Poe, J. M. Provenzale, K. C. Richards, J. Allison, S. Wood, D. A. Wenger, D. Pietryga, D. Wall, and M. Champagne. 2005. Transplantation of umbilical-cord blood in babies with infantile Krabbe's disease. New England Journal of Medicine 352(20): 2069-2081.

Eviatar, L., S. L. Sklower, K. Wisniewski, R. S. Feldman, and A. Gochoco. 1986. Farber lipogranulomatosis: an unusual presentation in a black child. Pediatric neurology 2(6): 371-374.

288

Fahy, E., S. Subramaniam, R. C. Murphy, M. Nishijima, C. R. Raetz, T. Shimizu, F. Spener, G. van Meer, M. J. Wakelam, and E. A. Dennis. 2009. Update of the LIPID MAPS comprehensive classification system for lipids. Journal of lipid research 50(Supplement): S14.

Farber, S. 1952. A lipid metabolic disorder: disseminated lipogranulomatosis; a syndrome with similarity to, and important difference from, Niemann-Pick and Hand-Schuller-Christian disease. A.M.A.American journal of diseases of children 84(4): 499-500.

Farber, S., J. Cohen, and L. L. Uzman. 1957. Lipogranulomatosis; a new lipo-glycoprotein storage disease. Journal of the Mount Sinai Hospital, New York 24(6): 816-837.

Farina, F., F. Cappello, M. Todaro, F. Bucchieri, G. Peri, G. Zummo, and G. Stassi. 2000. Involvement of caspase-3 and GD3 ganglioside in ceramide-induced apoptosis in Farber disease. Journal of Histochemistry & Cytochemistry 48(1): 57-62.

Feingold, K. R. 2007. Thematic review series: skin lipids. The role of epidermal lipids in cutaneous permeability barrier homeostasis. Journal of lipid research 48(12): 2531-2546.

Feldhammer, M., S. Durand, and A. V. Pshezhetsky. 2009. Protein misfolding as an underlying molecular defect in mucopolysaccharidosis III type C. PloS one 4(10): e7434.

Fensom, A. H., B. Neville, A. Moser, P. F. Benson, H. W. Moser, and J. T. Dulaney. 1979. Prenatal diagnosis of Farber's disease. The Lancet 314(8150): 990-992.

Ferlazzo, E., D. Italiano, I. An, T. Calarese, V. Laguitton, P. Bramanti, P. Di Bella, and P. Genton. 2009. Description of a family with a novel progressive myoclonus epilepsy and cognitive impairment. Movement Disorders 24(7): 1016-1022.

Ferlinz, K., G. Kopal, K. Bernardo, T. Linke, J. Bar, B. Breiden, U. Neumann, F. Lang, E. H. Schuchman, and K. Sandhoff. 2001. Human acid ceramidase: processing, glycosylation, and lysosomal targeting. The Journal of biological chemistry 276(38): 35352-35360.

Ferreira, C. R., and W. A. Gahl. 2017. Lysosomal storage diseases. Translational Science of Rare Diseases 2(1-2): 1-71.

Filocamo, M., and A. Morrone. 2011. Lysosomal storage disorders: molecular basis and laboratory testing. Human genomics 5(3): 156.

Filosto, M., M. Aureli, B. Castellotti, F. Rinaldi, D. Schiumarini, M. Valsecchi, S. Lualdi, R. Mazzotti, V. Pensato, and S. Rota. 2016. ASAH1 variant causing a mild SMA phenotype with no myoclonic epilepsy: a clinical, biochemical and molecular study. European Journal of Human Genetics 24(11): 1578-1583.

Fiumara, A., F. Nigro, L. Pavone, and H. W. Moser. 1993. Farber disease with prolonged survival. Journal of inherited metabolic disease 16(5): 915-916.

Flory, C. M., M. L. Jones, and J. S. Warren. 1993. Pulmonary granuloma formation in the rat is partially dependent on monocyte chemoattractant protein 1. Laboratory investigation; a journal of technical methods and pathology 69(4): 396-404.

Fox, M. W. 1965. The visual cliff test for the study of visual depth perception in the mouse. Animal Behaviour 13(2-3): IN3.

Franci, C., L. M. Wong, J. Van Damme, P. Proost, and I. F. Charo. 1995. Monocyte chemoattractant protein-3, but not monocyte chemoattractant protein-2, is a functional ligand of the human monocyte

289 chemoattractant protein-1 receptor. Journal of immunology (Baltimore, Md.: 1950) 154(12): 6511- 6517.

Fratantoni, J. C., C. W. Hall, and E. F. Neufeld. 1968. Hurler and Hunter syndromes: mutual correction of the defect in cultured fibroblasts. Science (New York, N.Y.) 162(3853): 570-572.

Fujiwaki, T., S. Hamanaka, M. Koga, T. Ishihara, R. Nishikomori, E. Kinoshita, and K. Furusho. 1992. A case of Farber disease. Pediatrics International 34(1): 72-79.

Fuller, M., P. J. Meikle, and J. J. Hopwood. 2006. Epidemiology of lysosomal storage diseases: an overview.

Fusch, C., R. Huenges, H. W. Moser, A. C. Sewell, W. Roggendorf, B. Kustermann-Kuhn, A. Poulos, W. F. Carey, and K. Harzer. 1989. A case of combined Farber and Sandhoff disease. European journal of pediatrics 148(6): 558-562.

Futerman, A. H., and Y. A. Hannun. 2004. The complex life of simple sphingolipids. EMBO reports 5(8): 777-782.

Futerman, A. H., and G. Van Meer. 2004. The cell biology of lysosomal storage disorders. Nature reviews Molecular cell biology 5(7): 554-565.

Gan, J. J., V. Garcia, J. Tian, M. Tagliati, J. E. Parisi, J. M. Chung, R. Lewis, R. Baloh, T. Levade, and T. M. Pierson. 2015. Acid ceramidase deficiency associated with spinal muscular atrophy with progressive myoclonic epilepsy. Neuromuscular Disorders 25(12): 959-963.

Garbuzova-Davis, S., S. J. Gografe, C. D. Sanberg, A. E. Willing, S. Saporta, D. F. Cameron, T. Desjarlais, J. Daily, N. Kuzmin-Nichols, and W. Chamizo. 2006. Maternal transplantation of human umbilical cord blood cells provides prenatal therapy in Sanfilippo type B mouse model. The FASEB journal 20(3): 485-487.

García-Ruiz, C., A. Colell, M. Marí, A. Morales, M. Calvo, C. Enrich, and J. C. Fernández-Checa. 2003. Defective TNF-α–mediated hepatocellular apoptosis and liver damage in acidic sphingomyelinase knockout mice. The Journal of clinical investigation 111(2): 197-208.

Gasparin, F., B. S. Takahashi, M. R. Scolari, F. Gasparin, L. S. Pedral, and F. M. Damico. 2012. Experimental models of autoimmune inflammatory ocular diseases. Arquivos Brasileiros de Oftalmologia 75(2): 143-147.

Gatt, S. 1963. Enzymic Hydrolysis and Synthesis of Ceramides. The Journal of biological chemistry 238: 3131-3133.

Gaukel, E., C. Coquery, B. Sampey, E. Schuchman, and S. A. Wring. 2018a. Pharmacokinetics and tissue distribution of RVT-801, a recombinant human acid ceramidase, at efficacious doses in a murine model of Farber disease. Molecular genetics and metabolism 123(2): S50.

Gaukel, E., C. Coquery, B. Sampey, E. Schuchman, and S. A. Wring. 2018b. Prediction of the human equivalent dose (HED) of RVT-801, a recombinant human acid ceramidase, for the treatment of Farber disease. Molecular genetics and metabolism 123(2): S50.

Genovese, G., M. Fromer, E. A. Stahl, D. M. Ruderfer, K. Chambert, M. Landen, J. L. Moran, et al. 2016. Increased burden of ultra-rare protein-altering variants among 4,877 individuals with schizophrenia. Nature neuroscience 19(11): 1433-1441.

Germain, D. P. 2010. Fabry disease. Orphanet journal of rare diseases 5(1): 30.

290

Ghidoni, R., A. Caretti, and P. Signorelli. 2015. Role of sphingolipids in the pathobiology of lung inflammation. Mediators of inflammation 2015.

Gianetto, R., and C. De Duve. 1955. Tissue fractionation studies. 4. Comparative study of the binding of acid phosphatase, beta-glucuronidase and cathepsin by rat-liver particles. The Biochemical journal 59(3): 433-438.

Giráldez, B. G., R. Guerrero-López, L. Ortega-Moreno, A. Verdú, M. C. Carrascosa-Romero, Ó García-Campos, S. García-Muñozguren, J. M. Pardal-Fernández, and J. M. Serratosa. 2015. Uniparental disomy as a cause of spinal muscular atrophy and progressive myoclonic epilepsy: phenotypic homogeneity due to the homozygous c. 125C> T mutation in ASAH1. Neuromuscular Disorders 25(3): 222-224.

Glatz, J. F. 2011. Challenges in fatty acid and lipid physiology. Frontiers in physiology 2.

Göggel, R., S. Winoto-Morbach, G. Vielhaber, Y. Imai, K. Lindner, L. Brade, H. Brade, et al. 2004. PAF-mediated pulmonary edema: a new role for acid sphingomyelinase and ceramide. Nature medicine 10(2): 155-160.

Grabowski, G. A., M. Golembo, and Y. Shaaltiel. 2014. Taliglucerase alfa: an enzyme replacement therapy using plant cell expression technology. Molecular genetics and metabolism 112(1): 1-8.

Griese, M., F. Brasch, V. R. Aldana, M. M. Cabrera, U. Goelnitz, E. Ikonen, B. J. Karam, G. Liebisch, M. D. Linder, and P. Lohse. 2010. Respiratory disease in Niemann‐Pick type C2 is caused by pulmonary alveolar proteinosis. Clinical genetics 77(2): 119-130.

Grishchuk, Y., K. G. Stember, A. Matsunaga, A. M. Olivares, N. M. Cruz, V. E. King, D. M. Humphrey, S. L. Wang, A. Muzikansky, and R. A. Betensky. 2016. Retinal dystrophy and optic nerve pathology in the mouse model of mucolipidosis IV. The American journal of pathology 186(1): 199-209.

Groh, J., K. Heinl, B. Kohl, C. Wessig, J. Greeske, S. Fischer, and R. Martini. 2010. Attenuation of MCP-1/CCL2 expression ameliorates neuropathy in a mouse model for Charcot–Marie–Tooth 1X. Human molecular genetics 19(18): 3530-3543.

Gross, R. W. 2017. The evolution of lipidomics through space and time. Biochimica et Biophysica Acta (BBA)-Molecular and Cell Biology of Lipids 1862(8): 731-739.

Gu, M., J. L. Kerwin, J. D. Watts, and R. Aebersold. 1997. Ceramide profiling of complex lipid mixtures by electrospray ionization mass spectrometry. Analytical Biochemistry 244(2): 347-356.

Guffon, N., Y. Bertrand, I. Forest, A. Fouilhoux, and R. Froissart. 2009. Bone marrow transplantation in children with Hunter syndrome: outcome after 7 to 17 years. The Journal of pediatrics 154(5): 733- 737.

Guillou, H., D. Zadravec, P. G. Martin, and A. Jacobsson. 2010. The key roles of elongases and desaturases in mammalian fatty acid metabolism: Insights from transgenic mice. Progress in lipid research 49(2): 186-199.

Gunther, A., C. Siebert, R. Schmidt, S. Ziegler, F. Grimminger, M. Yabut, B. Temmesfeld, D. Walmrath, H. Morr, and W. Seeger. 1996. Surfactant alterations in severe pneumonia, acute respiratory distress syndrome, and cardiogenic lung edema. American journal of respiratory and critical care medicine 153(1): 176-184.

291

Haliloglu, G., A. Chattopadhyay, L. Skorodis, A. Manzur, E. Mercuri, B. Talim, Z. Akçören, Y. Renda, F. Muntoni, and H. Topaloğlu. 2002. Spinal muscular atrophy with progressive myoclonic epilepsy: report of new cases and review of the literature. Neuropediatrics 33(06): 314-319.

Hallman, M., R. Spragg, J. H. Harrell, K. M. Moser, and L. Gluck. 1982. Evidence of lung surfactant abnormality in respiratory failure. Study of bronchoalveolar lavage phospholipids, surface activity, phospholipase activity, and plasma myoinositol. The Journal of clinical investigation 70(3): 673-683.

Hamada, Y., H. Nagasaki, A. Fujiya, Y. Seino, Q. Shang, T. Suzuki, H. Hashimoto, and Y. Oiso. 2014. Involvement of de novo ceramide synthesis in pro-inflammatory adipokine secretion and adipocyte–macrophage interaction. The Journal of nutritional biochemistry 25(12): 1309-1316.

Han, X. 2002. Characterization and direct quantitation of ceramide molecular species from lipid extracts of biological samples by electrospray ionization tandem mass spectrometry. Analytical Biochemistry 302(2): 199-212.

Hannun, Y. A., and L. M. Obeid. 2011. Many ceramides. Journal of Biological Chemistry 286(32): 27855-27862.

Hannun, Y. A., and L. M. Obeid. 2008. Principles of bioactive lipid signalling: lessons from sphingolipids. Nature reviews Molecular cell biology 9(2): 139-150.

Hannun, Y. A., and L. M. Obeid. 2017. Sphingolipids and their metabolism in physiology and disease. Nature Reviews Molecular Cell Biology.

Haraoka, G., M. Muraoka, N. Yoshioka, S. Wakami, and I. Hayashi. 1997. First Case of Surgical Treatment of Farber's Diseas. Annals of Plastic Surgery 39(4): 405-410.

Harmatz, P., Z. Yu, R. Giugliani, I. V. D. Schwartz, N. Guffon, E. L. Teles, M. C. S. Miranda, et al. 2010. Enzyme replacement therapy for mucopolysaccharidosis VI: evaluation of long-term pulmonary function in patients treated with recombinant human N-acetylgalactosamine 4-sulfatase. Journal of inherited metabolic disease 33(1): 51-60.

Harzer, K., B. C. Paton, A. Poulos, B. Kustermann-Kuhn, W. Roggendorf, T. Grisar, and M. Popp. 1989. Sphingolipid activator protein deficiency in a 16-week-old atypical Gaucher disease patient and his fetal sibling: biochemical signs of combined sphingolipidoses. European journal of pediatrics 149(1): 31-39.

Hayasaka, S. 1983. Lysosomal enzymes in ocular tissues and diseases. Survey of ophthalmology 27(4): 245-258.

He, X., N. Okino, R. Dhami, A. Dagan, S. Gatt, H. Schulze, K. Sandhoff, and E. H. Schuchman. 2003. Purification and characterization of recombinant, human acid ceramidase. Catalytic reactions and interactions with acid sphingomyelinase. The Journal of biological chemistry 278(35): 32978-32986.

He, X., S. Dworski, C. Zhu, V. DeAngelis, A. Solyom, J. A. Medin, C. M. Simonaro, and E. H. Schuchman. 2017. Enzyme replacement therapy for Farber disease: Proof-of-concept studies in cells and mice. BBA Clinical 7: 85-96.

Hers, H. G. 1963. alpha-Glucosidase deficiency in generalized glycogenstorage disease (Pompe's disease). The Biochemical journal 86: 11-16.

Holland, W. L., J. T. Brozinick, L. Wang, E. D. Hawkins, K. M. Sargent, Y. Liu, K. Narra, K. L. Hoehn, T. A. Knotts, and A. Siesky. 2007. Inhibition of ceramide synthesis ameliorates glucocorticoid-, saturated-fat-, and obesity-induced insulin resistance. Cell metabolism 5(3): 167-179.

292

Holland, W. L., R. A. Miller, Z. V. Wang, K. Sun, B. M. Barth, H. H. Bui, K. E. Davis, B. T. Bikman, N. Halberg, and J. M. Rutkowski. 2011. Receptor-mediated activation of ceramidase activity initiates the pleiotropic actions of adiponectin. Nature medicine 17(1): 55-63.

Holm, B. A., and R. H. Notter. 1987. Effects of hemoglobin and cell membrane lipids on pulmonary surfactant activity. Journal of applied physiology (Bethesda, Md.: 1985) 63(4): 1434-1442.

Holm, B. A., R. H. Notter, and J. N. Finkelstein. 1985. Surface property changes from interactions of albumin with natural lung surfactant and extracted lung lipids. Chemistry and physics of lipids 38(3): 287-298.

Homaidan, F. R., M. E. El-Sabban, I. Chakroun, M. El-Sibai, and G. S. Dbaibo. 2002. IL-1 stimulates ceramide accumulation without inducing apoptosis in intestinal epithelial cells. Mediators of inflammation 11(1): 39-45.

Hoof, F. v., and H. G. Hers. 1968. The abnormalities of lysosomal enzymes in mucopolysaccharidoses. The FEBS Journal 7(1): 34-44.

Hopkins, P. V., C. Campbell, T. Klug, S. Rogers, J. Raburn-Miller, and J. Kiesling. 2015. Lysosomal storage disorder screening implementation: findings from the first six months of full population pilot testing in Missouri. The Journal of pediatrics 166(1): 172-177.

Hsu, J., D. Serrano, T. Bhowmick, K. Kumar, Y. Shen, Y. C. Kuo, C. Garnacho, and S. Muro. 2011. Enhanced endothelial delivery and biochemical effects of α-galactosidase by ICAM-1-targeted nanocarriers for Fabry disease. Journal of Controlled Release 149(3): 323-331.

Hu, W., R. Xu, W. Sun, Z. M. Szulc, J. Bielawski, L. M. Obeid, and C. Mao. 2010. Alkaline ceramidase 3 (ACER3) hydrolyzes unsaturated long-chain ceramides, and its down-regulation inhibits both cell proliferation and apoptosis. Journal of Biological Chemistry 285(11): 7964-7976.

Huang, J., A. Khan, B. C. Au, D. L. Barber, L. López-Vásquez, N. L. Prokopishyn, M. Boutin, et al. 2017. Lentivector Iterations and Pre-Clinical Scale-Up/Toxicity Testing: Targeting Mobilized CD34 Cells for Correction of Fabry Disease. Molecular Therapy-Methods & Clinical Development 5: 241- 258.

Hugle, B., L. Mueller, and T. Levade. 2014. Why Farber Disease may be misdiagnosed as juvenile idiopathic arthritis. Available at: www.the-rheumatologist.org/article/why-farber-disease-may-be- misdiagnosed-as-juvenile-idiopathic-arthritis/.Accessed June 1.

Hulkova, H., M. Cervenkova, J. Ledvinova, M. Tochackova, M. Hrebicek, H. Poupetova, A. Befekadu, L. Berna, B. C. Paton, and K. Harzer. 2001. A novel mutation in the coding region of the prosaposin gene leads to a complete deficiency of prosaposin and saposins, and is associated with a complex sphingolipidosis dominated by lactosylceramide accumulation. Human molecular genetics 10(9): 927- 940.

Hume, D. A., and S. Gordon. 1983. Mononuclear phagocyte system of the mouse defined by immunohistochemical localization of antigen F4/80. Identification of resident macrophages in renal medullary and cortical interstitium and the juxtaglomerular complex. Journal of Experimental Medicine 157(5): 1704-1709.

Ikegami, M., R. Dhami, and E. H. Schuchman. 2003. Alveolar lipoproteinosis in an acid sphingomyelinase-deficient mouse model of Niemann-Pick disease. American journal of physiology.Lung cellular and molecular physiology 284(3): 518.

293

Ikegami, M., J. A. Whitsett, Z. C. Chroneos, G. F. Ross, J. A. Reed, C. J. Bachurski, and A. H. Jobe. 2000. IL-4 increases surfactant and regulates metabolism in vivo. American journal of physiology.Lung cellular and molecular physiology 278(1): 75.

Iwamori, M., C. Costello, and H. W. Moser. 1979. Analysis and quantitation of free ceramide containing nonhydroxy and 2-hydroxy fatty acids, and phytosphingosine by high-performance liquid chromatography. Journal of lipid research 20(1): 86-96.

Iyonaga, K., M. Suga, T. Yamamoto, H. Ichiyasu, H. Miyakawa, and M. Ando. 1999. Elevated bronchoalveolar concentrations of MCP‐1 in patients with pulmonary alveolar proteinosis. European Respiratory Journal 14(2): 383-389.

Jankovic, J., and V. Rivera. 1978. Hereditary myoclonus and progressive muscular atrophy: a new syndrome. Transactions of the American Neurological Association 103: 116-118.

Jarisch, A., C. G. Steward, J. Sörensen, L. Porto, M. Kieslich, T. Klingebiel, and P. Bader. 2014. Odontoid infiltration and spinal compression in Farber Disease: reversal by haematopoietic stem cell transplantation. European journal of pediatrics 173(10): 1399-1403.

Jennemann, R., U. Rothermel, S. Wang, R. Sandhoff, S. Kaden, R. Out, T. J. van Berkel, J. M. Aerts, K. Ghauharali, and C. Sticht. 2010. Hepatic glycosphingolipid deficiency and liver function in mice. Hepatology 51(5): 1799-1809.

Jia, T., N. V. Serbina, K. Brandl, M. X. Zhong, I. M. Leiner, I. F. Charo, and E. G. Pamer. 2008. Additive roles for MCP-1 and MCP-3 in CCR2-mediated recruitment of inflammatory monocytes during Listeria monocytogenes infection. Journal of immunology (Baltimore, Md.: 1950) 180(10): 6846-6853.

Jmoudiak, M., and A. H. Futerman. 2005. Gaucher disease: pathological mechanisms and modern management. British journal of haematology 129(2): 178-188.

Johannsen, J., J. Heinemeyer, and J. Denecke. 2015. ASAH1 Mutation in a Boy with Non-5q SMA and Progressive Myoclonic Epilepsy. Neuropediatrics 46(S 01): 25.

Jones, E. E., T. W. Powers, B. A. Neely, L. H. Cazares, D. A. Troyer, A. S. Parker, and R. R. Drake. 2014. MALDI imaging mass spectrometry profiling of proteins and lipids in clear cell renal cell carcinoma. Proteomics 14(7-8): 924-935.

Kakkis, E. D., J. Muenzer, G. E. Tiller, L. Waber, J. Belmont, M. Passage, B. Izykowski, et al. 2001. Enzyme-replacement therapy in mucopolysaccharidosis I. New England Journal of Medicine 344(3): 182-188.

Kasumov, T., H. Huang, Y. Chung, R. Zhang, A. J. McCullough, and J. P. Kirwan. 2010. Quantification of ceramide species in biological samples by liquid chromatography electrospray ionization tandem mass spectrometry. Analytical Biochemistry 401(1): 154-161.

Kattner, E., A. Schäfer, and K. Harzer. 1997. Hydrops fetalis: manifestation in lysosomal storage diseases including Farber disease. European journal of pediatrics 156(4): 292-295.

Kawase, M., M. Watanabe, T. Kondo, T. Yabu, Y. Taguchi, H. Umehara, T. Uchiyama, K. Mizuno, and T. Okazaki. 2002. Increase of ceramide in adriamycin-induced HL-60 cell apoptosis: detection by a novel anti-ceramide antibody. Biochimica et Biophysica Acta (BBA)-Molecular and Cell Biology of Lipids 1584(2): 104-114.

294

Kernohan, K. D., L. Frésard, Z. Zappala, T. Hartley, K. S. Smith, J. Wagner, H. Xu, A. McBride, P. R. Bourque, and C. C. Consortium. 2017. Whole‐transcriptome sequencing in blood provides a diagnosis of spinal muscular atrophy with progressive myoclonic epilepsy. Human mutation 38(6): 611-614.

Kiernan, J. A. 1999. Histological and histochemical methods: theory and practice. Shock 12(6): 479.

Kim, S. Y., S. A. Choi, S. Lee, J. S. Lee, C. R. Hong, B. C. Lim, H. J. Kang, K. J. Kim, S. Park, and M. Choi. 2016. Atypical presentation of infantile‐onset farber disease with novel ASAH1 mutations. American Journal of Medical Genetics Part A 170(11): 3023-3027.

Kim, Y. J., S. J. Park, C. K. Park, S. H. Kim, and C. W. Lee. 1998. A case of Farber lipogranulomatosis. Journal of Korean medical science 13(1): 95-98.

Kirchhausen, T., D. Owen, and S. C. Harrison. 2014. Molecular structure, function, and dynamics of clathrin-mediated membrane traffic. Cold Spring Harbor perspectives in biology 6(5): a016725.

Kishnani, P. S., P. I. Dickson, L. Muldowney, J. J. Lee, A. Rosenberg, R. Abichandani, J. A. Bluestone, B. K. Burton, M. Dewey, and A. Freitas. 2016. Immune response to enzyme replacement therapies in lysosomal storage diseases and the role of immune tolerance induction. Molecular genetics and metabolism 117(2): 66-83.

Klingkowski, U., M. Beck, and F. Zépp. 1998. An 18-month-old girl with hoarseness, stiff joints and subcutaneous nodules. European journal of pediatrics 157(6): 515-516.

Koch, J., S. Gärtner, C. Li, L. E. Quintern, K. Bernardo, O. Levran, D. Schnabel, R. J. Desnick, E. H. Schuchman, and K. Sandhoff. 1996. Molecular cloning and characterization of a full-length complementary DNA encoding human acid ceramidase identification of the first molecular lesion causing Farber disease. Journal of Biological Chemistry 271(51): 33110-33115.

Koga, M., T. Ishihara, F. Uchino, and T. Fujiwaki. 1992. An autopsy case of Farber's lipogranulomatosis in a Japanese boy with gastrointestinal involvement. Pathology international 42(1): 42-48.

Kohl, B., S. Fischer, J. Groh, C. Wessig, and R. Martini. 2010. MCP-1/CCL2 modifies axon properties in a PMP22-overexpressing mouse model for Charcot-Marie-Tooth 1A neuropathy. The American journal of pathology 176(3): 1390-1399.

Kolesnick, R. N., and M. Krönke. 1998. Regulation of ceramide production and apoptosis. Annual Review of Physiology 60(1): 643-665.

Kondo, Y., J. M. Adams, M. T. Vanier, and I. D. Duncan. 2011. Macrophages counteract demyelination in a mouse model of globoid cell leukodystrophy. The Journal of neuroscience: the official journal of the Society for Neuroscience 31(10): 3610-3624.

Kondoh, K., T. Hineno, A. Sano, and Y. Kakimoto. 1991. Isolation and characterization of prosaposin from human milk. Biochemical and biophysical research communications 181(1): 286-292.

Kono, M., J. L. Dreier, J. M. Ellis, M. L. Allende, D. N. Kalkofen, K. M. Sanders, J. Bielawski, A. Bielawska, Y. A. Hannun, and R. L. Proia. 2006. Neutral ceramidase encoded by the Asah2 gene is essential for the intestinal degradation of sphingolipids. Journal of Biological Chemistry 281(11): 7324-7331.

295

Kostik, M. M., I. A. Chikova, V. V. Avramenko, L. I. Vasyakina, E. Le Trionnaire, V. G. Chasnyk, and T. Levade. 2013. Farber lipogranulomatosis with predominant joint involvement mimicking juvenile idiopathic arthritis. Journal of inherited metabolic disease 36(6): 1079-1080.

Kouskoff, V., A. Korganow, V. Duchatelle, C. Degott, C. Benoist, and D. Mathis. 1996. Organ-specific disease provoked by systemic autoimmunity. Cell 87(5): 811-822.

Krasnopolskaya, K. D., T. V. Mirenburg, E. L. Aronovich, T. V. Lebedeva, O. N. Odinokova, N. A. Demina, V. M. Kozlova, and M. I. Kuznetsov. 1993. Diagnosis and prevention of lysosomal storage diseases in Russia. Journal of inherited metabolic disease 16(6): 994-1002.

Krishnamurthy, K., S. Dasgupta, and E. Bieberich. 2007. Development and characterization of a novel anti-ceramide antibody. Journal of lipid research 48(4): 968-975.

Kudoh, T., and D. A. Wenger. 1982. Diagnosis of metachromatic leukodystrophy, Krabbe disease, and Farber disease after uptake of fatty acid-labeled cerebroside sulfate into cultured skin fibroblasts. The Journal of clinical investigation 70(1): 89-97.

Kuebler, W. M., Y. Yang, R. Samapati, and S. Uhlig. 2010. Vascular barrier regulation by PAF, ceramide, caveolae, and NO - an intricate signaling network with discrepant effects in the pulmonary and systemic vasculature. Cellular physiology and biochemistry: international journal of experimental cellular physiology, biochemistry, and pharmacology 26(1): 29-40.

Kurose, S., E. Ikeda, M. Tokiwa, N. Hikita, and M. Mochizuki. 2000. Effects of FTY720, a novel immunosuppressant, on experimental autoimmune uveoretinitis in rats. Experimental eye research 70(1): 7-15.

Lai, M. K., W. S. Chew, F. Torta, A. Rao, G. L. Harris, J. Chun, and D. R. Herr. 2016. Biological Effects of Naturally Occurring Sphingolipids, Uncommon Variants, and Their Analogs. Neuromolecular medicine 18(3): 396-414.

Lance, J. W., and W. A. Evans. 1984. Progressive myoclonic epilepsy, nerve deafness and spinal muscular atrophy. Clinical and experimental neurology 20: 141-151.

Landrum, M. J., J. M. Lee, M. Benson, G. Brown, C. Chao, S. Chitipiralla, B. Gu, J. Hart, D. Hoffman, and J. Hoover. 2015. ClinVar: public archive of interpretations of clinically relevant variants. Nucleic acids research 44(D1): D868.

Lawson, L. J., V. H. Perry, and S. Gordon. 1992. Turnover of resident microglia in the normal adult mouse brain. Neuroscience 48(2): 405-415.

Le Roy, C., and J. L. Wrana. 2005. Clathrin-and non-clathrin-mediated endocytic regulation of cell signalling. Nature reviews Molecular cell biology 6(2): 112-126.

Lee, S. M., C. Oh, S. Jung, and H. Kim. 2016. Farber Disease Misdiagnosed as Hemangioendothelioma. Journal of the Korean Association of Pediatric Surgeons 22(2): 54-58.

Lee, S. M., C. Schelcher, M. Demmel, M. Hauner, and W. E. Thasler. 2013. Isolation of human hepatocytes by a two-step collagenase perfusion procedure. Journal of visualized experiments: JoVE (79).

Lehnert, B. E., Y. E. Valdez, and L. M. Holland. 1985. Pulmonary macrophages: alveolar and interstitial populations. Experimental lung research 9(3-4): 177-190.

296

Levade, T., H. Enders, M. Schliephacke, and K. Harzer. 1995. A family with combined Farber and Sandhoff, isolated Sandhoff and isolated fetal Farber disease: postnatal exclusion and prenatal diagnosis of Farber disease using lipid loading tests on intact cultured cells. European journal of pediatrics 154(8): 643-648.

Levade, T., K. Sandhoff, H. Schulze, and J. A. Medin. 2014. Acid ceramidase deficiency: Farber lipogranulomatosis (143).

Levy, M., and A. H. Futerman. 2010. Mammalian ceramide synthases. IUBMB life 62(5): 347-356.

Li, C., S. Hong, G. Kopal, X. He, T. Linke, W. Hou, J. Koch, S. Gatt, K. Sandhoff, and E. H. Schuchman. 1998. Cloning and characterization of the full-length cDNA and genomic sequences encoding murine acid ceramidase. Genomics 50(2): 267-274.

Li, C., J. Park, X. He, B. Levy, F. Chen, K. Arai, D. A. Adler, C. M. Disteche, J. Koch, and K. Sandhoff. 1999. The human acid ceramidase gene (ASAH): structure, chromosomal location, mutation analysis, and expression. Genomics 62(2): 223-231.

Li, C., J. Park, C. M. Simonaro, X. He, R. E. Gordon, A. Friedman, D. Ehleiter, F. Paris, K. Manova, and S. Hepbiloikler. 2002. Insertional mutagenesis of the mouse acid ceramidase gene leads to early embryonic lethality in homozygotes and progressive lipid storage disease in heterozygotes. Genomics 79(2): 218-224.

Li, P., and E. M. Schwarz. 2003. The TNF-α transgenic mouse model of inflammatory arthritis 25(1): 19-33.

Liakath‐Ali, K., V. E. Vancollie, C. J. Lelliott, A. O. Speak, D. Lafont, H. J. Protheroe, C. Ingvorsen, A. Galli, A. Green, and D. Gleeson. 2016. Alkaline ceramidase 1 is essential for mammalian skin homeostasis and regulating whole‐body energy expenditure. The Journal of pathology 239(3): 374- 383.

Lin, H. Y., K. W. Chong, J. H. Hsu, H. C. Yu, C. C. Shih, C. H. Huang, S. J. Lin, et al. 2009. High incidence of the cardiac variant of Fabry disease revealed by newborn screening in the Taiwan Chinese population. Circulation.Cardiovascular genetics 2(5): 450-456.

Loiseau, N., A. Polizzi, A. Dupuy, N. Therville, M. Rakotonirainy, J. Loy, J. Viadere, A. Cossalter, J. Bailly, and O. Puel. 2015. New insights into the organ-specific adverse effects of fumonisin B1: comparison between lung and liver. Archives of Toxicology 89(9): 1619-1629.

Loizides‐Mangold, U. 2013. On the future of mass‐spectrometry‐based lipidomics. The FEBS journal 280(12): 2817-2829.

Lopez, M. E., A. D. Klein, J. Hong, U. J. Dimbil, and M. P. Scott. 2012. Neuronal and epithelial cell rescue resolves chronic systemic inflammation in the lipid storage disorder Niemann-Pick C. Human molecular genetics 21(13): 2946-2960.

Lopez-Vasquez, L., S. Dworski, R. Pouliot, T. Galbraith, M. A. Kamani, D. Lacroix, F. A. Auger, and J. A. Medin. 2016. Acid ceramidase deficiency leads to multiple skin abnormalities in a mouse model of Farber disease. Molecular genetics and metabolism 117(2): S76.

Lu, B., B. J. Rutledge, L. Gu, J. Fiorillo, N. W. Lukacs, S. L. Kunkel, R. North, C. Gerard, and B. J. Rollins. 1998. Abnormalities in monocyte recruitment and cytokine expression in monocyte chemoattractant protein 1-deficient mice. The Journal of experimental medicine 187(4): 601-608.

297

Luzio, J. P., Y. Hackmann, N. M. Dieckmann, and G. M. Griffiths. 2014. The biogenesis of lysosomes and lysosome-related organelles. Cold Spring Harbor perspectives in biology 6(9): a016840.

Luzio, J. P., P. R. Pryor, and N. A. Bright. 2007. Lysosomes: fusion and function. Nature reviews Molecular cell biology 8(8): 622-632.

Maceyka, M., S. G. Payne, S. Milstien, and S. Spiegel. 2002. Sphingosine kinase, sphingosine-1- phosphate, and apoptosis. Biochimica et Biophysica Acta (BBA)-Molecular and Cell Biology of Lipids 1585(2): 193-201.

Maceyka, M., and S. Spiegel. 2014. Sphingolipid metabolites in inflammatory disease. Nature 510(7503): 58-67.

Malm, G., A. M. Lund, J. Månsson, and A. Heiberg. 2008. Mucopolysaccharidoses in the Scandinavian countries: incidence and prevalence. Acta Paediatrica 97(11): 1577-1581.

Mango, R. L., L. Xu, M. S. Sands, C. Vogler, G. Seiler, T. Schwarz, M. E. Haskins, and K. P. Ponder. 2004. Neonatal retroviral vector-mediated hepatic gene therapy reduces bone, joint, and cartilage disease in mucopolysaccharidosis VII mice and dogs. Molecular genetics and metabolism 82(1): 4- 19.

Mao, C., and L. M. Obeid. 2008. Ceramidases: regulators of cellular responses mediated by ceramide, sphingosine, and sphingosine-1-phosphate. Biochimica et Biophysica Acta (BBA)- Molecular and Cell Biology of Lipids 1781(9): 424-434.

Mao, C., R. Xu, A. Bielawska, and L. M. Obeid. 2000. Cloning of an alkaline ceramidase from Saccharomyces cerevisiae an enzyme with reverse (CoA-independent) ceramide synthase activity. Journal of Biological Chemistry 275(10): 6876-6884.

Mao, C., R. Xu, A. Bielawska, Z. M. Szulc, and L. M. Obeid. 2000. Cloning and characterization of a Saccharomyces cerevisiae alkaline ceramidase with specificity for dihydroceramide. Journal of Biological Chemistry 275(40): 31369-31378.

Maor, G., S. Rencus-Lazar, M. Filocamo, H. Steller, D. Segal, and M. Horowitz. 2013. Unfolded protein response in Gaucher disease: from human to Drosophila. Orphanet journal of rare diseases 8(1): 140.

Marchesoni, C. L., N. Roa, A. M. Pardal, P. Neumann, G. Cáceres, P. Martínez, I. Kisinovsky, S. Bianchi, A. L. Tarabuso, and R. C. Reisin. 2010. Misdiagnosis in Fabry disease. The Journal of pediatrics 156(5): 828-831.

Mason, W. T. 1999. Fluorescent and luminescent probes for biological activity: a practical guide to technology for quantitative real-time analysis: Academic Press.

Masukawa, Y., H. Narita, E. Shimizu, N. Kondo, Y. Sugai, T. Oba, R. Homma, J. Ishikawa, Y. Takagi, and T. Kitahara. 2008. Characterization of overall ceramide species in human stratum corneum. Journal of lipid research 49(7): 1466-1476.

Matern, D., D. Gavrilov, D. Oglesbee, K. Raymond, P. Rinaldo, and S. Tortorelli. 2015. Newborn screening for lysosomal storage disorders 39(3): 206-216.

Matsuda, J., O. Suzuki, A. Oshima, A. Ogura, Y. Noguchi, Y. Yamamoto, T. Asano, K. Takimoto, K. Sukegawa, and Y. Suzuki. 1997. β-Galactosidase-deficient mouse as an animal model for GM1- gangliosidosis. Glycoconjugate journal 14(6): 729-736.

298

McNeill, A., G. Roberti, G. Lascaratos, D. Hughes, A. Mehta, D. F. Garway-Heath, and A. H. Schapira. 2013. Retinal thinning in Gaucher disease patients and carriers: results of a pilot study. Molecular genetics and metabolism 109(2): 221-223.

Mechtler, T. P., S. Stary, T. F. Metz, V. R. De Jesús, S. Greber-Platzer, A. Pollak, K. R. Herkner, B. Streubel, and D. C. Kasper. 2012. Neonatal screening for lysosomal storage disorders: feasibility and incidence from a nationwide study in Austria. The Lancet 379(9813): 335-341.

Medin, J. A., T. Takenaka, S. Carpentier, V. Garcia, J. Basile, B. Segui, N. Andrieu-Abadie, N. Auge, R. Salvayre, and T. Levade. 1999. Retrovirus-mediated correction of the metabolic defect in cultured Farber disease cells. Human Gene Therapy 10(8): 1321-1329.

Mehta, S., D. Blackinton, I. Omar, N. Kouttab, D. Myrick, J. Klostergaard, and H. Wanebo. 2000. Combined cytotoxic action of paclitaxel and ceramide against the human Tu138 head and neck squamous carcinoma cell line. Cancer chemotherapy and pharmacology 46(2): 85-92.

Mendell, J., S. Al-Zaidy, R. Shell, W. Arnold, L. Rodino-Klapac, T. Prior, L. Lowes, et al. 2017. AVXS- 101 phase 1 gene therapy clinical trial in SMA Type 1: end-of-Study event free survival and achievement of developmental milestones. Neuromuscular Disorders 27: S208.

Merrill, A. H., and M. C. Sullards. 2017. Opinion article on lipidomics: Inherent challenges of lipidomic analysis of sphingolipids. Biochimica et Biophysica Acta (BBA)-Molecular and Cell Biology of Lipids 1862(8): 774-776.

Mistry, P. K., J. Liu, M. Yang, T. Nottoli, J. McGrath, D. Jain, K. Zhang, et al. 2010. Glucocerebrosidase gene-deficient mouse recapitulates Gaucher disease displaying cellular and molecular dysregulation beyond the macrophage. Proceedings of the National Academy of Sciences of the United States of America 107(45): 19473-19478.

Mitchell, J., A. Solyom, B. Makay, N. Arslan, E. D. Batu, S. Ozen, B. Hügle, E. Schuchman, and B. Magnusson. 2016. Farber disease: Implications of anti-inflammatory treatment. Molecular genetics and metabolism 117(2): S82.

Molz, G. 1968. Farbersche Krankheit. Virchows Archiv 344(1): 86-99.

Momoi, T., Y. Ben-Yoseph, and H. L. Nadler. 1982. Substrate-specificities of acid and alkaline ceramidases in fibroblasts from patients with Farber disease and controls. The Biochemical journal 205(2): 419-425.

Mondal, R. K., M. Nandi, S. Datta, and M. Hira. 2009. Disseminated lipogranulomatosis. Indian pediatrics 46(2).

Morgan, R. A., D. Gray, A. Lomova, and D. B. Kohn. 2017. Hematopoietic Stem Cell Gene Therapy: Progress and Lessons Learned. Cell Stem Cell 21(5): 574-590.

Moritomo, H., T. Nakase, K. Maeda, T. Murase, and H. Yoshikawa. 2002. Surgical treatment of hand disorders in Farber's disease: A case report. The Journal of hand surgery 27(3): 503-507.

Moser, H. W. 1989. Ceramidase deficiency: Farber lipogranulomatosis. The metabolic basis of inherited disease.

Moser, H. W., A. L. Prensky, H. J. Wolfe, and N. P. Rosman. 1969. Farber's lipogranulomatosis: Report of a case and demonstration of an excess of free ceramide and ganglioside. The American Journal of Medicine 47(6): 869-890.

299

Moskot, M., K. Bocheńska, J. Jakóbkiewicz-Banecka, B. Banecki, and M. Gabig-Cimińska. 2018. Abnormal Sphingolipid World in Inflammation Specific for Lysosomal Storage Diseases and Skin Disorders. International Journal of Molecular Sciences 19(1): 247.

Mühlbauer, M., A. K. Bosserhoff, A. Hartmann, W. E. Thasler, T. S. Weiss, H. Herfarth, G. Lock, J. Schölmerich, and C. Hellerbrand. 2003. A novel MCP-1 gene polymorphism is associated with hepatic MCP-1 expression and severity of HCV-related liver disease. Gastroenterology 125(4): 1085- 1093.

Mullen, T. D., Y. A. Hannun, and L. M. Obeid. 2012. Ceramide synthases at the centre of sphingolipid metabolism and biology. Biochemical Journal 441(3): 789-802.

Mura, M., M. Binnie, B. Han, C. Li, C. F. Andrade, A. Shiozaki, Y. Zhang, N. Ferrara, D. Hwang, and T. K. Waddell. 2010. Functions of type II pneumocyte-derived vascular endothelial growth factor in alveolar structure, acute inflammation, and vascular permeability. The American journal of pathology 176(4): 1725-1734.

Muramatsu, T., N. Sakai, I. Yanagihara, M. Yamada, T. Nishigaki, C. Kokubu, H. Tsukamoto, M. Ito, and K. Inui. 2002. Mutation analysis of the acid ceramidase gene in Japanese patients with Farber disease. Journal of inherited metabolic disease 25(7): 585-592.

Muranjan, M., S. Agarwal, K. Lahiri, and M. Bashyam. 2012. Novel biochemical abnormalities and genotype in Farber disease. Indian Pediatr. 49(4): 320-322.

Naldini, L. 2011. Ex vivo gene transfer and correction for cell-based therapies. Nature Reviews Genetics 12(5).

Naldini, L. 2015. Gene therapy returns to centre stage. Nature 526(7573): 351-360.

Nasreen, A., A. K. Sunil, M. R. Pulak, and B. K. Sujit. 2017. Clinical Diagnosis of Farber’s Disease-A Rare Case Report. Int J Health Sci Res. 7(7): 388-390.

Navarro, C., C. Fachal, C. Rodriguez, L. L. Padro, and C. Dominguez. 1999. Lipoid proteinosis. A biochemical and ultrastructural investigation of two new cases. British Journal of Dermatology 141(2): 326-331.

Niu, J., and P. E. Kolattukudy. 2009. Role of MCP-1 in cardiovascular disease: molecular mechanisms and clinical implications. Clinical science (London, England : 1979) 117(3): 95-109.

Nivaggioni, V., A. Cano, I. Arnoux, G. Michel, and M. Loosveld. 2016. Early morphological diagnosis of Farber disease. British journal of haematology 175(2): 189.

Nowaczyk, M., A. Feigenbaum, M. M. Silver, J. Callahan, A. Levin, and V. Jay. 1996. Bone marrow involvement and obstructive jaundice in Farber lipogranulomatosis: clinical and autopsy report of a new case. Journal of inherited metabolic disease 19(5): 655-660.

Nyberg, L., A. Farooqi, L. Bläckberg, R. Duan, Å Nilsson, and O. Hernell. 1998. Digestion of ceramide by human milk bile salt-stimulated lipase. Journal of pediatric gastroenterology and nutrition 27(5): 560-567.

Ogretmen, B., and Y. A. Hannun. 2004. Biologically active sphingolipids in cancer pathogenesis and treatment. Nature Reviews Cancer 4(8): 604-616.

300

Oguz Akarsu, E., P. Tekturk, Z. Yapici, F. Tepgec, Z. O. Uyguner, and B. Baykan. 2016. Eyelid myoclonic status epilepticus: A rare phenotype in spinal muscular atrophy with progressive myoclonic epilepsy associated with ASAH1 gene mutation. Seizure 42: 49-51.

Ohfu, M., K. Abe, H. Hirano, H. Ogata, A. Mitsudome, S. Okada, K. Inui, and J. Nishimoto. 1987. Farber Lipogranulomatosis-A report of a case with nystagmus, myoclonus and convulsions. Brain Dev 9(2): 227.

Okino, N., X. He, S. Gatt, K. Sandhoff, M. Ito, and E. H. Schuchman. 2003. The reverse activity of human acid ceramidase. The Journal of biological chemistry 278(32): 29948-29953.

Ortolano, S., I. Viéitez, C. Navarro, and C. Spuch. 2014. Treatment of lysosomal storage diseases: recent patents and future strategies. Recent patents on endocrine, metabolic & immune drug discovery 8(1): 9-25.

Osawa, Y., H. Uchinami, J. Bielawski, R. F. Schwabe, Y. A. Hannun, and D. A. Brenner. 2005a. Roles for C16-ceramide and sphingosine 1-phosphate in regulating hepatocyte apoptosis in response to tumor necrosis factor-α. Journal of Biological Chemistry 280(30): 27879-27887.

Ozaki, H., M. Mizutani, H. Hayashi, E. Oka, S. Ohtahara, H. Kimoto, T. Tanaka, H. Hakozaki, K. Takahashi, and Y. Suzuki. 1978. Farber's disease (disseminated lipogranulomatosis): the first case reported in Japan. Acta Medica Okayama 32(1): 69-79.

Özkara, B., and F. Budak. 2017. Spinal Muscular Atrophy With Myoclonic Epilepsy. Epilepsi: Journal of the Turkish Epilepsi Society 23(1).

Pacienza, N., M. Yoshimitsu, N. Mizue, B. C. Au, J. C. Wang, X. Fan, T. Takenaka, and J. A. Medin. 2012. Lentivector transduction improves outcomes over transplantation of human HSCs alone in NOD/SCID/Fabry mice. Molecular Therapy 20(7): 1454-1461.

Parenti, G., G. Andria, and A. Ballabio. 2015. Lysosomal storage diseases: from pathophysiology to therapy. Annual Review of Medicine 66: 471-486.

Passini, M. A., J. C. Dodge, J. Bu, W. Yang, Q. Zhao, D. Sondhi, N. R. Hackett, S. M. Kaminsky, Q. Mao, and L. S. Shihabuddin. 2006. Intracranial delivery of CLN2 reduces brain pathology in a mouse model of classical late infantile neuronal ceroid lipofuscinosis. Journal of Neuroscience 26(5): 1334- 1342.

Patel, B. V., M. R. Wilson, K. P. O'Dea, and M. Takata. 2013. TNF-induced death signaling triggers alveolar epithelial dysfunction in acute lung injury. Journal of immunology (Baltimore, Md.: 1950) 190(8): 4274-4282.

Pavlova, E. V., P. B. Deegan, J. Tindall, I. McFarlane, A. Mehta, D. Hughes, J. E. Wraith, and T. M. Cox. 2011. Potential biomarkers of osteonecrosis in Gaucher disease. Blood Cells, Molecules, and Diseases 46(1): 27-33.

Pavone, L., H. W. Moser, F. Mollica, C. Reitano, and P. Durand. 1980. Farber's lipogranulomatosis: ceramidase deficiency and prolonged survival in three relatives. The Johns Hopkins medical journal 147(5): 193-196.

Pellissier, J. F., M. Berard-Badier, and N. Pinsard. 1986. Farber's disease in two siblings, sural nerve and subcutaneous biopsies by light and electron microscopy. Acta Neuropathologica 72(2): 178-188.

Perera, R. M., and R. Zoncu. 2016. The lysosome as a regulatory hub. Annual Review of Cell and Developmental Biology 32: 223-253.

301

Peters, C., E. G. Shapiro, J. Anderson, P. J. Henslee-Downey, M. R. Klemperer, M. J. Cowan, E. F. Saunders, et al. 1998. Hurler syndrome: II. Outcome of HLA-genotypically identical sibling and HLA- haploidentical related donor bone marrow transplantation in fifty-four children. The Storage Disease Collaborative Study Group. Blood 91(7): 2601-2608.

Petrache, I., and E. V. Berdyshev. 2016. Ceramide signaling and metabolism in pathophysiological states of the lung. Annual Review of Physiology 78: 463-480.

Petrache, I., K. Kamocki, C. Poirier, Y. Pewzner-Jung, E. L. Laviad, K. S. Schweitzer, M. Van Demark, M. J. Justice, W. C. Hubbard, and A. H. Futerman. 2013. Ceramide synthases expression and role of ceramide synthase-2 in the lung: insight from human lung cells and mouse models. PLoS One 8(5): e62968.

Pewzner-Jung, Y., S. Tavakoli Tabazavareh, H. Grassme, K. A. Becker, L. Japtok, J. Steinmann, T. Joseph, et al. 2014. Sphingoid long chain bases prevent lung infection by Pseudomonas aeruginosa. EMBO molecular medicine 6(9): 1205-1214.

Pewzner-Jung, Y., O. Brenner, S. Braun, E. L. Laviad, S. Ben-Dor, E. Feldmesser, S. Horn-Saban, D. Amann-Zalcenstein, C. Raanan, and T. Berkutzki. 2010. A critical role for ceramide synthase 2 in liver homeostasis II. Insights into molecular changes leading to hepatopathy. Journal of Biological Chemistry 285(14): 10911-10923.

Pinto, A. R., A. Ilinykh, M. J. Ivey, J. T. Kuwabara, M. D'Antoni, R. J. Debuque, A. Chandran, L. Wang, K. Arora, and N. Rosenthal. 2015. Revisiting cardiac cellular composition. Circulation research: CIRCRESAHA. 115.307778.

Platt, F. M., B. Boland, and van der Spoel, Aarnoud C. 2012. Lysosomal storage disorders: the cellular impact of lysosomal dysfunction. J Cell Biol 199(5): 723-734.

Platt, F. M., G. R. Neises, R. A. Dwek, and T. D. Butters. 1994. N-butyldeoxynojirimycin is a novel inhibitor of glycolipid biosynthesis. Journal of Biological Chemistry 269(11): 8362-8365.

Prensky, A. L., G. Ferreira, S. Carr, and H. W. Moser. 1967. Ceramide and Ganglioside Accumulation in Farber's Lipogranulomatosis.∗. Proceedings of the Society for Experimental Biology and Medicine 126(3): 725-728.

Pruett, S. T., A. Bushnev, K. Hagedorn, M. Adiga, C. A. Haynes, M. C. Sullards, D. C. Liotta, and A. H. Merrill. 2008. Thematic Review Series: Sphingolipids. Biodiversity of sphingoid bases (“sphingosines”) and related amino alcohols. Journal of lipid research 49(8): 1621-1639.

Qualman, S. J., H. W. Moser, D. Valle, A. E. Moser, S. E. Antonarakis, J. K. Boitnott, W. H. Zinkham, J. M. Opitz, and J. Bernstein. 1987. Farber disease: pathologic diagnosis in sibs with phenotypic variability. American Journal of Medical Genetics Part A 28(S3): 233-241.

Quehenberger, O., A. M. Armando, A. H. Brown, S. B. Milne, D. S. Myers, A. H. Merrill, S. Bandyopadhyay, K. N. Jones, S. Kelly, and R. L. Shaner. 2010. Lipidomics reveals a remarkable diversity of lipids in human plasma. Journal of lipid research 51(11): 3299-3305.

Radin, N. S. 1996. Treatment of Gaucher disease with an enzyme inhibitor. Glycoconjugate journal 13(2): 153-157.

Raith, K., and R. H. Neubert. 2000. Liquid chromatography–electrospray mass spectrometry and tandem mass spectrometry of ceramides. Analytica Chimica Acta 403(1): 295-303.

302

Raith, K., R. Wolf, J. Wagner, and R. H. Neubert. 1998. Separation of phospholipids by nonaqueous capillary electrophoresis with electrospray ionisation mass spectrometry. Journal of Chromatography A 802(1): 185-188.

Ramsubir, S., T. Nonaka, C. B. Girbés, S. Carpentier, T. Levade, and J. A. Medin. 2008. In vivo delivery of human acid ceramidase via cord blood transplantation and direct injection of lentivirus as novel treatment approaches for Farber disease. Molecular genetics and metabolism 95(3): 133-141.

Rastall, D. P., and A. Amalfitano. 2015. Recent advances in gene therapy for lysosomal storage disorders. The application of clinical genetics 8: 157.

Rauch, H. J., and L. Auböck. 1983. " Banana bodies" in disseminated lipogranulomatosis (Farber's disease). The American Journal of Dermatopathology 5(3): 263-266.

Rauvala, H., and M. Hallman. 1984. Glycolipid accumulation in bronchoalveolar space in adult respiratory distress syndrome. Journal of lipid research 25(11): 1257-1262.

Raveney, B. J., D. A. Copland, L. B. Nicholson, and A. D. Dick. 2008. Fingolimod (FTY720) as an acute rescue therapy for intraocular inflammatory disease. Archives of Ophthalmology 126(10): 1390- 1395.

Rivel, J., C. Vital, J. Battin, J. P. Heheunstre, and H. Leger. 1977. La lipogranulomatose disséminée de Farber études anatomoclinique et ultrastructurale de deux observations familiales. Arch Anat Cytol Pathol 25: 37-42.

Rivera, J., R. L. Proia, and A. Olivera. 2008. The alliance of sphingosine-1-phosphate and its receptors in immunity. Nature Reviews Immunology 8(10): 753.

Robert, K. Y., Y. Nakatani, and M. Yanagisawa. 2009. The role of glycosphingolipid metabolism in the developing brain. Journal of lipid research 50(Supplement): S445.

Rombach, S. M., B. E. Smid, M. G. Bouwman, G. E. Linthorst, M. G. Dijkgraaf, and C. E. Hollak. 2013. Long term enzyme replacement therapy for Fabry disease: effectiveness on kidney, heart and brain. Orphanet journal of rare diseases 8(1): 47.

Rubboli, G., P. Veggiotti, A. Pini, A. Berardinelli, G. Cantalupo, E. Bertini, F. D. Tiziano, A. D'amico, E. Piazza, and E. Abiusi. 2015. Spinal muscular atrophy associated with progressive myoclonic epilepsy: a rare condition caused by mutations in ASAH1. Epilepsia 56(5): 692-698.

Ruckhaberle, E., U. Holtrich, K. Engels, L. Hanker, R. Gatje, D. Metzler, T. Karn, M. Kaufmann, and A. Rody. 2009. Acid ceramidase 1 expression correlates with a better prognosis in ER-positive breast cancer. Climacteric 12(6): 502-513.

Ruhwald, M., M. Bjerregaard-Andersen, P. Rabna, J. Eugen-Olsen, and P. Ravn. 2009. IP-10, MCP- 1, MCP-2, MCP-3, and IL-1RA hold promise as biomarkers for infection with M. tuberculosis in a whole blood-based T-cell assay. BMC research notes 2(1): 19.

Rutsaert, J., M. Tondeur, E. Vamos-Hurwitz, and P. Dustin. 1977. The cellular lesions of Farber's disease and their experimental reproduction in tissue culture. Laboratory investigation; a journal of technical methods and pathology 36(5): 474-480.

Ryan, A. J., D. M. McCoy, S. E. McGowan, R. G. Salome, and R. K. Mallampalli. 2003. Alveolar sphingolipids generated in response to TNF-alpha modifies surfactant biophysical activity. Journal of applied physiology (Bethesda, Md.: 1985) 94(1): 253-258.

303

Saddoughi, S. A., P. Song, and B. Ogretmen. 2008. Roles of bioactive sphingolipids in cancer biology and therapeutics: 413-440.

Salo, M. K., R. Karikoski, M. Hällström, H. Jalanko, and C. Holmberg. 2003. Farber Disease diagnosed after liver transplantation. Journal of pediatric gastroenterology and nutrition 36(2): 274- 277.

Samie, M. A., and H. Xu. 2014. Lysosomal exocytosis and lipid storage disorders. Journal of lipid research 55(6): 995-1009.

Samuelsson, K., R. Zetterström, and B. I. Ivemark. 1972. Studies on a case of lipogranulomatosis (Farber’s disease) with protracted course: 533-548.

Samuelsson, K., and R. Zetterström. 1971. Ceramides in a patient with lipogranulomatosis (Farber's disease) with chronic course. Scandinavian Journal of Clinical and Laboratory Investigation 27(4): 393-405.

Sands, M. S., and B. L. Davidson. 2006. Gene therapy for lysosomal storage diseases. Molecular Therapy 13(5): 839-849.

Sango, K., M. Takano, K. Ajiki, A. Tokashiki, N. Arai, H. Kawano, H. Horie, and S. Yamanaka. 2005. Impaired neurite outgrowth in the retina of a murine model of Sandhoff disease. Investigative ophthalmology & visual science 46(9): 3420-3425.

Santos-Cortez, R. L. P., Y. Hu, F. Sun, F. Benahmed-Miniuk, J. Tao, J. K. Kanaujiya, S. Ademola, et al. 2017. Identification of ASAH1 as a susceptibility gene for familial keloids. European journal of human genetics: EJHG 25(10): 1155-1161.

Sanvicens, N., and T. G. Cotter. 2006. Ceramide is the key mediator of oxidative stress‐induced apoptosis in retinal photoreceptor cells. Journal of neurochemistry 98(5): 1432-1444.

Sarafi, M. N., E. A. Garcia-Zepeda, J. A. MacLean, I. F. Charo, and A. D. Luster. 1997. Murine monocyte chemoattractant protein (MCP)-5: a novel CC chemokine that is a structural and functional homologue of human MCP-1. The Journal of experimental medicine 185(1): 99-109.

Sassa, T., and A. Kihara. 2014. Metabolism of very long-chain fatty acids: genes and pathophysiology. Biomolecules & therapeutics 22(2): 83.

Sassa, T., S. Suto, Y. Okayasu, and A. Kihara. 2012. A shift in sphingolipid composition from C24 to C16 increases susceptibility to apoptosis in HeLa cells. Biochimica et Biophysica Acta (BBA)- Molecular and Cell Biology of Lipids 1821(7): 1031-1037.

Sathe, S., and T. Pearson. 2014. Phenotypic characterization of the spinal muscular atrophy with progressive myoclonus epilepsy syndrome caused by ASAH1 mutations. Molecular genetics and metabolism 111(2): S93.

Sayah, A., and J. C. English. 2005. Rheumatoid arthritis: a review of the cutaneous manifestations. Journal of the American Academy of Dermatology 53(2): 191-209.

Saygi, S., Z. Haytoglu, T. Savas, O. Alkan, and I. Erol. 2015. P78–2554: Farber disease in a child with a novel homozygous c. 92G> T mutation. European Journal of Paediatric Neurology 19: S116.

Schafer, A., K. Harzer, E. Kattner, H. J. Schafer, G. Stoltenburg, and H. Lietz. 1996. Disseminated lipogranulomatosis (Farber disease) with hydrops fetalis. Der Pathologe 17(2): 145-149.

304

Schanche, A. F., S. M. Bierman, R. L. Sopher, and B. J. O'loughlin. 1964. Disseminated lipogranulomatosis: early roentgenographic changes. Radiology 82(4): 675-678.

Schiffmann, R., J. B. Kopp, H. A. Austin III, S. Sabnis, D. F. Moore, T. Weibel, J. E. Balow, and R. O. Brady. 2001. Enzyme replacement therapy in Fabry disease: a randomized controlled trial. Jama 285(21): 2743-2749.

Schlondorff, D. 1987. The glomerular mesangial cell: an expanding role for a specialized pericyte. The FASEB Journal 1(4): 272-281.

Schmoeckel, C. 1980. Subtle clues to diagnosis of skin diseases by electron microscopy:" Farber bodies" in disseminated lipogranulomatosis (Farber's disease). The American Journal of Dermatopathology 2(2): 153-154.

Schmoeckel, C., and M. Hohlfed. 1979. A specific ultrastructural marker for disseminated lipogranulomatosis (Farber). Archives of Dermatological Research 266(2): 187-196.

Schnabel, D., M. Schroder, W. Furst, A. Klein, R. Hurwitz, T. Zenk, J. Weber, K. Harzer, B. C. Paton, and A. Poulos. 1992. Simultaneous deficiency of sphingolipid activator proteins 1 and 2 is caused by a mutation in the initiation codon of their common gene. The Journal of biological chemistry 267(5): 3312-3315.

Schneider, E. L., C. J. Epstein, M. J. Kaback, and D. Brandes. 1977. Severe pulmonary involvement in adult Gaucher's disease: report of three cases and review of the literature. The American Journal of Medicine 63(3): 475-480.

Schuchman, E. H. 2016. Acid ceramidase and the treatment of ceramide diseases: The expanding role of enzyme replacement therapy. Biochimica et Biophysica Acta (BBA)-Molecular Basis of Disease 1862(9): 1459-1471.

Schuchman, E. H., J. Clancy, E. Dimango, I. Petrache, E. Gulbins, and X. He. 2015. Novel use of the lysosomal enzyme acid ceramidase for the treatment of inflammatory lung diseases, including cystic fibrosis. Molecular genetics and metabolism 114(2): S105.

Schuchman, E. H., M. Frohbergh, J. M. Guevara, X. He, V. A. DeAngelis, and C. M. Simonaro. 2016. Cartilage and bone disease in a mouse model of Farber lipogranulomatosis and response to treatment. Molecular genetics and metabolism 117(2): S103.

Schuchman, E. H., J. Mitchell, and A. Solyom. 2017. Morbidity and mortality associated with Farber disease and prospects for therapy. Expert Opinion on Orphan Drugs 5(9): 717-726.

Schultze, G., and E. K. Lang. 1960. Disseminated lipogranulomatosis: report of a case. Radiology 74(3): 428-431.

Scott, C. T., and L. DeFrancesco. 2016. Gene therapy's out-of-body experience. Nature biotechnology 34(6): 600-607.

Scott, R., S. Elliott, N. Buroker, L. I. Thomas, J. Keutzer, M. Glass, M. H. Gelb, and F. Turecek. 2013. Identification of infants at risk for developing Fabry, Pompe, or mucopolysaccharidosis-I from newborn blood spots by tandem mass spectrometry. The Journal of pediatrics 163(2): 498-503.

Seeger, W., G. Stohr, H. R. Wolf, and H. Neuhof. 1985. Alteration of surfactant function due to protein leakage: special interaction with fibrin monomer. Journal of applied physiology (Bethesda, Md.: 1985) 58(2): 326-338.

305

Seelan, R. S., C. Qian, A. Yokomizo, D. G. Bostwick, D. I. Smith, and W. Liu. 2000. Human acid ceramidase is overexpressed but not mutated in prostate cancer. Genes, Chromosomes and Cancer 29(2): 137-146.

Seidova, S., K. Kotliar, F. Foerger, M. Klopfer, and I. Lanzl. 2009. Functional retinal changes in Gaucher disease. Documenta ophthalmologica 118(2): 151-154.

Sessa, M., L. Lorioli, F. Fumagalli, S. Acquati, D. Redaelli, C. Baldoli, S. Canale, et al. 2016. Lentiviral haemopoietic stem-cell gene therapy in early-onset metachromatic leukodystrophy: an ad-hoc analysis of a non-randomised, open-label, phase 1/2 trial. The Lancet 388(10043): 476-487.

Settembre, C., C. Di Malta, V. A. Polito, M. Garcia Arencibia, F. Vetrini, S. Erdin, S. U. Erdin, et al. 2011. TFEB links autophagy to lysosomal biogenesis. Science (New York, N.Y.) 332(6036): 1429- 1433.

Seymour, J. F., and J. J. Presneill. 2002. Pulmonary alveolar proteinosis: progress in the first 44 years. American journal of respiratory and critical care medicine 166(2): 215-235.

Shaaltiel, Y., S. Gingis–Velitski, S. Tzaban, N. Fiks, Y. Tekoah, and D. Aviezer. 2015. Plant‐based oral delivery of β‐glucocerebrosidase as an enzyme replacement therapy for Gaucher's disease. Plant biotechnology journal 13(8): 1033-1040.

Shell, R., S. Al-Zaidy, W. D. Arnold, L. Rodino-Klapac, T. W. Prior, K. Kotha, L. Lowes, et al. 2017. AVXS-101 Phase 1 Gene Therapy Clinical Trial In SMA Type 1: Interim Data Demonstrates Improvements In Survival, Motor Function, And Decreased Dependence On Ventilator Support: A2648.

Sheth, J., J. J. Joseph, K. Shah, M. Muranjan, M. Mistri, and F. Sheth. 2017. Pulmonary manifestations in Niemann-Pick type C disease with mutations in NPC2 gene: case report and review of literature. BMC Medical Genetics 18(1): 5.

Shtraizent, N., E. Eliyahu, J. H. Park, X. He, R. Shalgi, and E. H. Schuchman. 2008. Autoproteolytic cleavage and activation of human acid ceramidase. The Journal of biological chemistry 283(17): 11253-11259.

Sideris, G. A., and M. Josephson. 2016. Pulmonary alveolar proteinosis and Niemann Pick disease type B: An unexpected combination. Respiratory Medicine Case Reports 19: 37-39.

Sikora, J., S. Dworski, E. E. Jones, M. A. Kamani, M. C. Micsenyi, T. Sawada, P. Le Faouder, et al. 2017. Acid Ceramidase Deficiency in Mice Results in a Broad Range of Central Nervous System Abnormalities. The American Journal of Pathology 187(4): 864-883.

Sirrs, S., C. Hollak, M. Merkel, A. Sechi, E. Glamuzina, M. C. Janssen, R. Lachmann, J. Langendonk, M. Scarpelli, and T. B. Omran. 2015. The frequencies of different inborn errors of metabolism in adult metabolic centres: Report from the ssiem adult metabolic physicians group: 85-91.

Snider, A. J., B. X. Wu, R. W. Jenkins, J. A. Sticca, T. Kawamori, Y. A. Hannun, and L. M. Obeid. 2012. Loss of neutral ceramidase increases inflammation in a mouse model of inflammatory bowel disease. Prostaglandins & other lipid mediators 99(3): 124-130.

Solyom, A., B. Huegle, B. Magnusson, B. Makay, N. Arslan, J. Mitchell, P. Tanpaiboon, N. Guelbert, R. Puri, and L. Jung. 2015. Farber Disease: Important Differential Diagnostic Information for Jia and Other Inflammatory Arthritis Phenotypes Is Revealed By Data from the Largest Clinical Cohort to Date. Arthritis & Rheumatology 67: 1838-1839.

306

Solyom, A., B. Johnson, and A. Shih. 2018. Clinical study design for enzyme replacement therapy in acid ceramidase deficiency presenting as Farber disease, an ultra-rare condition with broad phenotypic heterogeneity. Molecular genetics and metabolism 123(2): S134.

Solyom, A., J. Mitchell, M. Beck, B. Hügle, and E. H. Schuchman. 2017. Farber disease: design of the first observational and cross-sectional cohort study capturing retrospective and prospective data on the natural history and phenotypic spectrum of patients, including novel methodologies for assessment of disease-specific symptoms. Molecular genetics and metabolism 120(1): S125.

Solyom, A., C. M. Simonaro, X. He, and E. H. Schuchman. 2015. Acid ceramidase deficiency: Clinical implications of an emerging phenotypic spectrum and potential therapies. Molecular genetics and metabolism 2(114): S109.

Sommer, U., H. Herscovitz, F. K. Welty, and C. E. Costello. 2006. LC-MS-based method for the qualitative and quantitative analysis of complex lipid mixtures. Journal of lipid research 47(4): 804- 814.

Souillet, G., P. Guiband, A. H. Fensom, I. Maire, and M. T. Zabot. 1989. Outcome of displacement bone marrow transplantation in Farber’s disease: a report of a case. Correction of certain genetic diseases by transplantation. London: COGENT 1989: 137-141.

Spence, M. W., S. Beed, and H. W. Cook. 1986. Acid and alkaline ceramidases of rat tissues. Biochemistry and Cell Biology 64(5): 400-404.

Spener, F., M. Lagarde, A. Geloen, and M. Record. 2003. What is lipidomics?. European journal of lipid science and technology 105(9): 481-482.

Standiford, T. J., S. L. Kunkel, N. W. Lukacs, M. J. Greenberger, J. M. Danforth, R. G. Kunkel, and R. M. Strieter. 1995. Macrophage inflammatory protein-1 alpha mediates lung leukocyte recruitment, lung capillary leak, and early mortality in murine endotoxemia. Journal of immunology (Baltimore, Md.: 1950) 155(3): 1515-1524.

Staretz-Chacham, O., T. C. Lang, M. E. LaMarca, D. Krasnewich, and E. Sidransky. 2009. Lysosomal storage disorders in the newborn. Pediatrics 123(4): 1191-1207.

Stenson, P. D., M. Mort, E. V. Ball, K. Howells, A. D. Phillips, N. S. Thomas, and D. N. Cooper. 2009. The human gene mutation database: 2008 update. Genome medicine 1(1): 13.

Strelow, A., K. Bernardo, S. Adam-Klages, T. Linke, K. Sandhoff, M. Krönke, and D. Adam. 2000. Overexpression of acid ceramidase protects from tumor necrosis factor–induced cell death. Journal of Experimental Medicine 192(5): 601-612.

Strettoi, E., C. Gargini, E. Novelli, G. Sala, I. Piano, P. Gasco, and R. Ghidoni. 2010. Inhibition of ceramide biosynthesis preserves photoreceptor structure and function in a mouse model of retinitis pigmentosa. Proceedings of the National Academy of Sciences of the United States of America 107(43): 18706-18711.

Suga, M., K. Iyonaga, H. Ichiyasu, N. Saita, H. Yamasaki, and M. Ando. 1999. Clinical significance of MCP-1 levels in BALF and serum in patients with interstitial lung diseases. The European respiratory journal 14(2): 376-382.

Sugita, M., J. T. Dulaney, and H. W. Moser. 1972. Ceramidase deficiency in Farber's disease (lipogranulomatosis). Science (New York, N.Y.) 178(4065): 1100-1102.

307

Sugita, M., M. Iwamori, J. Evans, R. H. McCluer, J. T. Dulaney, and H. W. Moser. 1974. High performance liquid chromatography of ceramides: application to analysis in human tissues and demonstration of ceramide excess in Farber's disease. Journal of lipid research 15(3): 223-226.

Sullards, M. C. 2000. Analysis of sphingomyelin, glucosylceramide, ceramide, sphingosine, and sphingosine 1-phosphate by tandem mass spectrometry. Methods in enzymology 312: 32-45.

Sullards, M. C., Y. Liu, Y. Chen, and A. H. Merrill. 2011. Analysis of mammalian sphingolipids by liquid chromatography tandem mass spectrometry (LC-MS/MS) and tissue imaging mass spectrometry (TIMS). Biochimica et Biophysica Acta (BBA)-Molecular and Cell Biology of Lipids 1811(11): 838-853.

Summers, S. A., and D. H. Nelson. 2005. A role for sphingolipids in producing the common features of type 2 diabetes, metabolic syndrome X, and Cushing’s syndrome. Diabetes 54(3): 591-602.

Sun, W., J. Jin, R. Xu, W. Hu, Z. M. Szulc, J. Bielawski, L. M. Obeid, and C. Mao. 2010. Substrate specificity, membrane topology, and activity regulation of human alkaline ceramidase 2 (ACER2). Journal of Biological Chemistry 285(12): 8995-9007.

Taglioli, M., S. Bartolini, G. Volpi, G. Alberti, and G. Ambrosetto. 1990. Progressive familial myoclonic epilepsy with bulbo-spinal amyotrophy. Clinical, electrophysiological study, and biopsy of a case. Rivista di neurologia 60(5): 201-206.

Tanaka, T., K. Takahashi, H. Hakozaki, H. Kimoto, and Y. Suzuki. 1979. Farber's Disease (Disseminated Lipogranulomatosis) A Pathological, Histochemical and Ultrastructural Study. Pathology international 29(1): 135-155.

Tani, M., N. Okino, S. Mitsutake, and M. Ito. 1999. Specific and sensitive assay for alkaline and neutral ceramidases involving C12-NBD-ceramide. The Journal of Biochemistry 125(4): 746-749.

Tardieu, M., M. Zérah, M. Gougeon, J. Ausseil, S. de Bournonville, B. Husson, D. Zafeiriou, et al. 2017. Intracerebral gene therapy in children with mucopolysaccharidosis type IIIB syndrome: an uncontrolled phase 1/2 clinical trial. The Lancet Neurology 16(9): 712-720.

Teoh, H. L., A. Solyom, E. H. Schuchman, D. Mowat, T. Roscioli, M. Farrar, and H. Sampaio. 2016. Polyarticular Arthritis and Spinal Muscular Atrophy in Acid Ceramidase Deficiency. Pediatrics 138(4): e20161068. Epub 2016 Sep 20.

Tesch, G. H. 2008. MCP-1/CCL2: a new diagnostic marker and therapeutic target for progressive renal injury in diabetic nephropathy. American journal of physiology. Renal physiology 294(4): 697.

Thomas, R. L., C. M. Matsko, M. T. Lotze, and A. A. Amoscato. 1999. Mass spectrometric identification of increased C16 ceramide levels during apoptosis. Journal of Biological Chemistry 274(43): 30580-30588.

Thurberg, B. L., C. L. Maloney, C. Vaccaro, K. Afonso, A. C. Tsai, E. Bossen, P. S. Kishnani, and M. O'Callaghan. 2006. Characterization of pre-and post-treatment pathology after enzyme replacement therapy for Pompe disease. Laboratory investigation 86(12): 1208-1220.

Tibboel, J., I. Reiss, J. C. de Jongste, and M. Post. 2014. Sphingolipids in lung growth and repair. CHEST Journal 145(1): 120-128.

Topaloglu, H., and J. Melki. 2016. Spinal muscular atrophy associated with progressive myoclonus epilepsy. Epileptic Disorders 18(s2): 128-134.

308

Toppet, M., E. Vamos‐Hurwitz, G. Jonniaux, N. Cremer, M. Tondeur, and S. Pelc. 1978. Farber's disease as a ceramidosis: clinical, radiological and biochemical aspects. Acta Paediatrica 67(1): 113- 119.

Torcoletti, M., A. Petaccia, R. M. Pinto, U. Hladnik, F. Locatelli, C. Agostoni, and F. Corona. 2014. Farber disease in infancy resembling juvenile idiopathic arthritis: identification of two new mutations and a good early response to allogeneic haematopoietic stem cell transplantation. Rheumatology 53(8): 1533-1534.

Triggs-Raine, B. L., A. S. Feigenbaum, M. Natowicz, M. Skomorowski, S. M. Schuster, J. T. Clarke, D. J. Mahuran, E. H. Kolodny, and R. A. Gravel. 1990. Screening for carriers of Tay-Sachs disease among Ashkenazi Jews: a comparison of DNA-based and enzyme-based tests. New England Journal of Medicine 323(1): 6-12.

Tsugawa, H., K. Ikeda, W. Tanaka, Y. Senoo, M. Arita, and M. Arita. 2017. Comprehensive identification of sphingolipid species by in silico retention time and tandem mass spectral library. Journal of cheminformatics 9(1): 19.

Turner, L. S., J. C. Cheng, T. H. Beckham, T. E. Keane, J. S. Norris, and X. Liu. 2011. Autophagy is increased in prostate cancer cells overexpressing acid ceramidase and enhances resistance to C6 ceramide. Prostate cancer and prostatic diseases 14(1): 30-37.

Uhlig, S., and E. Gulbins. 2008. Sphingolipids in the lungs. American journal of respiratory and critical care medicine 178(11): 1100-1114.

Van Breemen, M. J., M. De Fost, J. S. Voerman, J. D. Laman, R. G. Boot, M. Maas, C. E. Hollak, J. M. Aerts, and F. Rezaee. 2007. Increased plasma macrophage inflammatory protein (MIP)-1α and MIP-1β levels in type 1 Gaucher disease. Biochimica et Biophysica Acta (BBA)-Molecular Basis of Disease 1772(7): 788-796.

Van den Hout, J., A. Reuser, J. De Klerk, W. F. Arts, J. Smeitink, and Van der Ploeg, A T. 2001. Enzyme therapy for Pompe disease with recombinant human α-glucosidase from rabbit milk. Journal of inherited metabolic disease 24(2): 266-274. van Doorn, R., P. G. Nijland, N. Dekker, M. E. Witte, M. A. Lopes-Pinheiro, B. van het Hof, G. Kooij, A. Reijerkerk, C. Dijkstra, and van van der Valk, Paul. 2012. Fingolimod attenuates ceramide-induced blood–brain barrier dysfunction in multiple sclerosis by targeting reactive astrocytes. Acta Neuropathologica: 1-14. van Echten-Deckert, G., A. Klein, T. Linke, T. Heinemann, J. Weisgerber, and K. Sandhoff. 1997. Turnover of endogenous ceramide in cultured normal and Farber fibroblasts. Journal of lipid research 38(12): 2569-2579. van Lijnschoten, G., J. E. Groener, S. M. Maas, Y. Ben-Yoseph, K. P. Dingemans, and G. J. A. Offerhaus. 2000. Intrauterine fetal death due to Farber disease: case report. Pediatric and Developmental Pathology 3(6): 597-602.

Van Veldhoven, P. P., W. R. Bishop, D. A. Yurivich, and R. M. Bell. 1995. Ceramide quantitation: evaluation of a mixed micellar assay using E. coli diacylglycerol kinase. Biochemistry and molecular biology international 36(1): 21-30.

Vejselova, D., H. M. Kutlu, and G. Kuş. 2016. Examining impacts of ceranib-2 on the proliferation, morphology and ultrastructure of human breast cancer cells. Cytotechnology 68(6): 2721-2728.

309

Vielhaber, G., L. Brade, B. Lindner, S. Pfeiffer, R. Wepf, U. Hintze, K. Wittern, and H. Brade. 2001. Mouse anti-ceramide antiserum: a specific tool for the detection of endogenous ceramide. Glycobiology 11(6): 451-457.

Vogel, P., M. S. Donoviel, R. Read, G. M. Hansen, J. Hazlewood, S. J. Anderson, W. Sun, J. Swaffield, and T. Oravecz. 2009. Incomplete inhibition of sphingosine 1-phosphate lyase modulates immune system function yet prevents early lethality and non-lymphoid lesions. PloS one 4(1): e4112.

Vormoor, J., K. Ehlert, A. H. Groll, H. Koch, M. Frosch, and J. Roth. 2004. Successful hematopoietic stem cell transplantation in Farber disease. The Journal of pediatrics 144(1): 132-134.

Walia, J. S., A. Neschadim, O. Lopez-Perez, A. Alayoubi, X. Fan, S. Carpentier, M. Madden, C. Lee, F. Cheung, and D. A. Jaffray. 2011. Autologous Transplantation of Lentivector/Acid Ceramidase– Transduced Hematopoietic Cells in Nonhuman Primates. Human Gene Therapy 22(6): 679-687.

Wang, H. 2017. Crosslink between lipids and uveitis: a lipidomic analysis. Investigative ophthalmology & visual science 58(8): 5758.

Wang, K., R. Xu, A. J. Snider, J. Schrandt, Y. Li, A. B. Bialkowska, M. Li, J. Zhou, Y. A. Hannun, and L. M. Obeid. 2016. Alkaline ceramidase 3 deficiency aggravates colitis and colitis-associated tumorigenesis in mice by hyperactivating the innate immune system. Cell death & disease 7(3): e2124.

Wang, L., and S. M. Dudek. 2009. Regulation of vascular permeability by sphingosine 1-phosphate. Microvascular research 77(1): 39-45.

Wang, R. Y., A. Lelis, J. Mirocha, and W. R. Wilcox. 2007. Heterozygous Fabry women are not just carriers, but have a significant burden of disease and impaired quality of life. Genetics in Medicine 9(1): 34-45.

Weaver, T. E., C. Na, and M. Stahlman. 2002. Biogenesis of lamellar bodies, lysosome-related organelles involved in storage and secretion of pulmonary surfactant 13(4): 263-270.

Weber, C., A. Krueger, A. Munk, C. Bode, P. P. Van Veldhoven, and M. H. Graler. 2009. Discontinued postnatal thymocyte development in sphingosine 1-phosphate-lyase-deficient mice. Journal of immunology (Baltimore, Md.: 1950) 183(7): 4292-4301.

Wenger, D. A., S. Coppola, and S. Liu. 2003. Insights into the diagnosis and treatment of lysosomal storage diseases. Archives of Neurology 60(3): 322-328.

Wevers, R. A., and V. Gieselmann. 2005. Lysosomes and lysosomal disorders. Magnetic Resonance of Myelination and Myelin Disorders: 66-73.

Willis, A., C. VanHuse, K. P. Newton, M. Wasserstein, and R. A. Morotti. 2008. Farber's disease type IV presenting with cholestasis and neonatal liver failure: report of two cases. Pediatric and Developmental Pathology 11(4): 305-308.

Wittmann, J., E. Karg, S. Turi, E. Legnini, G. Wittmann, A. Giese, J. Lukas, U. Gölnitz, M. Klingenhäger, and O. Bodamer. 2012. Newborn screening for lysosomal storage disorders in Hungary: 117-125.

Worgall, S., D. Sondhi, N. R. Hackett, B. Kosofsky, M. V. Kekatpure, N. Neyzi, J. P. Dyke, D. Ballon, L. Heier, and B. M. Greenwald. 2008. Treatment of late infantile neuronal ceroid lipofuscinosis by CNS administration of a serotype 2 adeno-associated virus expressing CLN2 cDNA. Human Gene Therapy 19(5): 463-474.

310

Wraith, J. E. 2004. The clinical presentation of lysosomal storage disorders. Acta Neurologica Taiwanica 13: 101-106.

Wu, B. X., J. Fan, N. P. Boyer, R. W. Jenkins, Y. Koutalos, Y. A. Hannun, and C. E. Crosson. 2015. Lack of Acid Sphingomyelinase Induces Age-Related Retinal Degeneration. PloS one 10(7): e0133032.

Wu, Y. P., and R. L. Proia. 2004. Deletion of macrophage-inflammatory protein 1 alpha retards neurodegeneration in Sandhoff disease mice. Proceedings of the National Academy of Sciences of the United States of America 101(22): 8425-8430.

Wymann, M. P., and R. Schneiter. 2008. Lipid signalling in disease. Nature Reviews Molecular Cell Biology 9(2): 162-176.

Xia, J. Y., W. L. Holland, C. M. Kusminski, K. Sun, A. X. Sharma, M. J. Pearson, A. J. Sifuentes, J. G. McDonald, R. Gordillo, and P. E. Scherer. 2015. Targeted induction of ceramide degradation leads to improved systemic metabolism and reduced hepatic steatosis. Cell metabolism 22(2): 266-278.

Xu, H., and D. Ren. 2015. Lysosomal physiology. Annual Review of Physiology 77: 57-80.

Xu, R., J. Jin, W. Hu, W. Sun, J. Bielawski, Z. Szulc, T. Taha, L. M. Obeid, and C. Mao. 2006. Golgi alkaline ceramidase regulates cell proliferation and survival by controlling levels of sphingosine and S1P. The FASEB Journal 20(11): 1813-1825.

Xu, R., K. Wang, I. Mileva, Y. A. Hannun, L. M. Obeid, and C. Mao. 2016. Alkaline ceramidase 2 and its bioactive product sphingosine are novel regulators of the DNA damage response. Oncotarget 7(14): 18440.

Yang, G., L. Badeanlou, J. Bielawski, A. J. Roberts, Y. A. Hannun, and F. Samad. 2009. Central role of ceramide biosynthesis in body weight regulation, energy metabolism, and the metabolic syndrome. American journal of physiology. Endocrinology and metabolism 297(1): 211.

Yang, Y., J. Yin, W. Baumgartner, R. Samapati, E. A. Solymosi, E. Reppien, W. M. Kuebler, and S. Uhlig. 2010. Platelet-activating factor reduces endothelial nitric oxide production: role of acid sphingomyelinase. The European respiratory journal 36(2): 417-427.

Yang, Y., and S. Uhlig. 2011. The role of sphingolipids in respiratory disease. Therapeutic advances in respiratory disease 5(5): 325-344.

Yanovitch, T. L., S. G. Banugaria, A. D. Proia, and P. S. Kishnani. 2010. Clinical and histologic ocular findings in pompe disease. Journal of pediatric ophthalmology and strabismus 47(1): 34-40.

Yasugi, E., and K. Watanabe. 2002. LIPIDBANK for Web, the newly developed lipid database. Tanpakushitsu kakusan koso. Protein, nucleic acid, enzyme 47(7): 837-841.

Yeager, A. M., K. Armfield Uhas, C. D. Coles, P. C. Davis, W. L. Krause, and H. W. Moser. 2000. Bone marrow transplantation for infantile ceramidase deficiency (Farber disease). Bone marrow transplantation 26(3).

Yildiz, E. P., G. Yesil, G. Bektas, M. Caliskan, B. Tatlı, N. Aydinli, and M. Ozmen. 2018. Spinal muscular atrophy with progressive myoclonic epilepsy linked to mutations in ASAH1. Clinical neurology and neurosurgery 164: 47-49.

Yoshimura, T., E. A. Robinson, S. Tanaka, E. Appella, and E. J. Leonard. 1989. Purification and amino acid analysis of two human monocyte chemoattractants produced by phytohemagglutinin-

311 stimulated human blood mononuclear leukocytes. Journal of immunology (Baltimore, Md.: 1950) 142(6): 1956-1962.

Yu, F. P. S., S. Amintas, T. Levade, and J. A. Medin. 2018. Acid ceramidase deficiency: Farber disease and SMA-PME. Orphanet Journal of Rare Diseases 13(1).

Yu, F. P. S., S. Dworski, and J. A. Medin. 2018. Deletion of MCP-1 Impedes Pathogenesis of Acid Ceramidase Deficiency. Scientific reports 8(1): 6.

Yu, F. P., D. Islam, J. Sikora, S. Dworski, J. Gurka', L. Lopez-Vasquez, M. Liu, et al. 2017. Chronic lung injury and impaired pulmonary function in a mouse model of acid ceramidase deficiency. American journal of physiology. Lung cellular and molecular physiology: ajplung.00223.2017.

Zappatini-Tommasi, L., C. Dumontel, P. Guibaud, and C. Girod. 1992. Farber disease: an ultrastructural study. Virchows Archiv A 420(3): 281-290.

Zarbin, M. A., W. R. Green, A. B. Moser, and C. Tiffany. 1988. Increased levels of ceramide in the retina of a patient with Farber's disease. Archives of Ophthalmology 106(9): 1163.

Zarbin, M. A., W. R. Green, H. W. Moser, and S. J. Morton. 1985. Farber's disease: light and electron microscopic study of the eye. Archives of Ophthalmology 103(1): 73-80.

Zetterström, R. 1958. Disseminated lipogranulomatosis (Farber's disease). Acta paediatrica 47(5): 501-510.

Zhang, H., D. Li, Y. Su, S. Jiang, Y. Xu, K. Jiang, and D. Cui. 2012. Identification of the N- acylsphingosine amidohydrolase 1 gene (ASAH1) for susceptibility to schizophrenia in a Han Chinese population. The World Journal of Biological Psychiatry 13(2): 106-113.

Zhang, Z., A. K. Mandal, A. Mital, N. Popescu, D. Zimonjic, A. Moser, H. Moser, and A. B. Mukherjee. 2000. Human acid ceramidase gene: novel mutations in Farber disease. Molecular genetics and metabolism 70(4): 301-309.

Zhou, J., M. Tawk, F. D. Tiziano, J. Veillet, M. Bayes, F. Nolent, V. Garcia, S. Servidei, E. Bertini, and F. Castro-Giner. 2012. Spinal muscular atrophy associated with progressive myoclonic epilepsy is caused by mutations in ASAH1. The American Journal of Human Genetics 91(1): 5-14.

Ziegler, R. J., C. Li, M. Cherry, Y. Zhu, D. Hempel, N. van Rooijen, Y. A. Ioannou, R. J. Desnick, M. A. Goldberg, and N. S. Yew. 2002. Correction of the nonlinear dose response improves the viability of adenoviral vectors for gene therapy of Fabry disease. Human Gene Therapy 13(8): 935-945.

Zielonka, M., S. F. Garbade, S. Kölker, G. F. Hoffmann, and M. Ries. 2017. A cross-sectional quantitative analysis of the natural history of Farber disease: an ultra-orphan condition with rheumatologic and neurological cardinal disease features. Genetics in Medicine.

Zoja, C., D. Corna, M. Locatelli, D. Rottoli, A. Pezzotta, M. Morigi, C. Zanchi, et al. 2015. Effects of MCP-1 inhibition by bindarit therapy in a rat model of polycystic kidney disease. Nephron 129(1): 52- 61.

312

Chapter 1 Permission to reuse the article “Acid ceramidase deficiency: Farber disease and SMA-PME” was granted under the Creative Commons Attribution 4.0 International Licence.

Chapter 4 Permission to reuse the article “Chronic lung injury and impaired pulmonary function in a mouse model of acid ceramidase deficiency” was granted November 22m 2017 by Joseph Girouard on behalf of the American Journal of Physiology- Lung Molecular and Cellular Physiology.

Chapter 6 Permission to reuse the article “Deletion of MCP-1 Impedes Pathogenesis of acid Ceramidase Deficiency” was granted under the Creative Commons Attribution 4.0 International License.