Alveolar Epithelial Heparan Sulfate and Chondroitin Sulfate in The

ALVEOLAR EPITHELIAL HEPARAN SULFATE AND CHONDROITIN SULFATE IN THE

HEALTHY AND INJURED LUNG

SARAH MARIE HAEGER

B.S. University of Colorado, 2011

A thesis submitted to the

Faculty of the Graduate School of the

University of Colorado in partial fulfillment

of the requirements for the degree of

Doctor of Philosophy

Pharmacology Program

2018

This thesis for the Doctor of Philosophy degree by

Sarah Marie Haeger

has been approved for the

Pharmacology Program

Peter M. Henson, Chair

Anthony N. Gerber

Mary Weiser-Evans

Arthur Gutierrez-Hartmann

Christopher M. Evans

Rubin M. Tuder, Advisor

Eric P. Schmidt, Co-Advisor

Date: 05/18/2018

Haeger, Sarah Marie (Ph.D., Pharmacology)

Alveolar Epithelial Heparan Sulfate and Chondroitin Sulfate in the Healthy and Injured Lung

Thesis directed by Professor Rubin M. Tuder

ABSTRACT1

The lung epithelial glycocalyx is a carbohydrate-enriched layer lining the pulmonary epithelial surface. Although epithelial glycocalyx visualization has been reported in vivo, its composition and function remain unknown. Furthermore, while sulfated glycosaminoglycans

(GAGs), including heparan sulfate (HS) and chondroitin sulfate (CS), are known to critically contribute to the structure and function of the glycocalyx of numerous cell types, the contribution of HS and CS to the lung epithelial glycocalyx and its function has not been studied.

Disruption of the cell glycocalyx can occur during tissue inflammation and injury. In vitro and in vivo models of lung epithelial injury have demonstrated shedding of the heparan sulfate and chondroitin sulfate proteoglycans syndecan-1 and syndecan-4; however, the release of epithelial HS and CS, and its effect on lung injury and repair has not been studied. In this dissertation a variety of approaches were utilized to determine the existence, structure, and function of alveolar epithelial HS and CS in the healthy lung, to identify and elucidate the mechanism of epithelial HS and CS shedding during intratracheal-induced lung injury in mice and in patients with the acute respiratory distress syndrome (ARDS), and to determine the effect of soluble HS and CS on lung injury and repair during intratracheal-induced lung injury in mice.

Using immunofluorescence and mass spectrometry, we identified heparan sulfate (HS) and chondroitin sulfate (CS) within the lung epithelial glycocalyx. In vivo selective enzymatic

1Portions of this abstract were previously published in the American journal of respiratory cell and molecular biology. Reprinted with permission of the American Thoracic Society. Copyright © 2018 American Thoracic Society. (1, 2)

iii degradation of epithelial HS, but not CS, increased baseline lung permeability. Using mass spectrometry and gel electrophoresis approaches to determine the fate of epithelial HS and CS during lung injury, we detected shedding of ≥ β0 saccharide-long HS and CS into bronchoalveolar lavage fluid in intratracheal LPS-treated mice. Further, airspace HS and CS in clinical samples from acute respiratory distress syndrome patients correlated with indices of alveolar permeability, reflecting the translational relevance of these findings. Using pharmacologic and transgenic animal approaches, we determined that matrix metalloproteinases

(MMPs) partially mediate HS shedding during intratracheal LPS-induced lung injury. We found a trend towards decreased alveolar permeability after treatment with the MMP inhibitor doxycycline; however, this did not reach statistical significance. HS shed into the alveolar airspace during lung injury was discovered to be heavily un-sulfated, while shed alveolar CS was enriched in 4-O sulfation. As sulfation is essential for most HS functions, it is unlikely that alveolar HS binds other mediators, and as such, may have no effect on lung injury development or subsequent repair. In contrast, we discovered that exogenous 4-O sulfated CS (CS-A) can induce mild inflammation itself and enhance LPS-induced alveolar inflammation, indicating that alveolar 4-O sulfated CS may contribute to alveolar inflammation during injury.

These studies suggest that epithelial HS contributes to the lung epithelial barrier and its degradation is sufficient to increase lung permeability. Furthermore, shedding of epithelial HS and CS into the alveolar airspace occurs during lung injury and correlates with alveolar permeability. While alveolar HS may have no effect on lung injury or repair, alveolar CS may enhance alveolar inflammation during lung injury.

The form and content of this abstract are approved. I recommend its publication.

Approved: Rubin M. Tuder

DEDICATION

To my family – Mary, Kurt, Kristen, Adam, and Vanessa. Thank you for embracing and fostering my sense of curiosity and adventure. Your support replenishes me and allows me to be persistent and continue to strive for success.

ACKNOWLEDGEMENTS

First and foremost, I would like to thank my mentors, Drs. Eric Schmidt and Rubin

Tuder. I’m honored to have been able to learn from such accomplished, enthusiastic, and caring physician scientists. You have invested so much time into my training and progress, and have given me numerous opportunities to both learn from and share my work with others. I am very thankful to have been mentored by people that care as much as you do about my success. I would also like to thank all the members of the Tuder-Graham-Schmidt lab. I gained a second family when I joined the lab and am very thankful to have the support and friendship of everyone. I am especially thankful for Drs. Kaori Oshima and Sarah McMurtry; your friendship will be life-long.

I would like to acknowledge Dr. Robert Linhardt and all the members of the Linhardt lab.

The surface plasmon resonance and GAG mass spectrometry included in this thesis were performed by members of the Linhardt lab. I would also like to personally thank the Linhardt lab for their hospitality during my visits and for the time spent teaching me about the GAG isolation and mass spectrometry techniques used for this work.

I would also like to acknowledge Dr. Rachel Zemans for her advice and technical support. Thank you for taking the time and effort to listen and provide advice about my project, and for teaching me how to isolate rat primary type II cells.

Some reagents and samples for the work described in this thesis were provided by Drs.

Eva Nozik-Grayck and Julie Bastarache. Thank you Dr. Grayck for providing us with extracellular superoxide dismutase overexpressing mice. Thank you Dr. Bastarache for providing us with heat and moisture exchanger fluid from ARDS patients; I have always desired for this work to be truly translational.

I am very grateful for the students and leadership of the University of Colorado MSTP and Pharmacology Graduate Program. A special thank you to my MSTP class; I am very appreciative of your continued support and friendship. It is rare to find a group of people that share such infinite warmth and care.

Finally, I am extremely appreciative to have such a wonderful family and fiancée. I am very grateful to have parents who taught me work ethic, persistence, and humility. Through you

I have learned to give 100% effort to everything I commit, and to never dwell on my failures or boast in my successes. I’m lucky to have such an amazing sister and friend, Kristen, with whom

I’ve laughed through life with ever since I can remember. Lastly, I am so thankful for my fiancée and best friend, Vanessa. You are so supportive of anything I want to do. I can’t wait for all of our adventures ahead.

vii

TABLE OF CONTENTS

CHAPTER Page

I. HEPARAN SULFATE AND CHONDROITIN SULFATE 1

Heparan Sulfate and Chondroitin Sulfate Structure and Biosynthesis 1

General Functions of Heparan Sulfate and Chondroitin Sulfate 5

Degradation, Turnover, and Shedding of Heparan Sulfate and Chondroitin Sulfate 8

Endoglycosidases 9

Proteases 11

Matrix metalloproteinases 11

A disintegrin and metalloproteinases and a disintigren and metalloproteinase 12 with thrombospondin motifs

Reactive Oxygen Species 13

Expression, Location, and Importance of Heparan Sulfate and Chondroitin Sulfate in the 14 Lung

Expression and Location of Heparan Sulfate and Chondroitin Sulfate 14

Importance of Heparan Sulfate and Chondroitin Sulfate 16

II. ALVEOLAR STRUCTURE AND FUCNTION 18

Overview of Lung Structure and Function 18

The Alveolar Endothelium 19

The Alveolar Type I Epithelium 21

The Alveolar Type II Epithelium 24

Pulmonary Surfactant Production 25

Tight Barrier Formation 26

Alveolar Fluid Clearance 26

viii

Alveolar Epithelial Progenitor Capacity 28

Alveolar Epithelial Type II Heparan Sulfate and Chondroitin Sulfate 28

III. ACUTE LUNG INJURY AND THE ACUTE RESPIRATORY DISTRESS 30 SYNDROME

Acute Respiratory Distress Syndrome Definition and Epidemiology 30

Pathophysiology of ARDS 31

Pathogenesis of ARDS 32

Indirect Lung Injury 33

Direct Lung Injury 35

Treatment of ARDS 38

IV. THESIS STATEMENT 40

V. ALVEOLAR EPITHELIAL HEPARAN SULFATE AND CHONDROITIN 42 SULFATE IN THE HEALTHY LUNG

Introduction 42

Objectives 43

Materials and Methods 43

Materials 43

Animals 43

Alveolar Epithelial HS and CS Degradation 43

Isolation and Quantification of BAL Fluid and Plasma Heparan Sulfate 44

Lung/Alveolar Permeability, Edema, and Inflammation Quantification 45

Immunohistochemistry 45

RNA Isolation, cDNA Synthesis, and qRT-PCR 46

Statistical Analyses 46

Results 47

Epithelial Glycocalyx Heparan Sulfate Contributes to Epithelial Barrier Function 47

Epithelial Glycocalyx Chondroitin Sulfate Does Not Contribute to Epithelial Barrier 49 Function

Heparinase I/III-Generated Heparan Sulfate Fragments do not Increase Lung 51 Epithelial Permeability

Intratracheal Heparinase I/III does not Change Alveolar ZO-1 or Claudin-18 53 Expression

Summary of Results 54

Discussion 55

VI. ALVEOLAR EPITHELIAL HEPARAN SULFATE AND CHONDROITIN 57 SULFATE ARE SHED DURING DIRECT LUNG INJURY

Introduction 57

Objectives 58

Materials and Methods 58

Materials 58

Animals 59

Intratracheal LPS-Induced Lung Injury 59

HME collection 60

Isolation, Quantification, and Size Determination of HME Fluid, BAL Fluid and 60 Plasma Heparan Sulfate and Chondroitin Sulfate

Lung/Alveolar Permeability and Inflammation Quantification 60

Western Blotting 60

Zymography 61

RNA Isolation, cDNA Synthesis, and qRT-PCR 61

Statistical Analyses 61

Results 61

Heparan Sulfate and Chondroitin Sulfate are Released into the Airspace During 61 Intratracheal LPS-Induced Lung Injury in Mice

Heparan Sulfate and Chondroitin Sulfate are Released into the Airspace During the 63 Acute Respiratory Distress Syndrome in Humans

Heparan Sulfate and Chondroitin Sulfate Shed During Intratracheal LPS-Induced 65 Lung Injury are Long and are Accompanied by Shedding of Epithelial Syndecan-1 and Syndecan-4

Matrix Metalloproteinases Mediate Alveolar Heparan Sulfate Shedding During 67 Intratracheal LPS-Induced Lung Injury

MMP-9 Knockout Mice are not Protected Against Alveolar Heparan Sulfate or 71 Chondroitin Sulfate Shedding During Intratracheal LPS-Induced Lung Injury

Summary of Results 73

Discussion 75

VII. SOLUBLE ALVEOLAR CHONDROITIN SULFATE ENHANCES LUNG 77 INJURY, WHILE HEPARAN SULFATE DOES NOT EFFECT LUNG REPAIR

Introduction 77

Objectives 78

Materials and Methods 78

Materials 78

Animals 79

Intratracheal LPS-Induced Lung Injury 79

Determination of Heparan Sulfate and Chondroitin Sulfate Sulfation 79

Surface Plasmon Resonance 79

Lung/Alveolar Permeability and Inflammation Quantification 80

Statistical Analyses 80

Results 80

Alveolar Heparan Sulfate is Relatively Un-Sulfated in Mice and Patients with Direct 80 Lung Injury

Heparan Sulfate Sulfation and Size Characteristics Necessary for Binding to 81 Epithelial Reparative Growth Factors

Exogenous Heparin Octosaccharides or Full-Length Heparan Sulfate do not Affect 82 Repair After Intratracheal LPS-Induced Lung Injury

Alveolar Chondroitin Sulfate is 4-O Sulfated in Mice and Patients with Direct Lung 84 Injury

Intratracheal Protamine does not Affect the Severity of or Recovery from Lung Injury 85

Exogenous CS-A Induces Mild Inflammation and Enhances Intratracheal 87 LPS-Induced Lung Injury

Summary of Results 90

Discussion 90

VIII. CONCLUSIONS AND FUTURE DIRECTIONS 92

Perspective 92

Summary of Findings 92

Chapter V 93

Chapter VI 94

Chapter VII 94

Future Directions 95

Mechanism by which Epithelial Heparan Sulfate Contributes to Alveolar Barrier 95 Function

Differences in Sulfation of Alveolar Heparan Sulfate 97

Evolutionary Benefit of Redundant Heparan Sulfate Sheddases during Lung Injury 98

Alveolar Heparan Sulfate and Chondroitin Sulfate as Biomarkers of Direct Lung 99 Injury

Determination of Alveolar Chondroitin Sulfate Subunit Composition and Function 100

xii

REFERENCES 102

APPENDIX 121

A. HEPARAN SULFATE MODULATES HEPATOCYTE GROWTH 121 FACTOR SIGNALING IN ALVEOLAR TYPE II CELLS

xiii

LIST OF TABLES

TABLE Page

I-1 Chondroitin Sulfate Subtypes 5

I-2 Heparan Sulfate and Chondroitin Sulfate Binding Proteins 8

I-3 Lung Developmental Defects in Heparan Sulfate and Chondroitin Sulfate 17 Biosynthesis Genetic Mutants

III-1 The ARDS Berlin Definition 30

VI-1 ARDS HME Fluid Patient Demographics 65

xiv

LIST OF FIGURES

FIGURE Page

I-1 Heparan Sulfate and Chondroitin Sulfate Biosynthesis 4

I-2 Functions of Heparan Sulfate and Chondroitin Sulfate on the Cell Surface and 6 in the Extracellular Matrix

I-3 Degradation and Shedding of Heparan Sulfate, Chondroitin Sulfate, and 9 Proteoglycan Core Proteins

II-1 Alveolar Structure and Epithelial Function 22

IV-1 Thesis Aims 41

V-1 Intratracheal Heparinase I/III Removes Alveolar Epithelial Heparan Sulfate 48 Releasing Alveolar N- and 2-O Sulfated Heparan Sulfate Fragments

V-2 Intratracheal Heparinase I/III Increases Alveolar Permeability to Protein, but 50 not Lung Edema or Alveolar Inflammation

V-3 Intratracheal Chondroitinase ABC Removes Alveolar Epithelial Chondroitin 51 Sulfate but does not Increase Lung Permeability or Inflammation

V-4 Exogenous Heparinase I/III-Generated Heparan Sulfate Fragments do not 53 Increase Lung Permeability or Inflammation

V-5 Intratracheal Heparinase I/III does not Alter Alveolar ZO-1 or Claudin-18 54 Expression

VI-1 Increased Airspace Heparan Sulfate and Chondroitin Sulfate are Detected 64 During Intratracheal LPS-Induced Lung Injury in Mice

VI-2 HME Heparan Sulfate and Chondroitin Sulfate Correlate with HME Protein in 66 ARDS Patients

VI-3 Increased Airspace Heparan Sulfate and Chondroitin Sulfate After Intratracheal 67 LPS Instillation is Long and is Accompanied by Increased Airspace Syndecan-1 and Syndecan-4

VI-4 EC-SOD Overexpression and Intratracheal TAPI-2 Instillation do not Attenuate 69 the Increased Airspace Heparan Sulfate, Chondroitin Sulfate or Protein after Intratracheal LPS Instillation

VI-5 Lung MMP-9 mRNA Expression, and BAL MMP-2 and MMP-9 Protein and 70 Activity are Increased After Intratracheal LPS Instillation

VI-6 Doxycycline Partially Inhibits the Increase in Airspace Heparan Sulfate and 72 Syndecan-1, but not Chondroitin Sulfate, Syndecan-4 or BAL Protein, After Intratracheal LPS Instillation

VI-7 MMP-9 Knockout Mice are not Protected from Alveolar Heparan Sulfate, 74 Chondroitin Sulfate, Syndecan-1, or Syndecan-4 Shedding, or BAL Protein or Neutrophilia, and Exhibit Increased BAL MMP-2 Protein After Intratracheal LPS Instillation

VII-1 Alveolar Heparan Sulfate is Heavily Un-Sulfated in Mice During Intratracheal 81 LPS-Induced Lung Injury and in ARDS Patients

VII-2 Size and Sulfation of Heparin Necessary to Bind Keratinocyte Growth Factor and 83 Hepatocyte Growth Factor

VII-3 Intratracheal Heparin Octasaccharides and Full Length Heparan Sulfate do not 84 Affect Alveolar Permeability or Inflammation on Day 5 After Intratracheal LPS

VII-4 Alveolar Chondroitin Sulfate is Enriched in 4-O Sulfation in Mice During 85 Intratracheal LPS-Induced Lung Injury and in ARDS Patients

VII-5 Intratracheal Protamine does not Affect Alveolar Permeability or Inflammation 88 on Day 2 or Day 5 After Intratracheal LPS

VII-6 High Dose Intratracheal Chondroitin Sulfate-A Enhances LPS-Induced 89 Permeability and Inflammation and Induces Modest Inflammation in the Absence of a Prior Stimulation

VIII-1 Thesis Findings 93

VIII-2 HME Total Sulfated Heparan Sulfate and Chondroitin Sulfate in ARDS Patients 100

A-1 Heparan Sulfate may Inhibit Hepatocyte Growth Factor Signaling in ATII Cells 125

A-2 Heparan Sulfate may not Inhibit Hepatocyte Growth Factor Signaling in ATII 127 Cells Devoid of Cell-Surface Heparan Sulfate

A-3 Hepatocyte Growth Factor may not Accelerate MLE-12 Scratch Closure 129

xvi

LIST OF ABBREVIATIONS

Abbreviation Meaning

ADAM A disintegrin and metalloproteinase

ADAMTS A disintegrin and metalloproteinase with thrombospondin motifs

APACHE Acute physiology and chronic health evaluation

ARDS the Acute respiratory distress syndrome

ATI Alveolar type I

ATII Alveolar type II

BAL Bronchoalveolar lavage

CFTR Cystic fibrosis transmembrane regulator

CS Chondroitin sulfate

DAMP Damage-associated molecular pattern

DAPI 4’,6-diamidino-2-phenylindole

DMMB 1,9-Dimethylmethylene blue

DS Dermatan sulfate

ECM Extracellular matrix

EC-SOD Extracellular superoxide dismutase

ENaC Epithelial sodium channel

ESL Endothelial surface layer

FGF Fibroblast growth factor

FITC Fluorescein isothiocyanate

GAG Glycosaminoglycan

GalNAc N-acetyl galactosamine

xvii

GlcA Glucuronic acid

GlcNAc N-acetyl glucosamine

HGF Hepatocyte growth factor

HI Heat-inactivated

HME Heat and moisture exchanger

HS Heparan sulfate

ICAM Intercellular adhesion molecule

ICU Intensive care unit

IdoA Iduronic acid

IL Interleukin

JAM Junctional adhesion molecule

KGF Keratinocyte growth factor

LC-MS/MS MRM Liquid chromatography-tandem mass spectrometry multiple reaction monitoring

LEA Lycopersicon esculentum

LOS Length of stay

LPS Lipopolysaccharide

MMP Matrix metalloproteinase

MMP9ko Matrix metalloproteinase-9 knockout

MPO Myeloperoxidase

MWCO Molecular weight cutoff

NADPH Nicatinamide adenine dinucleotide phosphate

NO Nitric oxide

OCT Optimal cutting temperature compound

xviii

PAPS γ’-Phosphoadenosine-5’-phosphosulfate

PBS Phosphate-buffered saline

PEEP Positive end-expiratory pressure

Q Perfusion

ROS Reactive oxygen species

SP Surfactant protein

Th1 T-helper 1

TIMP Tissue inhibitor of metalloproteinase

TLR Toll-like receptor

TNF Tumor necrosis factor

UDP Uridine diphosphate

V Ventilation

VILI Ventilator-induced lung injury

XOR Xanthine oxidoreductase

ZO Zonula occludens

xix

CHAPTER I

HEPARAN SULFATE AND CHONDROITIN SULFATE2

Heparan Sulfate and Chondroitin Sulfate Structure and Biosynthesis

Heparan sulfate (HS) and chondroitin sulfate (CS) are linear glycosaminoglycans (GAGs) composed of repeating disaccharide units of a hexuronic acid (glucuronic acid or its epimer, iduronic acid), and glucosamine (in HS) or galactosamine (in CS). HS and CS are synthesized attached to core proteins; these GAG-protein complexes are termed HS, CS, or combined HS/CS proteoglycans (2, 3).

Synthesis of HS and CS both occur within the Golgi apparatus, beginning from a common pathway. After translation in the endoplasmic reticulum, core proteins that are destined to become HS and/or CS proteoglycans traffic to the Golgi where a tetrasaccharide sequence

(xylose-galactose-galactose-glucuronic acid) is covalently O-linked to distinct serine residues.

Following addition of the tetrasaccharide linker, polymerization of HS and CS diverges (Figure

I-1). If the tetrasaccharide linker galactose residues are sulfated, CS polymerization will occur, and N-acetylgalactosamine (GalNAc) is added to the end of the tetrasaccharide link. In contrast, if the tetrasaccharide linker galactose residues are un-sulfated and there is a specific sequence of amino acids surrounding the O-linked serine, HS polymerization will occur and N- acetylglucosamine (GlcNAc) is the fifth saccharide added by exostosin like glycosyltransferase 3

(EXTL3) (3, 4).

After addition of the first GlcNAc, HS chain polymerization continues as GlcNAc and glucuronic acid (GlcA) residues are sequentially added by the complexed exostosin glycosyltransferase 1 (EXT1) and EXT2 (4). Select GlcNAc residues within the recently

2Portions of this chapter were previously published in the American journal of respiratory cell and molecular biology. Reprinted with permission of the American Thoracic Society. Copyright © 2018 American Thoracic Society. (1, 2)

1 polymerized HS chain are then deacetylated and sulfated at the amino position (GlcNS) by N- deacetylase/N-sulfotransferase. Sulfation at the N-position primes nearby GlcA residues to be epimerized to iduronic acid (IdoA) by C5-epimerase and sulfated at the 2-O position by 2-O- sulfotransferase. Further sulfation of GlcNAc/GlcNS at the 6-O position by 6-O- sulfotransferases, and at the 3-O position by 3-O-sulfotransferases also occurs at a subset of residues. Of note, 3-O sulfation of HS is very rare and usually only occurs in heparin synthesized in mast cells via this same biosynthetic pathway. Once synthesized, HS will contain clusters of heavily sulfated regions interspersed between regions that are relatively un-sulfated.

As such, subdomains of HS are often referred to according to their overall amount of sulfation:

NS domains being heavily sulfated, NA domains being relatively un-sulfated, and NS/NA domains being mixed (3).

CS polymerization and modification occurs via similar steps as HS synthesis.

After addition of the first GalNAc, chondroitin synthases 1 and 2, in concert with chondroitin polymerizing factor, elongate the CS chain sequentially adding GalNAc and GlcA residues (5).

Select GlcA residues are then epimerized to IdoA, again by C5-epimerase. Modification of CS with sulfation of GalNAc at the 4-O and 6-O positions, and IdoA sulfation at the 2-O position, can then be performed by 4-O- and 6-O-sulfotransferases, and 2-O sulfotransferase respectively.

In contrast to HS, once CS is synthesized, it is more likely to contain sulfation throughout the entire structure and is less likely to have distinct subdomains that vary in overall sulfation; however, the specific sites at which CS is sulfated (the CS sulfation pattern) varies throughout the chain (3). While the sulfation of specific HS dissacharides are referred to by their full sulfation pattern (e.g. NS2S HS being an HS disaccharide sulfated at the N-position of Gal and

2-position of IdoA), CS disaccharide sulfation patterns have distinct subtype names, CS A-E

(Table I-1) (3).

To synthesize both HS and CS, the cell utilizes activated sugar and sulfate donors to transfer monosaccharides and sulfates to the developing GAG chain. In the cytoplasm, monosaccharides are activated by a kinase or nucleotide exchange reaction to create high energy nucleoside diphosphate-sugars, most often uracil diphosphate-sugars (UDP-monosaccharides)

(6). Also in the cytoplasm, synthesis of the ubiquitous sulfate donor γ’-phosphoadenosine-

5’phosphosulfate (PAPS) occurs through a two-step reaction requiring 2 molecules of ATP and inorganic sulfate (7, 8). As both of these reactions require energy, glycosaminoglycan synthesis is energy-dependent. Once synthesized, UDP-monosaccharides and PAPS are transported to the

Golgi by energy-independent antiporters and transporters, where they are then used by glycosyltransferases and sulfotransferases to elongate and sulfate developing GAG chains (6).

While both the proteoglycan core protein structure and sulfation of the tetrasaccharide linker regulate the synthesis of HS and CS, the availability of UDP-monosaccharides and PAPS also govern the synthesis and degree of sulfation of HS and CS (7, 8).

After complete synthesis and modification, HS/CS proteoglycans may be decorated with as few as 1 HS/CS chain and up to as many as 100 CS chains, like in the CS proteoglycan aggrecan. At this time the HS/CS proteoglycans are then ready for export/trafficking to their final destination. According to the core protein structure, proteoglycans are either inserted into the plasma membrane, secreted into the extracellular matrix/basement membrane, or stored in secretory vesicles (3).

sulfate chondroitin (HS) are (CS) and sulfate Heparan

N-acetylgalactosamine and iduronic acid by 2-O, 4-O, and 6-O sulfotransferases. Figure modified from Haeger, et 2016 et modified Haeger, Figure al, sulfotransferases. from 6-O and 2-O, by (2).iduronic and 4-O, N-acetylgalactosamine acid

chondroitin sulfate polymerizing factor (CSS1-3), epimerization of glucuronic acid to iduronic acid by C5 epimerase, and sulfation and of glucuronic of acid to C5 by iduronic (CSS1-3), epimerase, factor acid polymerizing epimerization sulfate chondroitin

polymerization (addition of N-acetylgalactosamine and glucuronic acid repeating disaccharides) by chondroitin sulfate synthases and and synthases by repeating sulfate chondroitin glucuronic acid disaccharides) and N-acetylgalactosamine of (addition polymerization

N-deacetylases/N-sulfotransferases (NDST 1-4) and 2-O, 3-O, and 6-O (NDST CS chain with 1-4) and sulfotransferases. synthesis continues and 3-O, N-deacetylases/N-sulfotransferases 2-O,

EXT2, epimerization of glucuronic acid to iduronic acid by C5 epimerase, and sulfation of N-acetylglucosamine C5 by of iduronic to of acid sulfation epimerase, by epimerization iduronic and glucuronic EXT2, acid and acid

chain polymerization (addition of N-acetylglucosamine and glucuronic acid repeating disaccharides) and disaccharides) (addition acid 1 exostosin by (EXT1) polymerization glucuronic and repeating chain N-acetylglucosamine of

synthesized via a common tetrasaccharide linker covalently added to proteoglycan core proteins. HS proteins. core proteoglycan to with added synthesis continues then covalently linker a via tetrasaccharide common synthesized Figure I-1: Heparan Sulfate and Sulfate Biosynthesis. Heparan Chondroitin I-1: Figure Sulfate

Table I-1: Chondroitin Sulfate Subtypes. Disaccharide structure of chondroitin sulfate subtypes. Modified from Esko, et al, 2009 (3).

Chondroitin Sulfate Subtype Disaccharide Structure A GlcA-GalNAc4S B (or dermatan sulfate (DS)) IdoA-GalNAc4S C GlcA-GalNAc6S D GlcA2S-GalNAc6S E GlcA-GalNAc4S6S

General Functions of Heparan Sulfate and Chondroitin Sulfate

The general functions of HS and CS are largely dependent on both the overall sulfation

and the specific sulfation pattern of each GAG chain. Sulfation imparts HS and CS with a

landscape of negative charges facilitating their binding to positively charged molecules or

positively charged moieties within a net-uncharged molecule. HS and CS may function as

structural molecules in a manner independent of the sequence of sulfation. Sulfation of HS and

CS generally allows these GAGs to bind cations, creating an osmotic force to sequester water

and form a hydrated rigid gel-like material that absorbs compression (2, 3). As such, HS and CS

within the extracellular matrix (ECM) form a hydrated scaffold for cells to embed within, and

both within the ECM and on the cell surface, provide a durable structure that can withstand high

compressive and shear forces. Furthermore, the net negative charge of HS and CS allows them

to form a charged meshwork or charged molecular sieve that resists the transflux of proteins and

other molecules (2, 3).

In contrast to the sulfation sequence-independent structural function of HS and CS, HS

and CS may function in a sulfation sequence-dependent manner to regulate cellular signaling by

binding positively charged residues on bioactive ligands and/or their receptors (Figure I-2). It is

thought that precise sequences of sulfation create a three-dimensional landscape of negative

charges that allows HS and CS to bind distinct ligands and/or receptors with specificity. The

exact sulfation sequence of HS/CS that certain proteins bind is known for very few proteins;

however, enrichment of certain HS/CS disaccharide sulfation patterns necessary for binding (i.e.

NS2S HS, CS-B, etc.) is well described for several proteins (9). IdoA-rich GAG sequences

(IdoA-enriched HS and CS-B(dermatan sulfate, DS)) exhibit increased conformational flexibility

and are commonly found in HS/CS moieties that bind signaling proteins (3). Nevertheless, CS

chains that lack IdoA residues but contain other sulfated moieties (CS-A, C-E) bind to and

regulate the function of many proteins as well (10-12).

Figure I-2: Functions of Heparan Sulfate and Chondroitin Sulfate on the Cell Surface and in the Extracellular Matrix. Heparan sulfate (HS) and chondroitin sulfate (CS) on the cell surface function as scaffolding molecules/co-receptors for growth factor ligand-receptor binding and integrin-ECM interactions. HS and CS in the extracellular matrix (ECM) bind and store signaling ligands that can be released upon tissue damage and ECM degradation. Additionally, cell surface and ECM HS and CS bind chemokines and cytokines preventing their proteolytic degradation and creating concentration gradients for inflammatory cell recruitment and activation. Figure modified from Sarrazin, et al, 2011(9).

Through this mechanism, HS and CS can modulate numerous types of signaling pathways including growth factor, cell adhesion, and cytokine/chemokine signaling (Figure I-2 and Table I-2). Both membrane-bound and ECM HS/CS modulate growth factor signaling.

Cell surface HS/CS bind growth factor ligands, and their receptors, providing a cis-scaffold for ligand–receptor binding. While some signaling pathways are only enhanced by HS/CS stabilizing ligand-receptor binding, HS is required for other growth factor ligand-receptor interactions (9). Within the ECM, HS and CS bind and store signaling ligands, which can be released upon ECM degradation during injury (3, 9). Cell-cell and cell-matrix adhesion are also regulated by cell-surface HS and CS. Syndecan-bound HS and CS on the cell surface can bind

ECM proteins and, through interactions with the syndecan core protein, can stabilize integrin-

ECM binding to enhance cell-matrix adhesion. Furthermore, HS/CS on the cell surface can facilitate cell-cell adhesion as endothelial HS/CS has been shown to bind to L- and E-selectin on leukocytes enabling endothelial-leukocyte adhesion (9, 13). Finally, HS and CS, both membrane-bound and within the ECM, bind to cytokines and chemokines. By binding cytokines and chemokines, HS/CS can prevent their proteolysis and establish concentration gradients needed for the recruitment and activation of inflammatory cells (9, 14).

Table I-2: Heparan Sulfate and Chondroitin Sulfate Binding Proteins. A selective list of heparan sulfate and chondroitin sulfate binding growth factor ligands, cytokines/chemokines, and cell-cell and cell-matrix adhesion proteins (15-20).

Protein GAG Binding Growth Factors FGF2, KGF (FGF-7), FGF-10 HS, DS/CS-B (weak), CS-E (weak) HGF HS, DS/CS-B, CS-E, CS-D (weak)

VEGF HS, DS/CS-B Cytokines/Chemokines IL-8 HS > DS/CS-B > CS IFN HS, CS, DS/CS-B CXCL1 HS TNFα HS, CS-E RANTES DS/CS-B > HS > CS Adhesion Proteins Fibronectin HS > CS-C > CS-A Laminins HS > CS-C = CS-A L-selectin HS, DS/CS-B

Degradation, Turnover, and Shedding of Heparan Sulfate and Chondroitin Sulfate

The functions of proteoglycans and their associated HS and CS GAGs are regulated by

both homeostatic and pathogenic turnover. HS, CS, and proteoglycan core proteins can be

recycled, degraded, and/or shed by a variety of different mechanisms. During homeostasis,

membrane-bound proteoglycans are endocytosed, degraded in the lysosome to their original

monosaccharide and sulfate constituents, and recycled for use in the synthesis of new

proteoglycans. In addition, during inflammation, injury, and tissue remodeling, both membrane-

bound and ECM proteoglycans can be degraded and/or shed by several enzymes and reactive

molecules that can act on several targets within the overall proteoglycan (Figure I-3). Below we

will discuss the activity and role of endoglycosidases, proteases, and reactive oxygen species in

degrading and shedding proteoglycans and/or their associated HS/CS GAG chains.

Figure I-3: Degradation and Shedding of Heparan Sulfate, Chondroitin Sulfate, and Proteoglycan Core Proteins. Several enzymes and reactive molecules cleave/shed proteoglycans and/or their associated GAG chains. Heparanase specifically cleaves HS and hyaluronidase-1, -4, and testicular hyaluronidase specifically cleave CS, or CS in addition to hyaluronic acid. Reactive oxygen species (ROS) can fragment both HS and CS in addition to the proteoglycan core protein. Proteases cleave the proteoglycan core protein. While a cell-surface proteoglycan is depicted in this figure, these enzyzmes and reactive molecules can similarly degrade/shed ECM HS, CS, and HS/CS proteoglycan core proteins. Figure modified from Haeger, et al, 2018 (1). Endoglycosidases

GAG endoglycosidases are hydrolases that cleave GAGs at certain positions within the

polysaccharide, releasing oligosaccharide fragments. These enzymes differ from lysosomal

exoglycosidases involved in complete GAG degradation and recycling in that endoglycosidases

can cleave GAGs at several positions within the GAG chain, not just at the non-reducing end as

occurs with exoglycosidases (3). HS can be specifically cleaved by heparanase, an

endoglycosidase that recognizes specific HS moieties and hydrolyses GlcNAc/GlcNS-GlcA

bonds. Two mammalian heparanase genes have been identified; however, only heparanse-1

(referred to herein as heparanase) contains enzymatic activity. Having no endoglycosidase

activity, heparanase-2 is instead thought to be an inhibitor of heparanase-1, binding HS and

inhibiting its cleavage by heparanase-1 (21, 22).

Heparanase is translated as a pre-proenzyme that is cleaved by signal peptidase and further modified by Cathepsin L to generate a heterodimer of a 50kDa and 8kDa subunit that is enzymatically active. While it is not completely understood which specific HS domains are recognized by heparanase, 6-O sulfation of HS appears necessary for its binding to and cleavage by heparanase (22). Heparanase is often upregulated during tissue injury, inflammation, and cancer. Cleavage of HS by heparanase has been shown to generate low molecular weight oligosaccharides that are heavily sulfated and biologically active, containing the proper sulfation to bind growth factor ligands/receptors and potentiate signaling (22, 23). Furthermore, cleavage of HS in the ECM by heparanase releases growth factors, cytokines, and chemokines increasing cell proliferation, migration, and inflammation (9). Heparanase additionally can induce and propagate inflammation by degrading endothelial cell-surface HS, thereby facilitating endothelial-leukocyte adhesion and extravasation, and by generating damage-associated molecular pattern (DAMP)-like HS fragments that bind and activate toll like receptor 4 (TLR-4)

(24, 25).

While it was originally thought that no mammalian CS-specific extracellular endoglycosidases existed, recent studies have shown that some hyaluronidases (hyaluronic acid endoglycosidases) exhibit enzymatic activity towards CS in addition to hyaluronic acid (HA). In fact, hyaluronidase-4, a hydrolase that recognizes and degrades CS-D subunits, is thought to have CS-specific hydrolase activity as it exhibits very little HA hydrolytic activity (26).

Furthermore, two other hyaluronidases, hyaluronidase-1 and testicular hyaluronidase also exhibit

CS hydrolytic activity. These hyaluronidases recognize and degrade CS-A subunits within full

CS chains at an equal or even greater rate than HA (27). Despite these discoveries, the role of

10 these hyaluronidases in CS degradation in vivo and the effect of hyaluronidase-induced CS fragments remain poorly understood.

Proteases

Proteases cleave the proteoglycan core protein, shedding a portion of the core protein containing bound HS and CS. Several proteases are known to cleave HS and CS proteoglycans; however, two families of proteases, matrix metalloproteinases (MMPs) and a disintegrin and metalloproteinases (ADAMs)/ADAM with thrombospondin motifs (ADAMTS’) have been heavily described and are well understood.

Matrix metalloproteinases. The MMPs are a family of 23 proteins that recognize and cleave a variety of ECM, extracellular, and membrane-bound proteins. MMPs can be stratified into two subgroups based upon tissue localization: secreted and membrane bound/anchored proteinases. MMPs are translated as proenzymes, containing conserved pro- and catalytic domains, that originate in an inactive form via interactions between the pro-domain thiol group and Zn2+ in the catalytic domain (referred to as the cysteine switch). MMPs are activated by cleavage/proteolysis of the pro-domain, thereby disassembling the inhibitory thiol-Zn2+ interaction. Several enzymes and reactive molecules can remove the MMP pro-domain including furin, plasmin, reactive oxygen species, and other active MMPs; however, activation of

MMPs is complex often requiring multi-step proteolytic processes (28, 29). MMPs are additionally regulated by a family of endogenous inhibitors, the tissue inhibitors of metalloproteinases (TIMPs). All TIMPs (TIMPs 1-4) can inhibit each MMP, though tissue localization and variable inhibitory efficacy provides some specificity with which MMPs each

TIMP inhibits in vivo. Although TIMPs have other indirect affects, it is generally understood

11 that the local ratio of MMP/TIMP activity controls the overall net proteolytic activity of each

MMP (30).

MMP-2, -7, -9, -14 (MT1-MMP), and -16 (MT3-MMP) are all known to cleave at least one member of the syndecan family, a family of membrane-bound HS/CS proteoglycans. While all of these MMPs have the ability to cleave syndecan-1, MMP-2 and MMP-9 can additionally cleave syndecans-2 and -4 (31). The gelatinases, MMP-2 and MMP-9, are secreted MMPs with overlapping and redundant activity (32). MMP-2 is activated by MT1-MMP and (paradoxically)

TIMP-2; in contrast, the mechanisms of MMP-9 activation remain poorly understood (28).

Activated MMP-2 and MMP-9 are able to cleave the extracellular domains of syndecan-1, -2, and -4 near the plasma membrane, releasing syndecan ectodomains containing all bound GAG chains (31). MMP-7 is also a secreted proteinase, but binds to cell-surface proteoglycans and, therefore, is localized to the plasma membrane like membrane-bound MT1-MMP and MT3-

MMP. MMP-7, MT1-MMP, and MT3-MMP all cleave syndecan-1, also near the plasma membrane releasing GAG-bound syndecan-1 ectodomains (31). MT1-MMP and MT3-MMP are both activated intracellularly by furin and, once inserted into the plasma membrane, their constitutive activity is inhibited by TIMPs and endocytic removal from the membrane. Similarly to MMP-9, the mechanism of activation of MMP-7 remains unknown (28, 29).

A disintegrin and metalloproteinases and a disintigren and metalloproteinase with thrombospondin motifs. Together with MMPs, A disintegrin and metalloproteinases

(ADAMs)/ADAM with thrombospondin motifs (ADAMTS’) are metzincins (zinc-bearing proteinases). As such, ADAMs/ADAMTS’ are structurally and functionally similar to MMPs.

Like MMPs, ADAMs/ADAMTS’ are translated as proenzymes that contain a cysteine switch to regulate catalytic activity (33, 34). In contrast to MMPs, ADAMs and ADAMTS’ contain

12 disintegrin domains that, separate from their proteolytic function, allow binding to integrins affecting cell-cell and cell-matrix adhesion (35).

Most ADAM proteases are integral membrane proteins that are activated intracellularly by furin before insertion in the plasma membrane. Once on the cell surface ADAMs are also regulated by TIMPs and endocytic removal from the plasma membrane (33). ADAMTS proteases are all secreted proteins; however, they interact closely with the cell membrane and

ECM by binding sulfated glycosaminoglycans via their type I thrombospondin repeat domains.

Similar to ADAMs, ADAMTS’ are also activated intracellularly by furin and can be inhibited by

TIMP-3 (34). Of the 1γ proteolytically functional ADAMs and 19 ADAMTS’, only ADAM17,

ADAMTS-1 and ADAMTS-4 have been shown to cleave syndecans. ADAM17 is known to cleave both syndecan-1 and syndecan-4, and like MMPs, releases syndecan ectodomains containing all bound GAG chains. In contrast, ADAMTS-1 and ADAMTS-4 only cleave syndecan-4 and do so near the first most distal GAG, releasing only a small portion of the proteoglycan ectodomain containing one GAG chain (31). While only these three

ADAMs/ADAMTS’ are known to cleave syndecans, several other ADAMTS’ are known to cleave other extracellular CS proteoglycans including aggrecan, versican, neurocan, and brevican

(36).

Reactive Oxygen Species

Reactive oxygen species (ROS) are a family of oxygen radicals and nonradical oxidizing agents that can fragment both HS/CS chains and the proteoglycan core protein releasing partial protein and GAG fragments (37, 38). Reactive oxygen species are generated in all cells and are needed for several cellular processes; however, a burst of additional ROS is often generated during tissue infection, inflammation, and injury. Neutrophils and other leukocytes generate

ROS by nicotinamide adenine dinucleotide phosphate (NADPH) oxidase, xanthine oxidoreductase (XOR), and myeloperoxidase (MPO) as an antimicrobial response, but in addition to killing bacteria, the generated ROS can also damage several host tissue structures

- (39). Superoxide radicals (O2 ) produced by NADPH oxidase and XOR, in addition to other

ROS, have been shown to fragment HS, CS, and proteoglycan core proteins in vitro (37, 38).

The degree of GAG depolymerization by ROS appears to be inversely proportional to the amount of GAG sulfation, suggesting that heavily sulfated GAGs are more resistant to fragmentation by ROS (40). In certain models of tissue injury in vivo, superoxide dismutase

- (SOD), an enzyme that converts O2 to the less reactive hydrogen peroxide, protects against fragmentation of both syndecan-1 and HS, indicating that ROS can fragment GAGs and proteoglycans in vivo as well (41, 42).

Expression, Location, and Importance of Heparan Sulfate and Chondroitin Sulfate in the

Lung

Expression and Location of Heparan Sulfate and Chondroitin Sulfate

Consistent with their critical general functions, HS and CS are widely expressed within the adult lung. CS is the most abundant GAG in the lung, with 4-O and 6-O sulfated CS and DS totaling approximately 50% of all lung GAGs. Of all lung CS, 4-O sulfated, CS-A unit containing CS is the most readily expressed (50% of all lung CS). HS is also heavily expressed within the adult lung, representing approximately 40% of all lung GAGs (43). Localization of

CS and HS within rough anatomical compartments of the lung has also been investigated. The relative contributions of GAGs in the different anatomical compartments vary greatly. Bronchi and large vessels (both arteries and veins) are enriched in CS, 6-O sulfated CS in bronchi and both 6-O sulfated and DS-unit containing CS in large arteries and veins. In contrast, alveoli and

14 pleura are enriched in DS-unit containing CS and HS (44). The contributions of HS and CS to precise microscopic structures within the lung are known for some cellular and matrix structures including the alveolar endothelial cell surface and alveolar basement membrane, but remain unknown for many other structures (24, 45).

Given the widespread expression of HS and CS within the lung, both membrane-bound and secreted proteoglycans are present within the adult normal lung. Membrane-bound HS/CS proteoglycans can be subdivided into two categories based upon their association within the plasma membrane. One subset of HS/CS proteoglycans are integral membrane proteins and are thus incorporated within the plasma membrane, while the others are glycosylphosphatidylinositol

(GPI)-anchored proteins tethered to the plasma membrane. Both syndecans and glycpicans have been identified within the lung. Syndecans (types 1-4) are integral transmembrane HS/CS proteoglycans, of which syndecan-1 and syndecan-3 are decorated with both HS and CS, while syndecan-2 and syndecan-4 are decorated with HS only. Expression of specific syndecan isoforms is largely cell and tissue type dependent, except for syndecan-4 which is expressed in several cell and tissue types. Syndecan-1 is expressed in epithelium, syndecan-2 in endothelium and fibroblasts, and syndecan-3 in neuronal and musculoskeletal tissues (31). Consistent with these findings, the adult lung expresses syndecan-1, -2, and -4 (46). Glypicans are GPI-anchored

HS proteoglycans that are highly expressed during development and thought to have decreased expression in adult tissues (47). As such, glypican expression in the normal adult lung remains relatively understudied; however, glypican-3 and glypican-5 are thought to be overexpressed in certain types of lung cancer and play a role in tumor development and progression (14, 48, 49).

HS/CS proteoglycans, additionally, are large constituents of the ECM and basement membrane within the lung. The CS proteoglycans decorin and biglycan, and the HS

15 proteoglycans perlecan, collagen XVIII, and agrin are all found within the normal adult lung

(14). Decorin and biglycan are small leucine-rich proteoglycans and are found within the ECM and contain 1-2 CS chains per proteoglycan. Perlecan, collagen XVIII and agrin are basement membrane proteoglycans that each contain 1-3 HS chains (3).

Importance of Heparan Sulfate and Chondroitin Sulfate

Transgenic animal studies have demonstrated the importance of HS and CS to proper lung development. As summarized in Table I-3, the most severe pulmonary-relevant phenotypes described (embryonic lethal) arise from null mutants of Glact1, Ext1, and Ext2, which encode enzymes that build the initial HS/CS tetrasaccharide linker and polymerize the HS chain. In addition, genes involved in the further modification of HS also affect lung development. Null allele mutants of Ndst1, which encodes an enzyme that sulfates glucosamine at the N position, suffer from neonatal respiratory distress due to lung hypoplasia. Genetic manipulations that result in HS containing less iduronic acid (Glce/Hsepi) or 6-O sulfation

(H6st1) also produce animals that lack proper alveolar development (2).

Surprisingly, mice deficient in chondroitin sulfate synthase 1 or chondroitin sulfate polymerizing factor (enzymes involved in CS chain polymerization) do not exhibit any pulmonary defects; however, CS synthesis is not completely abrogated in these animals as they only exhibit differences in CS chain length or sulfation (5, 50). The only genetic mutant of CS biosynthetic machinery that affects lung development is a gene trap mutant of chondroitin-4- sulfotransferase 1 (Chst11). These mice develop neonatal respiratory distress and die within 6 hours after birth. Although, the cause of respiratory distress and any alterations in lung histology within Chst11 mutant mice remain unknown, these findings indicate that 4-O sulfation of CS may be important for lung development (51).

Table I-3: Lung Developmental Defects in Heparan Sulfate and Chondroitin Sulfate Biosynthesis Genetic Mutants. Modified from Haeger, et al, 2016 (2).

Gene Role in HS/CS Biosynthesis Effect on Lung Development Glact1 Forms HS/CS-protein tetrasaccharide link Null allele: embryonic lethal Ext1/Ext2 Elongates HS chain Null allele: embryonic lethal; aberrant endoderm development Ndst1 Deacetylates and sulfates HS Null allele: perinatal lethal; pulmonary hypoplasia glucosamine residues at the N-position and neonatal respiratory distress H6st1 Sulfates HS glucosamine residues at the Null allele: embryonic/perinatal lethal; enlarged 6-position alvleoli Glce (Hsepi) Epimerizes HS glucuronic acid to Targeted disruption: perinatal lethal; neonatal iduronic acid respiratory distress and thickened poorly inflated alveoli Chst11 Sulfates CS galactosamine residues at the Gene trap: neonatal respiratory distress 4-position

In this first chapter, we have discussed the biosynthesis, general function, shedding/degradation, and lung expression of HS, CS, and their associated proteoglycans. In

Chapters II and III we will discuss the structure and function of the alveolus, and the development and pathogenesis of ARDS. While we briefly highlighted the importance of HS and CS in lung development in this first chapter, we will further discuss the known contributions of HS and CS to lung homeostasis and injury, as they are applicable, in Chapters II and III.

CHAPTER II

ALVEOLAR STRUCTURE AND FUCNTION3

Overview of Lung Structure and Function

While the mammalian lung participates in multiple homeostatic processes, its primary function is to provide an interface for gas exchange between environmental air and the circulation. This bidirectional gas exchange is essential to host survival, allowing not only for blood oxygenation (and subsequent oxygen delivery to tissues), but also the clearance of the byproducts of cellular respiration (such as carbon dioxide). To facilitate this critical homeostatic function, the lung has evolved as a complex structure of alveolar airspaces and accompanying capillaries that maximizes the surface area for gas exchange.

Both the lung airspace and vasculature are fractal structures that arise as a single vessel that undergoes several generations of bifurcations to reach the final gas exchange unit, the alveolus. The lung airspace originates as a single airway, the trachea, that bifurcates into the right and left main bronchus, that then continue to bifurcate into smaller airways, giving rise to bronchioles, respiratory bronchioles, alveolar ducts, and, ultimately, alveoli. It is estimated that the airways undergo 23 generations of bifurcations between the trachea and the alveolus (52).

This airspace is lined by a transitioning epithelial layer, beginning as a pseudostratified columnar epithelium in the trachea and thinning into a largely simple squamous epithelium lining the alveolus (53). Similarly, the lung vasculature originates as a single pulmonary artery and undergoes an estimated 28 generations of bifurcations before reaching the pre-capillaries in the alveolus (52). This bifurcating fractal structure, in addition to the cytological composition of the airspace and vasculature, is optimized for efficient gas exchange maximizing the surface area

3Portions of this chapter were previously published in the American journal of respiratory cell and molecular biology. Reprinted with permission of the American Thoracic Society. Copyright © 2018 American Thoracic Society. (1, 2)

18 and minimizing the thickness that carbon dioxide and oxygen must diffuse across.

The alveoli and additional airspaces capable of gas exchange (e.g. the alveolar ducts and respiratory bronchioles) have an estimated volume of 4.2 liters, accounting for 96% of the entire volume of the lung. This environmental air is enclosed within a 130 m2 epithelial surface area,

88% (115 m2) of which directly interfaces with capillaries and serves as the interface for oxygen and carbon dioxide diffusion (54). To ensure optimal gas exchange, the lung must both maintain matched perfusion and ventilation along the surface area of the alveolar space while ensuring a thin and tight barrier (“alveolar septum”) preventing capillary fluid extravasation and consequent alveolar flooding. Under conditions when perfusion (“Q”) and/or ventilation (“V”) in areas of the lung are sub-optimally matched (V/Q mismatch), the lung uses compensatory mechanisms, including hypoxic vasoconstriction and hypocapnic bronchoconstriction, to correct the uncoupling and maintain matched ventilation and perfusion (55).

To ensure adequate alveolar barrier function, the alveolar septum is comprised of alveolar endothelial, alveolar type I epithelial, and alveolar type II epithelial cells (and in several areas, a shared basement membrane), forming a nearly 2µm barrier that restricts fluid extravasation from the vasculature but allows gas exchange (56). Each of these cell types, as described below, exhibit specialized functions that contribute to this barrier, as well as other facets of lung homeostasis.

The Alveolar Endothelium

The capillaries within the alveolus are comprised of a thin layer of continuous endothelium. Each alveolus is estimated to contain 170 endothelial cells that extend cytoplasmic projections to each cover approximately 1350µm2 (56). The microvasculature within the alveolar capillaries exhibit distinctly different properties than those of the macrovasculature in

19 the larger segments of the pulmonary artery. These distinct properties reflect the specialized functions of the alveolar endothelium, including facilitation of gas exchange and maintenance of a tight barrier opposing fluid and protein extravasation.

Alveolar endothelial cells are on average 0.46µm thick, and nearly directly interface with the alveolar epithelium in over half of the alveolus, where the endothelium and epithelium share a fused basement membrane (56, 57). This extremely thin layer allows for efficient exchange of oxygen and carbon dioxide. While permitting transalveolar gas exchange, the endothelium must simultaneously maintain a tight barrier opposing fluid and protein flux, thereby preventing alveolar edema and flooding. Accordingly, the pulmonary alveolar capillary endothelium forms a tighter barrier (exhibiting approximately 100-times increased resistance to fluid and protein) than that of the macrovascular endothelium. Microvascular endothelial cells also express a distinct transcriptomic profile in comparison to macrovascular endothelial cells that includes differences in junctional components like CD166, N-cadherin, occludin, and zonula occludens-2

(ZO-2) (58).

In addition to cell-cell junctions, the pulmonary microvascular endothelium express apical cell-surface HS and CS that form a luminal lining, termed the endothelial glycocalyx or endothelial surface layer (ESL) (2). By forming a rigid gel-like layer, the pulmonary endothelial glycocalyx contributes to the endothelial barrier to fluid and protein, transduces vascular shear stress into endothelial nitric oxide (NO) synthesis, and regulates the availability of cell membrane adhesion molecules to circulating leukocytes (59). The contribution of HS to these functions of the endothelial glycocalyx have been heavily studied; however, the role of CS in endothelial glycocalyx function is poorly understood. Enzymatic HS degradation causes collapse of pulmonary ESL thickness (24), leading to lung edema (60, 61), aberrant pressure-

20 induced endothelial NO synthesis (60, 61), and lung inflammation (24). While degradation of

CS in the mesentery vasculature induces collapse of ESL thickness (62), degradation of endothelial CS does not affect shear-induced NO synthesis (63).

The contribution of HS to endothelial barrier function has been attributed to its physical presence as a charged meshwork overlying endothelial cells (59). In contrast, the mechanism by which an intact ESL regulates NO synthesis is less certain, with investigators speculating the importance of interactions between HS proteoglycans (glypican) and endothelial NO synthase within endothelial caveoli (63). Furthermore, the impact of HS on leukocyte adhesion is complex. Loss of ESL thickness exposes endothelial surface adhesion molecules, facilitating neutrophil adhesion within alveolar capillaries (24). In contrast, endothelial surface HS can serve as an L-selectin ligand and regulator for chemotactic agent availability (64); as such, aberrance of pulmonary ESL HS structure or sulfation might be expected to decrease L-selectin– mediated neutrophil–endothelial interaction. It remains unclear how these seemingly disparate roles of ESL HS on leukocyte diapedesis are reconciled in vivo.

The Alveolar Type I Epithelium

Similar to the alveolar endothelium, the epithelium covering the majority of the alveolar surface area must be sufficiently thin to allow efficient gas exchange yet adequately robust to oppose fluid and protein flux from the vasculature/interstitium into the airspace. Alveolar type I epithelial (ATI) cells comprise 95% of this epithelial surface area and accordingly must satisfy these functions (56). The alveolar epithelium is approximately 0.36µm thick and, like the endothelium, extends cytoplasmic projections to each cover a large surface area of over 5000µm2

(56, 57). In addition, the alveolar epithelium forms a tight barrier to fluid and protein, likely even more robust than the alveolar endothelium. Indeed, others have shown that aberrations

21 within the alveolar epithelial barrier alone are sufficient to increase alveolar permeability to albumin (65). Given this important function, the alveolar epithelium utilizes several mechanisms to maintain this tight barrier (Figure II-1).

Figure II-1: Alveolar Structure and Epithelial Function. Alveolar epithelial type I (ATI) cells (an example shown in yellow) form a thin lining that covers the majority of the alveolar airspace surface area (green). ATI cells extend cytoplasmic projections that nearly directly interface the alveolar endothelium to facilitate gas exchange (red). ATI cells utilize tight junctions, ion channels/pumps and water channels, to maintain a tight alveolar epithelial barrier. Cuboidal alveolar type II (ATII) cells (blue), also present within the alveolus, synthesize and secrete surfactant, and exhibit alveolar epithelial regenerative capacity, in addition to utilizing tight junctions and alveolar fluid clearance mechanisms like ATI cells. The existence and function of HS and CS within an alveolar epithelial glycocalyx remains unknown. Figure adapted from Weibel, 2017 (56).

Alveolar epithelial cells form a series of two different apical junction complexes that restrict paracellular transport of fluid and proteins from the vasculature/interstitium into the airspace. Adherens junctions form more basolaterally and are comprised of membrane spanning

E-cadherins that bind to additional E-cadherins on adjacent epithelial cells facilitating cell-cell adhesion. Intracellularly, E-cadherin is anchored to the cytoskeleton via to α-, -, and p120 catenins, maintaining adherens junction localization at the plasma membrane (66). While adherens junctions have been shown to play an important role in the airway epithelium during asthma and allergic inflammation, little is known about adherens junctions in the alveolar

22 epithelium (67). Loss of p120 catenin does exacerbate ventilator-induced lung injury (VILI), a lung injury model characterized by increased alveolar permeability; however, the contribution of adherens junctions to the intact alveolar epithelial barrier has been relatively unstudied (68).

In contrast to adherens junctions, alveolar epithelial tight junctions have been extensively studied. Tight junctions arrange into strands near the apical cell surface and are similarly comprised of membrane spanning proteins that bind intracellular mediators for anchorage to the actin cytoskeleton (69). Three types of membrane spanning proteins are present within tight junctions: claudins, occludin, and junctional adhesion molecules (JAMs). These membrane spanning proteins bind members of the zonula occludens (ZO) family which link them to proteins that directly associate with the actin cytoskeleton (66). In comparison to claudins, less is known about the role of occludin and JAMs in alveolar barrier function; however, a reduction of occludin has been observed in models of epithelial barrier disruption and JAM-A is thought to effect alveolar leukocyte migration (69).

The expression and function of claudins within the respiratory epithelium have been extensively studied. Claudins can be organized into two categories according to their function; pore forming claudins that facilitate cation movement through tight junctions and sealing claudins that restrict paracellular protein flux (70, 71). ATI cells express primarily the sealing claudins-4 and -18 (72). Deficiencies in either claudin-4 or -18 result in only mild defects in alveolar barrier function due to varying mechanisms involving alveolar fluid clearance. Claudin-

4 deficiency decreases epithelial Na+/K+-ATPase activity and claudin-4 deficient mice exhibit exaggerated barrier dysfunction during lung injury. In contrast, claudin-18 deficiency results in a compensatory increase in epithelial sodium channel (ENaC) and Na+/K+-ATPase activity and claudin-18 deficient mice are therefore protected against lung injury (71).

While tight junctions may contribute most significantly to the barrier that ATI cells form,

ATI cells also express channels and pumps that function to clear fluid that may accumulate in the alveolar airspace. ATI epithelial cells express aquaporin 5 that allows ATI cells to reabsorb alveolar airspace water along, but not against, an osmotic gradient (73). It was originally thought that alveolar fluid clearance was performed primarily by alveolar epithelial type II (ATII) cells; however, ATI cells also express the epithelial sodium channel (ENaC) and Na+/K+-ATPase, and knockout of the 1 subunit of the Na+/K+-ATPase in ATI cells decreases alveolar fluid clearance in mice, although not to the same extent as knockout of the same protein in ATII cells (74).

In contrast to the endothelial glycocalyx, the existence, function, and contribution of HS and CS to the alveolar epithelial glycocalyx remains understudied. Using ruthenium red staining and electron microscopy, a negatively charged glycocalyx lining the luminal surface the of type I alveolar epithelium can be visualized (75). Furthermore, expression of syndecan-1 and syndecan-4 have been detected in alveolar epithelial cells (76). Despite these findings, the presence and function of HS and/or CS within the ATI epithelial glycocalyx is unknown.

The Alveolar Type II Epithelium

In comparison to ATI cells, ATII cells are cuboidal and contain microvilli, and thus are much thicker than ATI cells. While the ATII epithelium only covers 5% of the alveolar airspace surface area, there are almost twice as many ATII cells than type I cells present in each alveolus

(56). As such, the alveolar type II epithelium is essential for alveolar function, performing essential homeostatic roles such as surfactant production, forming a tight barrier with neighboring ATI cells, clearing edema fluid from the alveolus, and serving as alveolar epithelial progenitor cells (Figure II-1) (77, 78). Each of these important functions contribute to the overall role of the alveolus to facilitate gas exchange.

Pulmonary Surfactant Production

First discovered in the 1950’s, surfactant is a lipid-protein mixture that coats the surface of the alveolar epithelium. Comprised of approximately 90% lipid (largely phosphatidylcholine and phosphatidylglycerol) and 10% protein (surfactant proteins A-D), surfactant reduces alveolar surface tension, allowing for patency of alveoli of heterogeneous size and thereby preventing atelectasis. Additionally, surfactant functions to aid in immunity against pathogens (77, 78).

After being synthesized, surfactant is stored in lamellar bodies and released by exocytosis upon mechanical stretch or alternative stimuli (i.e. -adrenergic receptor activation of adenylate cyclase, protein kinase-C activation, or gap junction-mediated Ca2+ uptake). After secretion, surfactant is remodeled into a crystal-like structure of phospholipids surrounding surfactant protein-A, termed tubular myelin. It is then re-organized into a film that is positioned on top of a thin aqueous layer (the alveolar aqueous hypophase) above the alveolar epithelial surface (79).

During compression of the alveolus, the surfactant film can further re-organize into regions that are enriched in a specific phosphatidylcholine, dipalmitoylphosphatidylcholine, which is thought to be the critical phospholipid to reduce alveolar surface tension (80). This surfactant film not only reduces surface tension of the alveolar surface, thereby preventing atelectasis and damage to the alveolar epithelium, but is also thought to restrict formation of alveolar edema by producing a force perpendicular towards the alveolar epithelial surface (77).

Surfactant proteins help organize the surfactant lipid mixture into both its tubular myelin and film structures, however; surfactant proteins A (SP-A) and D (SP-D) exhibit additional immunoregulatory functions as well (81). In comparison to the hydrophobic surfactant proteins

B (SP-B) and C (SP-C), SP-A and SP-D are hydrophilic self-oligomerizing c-type lectins. SP-A and SP-D are best known for being able to opsonize pathogens to both directly kill them and

25 increase their phagocytic clearance. In addition, SP-A and SP-D are also thought to decrease T- cell proliferation and affect dendritic cell maturation, antigen uptake, and antigen presentation

(81). As such, pulmonary surfactant both prevents alveolar atelectasis and damage at baseline, but also protects against and potentially dampens the immune response to infection.

Tight Barrier Formation

Similar to ATI cells, ATII cells also express tight junction proteins that help maintain a tight barrier at type I-type II cell junctions. In addition to claudin-4 and -18, which are likewise expressed in ATI cells, type II cells also express claudin-3 (72). Much of the information known regarding the function of claudin-3 to alveolar permeability comes from simplified in vitro studies in which claudin-3 is overexpressed. In these studies, overexpression of claudin-3 both enhances and reduces epithelial barrier function depending on the cell type studied. While claudin-3 overexpression increases trans-epithelial resistance in non-alveolar epithelial cells, consistent with its characterization as a sealing-type claudin, claudin-3 overexpression in alveolar epithelial (type I) cells reduces trans-epithelial resistance (72). Given these divergent findings, in which it is difficult to interpret overexpression data in alveolar epithelial cells that do not express claudin-3 at baseline, additional investigations are warranted to determine the effect of knockdown/knockout of claudin-3 in ATII cells both in vitro and in vivo.

Alveolar Fluid Clearance

ATII cells are thought to be responsible for the majority of alveolar fluid/edema clearance; however, as mentioned above, type I cells have also been implicated (73). By transporting ions across the alveolar epithelium thereby driving fluid clearance, several ion channels and pumps are thought to aid in clearing fluid from the alveolar airspace. ENaC on the apical type II cell surface and the basolateral Na+/K+-ATPase have been extensively studied and

26 are known to facilitate alveolar fluid clearance. ENaC and Na+/K+-ATPase both transport sodium across the epithelium, ENaC driving epithelial Na+ uptake from the alveolar airspace, and Na+/K+-ATPase exporting Na+, in exchange for K+, across the basolateral membrane into the interstitium (82). In addition, the cystic fibrosis transmembrane conductance regulator (CFTR) and other channels that have yet to be characterized are also thought to play a role in alveolar fluid clearance (73).

Several signaling mediators modulate the rate of alveolar fluid clearance by regulating the expression and/or activity of both ENaC and Na+/K+-ATPase. -adrenergic stimulation, steroids, and keratinocyte growth factor (KGF, also known as fibroblast growth factor 7), amongst other signaling mediators, all increase the rate of alveolar fluid clearance, although via differing mechanisms (82). By increasing the number of ENaC channels and Na+/K+-ATPase pumps in the membrane, in addition to increasing Cl- transport, -adrenergic-mediated cAMP signaling increases alveolar fluid clearance. Glucocorticoids increase the transcription of various subunits of both ENaC and Na+/K+-ATPase, and additionally increase their activity post- transcriptionally. The mineralocorticoid aldosterone similarly transcriptionally and post- transcriptionally increases Na+/K+-ATPase expression and activity; however, less is known about the effect of aldosterone on ENaC expression and activity. As a mitogen, KGF increases alveolar fluid clearance by inducing ATII cell proliferation but is thought to also increase

Na+/K+-ATPase expression independently of its mitogenic effect. Utilizing these mechanisms, alveolar fluid clearance rate can be modulated both endogenously and exogenously to maintain a thin alveolar aqueous hypophase and prevent pulmonary edema formation.

Alveolar Epithelial Progenitor Capacity

At homeostasis and during lung injury the alveolar epithelium repairs and regenerates to maintain a continuous alveolar epithelial barrier. ATI cells are thought to be terminally differentiated and to have no regenerative capacity; however recent studies show that these cells are indeed plastic and may have some ability to regenerate both ATI and ATII cells (83).

Despite these recent findings, a great deal of evidence shows that ATII cells and cells within the bronchoalveolar duct junction are the main progenitor cells of the alveolar epithelium (84, 85).

During development and in response to injury, ATII cells proliferate, spread, and differentiate into type I cells to repair defects in the alveolar epithelium. Several mediators are known to effect type II cell regenerative capacity; however, complete regulation of type II cell-induced alveolar repair is complex and not fully understood.

Three growth factors, in particular, have been implicated in regulating ATII cell- mediated regeneration. KGF, fibroblast growth factor 10 (FGF-10), and hepatocyte growth factor (HGF) have all been shown to increase ATII cell proliferation, migration/spreading, and/or

ATI differentiation (86, 87). Furthermore, exogenous KGF, FGF-10, and HGF are all protective in several models of acute lung injury (86, 88). As these three factors are all made in and secreted from mesenchymal cells, mesenchymal stem cell treatment has gained increasing attention as a potential therapy in acute lung injury and the acute respiratory distress syndrome

(ARDS) to increase the regenerative capacity of ATII cells (89).

Alveolar Epithelial Type II Heparan Sulfate and Chondroitin Sulfate

Similar to the alveolar type I epithelium, the existence and GAG composition of an alveolar type II epithelial glycocalyx has been understudied. Others have demonstrated expression of HS/CS proteoglycans in ATII cells (90, 91), and have studied the effect of HS

28 modifying enzymes on lung epithelial cell viability and function (92). In addition, soluble HS has been shown to influence alveolar epithelial cell signaling and phenotype, as soluble heparin

(a heavily sulfated from of HS), in combination with FGF-1, enhances ATII cell SP-B and aquaporin-5 mRNA expression (93). However, despite these findings, whether cell-surface HS and/or CS contribute to a type II alveolar epithelial glycocalyx and surface layer remains unknown.

CHAPTER III

ACUTE LUNG INJURY AND THE ACUTE RESPIRATORY DISTRESS SYNDROME4

Acute Respiratory Distress Syndrome Definition and Epidemiology

The acute respiratory distress syndrome (ARDS) was first identified in 1967 by

Ashbaugh and colleagues, who described and characterized respiratory failure in 12 adult

patients that closely resembled respiratory distress syndrome in newborns. Respiratory failure in

these patients followed a recent insult/illness, was refractory to oxygen, and was accompanied by

bilateral infiltrates seen on chest radiography (95). These characteristics described in 1967 are

still recognized as hallmarks of ARDS and are included in the current Berlin Definition used for

clinical diagnosis (Table III-1). The Berlin Definition establishes ARDS as acute hypoxic

respiratory failure within one week of an initial insult that is accompanied by bilateral opacities

on chest radiograph that cannot be fully attributed to heart failure or volume overload. The

definition further stratifies patients into mild, moderate, and severe categories according to their

severity of hypoxia (94). Table III-1: The ARDS Berlin Definition. Table obtained from ARDS Definition Task Force, 2012 (94).

4Portions of this chapter were previously published in the American journal of respiratory cell and molecular biology. Reprinted with permission of the American Thoracic Society. Copyright © 2018 American Thoracic Society. (1, 2)

ARDS affects nearly 200,000 people in the United States each year and has a mortality rate of 20-40% (96). Despite a clear and widely recognized definition, ARDS is very heterogeneous. As such, the incidence, mortality, and pathogenesis of ARDS vary greatly.

Clinicians and researchers are increasingly interested in developing methods to identify those who are at high risk for ARDS and classify ARDS patients into sub-phenotypes based upon clinical biomarkers and mode of lung injury (97-100). By understanding what genetic factors increase ARDS risk and what biomarkers distinguish sub-phenotypes of ARDS patients, the pathogenesis of ARDS in different sub-phenotypes of patients may be more easily recognized and differentiated, which may lead to more effective therapies in these different subsets of patients.

Pathophysiology of ARDS

Although the pathogenesis of ARDS may differ in separate subsets of patients, there are several pathophysiologic findings that are common amongst most patients. Acutely, during the first six days after onset, alveolar inflammation (neutrophils and macrophages), edema, and hyaline membranes develop (101, 102). These findings can be observed on lung histology as a diffuse alveolar damage pattern; however, biopsy is not commonly obtained due to unnecessary risk. Instead, these findings are usually clinically observed on chest radiograph or computed tomography (CT) as alveolar opacities/infiltrates and atelectasis (94). Physiologically, the formation of alveolar edema, inflammation, and atelectasis during ARDS decreases ventilation to affected areas of the lung. To compensate for the loss of ventilation, the lung reduces perfusion to these under-ventilated regions via hypoxic vasoconstriction; however, compensatory hypoxic vasoconstriction is often impaired in ARDS resulting in intrapulmonary shunting and hypoxemia

(103, 104).

Following acute inflammation and injury, a subset of patients may develop fibroproliferative ARDS. As alveolar edema and inflammation resolve, and the lung attempts to repair and regenerate, type II cells proliferate and fibroblasts infiltrate the alveolus to remodel the alveolus, depositing collagen leading to some fibrosis (101). If severe enough, these patients can develop residual fibrosis and have diminished lung function long after recovery from their initial insult (105).

Pathogenesis of ARDS

As the pathogenesis of ARDS is variable amongst patients, recent studies have aimed to identify sub-phenotypes of ARDS in which clear and distinct mechanisms that underlie the development of ARDS can be distinguished. Using plasma biomarkers and clinical data, Calfee and colleagues have shown that two distinct subtypes of ARDS patients exist. These two sub- phenotypes differ greatly in the level of plasma inflammatory cytokines, indicating that one sub- phenotype represents a “hyper-inflammatory” ARDS while the other appears to be “hypo- inflammatory” (97). Although inflammation, in particular neutrophilic inflammation, is thought to be critical for the development of lung injury during ARDS, neutropenic patients still develop

ARDS providing further validity to this sub-phenotype analysis (106).

While the pathogenesis of a “hypo-inflammatory” ARDS remains unknown, clinical data and animal models show that neutrophils are critical for the development and correlate with the severity of ARDS in patients and acute lung injury in animals. In several diverse animal models of acute lung injury, neutrophil depletion is protective, decreasing lung edema, permeability, and inflammation (106, 107). Furthermore, in ARDS patients with sepsis, a high concentration of bronchoalveolar lavage (BAL) neutrophils predicts mortality (108).

During ARDS, neutrophils can injure the alveolar endothelium and/or epithelium by producing reactive oxygen species and proteolytic enzymes that, while intended to damage and kill pathogens, also damage “innocent bystander” host tissue. As ARDS and acute lung injury are heterogeneous entities representing the end result of any number of intrapulmonary or systemic insults, the relative degree of injury to the alveolar endothelium versus the epithelium by neutrophils, and other sources, varies greatly (109). As such, investigations have targeted to better understand the pathogenesis of and identify biomarkers in ARDS/acute lung injury due to intrapulmonary/airspace insults (direct lung injury) versus systemic insults (indirect lung injury).

These findings would not only allow physicians to determine the injury driving insult in ARDS patients in real time, but may also lead to the development of targeted therapies that are more effective in patients with ARDS driven by endothelial injury versus patients with ARDS driven by epithelial injury.

Indirect Lung Injury

ARDS that occurs due to a systemic extra-pulmonary insult, including sepsis, shock, trauma, and pancreatitis, is termed “indirect lung injury”. After an initial systemic insult, the alveolar endothelium becomes activated and inflamed. Activation of the endothelium via numerous mediators and pathways, including TLR signaling, TNFα, and IL-1, from the initial insult enhances inflammation, promotes thrombosis, and facilitates lung endothelial-neutrophil adhesion (110, 111). TNFα and IL-1 induce endothelial production of IL-6, IL-1, and IL-8, and increase endothelial cell-surface ICAM-1 expression (110, 111). Activation of neutrophils by these factors induces actin polymerization, increased cell rigidity and upregulation of adhesive factors CD18, α4, and α5 integrins. As a result, neutrophils experience prolonged transit time through the narrow pulmonary capillaries and adhere to the activated endothelium expressing

33 intracellular adhesion molecule-1 (ICAM-1) (112). Adhesion of neutrophils increases the endothelial production and release of angiopoietin-2, which induces vascular de-stabilization and can cause endothelial cell apoptosis (113, 114). Moreover, neutrophils themselves secrete ROS, neutrophil elastase, and MMPs that then further induce endothelial damage (115). Through these mechanisms, many of which are feed-forward, injury to the alveolar endothelium induces gap formation and apoptosis resulting in enhanced alveolar endothelial permeability and interstitial and alveolar edema (114).

In addition to these pathogenic events and mechanisms, the fate and role of the endothelial glycocalyx during indirect lung injury has also been extensively studied. In an animal model of indirect lung injury, induction of endothelial heparanase expression by TNFα leads to endothelial glycocalyx HS degradation and collapse of the ESL allowing neutrophils to access their cognate endothelial adhesion molecules, adhere, and cause inflammatory lung injury

(24). In ARDS patients with indirect lung injury, but not direct lung injury, there is an increase in circulating HS, indicating endothelial glycocalyx degradation likely occurs in patients with

ARDS due to indirect lung injury (116, 117). No increase in circulating CS was detected in

ARDS patients with indirect lung injury, indicating that CS degradation from the endothelial glycocalyx may not occur in these patients (117). However, further investigations are needed to determine the fate and role endothelial glycocalyx CS during indirect lung injury.

While these findings suggest that plasma HS may serve as a biomarker of indirect lung injury and a measure of endothelial injury severity, plasma HS concentration correlated with intensive care unit (ICU) length of stay (LOS), suggesting that HS fragments may additionally contribute to indirect lung injury (117). HS fragments released by enzymatic heparanase degradation may themselves propagate injury as a damage-associated molecular pattern (25).

However, paradoxically, soluble heparin (a highly-sulfated HS) has been shown to be anti- inflammatory. Soluble heparin decreases inflammation by inhibiting LPS-induced NF-κB signaling and IL-8 secretion from pulmonary endothelium, and by binding and sequestering other circulating damage-associated molecular patterns, such as extracellular nuclear proteins, high-mobility group box 1 protein, and histones (118-121). While the degradation of endothelial glycocalyx-HS clearly contributes to indirect lung injury by facilitating endothelial-leukocyte adhesion, the functions of the released circulating HS may be more complex and remains poorly understood.

Direct Lung Injury

In contrast to indirect lung injury, ARDS/acute lung injury that occurs due to intrapulmonary insults including pneumonia, inhalational injury, pulmonary contusion, or aspiration, is termed “direct lung injury”. In response to physical damage, TLR signaling, or cytokine release from an initial insult, the alveolar epithelium and alveolar macrophages become activated and secrete numerous inflammatory cytokines including TNFα, IL-1, and IL-8 (122,

123). These inflammatory cytokines recruit circulating neutrophils to the lung where, after endothelial adhesion and extravasation, they adhere to the basolateral epithelial cell surface and transmigrate across the epithelium into the airspace. While the mechanisms by which neutrophils adhere and migrate across the alveolar epithelium are not completely understood, neutrophil 2 integrin binding to JAM-C appears to mediate neutrophil-epithelial adhesion, and following adhesion, both neutrophil and epithelial CD47 facilitate neutrophil transmigration into the airspace (124). Once in the airspace, neutrophils secrete proteases and ROS, which in addition to their antimicrobial properties, cause further damage and injury to the alveolar epithelium. This damage results in abrogation of tight junction complex formation, impaired

35 surfactant production and function, and decreased fluid clearance, all which enhance alveolar edema formation (125).

During direct lung injury there is an increase in alveolar epithelial permeability that is associated with breakdown of alveolar epithelial tight junctions. Both in vitro and in vivo studies show altered expression and localization of key tight junction proteins during epithelial inflammation. In vivo, claudin-4 is up-regulated, while claudin-18 is downregulated. As both claudin-4 and -18 enhance alveolar barrier function, a decrease in claudin-18 is thought to partially contribute to the increase in alveolar permeability during direct lung injury, while an increase in claudin-4 may be a compensatory action aiding in epithelial repair (126).

Furthermore, viral infection and pro-inflammatory cytokine stimulation of lung epithelial cells decreases ZO-1 expression and membrane junction localization of ZO-1 and JAM, and is associated with increased epithelial monolayer permeability (127, 128).

Surfactant production and function are additionally impaired during direct lung injury. In patients with ARDS due to direct lung injury, there is decreased lavagable bronchoalveolar surfactant phospholipids and proteins (SP-A, -B, and -C) (129, 130). This decrease in alveolar surfactant is thought to be due to both decreased surfactant production (diminished surfactant protein transcription), and increased surfactant degradation by ROS-mediated phospholipid and protein oxidation, and phospholipase A2-dependent phospholipid hydrolysis (129, 131). Patients with ARDS also exhibit alterations in BAL surfactant phospholipid composition, containing decreased percentages of the main surfactant phospholipid contributors phosphatidylcholine and phosphatidylgylcerol (131). Furthermore, not only is alveolar surfactant concentration and composition altered in direct lung injury, surfactant activity is also impaired. Extravasation of

36 proteins from the vasculature into the alveolar airspace impairs both surfactant adsorption and activity, thereby leading to atelectasis and further alveolar injury (125, 131).

Alveolar edema and flooding during direct lung injury occurs not only due to increased edema formation via epithelial barrier and surfactant dysfunction, but also due to decreased alveolar fluid clearance. Impaired/submaximal alveolar fluid clearance is common, occurring in

69% of ARDS patients. As functional alveolar fluid clearance is thought to be mediated by ion

(mainly Na+) transport across the alveolar epithelium, impaired alveolar fluid clearance is more common in patients with ARDS due to factors other than sepsis (likely direct lung injury) (132).

Decreased alveolar fluid clearance during direct lung injury is likely due to both death of ATI cells that leave behind a denuded basement membrane and alterations in epithelial Na+ transporters (73, 133). In animal models of direct lung injury, there is decreased expression and activity of ENaC and Na+/K+-ATPase, and rescue of Na+/K+-ATPase in the alveolar epithelium by genetic overexpression can help to restore alveolar fluid clearance and decrease alveolar edema (133, 134).

In contrast to indirect lung injury, the fate and role of an alveolar epithelial glycocalyx in direct lung injury remains poorly understood. Limited in vitro and in vivo studies show that

MMP-7 mediates syndecan-1 shedding from the airway epithelium following intratracheal bleomycin, and that ADAM-17 induces syndecan-1 and syndecan-4 shedding in vitro following pro-inflammatory cytokine stimulation (76, 135). Despite this previous literature, the existence and role of alveolar epithelial HS and CS shedding, and the specific effect of the shed HS and CS fragments on lung injury and epithelial permeability during direct lung injury in vivo has not been studied.

Treatment of ARDS

Despite an abundance of clinical trials, treatment for ARDS consists largely of supportive care strategies as no targeted therapies have proven effective (136, 137). Both ventilatory and fluid management strategies have been shown effective to improve mortality. Multiple studies and a meta-analysis have shown that high positive end expiratory pressure (PEEP) and low tidal volume ventilation improve ARDS mortality, by increasing and maintaining the recruitment of alveoli during expiration and by reducing barotrauma to the already injured lung (138, 139).

Furthermore, prone positioning also decreases mortality by increasing ventilation to areas of the lung that are normally collapsed in the supine position, thereby restoring V/Q matching (140).

Although no decrease in mortality was measured, conservative fluid management shortened the duration of mechanical ventilation in ARDS patients, improving lung function by limiting fluid- induced pulmonary edema (141). While these supportive care strategies all improve ARDS outcomes, clinical trials of therapies targeting surfactant, ROS, and inflammation (all of which are implicated in the pathogenesis of ARDS) have not been effective at decreasing ARDS mortality (136).

As previously discussed, ARDS is heterogeneous, and although these failed targeted therapies did not improve mortality in ARDS all-comers, stratification of patients into sub- phenotypes may allow for the identification of therapies that are more effective in certain ARDS sub-phenotypes. Indeed, response to both already-effective supportive strategies and failed targeted therapies vary in patients with direct versus indirect lung injury (99). ARDS patients with indirect lung injury exhibit a greater response to high PEEP ventilation and prone positioning, as it is thought that alveolar recruitment in patients with direct lung injury will not improve oxygenation as many affected alveoli are flooded or consolidated. Furthermore,

38 patients with direct lung injury exhibit an increased response to exogenous surfactant treatment, likely due to the increased alveolar epithelial injury and surfactant dysfunction in direct lung injury versus indirect lung injury (99). Consistent with these findings, determining strategies to stratify ARDS patients may not only increase what is known about distinct pathogenic mechanisms, but may also lead to the development and approval of effective therapies in certain sub-phenotypes of patients.

CHAPTER IV

THESIS STATEMENT5

The glycocalyx, a carbohydrate-enriched layer lining the cell surface, has long been an understudied cellular structural element. While visualization of the alveolar epithelial glycocalyx has been achieved by ruthenium red staining and electron microscopy (75), the structure and function of the epithelial glycocalyx remains poorly understood, particularly in comparison to the endothelial glycocalyx. The pulmonary endothelial glycocalyx is enriched in

HS and contributes to critical vascular functions such as endothelial barrier integrity, mechanotransduction of shear stress, regulation of vascular tone, and inhibition of leukocyte adhesion (24, 59-61). Furthermore, TNFα-induced heparanase expression is known to cause endothelial glycocalyx-HS degradation, ESL collapse, and endothelial neutrophil adhesion (24).

In contrast, the structure and function of the lung epithelial glycocalyx, and its fate during lung injury, remain nearly entirely unknown. While others have described shedding of epithelial syndecan-1 and syndecan-4 during lung injury and inflammation in vitro and in vivo (76, 135), the GAG composition of the epithelial glycocalyx in the healthy adult lung, the existence and role of alveolar epithelial HS and CS shedding during direct lung injury, and the specific effect of the shed HS and CS fragments on lung injury and epithelial permeability in vivo has not been studied. As such, the studies in this thesis aim to better understand the GAG composition and function of the epithelial glycocalyx in the healthy lung, and determine its fate and role in direct lung injury. In Chapter V we will discuss the GAG composition and function of the epithelial glycocalyx in the healthy adult lung. We will then demonstrate and determine the role of epithelial HS and CS shedding during direct lung injury in Chapter VI, and discuss possible

5Portions of this chapter were previously published in the American journal of respiratory cell and molecular biology. Reprinted with permission of the American Thoracic Society. Copyright © 2018 American Thoracic Society. (1, 2)

effects of the shed alveolar HS and CS on lung injury and repair in Chapter VII (Figure IV-1).

Figure IV-1: Thesis Aims. In this thesis, we will identify the GAG composition and function in the healthy lung (Chapter V), demonstrate the existence and role of epithelial HS and CS shedding during direct lung injury (Chapter VI), and discuss the function of shed alveolar HS and CS on lung injury and repair (Chapter VII). Modified from Haeger, et al, 2016 (2).

CHAPTER V

ALVEOLAR EPITHELIAL HEPARAN SULFATE AND CHONDROITIN SULFATE IN

THE HEALTHY LUNG6

Introduction

The glycocalyx, a carbohydrate-enriched layer lining the cell surface, has long been an understudied cellular structural element. Artifactual degradation of the glycocalyx during tissue fixation has made visualizing and studying cell glycocalyces difficult, especially in whole tissue preparations (142). The use of periodic acid-Schiff and colloidal iron staining allowed for the first description of the glycocalyx by identifying negatively charged carbohydrates on the surface of numerous cell types (143). Since these initial studies, electron microscopy and immunohistochemistry have been increasingly used to visualize glycocalyces on numerous cell types in several different organs, including on the pulmonary endothelial and epithelial surfaces

(75, 144). The pulmonary endothelial glycocalyx is known to contribute to critical vascular functions such as endothelial barrier integrity, mechanotransduction of shear stress, regulation of vascular tone, and inhibition of leukocyte adhesion (24, 59-61). In contrast, the structure and function of the lung epithelial glycocalyx remain nearly entirely unknown. As such, we aimed to confirm the presence and determine the function of alveolar epithelial HS and CS within the healthy adult lung.

6Portions of this chapter were previously published in the American journal of respiratory cell and molecular biology. Reprinted with permission of the American Thoracic Society. Copyright © 2018 American Thoracic Society. (1, 2)

Objectives

• Confirm the presence of alveolar epithelial HS and CS in the healthy adult lung.

• Determine the sulfation pattern of alveolar epithelial cell-surface HS.

• Determine if alveolar epithelial HS or CS contributes to alveolar epithelial barrier

function.

Materials and Methods

Materials

Heparinase I/III blend from Flavobacterium heparinum and chondroitinase ABC from

Proteus vulgaris were purchased from Sigma-Aldrich and reconstituted in PBS. Heparinase I/III and chondroitinase ABC were stored at 4°C and used within 2 weeks of reconstitution. As controls, heparinase I/III and chondroitinase ABC were heat inactivated at 100°C for 15 minutes.

Animals

Experiments were approved by the University of Colorado Institutional Animal Care and

Use Committee and conducted in accordance with the National Institutes of Health guidelines.

Wild-type 8-12 week old male C57BL/6 mice purchased from Jackson Laboratories (Bar Harbor,

ME) were used for experiments.

Alveolar Epithelial HS and CS Degradation

To confirm the presence, and determine the effect of epithelial HS and CS degradation, we anesthetized mice with inhaled isoflurane and using a mouse laryngoscope intratracheally instilled 15 units active or heat-inactivated (HI) heparinase I/III (in 40µl) to degrade epithelial

HS, or 2 units active or HI chondroitinase ABC (in 40µl) to degrade epithelial CS. For experiments in which mice were treated with heparinase I/III-generated HS fragments, 8.75 µg full length HS (Celsus Laboratories; Cincinnati, OH, USA), was treated with 4.375 units active

43 or HI heparinase I/III for 3 hours at 37°C. The heparinase-HS mixture was then heat inactivated at 100°C for 15 minutes and intratracheally instilled.

Animals were euthanized (using ketamine/xylazine) 12, 24, and 72 hours after instillation. Blood was collected from the inferior vena cava followed by plasma isolation.

Plasma HS was measured as described below. Three – one milliliter BALs were performed using ice cold PBS. BAL total protein, mouse albumin, HS, and neutrophils were measured as described below. The lung was then inflated with 50%/50% PBS/OCT and snap frozen in liquid nitrogen for immunofluorescence. For experiments in which lung wet/dry ratio was measured, after blood collection, the entire un-lavaged lung was extracted and immediately weighed (wet weight) and weighed again after being placed in a 60°C drying oven for 24 hours (dry weight).

To determine the effect of heparinase I/III-generated HS fragments, in mice treated with active/HI heparinase I/III-treated HS, 24 hours after intratracheal instillation, plasma and BAL were collected as detailed above, and the lung was perfused and snap frozen in liquid nitrogen for RNA isolation. BAL total protein, BAL neutrophils, and lung TNFα, IL-6, and IL-1 mRNA expression were measured as detailed below.

Isolation and Quantification of BAL Fluid and Plasma Heparan Sulfate

HS was isolated from plasma and BAL fluid from mice at several timepoints after intratracheal instillation. BAL fluid and plasma were loaded into a 3kDa molecular weight cutoff (MWCO) column and HS disaccharides were generated and isolated as previously described (145). Disaccharide analyses were then performed by liquid chromatography-tandem mass spectrometry multiple reaction monitoring (LC-MS/MS MRM) as detailed previously

(145). BAL HS concentration was converted to alveolar lining fluid HS concentration by a urea

44 dilution correction factor determined in separate samples by measuring urea in both BAL fluid and plasma (146).

Lung/Alveolar Permeability, Edema, and Inflammation Quantification

Total BAL cells were determined on freshly collected BAL fluid using a hemocytometer.

The BAL fluid was then centrifuged at 1000 RPM for 5 minutes. Following centrifugation, the supernatant was collected for BAL total protein and albumin quantification (in addition to

HS/CS quantification). The pellet was resuspended in 400µl PBS for BAL differential cell counts. Differential BAL cell counts were performed by creating BAL cell cytospins, staining the resulting cytospin slides with the PROTOCOL Hema 3 staining kit (Fisher Scientific;

Hampton, NH, USA), and counting the percentage of neutrophils, macrophages/monocytes, and lymphocytes. Total BAL protein was quantified using the Bradford Protein Assay purchased from Bio-Rad (Hercules, CA, USA) as instructed in 96-well plates using 20µl of each sample/standard. Absorbance was measured at 600nm using the Promega GloMax® machine.

BAL mouse albumin was quantified by mouse albumin ELISA using 1:1500 diluted BAL fluid following the manufacturer’s instructions (Bethyl, Montgomery, TX, USA). Lung wet/dry ratio was calculated by dividing the lung wet weight by the lung dry weight, obtained as described above.

Immunohistochemistry

Staining for HS and CS was performed on unfixed, inflated, and frozen tissue to preserve epithelial glycosaminoglycans. 0.7 µm thick lung sections were then blocked with the M.O.M staining kit (Vector; Burlingame, CA, USA) and stained for 45 minutes for HS using the mouse

HS 10E4 antibody (AmsBio; Cambridge, MA, USA) and for CS using the mouse CS-56 antibody (ThermoFisher; Waltham, MA, USA). Following primary antibody application,

45 sections were treated with the M.O.M. biotinylated secondary antibody, fluorescein-streptavidin conjugate, and DyLight 594-lycopersicon esculentum (LEA) lectin (Vector). The sections were then mounted with DAPI mounting media (Vector), coverslipped, and imaged.

Staining for ZO-1 and claudin-18 were performed on inflated and frozen tissue that was fixed with 4% paraformaldehyde for 15 minutes. 0.7µm thick lung sections were then blocked with 10% serum of the secondary antibody host and stained for 1 hour for ZO-1 (ThermoFisher) or claudin-18 (ThermoFisher). Following primary antibody application, the sections were treated with FITC-secondary antibody and mounted with DAPI mounting media (Vector), coverslipped, and imaged.

RNA Isolation, cDNA Synthesis, and qRT-PCR

Total RNA was isolated from whole lung tissue using QIAzol and the RNeasy mini kit per manufacturer’s instructions (Qiagen; Germantown, MD, USA). cDNA was synthesized using the iScript cDNA synthesis kit (BioRad). Taqman probes were then used to perform qRT-

PCR using the Applied Biosystems 7300 Real-Time PCR System. Cyclophilin was used as a housekeeping gene and data were analyzed using the 2-ΔCt method.

Statistical Analyses

Animals were randomized to treatment or control groups. Single comparisons were made using unpaired, two-tailed t-tests with Welch’s correction, and given small sample sizes confirmed with non-parametric Mann-Whitney tests. Graphs show all experimental replicates with means and standard deviations. Analyses were performed using Prism (GraphPad; La Jolla,

CA, USA) and results were considered statistically significant if p<0.05 (significance by t-test and Mann-Whitney test noted by *, by Mann-Whitney test only by #).

Results

Epithelial Glycocalyx Heparan Sulfate Contributes to Epithelial Barrier Function

To confirm the existence of HS within the pulmonary epithelial glycocalyx, we treated mice with intratracheal heparinase I/III, a combination of recombinant bacterial enzymes that specifically degrade HS, to remove any epithelial HS in vivo (Figure V-1A). We then stained inflated, frozen, unfixed mouse lung tissue for HS. Mice treated with enzymatically-inactive

(heat-inactivated (HI)) heparinase I/III demonstrated diffuse expression of alveolar epithelial HS, as visualized by HS (HS 10E4 antibody) colocalization with the peripheral airway/alveolar epithelial marker LEA lectin (Figure V-1B). In contrast, mice treated with enzymatically- active heparinase I/III demonstrated a patchy loss of epithelial HS expression, consistent with the patchy distribution of intratracheal instillation, confirming antibody specificity and epithelial HS loss. HS degradation by intratracheal heparinase I/III instillation was constrained to the epithelium, as LC-MS/MS demonstrated HS shedding into the alveolar airspace (Figure V-1C) but not the circulation (Figure V-1D). Mass spectrometry revealed that HS released from the lung epithelium after heparinase I/III was heavily N- and 2-O sulfated (NS2S), with 55% of HS disaccharides being jointly sulfated at these two sites, indicating the alveolar epithelial glycocalyx HS is enriched with sulfation at the N- and 2-O positions (Figure V-1E).

There was minimal airway HS expression by HS 10E4 immunofluorescence (Figure V-

1F), suggesting that HS fragments released into the BAL after heparinase I/III largely reflect alveolar epithelial HS. Any contribution of airway mucus HS (i.e. not epitheilial-anchored) is likely reflected in the BAL HS of mice treated with HI heparinase I/III. Taken together, these data suggest that the increase in BAL HS measured after intratracheal heparinase I/III instillation largely reflects fragmentation of alveolar epithelial cell-surface HS. However, we cannot

exclude the epithelial basement membrane/extracellular matrix as an additional source of BAL

HS after heparinase I/III. Furthermore, we cannot exclude the contribution of airway epithelial

HS that may not be detected by the 10E4 antibody (147).

Figure V-1: Intratracheal Heparinase I/III Removes Alveolar Epithelial Heparan Sulfate Releasing Alveolar N- and 2-O Sulfated Heparan Sulfate Fragments. A) Mice were instilled with 15U intratracheal (IT) active or HI heparinase I/III (hep I/III or HI hep I/III). B) Twelve hours after instillation frozen lungs were sectioned and stained for heparan sulfate (HS 10E4, green), and the peripheral airway/alveolar epithelium (LEA lectin, red). Images of the alveolus were taken. Twelve, 24, and 72 hours after instillation BAL fluid and plasma were obtained and C) BAL and D) plasma HS concentration were measured. E) HS chemical structure with sulfation sites and sulfation of BAL HS 12 hours after 15U intratracheal heparinase I/III instillation. F) Peripheral airway staining heparan sulfate (HS 10E4, green), and the peripheral airway/alveolar epithelium (LEA lectin, red). Modified from Haeger, et al, 2018 (1).

We measured BAL protein, BAL albumin, lung wet/dry ratio, and performed BAL neutrophil cell counts in mice treated with intratracheal active or HI heparinase I/III to determine if epithelial HS contributes to the maintenance of alveolar epithelial barrier function or if epithelial HS degradation increases lung inflammation. Intratracheal enzymatically-active heparinase I/III treatment increased BAL total protein and albumin 12, 24, and 72 hours after instillation, in comparison to HI heparinase I/III treatment (Figure V-2A and V-2B), indicating that epithelial HS contributes to the lung epithelial barrier to protein. In contrast, there was no change in lung wet/dry ratio in mice treated with active heparinase I/III in comparison to heat- inactived heparinase I/III, suggesting that epithelial HS degradation may not contribute to the lung epithelial barrier to fluid, or that lung wet/dry ratio is not sensitive enough to measure any mild edema induced by enzymatically-active heparinase I/III instillation (Figure V-2C). We detected no increase in BAL neutrophils after heparinase I/III treatment (Figure V-2D), indicating that the increase in protein permeability after heparinase I/III treatment was not due to alveolar inflammation.

Epithelial Glycocalyx Chondroitin Sulfate Does Not Contribute to Epithelial Barrier Function

To next confirm the existence of CS within the pulmonary epithelial glycocalyx, we treated mice with intratracheal chondroitinase ABC, a combination of recombinant bacterial chondroitin sulfate lyases, to remove any epithelial CS in vivo (Figure V-3A). We then stained inflated, frozen, unfixed mouse lung tissue for CS. Mice treated with enzymatically-inactive

(heat-inactivated (HI)) chondroitinase ABC demonstrated some diffuse expression of alveolar epithelial CS, as visualized by CS (CS-56 antibody) colocalization with the peripheral airway/alveolar epithelial marker LEA lectin (Figure V-3B). We measured BAL protein and performed BAL neutrophil cell counts in mice treated with intratracheal enzymatically active

versus HI chondroitinase ABC to determine if CS also contributes to the epithelial barrier to

protein. While treatment with active chondroitinases removed alveolar epithelial surface CS,

loss of CS did not result in an increase of BAL protein or BAL neutrophils (Figure V-3C and V-

3D), suggesting that not all sulfated glycosaminoglycans contribute to the lung epithelial barrier.

Figure V-2: Intratracheal Heparinase I/III Increases Alveolar Permeability to Protein, but not Lung Edema or Alveolar Inflammation. Twelve, 24, and 72 hours after heparinase I/III instillation A) BAL protein, B) BAL albumin, and D) BAL neutrophils were measured. Twenty- four hours after instillation C) lung wet/dry weight ratio was measured. Modified from Haeger, et al, 2018 (1).

Figure V-3: Intratracheal Chondroitinase ABC Removes Alveolar Epithelial Chondroitin Sulfate but does not Increase Lung Permeability or Inflammation. A) Mice were instilled with 2U IT active or heat-inactivated chondroitinase ABC (Ch ABC or HI Ch ABC). B) Twenty-four hours after instillation frozen lungs were sectioned and stained for CS, CS-56 (green), and the alveolar epithelium, LEA lectin (red). BAL fluid was also obtained 24 hours after instillation and BAL C) protein D) and neutrophils were measured. Modified from Haeger, et al, 2018 (1). Heparinase I/III-Generated Heparan Sulfate Fragments do not Increase Lung Epithelial

Permeability

Given that alveolar epithelial HS degradation increased alveolar permeability to protein, we next sought to determine the mechanism by which this occurs. We first aimed to determine whether HS fragments generated from heparinase I/III-mediated degradation led to the increased alveolar permeability after intratracheal heparinase I/III treatment, indicating that this increased permeability may be due to the released fragments, rather than removal of alveolar epithelial HS

51 that may function as part of the alveolar epithelial barrier. While others have shown that HS fragments can act as damage associated molecular patterns (DAMPs) and induce inflammation through TLR4 signaling, we detected no increase in BAL neutrophilia after heparinase I/III treatment (Figure V-1H) (25, 148). To confirm that heparinase I/III-generated HS fragments do not increase lung epithelial protein permeability by functioning as DAMPs, we treated mice with full-length HS pre-treated ex vivo with either heparinase I/III (at 37 ºC for 3 hours) to generate

HS fragments, or HI heparinase I/III as control. After ex-vivo HS fragmentation, the heparinase-

HS mixture was then heat-inactivated before intratracheal instillation (Figure V-4A). As HS is resistant to heat inactivation, this approach ceases heparinase I/III activity without interfering with any DAMP-like effect of HS (149).

We observed no increase in BAL protein or neutrophils after intratracheal instillation of heparinase I/III-treated HS in comparison to HI heparinase I/III-treated HS (Figure V-4B and

V-4C). Furthermore, we detected no increase in mRNA expression of pro-inflammatory cytokines IL-6 or IL-1 in whole lung tissue from mice treated with intratracheal heparinase

I/III-treated HS in comparison to HI heparinase I/III-treated HS (Figure V-4D). While there was increased TNFα expression (significance by Mann-Whitney test) in whole lung from mice treated with intratracheal heparinase I/III-treated HS (in comparison to HI heparinase I/III- treated HS), the overall absence of BAL neutrophilia in mice treated with heparinase I/III (Fig

V-1H) or heparinase-I/III-treated HS (Fig V-4C) is supporting evidence that intratracheal heparinase I/III does not significantly increase alveolar inflammation. Taken together, these results indicate that HS on the epithelial cell surface contributes to the lung epithelial barrier to protein and that the HS fragments generated after heparinase I/III treatment are not responsible for the increase in lung permeability observed after intratracheal heparinase I/III instillation.

Figure V-4: Exogenous Heparinase I/III-Generated Heparan Sulfate Fragments do not Increase Lung Permeability or Inflammation. A) Exogenous HS (8.75µg) was pre-treated with 4.375 units of active or HI hep I/III. The heparinase-HS mixture was then heat inactivated and intratracheally instilled. 24 hours after intratracheal instillation of active or HI hep I/III- treated heparan sulfate, B) BAL protein, C) BAL neutrophils, D) and whole lung TNFα, IL-6, and IL-1 mRNA expression was measured. Modified from Haeger, et al, 2018 (1). Intratracheal Heparinase I/III does not Change Alveolar ZO-1 or Claudin-18 Expression

Findings from the lung endothelium and the corneal epithelium suggest that cell glycocalyx-HS regulates the expression and localization of tight junction, adherens junction, and gap junction proteins (150, 151). Knockout of EXT1, and thereby HS, within the corneal epithelium resulted in decreased expression of the tight junction protein ZO-1. As tight junctions are critical to alveolar epithelial barrier function, we stained lung tissue for ZO-1 and claudin-18

53 from mice harvested 12 hours after intratracheal heparinase I/III instillation (Figure V-5C). We detected no difference in ZO-1 or claudin-18 immunofluorescence in mice treated with active versus HI heparinase I/III, indicating that epithelial HS degradation does not decrease alveolar expression of ZO-1 or claudin-18. Although these data suggest that epithelial HS degradation does not impact tight junction protein expression, further experimentation is needed to determine any role of epithelial HS in the expression of other tight junction proteins and their sub-cellular localization.

Figure V-5: Intratracheal Heparinase I/III does not Alter Alveolar ZO-1 or Claudin-18 Expression. Twelve hours after intratracheal heparinase I/III instillation frozen lungs were sectioned, fixed and stained for ZO-1 and Claudin-18 (green). Summary of Results

• HS and CS are present within the alveolar epithelium of the healthy adult human lung.

• Alveolar epithelial HS (but not CS) contributes to the epithelial barrier to protein and is N-

and 2-O sulfated.

• Alveolar epithelial HS degradation does not alter alveolar ZO-1 or claudin-18 expression.

Discussion

The epithelial glycocalyx remains a largely understudied component of the lung despite its discovery nearly 50 years ago (75). While the proteoglycans syndecan-1 and syndecan-4 are known to be expressed in the lung epithelium, this report identifies the structure and function of

HS on the surface of the healthy adult lung epithelium (76, 135). Using immunofluorescence we demonstrated the presence of HS and CS within the lung epithelium. We also identified that epithelial HS is enriched in disaccharides sulfated on the nitrogen of glucosamine and 2-position of iduronic acid (NS2S) using state-of-the-art LC-MS/MS MRM analyses (145). These patterns of sulfation influence the biological function of HS, as the molecular distribution of negative charge imparted by sulfate groups allow HS to bind (with specificity) to positively charged residues of proteins and regulate their function (23, 152).

Our findings additionally demonstrate that lung epithelial HS, but not CS, contributes to the lung epithelial barrier to protein. These observations are consistent with findings published by Saumon, et al. demonstrating that instillation of airway protamine, a positively charged protein that can bind to and neutralize the negative charge of sulfated glycosaminoglycans and other molecules, increases lung epithelial permeability and are similar to findings from the intestinal, bladder, and corneal epithelium (150, 153-156). The mechanisms by which lung epithelial HS promotes barrier integrity remain unknown. While findings from the lung endothelium and the corneal epithelium suggest that cell glycocalyx-HS regulates the expression and localization of tight junction, adherens junction, and gap junction proteins (150, 151), we observed no change in ZO-1 or claudin-18 expression in mice treated with intratracheal heparinase I/III. Alternatively, cell-surface HS may serve as a non-signaling, structural constituent of the epithelial barrier, similar to the “charged meshwork” of the endothelial

55 glycocalyx that opposes transvascular protein flux (157). Future studies are needed to determine relative contributions of epithelial HS to barrier-regulating, outside-in cellular signaling processes as well as to the physical barrier that opposes epithelial permeability.

CHAPTER VI

ALVEOLAR EPITHELIAL HEPARAN SULFATE AND CHONDROITIN SULFATE

ARE SHED DURING DIRECT LUNG INJURY7

Introduction

In Chapter V we determined that alveolar epithelial HS contributes to the epithelial barrier to protein. Given these findings, we hypothesized that degradation or shedding of epithelial HS may occur during lung injury, a disease state characterized by increased alveolar permeability. Indeed, it has been shown in other cell types that the structure and function of glycocalyx HS can be disrupted during inflammation and injury (24, 31, 158). Following non- pulmonary sepsis-induced (i.e. endothelial-targeted) lung injury, endothelial glycocalyx HS is degraded by heparanase, releasing small (≤ 8 saccharide-long) HS fragments into the circulation

(117). This septic HS degradation disrupts the critical vascular functions of the endothelial glycocalyx, resulting in lung edema and neutrophil adhesion—two pathogenic hallmarks of sepsis-induced acute lung injury and the acute respiratory distress syndrome (ARDS) (24). In contrast, epithelial insults do not appear to primarily target the endothelial glycocalyx, as human studies demonstrate a relative paucity of circulating HS and CS after direct pulmonary injury

(116, 117).

Instead, these direct pulmonary insults may degrade epithelial HS and or CS.

Bleomycin-induced lung injury induces matrix metalloproteinase 7 (MMP-7)-mediated shedding of syndecan-1 into the alveolar space, and in vitro models of alveolar epithelial injury are characterized by shedding of syndecans 1 and 4 via activation of disintegrin and metalloproteinase 17 (ADAM-17) (76, 135). Reactive oxygen species (ROS) may additionally

7Portions of this chapter were previously published in the American journal of respiratory cell and molecular biology. Reprinted with permission of the American Thoracic Society. Copyright © 2018 American Thoracic Society. (1)

serve as proteoglycan/GAG “sheddases” during acute lung injury, as proteoglycan and HS fragmentation can be prevented by extracellular superoxide dismutase (42). Despite this previous literature, the specific effect of HS and/or CS shedding on lung injury and epithelial permeability in vivo has not been studied. Therefore, we aimed to determine whether epithelial

HS and CS are shed into the alveolus during direct lung injury, and hypothesized that, as epithelial HS contributes to the alveolar barrier, shedding of HS from the lung epithelium contributes to the increased lung permeability during intratracheal injury.

Objectives

• Determine if alveolar epithelial HS and CS are shed into the airspace during direct lung

injury, in mice and in ARDS patients, and if the amount of HS and CS shed correlate with

alveolar permeability.

• Determine the sheddases that release alveolar epithelial HS and CS into the airspace during

direct lung injury in mice.

• Determine if alveolar epithelial HS sheddase inhibition reduces epithelial permeability

during direct lung injury in mice.

Materials and Methods

Materials

Lipopolysaccharide (LPS) from Escherichia coli O55:B5 was purchased from Sigma-

Aldrich (St. Louis, MO, USA) and dissolved in phosphate buffered saline (PBS). The ADAM17 inhibitor TAPI-2 was purchased from Millipore (Burlington, MA, USA) and reconstituted in

PBS. The broad-spectrum MMP inhibitor doxycycline hyclate was purchased from Sigma-

Aldrich and was reconstituted in water.

Animals

In addition to the animals and approval described in Chapter V, EC-SOD overexpressing mice, SPC promoter-driven human transgenic heterozygous EC-SOD overexpressing mice, which express 3-4 times more lung EC-SOD within type II alveolar epithelial cells were also used (159, 160). MMP-9 knockout/null (MMP9ko) mice were also purchased from Jackson.

Intratracheal LPS-Induced Lung Injury

To induce intratracheal LPS-mediated lung injury, animals were intratracheally instilled, as described previously, with 3mg/kg lipopolysaccharide (LPS) from Escherichia coli O55:B5 or

PBS (30µl) and euthanized 12 hours, 2 days, 4 days, and/or 6 days after instillation. For experiments containing TAPI-2, mice were treated concurrently with intratracheal LPS and

1mg/kg TAPI-2, a dose sufficient to inhibit ADAM-17 (161), and euthanized 2 days after instillation. For experiments containing doxycycline, mice were treated with 70mg/kg oral doxycycline hyclate by gavage (a dose sufficient to inhibit broad-spectrum MMP activity, (162)) every 24 hours beginning 3 days prior to intratracheal LPS until euthanasia 2 days after intratracheal LPS instillation. For experiments using EC-SOD over-expressing and MMP-9 knockout mice, animals were euthanized 2 days after intratracheal LPS instillation. After euthanasia, plasma and BAL were collected as detailed above. The entire lung was then perfused, and the right lung was snap frozen in liquid nitrogen for RNA isolation and the left lung inflated with 1% low-melt agarose and formalin fixed/paraffin embedded for histologic analysis. BAL neutrophils, total protein, HS and CS concentration and size, syndecan-1, syndecan-4, MMP protein and activity, plasma HS, and lung MMP mRNA expression were quantified as described below.

HME collection

The Vanderbilt University Medical Center Institutional Review Board approved this minimal risk study with a waiver of informed consent since heat and moisture exchanger (HME) filters are part of routine clinical care and used filters are typically discarded. Subjects were eligible if they were invasively mechanically ventilated and had bilateral radiographic infiltrates consistent ARDS (163). Endotracheal tube HME filters (AirLife Adult HME filter, CareFusion) were collected after being in position for up to 12 hours and transported to the laboratory on ice and centrifuged at 2,000 x g for 10 minutes to collect condensed fluid. HME fluid was then aliquoted and stored at -80°C for further analysis.

Isolation, Quantification, and Size Determination of HME Fluid, BAL Fluid and Plasma Heparan

Sulfate and Chondroitin Sulfate

HS and CS were isolated from plasma, bronchoalveolar lavage (BAL) fluid, and HME fluid as described in Chapter V. BAL sulfated GAG size was determined by first desalting BAL fluid with a 3kDa MWCO column followed by PAGE and alcian blue/silver staining as previously described (164).

Lung/Alveolar Permeability and Inflammation Quantification

Lung/alveolar permeability and inflammation were measured as described in Chapter V.

Western Blotting

5x Laemmli buffer was added to equal volumes of BAL fluid and heated at 100°C for 10 minutes. The BAL fluid was then loaded into 4-20% gradient Criterion Tris-HCl gels (BioRad) followed by gel electrophoresis and transfer to PVDF membrane. The membranes were blocked, stained for syndecan-1, syndecan-4 (Santa Cruz Biotechnology; Dallas, TX, USA), MMP-2

(Abcam, Cambridge, MA, USA), MMP-7, and/or MMP-9 (R&D Systems, Minneapolis, MN,

USA), and developed using enhanced chemiluminescence. The developed film was then scanned and band intensity was quantified by Image J.

Zymography

BAL MMP-2 and MMP-9 gelatinolytic activity was quantified by zymography. Equal volumes BAL fluid were added to 4x Laemmli buffer (without reducing agents) and loaded on

Novex™ 10% gelatin protein zymogram gels (Invitrogen, Carlsbad, CA, USA). Following electrophoresis, the gel was renatured, developed at 37°C overnight, and stained for 1 hour with

Bio-Safe™ Coomassie G-250 Stain (BioRad). The gel was imaged and band intensity quantified by Image J.

RNA Isolation, cDNA Synthesis, and qRT-PCR

RNA isolation, cDNA synthesis, and qRT-PCR were performed as described in Chapter

Statistical Analyses

Statistical analyses were performed as described in Chapter V. Significance by t-test and

Mann-Whitney test are noted by *, by t- test only noted by ǂ.

Results

Heparan Sulfate and Chondroitin Sulfate are Released into the Airspace During Intratracheal

LPS-Induced Lung Injury in Mice

Given the observed contributions of HS to the pulmonary epithelial barrier, and previous findings that syndecan-1 (a CS and HS-containing proteoglycan) and/or syndecan-4 are shed from the lung epithelium in vitro and in vivo, we sought to determine if epithelial HS and CS shedding/degradation occurs during acute lung injury. While live infection models of direct lung injury may best replicate ARDS in patients, HS is known to bind and effect the adherence of

61 bacteria to lung epithelial cells. As such we aimed to remove this confounding factor from our studies and therefore chose to induce lung epithelial injury via intratracheal instillation of lipopolysaccharide (LPS). We treated mice with intratracheal 3 mg/kg lipopolysaccharide

(LPS), a dose sufficient to induce alveolar hyper-permeability (as measured by BAL protein) and histologic injury (Figure VI-1A-C). Using LC-MS/MS MRM, we observed that during injury there was increased alveolar but not plasma HS 12 hours, 2 days, and 4 days after intratracheal

LPS administration, in comparison to PBS control (Figure VI-1D and E). The increase in alveolar HS after LPS instillation persisted after BAL high speed centrifugation (data not shown), suggesting that the HS detected in the BAL fluid is shed and not anchored to dead and sloughed epithelial cells. Consistent with our findings that epithelial HS functions to maintain epithelial barrier function, in animals treated with intratracheal LPS, alveolar HS shedding correlated with the degree of lung permeability as measured by BAL protein (Figure VI-1F).

Together these results indicate that HS shedding/degradation may contribute to the lung epithelial barrier dysfunction during intratracheal LPS-induced lung injury.

Similar to HS, an increase alveolar CS was detected 12 hours, and 2 days, 4 days, and 6 days after intratracheal LPS instillation (Figure VI-1G). Additionally, an increase in plasma CS was detected 4 days after intratracheal LPS instillation (Figure VI-1H). While at earlier timepoints (12 hours and 2 days) there was no increase in plasma CS in mice treated with LPS, in comparison to PBS, the plasma CS concentration was greatest 12 hours after intratracheal instillation and declined with time in both groups. These data suggest that CS is shed into the alveolar airspace during intratracheal LPS-mediated lung injury, and that leak of alveolar CS or shedding/degradation of CS into the vasculature occurs with intratracheal instillation and may be prolonged in mice treated with intratracheal LPS in comparison to PBS. Although alveolar

62 epithelial CS does not appear to contribute to alveolar epithelial barrier function (Chapter V), alveolar CS shedding/degradation correlated with the degree of lung permeability as measured by BAL protein (Figure VI-1I). As such, while alveolar HS shedding/degradation may contribute to the lung epithelial barrier dysfunction during intratracheal LPS-induced lung injury, alveolar CS shedding/degradation may represent a biomarker for the severity of lung injury.

Heparan Sulfate and Chondroitin Sulfate are Released into the Airspace During the Acute

Respiratory Distress Syndrome in Humans

To investigate the translational relevance of our findings in mice, we next sought to determine whether HS and CS shedding/degradation similarly occurs in human patients with

ARDS and if the amount of airspace HS and CS correlates with lung permeability. In patients with ARDS, alveolar edema fluid contents can be accurately measured in fluid collected from endotracheal tube heat and moisture exchangers (HME), allowing for a minimally invasive approach to airspace sampling in critically ill patients (165). We collected HME fluid from 15

ARDS patients and measured HME total protein, HS, and CS concentrations (Table VI-1). In accordance with our findings in acute lung injury in mice, HME HS and CS concentrations correlated with the total protein concentration in ARDS patients (Figure VI-2A and C). In addition, we detected a modest trend towards an increase in HME HS in ARDS patients with direct lung injury due to pneumonia/aspiration in comparison to patients with indirect lung injury due to sepsis (Figure VI-2B). There was a weaker trend, if any, detected in HME CS in patients with direct versus indirect lung injury (Figure VI-2D). To ensure that the three septic ARDS patients were not “uninjured” subjects skewing the correlation between HME HS and HME protein, we performed sensitivity analyses excluding these three patients as well as three

63 additional patients with similarly low HME HS, CS, and protein. After excluding these patients, the HME HS-protein and CS-protein correlations remained significant (Figure VI-2E and F).

Figure VI-1: Increased Airspace Heparan Sulfate and Chondroitin Sulfate are Detected During Intratracheal LPS-Induced Lung Injury in Mice. A) Mice were intratracheally instilled with 3mg/kg LPS or PBS as control. 12 hours, 2, 4, and 6 days after intratracheal instillation B) BAL protein was measured. C) H&E staining was performed on lung sections from day 2 after intratracheal LPS and PBS instillation. 12 hours, 2, 4, and 6 days after intratracheal instillation BAL and plasma D and E) HS and G and H) were measured. F and I) 12 hours, 2 and 4 days after intratracheal instillation, BAL protein-HS/CS correlations were analyzed by linear regression. Modified from Haeger, et al, 2018 (1).

Table VI-1: ARDS HME Fluid Patient Demographics. Modified from Haeger, et al, 2018 (1).

Study Variable Population

Total Population, n 15 Age (years), mean ± SD 49.1 ± 16.5 Gender, n (%)

Male 11 (73.3%) Female 4 (26.7%) ARDS Risk Factor, n (%)

Sepsis 3 (20.0%) Pneumonia/Aspiration 12 (80.0%) APACHE II, mean ± SD 29.9 ± 17.9

Duration of Ventilation (days), mean ± SD 10.5 ± 8.5 ICU Length of Stay (days), mean ± SD 12.8 ± 8.0 ICU Mortality, n (%) 5 (33.3%)

Hospital Length of Stay (days), mean ± SD 22.7 ± 18.9 Hospital Mortality, n (%) 6 (40.0%) 28 Day Mortality, n (%) 7 (46.7%)

Heparan Sulfate and Chondroitin Sulfate Shed During Intratracheal LPS-Induced Lung Injury are Long and are Accompanied by Shedding of Epithelial Syndecan-1 and Syndecan-4

We sought to identify the sheddase(s) responsible for alveolar HS and CS shedding and to determine if sheddase inhibition reduces lung permeability after intratracheal LPS. Several enzymes and reactive molecules have the ability to degrade HS and CS proteoglycans, either by fragmenting HS or CS into small oligosaccharides or by releasing full-length HS and CS through cleavage of the anchoring proteoglycan (Figure I-3) (31). Our group has previously shown that non-pulmonary sepsis induces heparanase-mediated degradation of endothelial HS into small (6-

8 saccharide) circulating fragments (24, 117). In contrast, after intratracheal LPS instillation, we observed shedding of long HS and CS polysaccharides (≥ β0 saccharides in length) into the airspace (Figure VI-3A), suggesting the presence of sheddases targeting the proteoglycan core

65 protein. Accordingly, we observed an increase in BAL syndecan-1 (7-fold) and syndecan-4

(20.4-fold) ectodomains from mice treated with LPS in comparison to PBS, further suggesting that release of alveolar HS and CS into the BAL after LPS occurs via cleavage of the proteoglycan core protein (Figure VI-3B and C).

Figure VI-2: HME Heparan Sulfate and Chondroitin Sulfate Correlate with HME Protein in ARDS Patients. Protein and HS were measured in HME fluid from ARDS patients. A) HME protein-HS correlation was analyzed by linear regression. B) HME HS in patients ARDS patients with pneumonia/aspiration versus sepsis. C) HME protein-CS correlation was analyzed by linear regression. D) HME CS in patients ARDS patients with pneumonia/aspiration versus sepsis. HME E) HS-protein and F) CS-protein correlations after patients with low protein were excluded. Modified from Haeger, et al, 2018 (1).

Figure VI-3: Increased Airspace Heparan Sulfate and Chondroitin Sulfate After Intratracheal LPS Instillation is Long and is Accompanied by Increased Airspace Syndecan-1 and Syndecan-4. A) PAGE and alcian blue/silver staining was performed on BAL fluid 2 and 4 days after intratracheal LPS instillation. B and C) BAL syndecan-1 and syndecan-4 were measured via western blot from mice 2 days after intratracheal LPS instillation. Modified from Haeger, et al, 2018 (1).

Matrix Metalloproteinases Mediate Alveolar Heparan Sulfate Shedding During Intratracheal

LPS-Induced Lung Injury

While several enzymes and reactive molecules are known to cleave proteoglycan core proteins,

ADAM-17, MMP-7, and ROS quenchable by EC-SOD have been shown to cleave syndecans from lung epithelial cells in vitro or in the airspace in vivo (42, 76, 135). To determine the mechanism of proteoglycan cleavage after intratracheal LPS, we accordingly employed in vivo

67 pharmacologic and transgenic inhibitory approaches to target EC-SOD quenchable ROS,

ADAM-17, and MMPs. We observed no reduction in LPS-induced alveolar HS shedding (or

BAL protein) in alveolar epithelial type II EC-SOD overexpressing mice (Figure VI-4A-C).

Similarly we detected no decrease in BAL HS, CS, or protein in wild-type mice treated with the

ADAM17 inhibitor TAPI-2 (at a dose demonstrated to prevent ADAM17 activity (161)) (Figure

VI-4D-G).

As MMP-2, -7, and -9 are also known to cleave HS and CS proteoglycans (31), we measured MMP-2, -7, and -9 mRNA expression in whole lung tissue, BAL MMP-2, -7, and -9 protein, and BAL MMP-2 and -9 activity by gelatin zymography in mice treated with intratracheal LPS. We observed an increase in whole lung MMP-9 mRNA expression in mice 2 days after intratracheal LPS instillation, while there was no change in whole lung MMP-2 mRNA and undetectable or very low MMP-7 mRNA (in both groups; data not shown) in comparison to PBS (Figure VI-5A). However, we detected an increase in both BAL MMP-2 and -9 protein (by western blot) and activity (by zymography) in mice treated with intratracheal

LPS (Figure VI-5B and C). Similar to whole lung MMP-7 mRNA expression, BAL MMP-7 protein was undetectable in animals treated with intratracheal LPS or PBS (data not shown).

To determine if MMPs, including MMP-2 and -9, mediate alveolar HS and/or CS shedding, we treated mice daily with oral doxycycline, a broad-spectrum MMP inhibitor, from 3 days prior to intratracheal LPS instillation until the animals were harvested 2 days after LPS instillation (Figure VI-6A). In mice treated with intratracheal LPS and oral doxycycline, we observed a 42% decrease in both alveolar HS and syndecan-1, but no decrease in syndecan-4, in comparison to mice treated with intratracheal LPS and oral water as control (Figures VI-6B, D, and E). While a decrease in alveolar CS in mice treated with intratracheal LPS and oral

68 doxycycline in comparison to oral water was not significant, there was a strong trend towards decreased alveolar CS in these mice (Figure VI-6C).

Figure VI-4: EC-SOD Overexpression and Intratracheal TAPI-2 Instillation do not Attenuate the Increased Airspace Heparan Sulfate, Chondroitin Sulfate or Protein after Intratracheal LPS Instillation. A) Wild type and EC-SOD overexpressing mice were treated with intratracheal LPS. B) BAL HS and C) protein were measured 2 days after intratracheal LPS instillation. D) Mice were instilled with 1mg/kg intratracheal TAPI-2 concurrently with LPS, and BAL E) HS, F) CS, and G) protein were measured 2 days after intratracheal instillation. Modified from Haeger, et al, 2018 (1).

Figure VI-5: Lung MMP-9 mRNA Expression, and BAL MMP-2 and MMP-9 Protein and Activity are Increased After Intratracheal LPS Instillation. A) Whole lung MMP-2 and MMP-9 mRNA expression was measured in mice 2 days after intratracheal LPS or PBS instillation. B) BAL MMP-2 and MMP-9 protein and C) gelatinolytic activity were measured by western blot and zymography in mice 12 hours and 2 days after intratracheal LPS or PBS instillation. Modified from Haeger, et al, 2018 (1).

Given our initial findings that epithelial HS degradation increases lung permeability we hypothesized that inhibition of HS shedding by doxycycline would partially restore alveolar barrier function and reduce BAL protein after intratracheal LPS instillation. In mice treated with doxycycline, we observed a nonsignificant trend towards protection against lung permeability

(Figure VI-6F). Additionally, in accordance with our findings that epithelial HS degradation does not increase lung inflammation, we observed no change in BAL neutrophils in mice treated with intratracheal LPS and oral doxycycline in comparison to mice treated with intratracheal

LPS and oral water (Figure VI-6F).

MMP-9 Knockout Mice are not Protected Against Alveolar Heparan Sulfate or Chondroitin

Sulfate Shedding During Intratracheal LPS-Induced Lung Injury

Considering that doxycycline partially attenuated alveolar HS and syndecan-1 shedding, and that MMP-9 protein and activity is abundantly increased in mice treated with intratracheal

LPS, we sought to determine if MMP-9 mediates alveolar HS and syndecan-1 shedding during intratracheal LPS-induced lung injury by utilizing MMP-9 knock out (MMP9ko) mice (Figure

VI-7A). After confirming a loss of BAL MMP-9 in MMP9ko mice (Figure VI-7B), we measured BAL HS, CS, syndecan-1, and syndecan-4 in MMP9ko mice treated with intratracheal

LPS in comparison to wild type mice treated with intratracheal LPS and MMP9ko mice treated with intratracheal PBS as control. In contrast to our hypothesis, we detected a trend towards increased BAL HS and syndecan-1, and increased BAL CS and syndecan-4 in MMP9ko mice treated with intratracheal LPS in comparison to wild type mice treated with LPS (Figure VI-7C-

F). There was also a trend towards increased BAL protein but no change in BAL neutrophils in

LPS-treated MMP9ko mice in comparison to wild type mice (Figure VI-7G and H).

Figure VI-6: Doxycycline Partially Inhibits the Increase in Airspace Heparan Sulfate and Syndecan-1, but not Chondroitin Sulfate, Syndecan-4 or BAL Protein, After Intratracheal LPS Instillation. A) Mice were treated with 70mg/kg oral doxycycline by gavage, or water as control every 24 hours starting 3 days before intratracheal instillation of LPS. Mice were harvested 2 days after intratracheal LPS instillation. BAL B) HS, C) CS, D) syndecan-1, E) syndecan-4, F) protein, G) and neutrophils were measured. Modified from Haeger, et al, 2018 (1).

To investigate this paradoxical effect, we measured BAL MMP-2 protein to determine if there was a compensatory increase of MMP-2 in MMP9ko mice after LPS, as has been observed in other disease models (32, 166). Indeed, we observed a trend towards increased BAL MMP-2 protein in MMP9ko mice treated with intratracheal LPS in comparison to wild type mice treated with intratracheal LPS (Figure VI-7I). These findings indicate that MMPs partially mediate alveolar HS and shedding during intratracheal LPS-induced lung injury and that inhibition of

MMP-9 alone does not protect against alveolar HS or CS shedding, potentially due to compensatory up-regulation of other proteoglycan sheddases. Of note, MMP-2 knockout mice have developmental defects in alveolization, preventing their use to specifically interrogate the role of MMP-2 in intratracheal LPS-induced HS and CS shedding (167).

Summary of Results

• HS and CS are shed into the alveolus during direct lung injury in mice and in ARDS

patients, and the concentration of alveolar HS and CS correlate with permeability.

• MMPs mediate partial shedding of HS and syndecan-1 into the alveolar airspace.

• Broad-spectrum inhibition of MMPs, that partially reduces HS and syndecan-1 shedding,

does not significantly reduce alveolar permeability.

• MMP-2 and MMP-9 expression and activity are increased during direct lung injury;

however, MMP-9 knockout does not protect against, and may (opposite to our

hypothesis) even enhance, HS, CS, syndecan-1, and syndecan-4 shedding.

Figure VI-7: MMP-9 Knockout Mice are not Protected from Alveolar Heparan Sulfate, Chondroitin Sulfate, Syndecan-1, or Syndecan-4 Shedding, or BAL Protein or Neutrophilia, and Exhibit Increased BAL MMP-2 Protein After Intratracheal LPS Instillation. A) MMP-9 knockout (MMP9ko) and wild type mice were treated with intratracheal LPS, and MMP9ko mice were treated with intratracheal PBS as control. B) A loss of BAL MMP-9 was confirmed by western blot in MMP9ko mice in comparison to wild type mice. BAL C) HS, D) syndean-1, E) syndecan-4, F) protein, G) neutrophils, H) and MMP-2 protein were measured. Modified from Haeger, et al, 2018 (1).

Discussion

In Chapter V we demonstrated that epithelial HS, but not CS, contributes to baseline alveolar epithelial barrier function. As such, in this chapter we sought to determine whether alveolar epithelial HS and CS are shed during direct lung injury, identify their sheddase(s), and discern whether epithelial HS shedding contributes to the alveolar epithelial barrier dysfunction that occurs during direct lung injury. During intratracheal LPS-induced lung injury in mice and

ARDS in humans, a disease state characterized by increased lung permeability, we detected shedding of HS and CS, accompanied by syndecan-1 and syndecan-4, into the airspace. These data complement the relative absence of an increase in circulating HS and CS observed in mice treated with intratracheal LPS or in ARDS patients with pneumonia versus non-pulmonary sepsis

(116, 117). These findings indicate that during ARDS, HS and CS shedding occur in an injury- inducible, compartment-specific fashion.

We also determined that MMPs partially mediate HS and syndcan-1 shedding during intratracheal LPS-induced lung injury as broad-spectrum MMP inhibition with doxycycline reduced their shedding. While we originally hypothesized that inhibition of HS shedding may decrease alveolar permeability during lung injury, we only detected a trend towards decreased

BAL protein in mice treated with doxycycline (which, given a small effect size, may prove significant with more animals). Whether complete inhibition of HS shedding will reduce alveolar permeability remains to be determined. We detected an increase in BAL MMP-2 and MMP-9 protein and activity in mice after intratracheal LPS instillation; however, knockout of MMP-9 did not protect against alveolar HS or syndecan-1 shedding, and enhanced CS and syndecan-4 shedding. As the susceptibilities of HS, CS, syndecan-1, and syndecan-4 shedding to doxycycline varied, and knockout of MMP-9 paradoxically may enhance their shedding, there

75 may be multiple partially-redundant HS and CS proteoglycan sheddases induced during direct lung injury.

CHAPTER VII

SOLUBLE ALVEOLAR CHONDROITIN SULFATE ENHANCES LUNG

INJURY, WHILE HEPARAN SULFATE DOES NOT EFFECT LUNG REPAIR8

Introduction

During intratracheal LPS-mediated lung injury, HS and CS are released into the alveolar space in concentrations proportionate to alveolar permeability (Chapter VI). While alveolar HS and CS may therefore serve as biomarkers for direct lung injury, these soluble alveolar GAGs may also be biologically active affecting lung injury induction, propagation and/or epithelial repair. As described in Chapter I, HS and CS, in a length and sulfation sequence-dependent manner, can bind signaling ligands and/or their receptors, thereby modulating several signaling pathways. For example, HS binds growth factor ligands and their receptors, and can either activate or inhibit growth factor signaling under distinct conditions. Cell-surface HS expressed in close proximity to growth factor receptors can act as a scaffolding molecule facilitating ligand-receptor binding, thereby enhancing signaling. In contrast, soluble HS polysaccharides can inhibit signaling by binding and sequestering growth factor ligands away from their receptors (168). This functional distinction between cell-surface HS and soluble HS is not absolute, however, as soluble HS can still activate growth factor signaling in some systems (e.g. the pulmonary endothelium) (23). Indeed, the effect of HS on growth factor signaling is complex and may be both cell-type and signaling protein specific.

CS can similarly affect cellular signaling. CS containing CS-B and CS-E disaccharide units bind many of the same signaling proteins that HS binds, thereby affecting several cellular functions (3, 169). While CS containing little or no CS-B or CS-E units are less likely to bind

8Portions of this chapter were previously published in the American journal of respiratory cell and molecular biology. Reprinted with permission of the American Thoracic Society. Copyright © 2018 American Thoracic Society. (1)

many of these signaling proteins, these CS chains may have other functions. CS-A and CS-C enriched CS chains have been shown to have immunomodulatory effects. Contradictory findings have shown that CS-A and CS-C can both increase and decrease inflammation by modulating

NFκB signaling and T-cell proliferation and activation. In both in vivo and relevant in vitro models of osteoarthritis, psoriasis, and inflammatory bowel disease, CS can improve disease severity by reducing NFκB signaling (170-172). However, in direct contradiction to these findings, CS-A induces TLR-4-dependent NFκB activation and can also enhance T-cell proliferation and activation to a Th1 phenotype (173-176).

Given the numerous and multifaceted (yet largely sulfation pattern-dependent) effects of soluble HS and CS on growth factor signaling and inflammation, we sought to determine the sulfation pattern of alveolar HS and CS in mice treated with intratracheal LPS and in HME fluid from ARDS patients and to determine any effect the soluble GAGs may have on lung injury and repair.

Objectives

• Determine the sulfation pattern of shed alveolar HS and CS during direct lung injury in

mice and in ARDS patients.

• Determine if shed alveolar HS and CS affect the severity of or recovery from lung injury.

Materials and Methods

Materials

In addition to the materials used in Chapters V and VI, heparin octasaccharides and full- length HS were purchased from Galen Laboratory Supplies (North Haven, CT, USA) and were reconstituted in PBS for intratracheal instillation. Human recombinant hepatocyte growth factor

(HGF) was purchased from Sigma and keratinocyte growth factor (KGF) was provided by Dr.

Moosa Mohammadi (New York University, NY, NY, USA). Protamine sulfate salt from salmon and chondroitin sulfate A from bovine trachea were purchased from Sigma-Aldrich and reconstituted in PBS.

Animals

Wild type mice were used as described in Chapters V and VI.

Intratracheal LPS-Induced Lung Injury

Intratracheal LPS-induced lung injury was performed as described in Chapter VI. For experiments in which mice were treated with heparin octasaccharides and full-length HS, mice were treated on day 4 after LPS instillation with 6µg intratracheal HS/heparin and euthanized the following day (day 5 after LPS). For experiments containing protamine, mice were treated with two instillations of intratracheal 10µg protamine (in 30µl PBS), either alone (not after LPS) at the start of the experiment and the following day (day 0 and day 1), or after intratracheal LPS (6 hours and 1 day after, or 3 and 4 days after) and euthanized the day after the second protamine instillation (on day 2 or day 5 of the experiment).

Determination of Heparan Sulfate and Chondroitin Sulfate Sulfation

Sulfation of HS and CS in BAL from mice treated with intratracheal LPS and in HME fluid from ARDS patients were obtained from LC-MS/MS MRM analyses, as described in

Chapter VI.

Surface Plasmon Resonance

Heparin oligosaccharides and selectively de-sulfated heparin were prepared as previously described (23). Binding of HGF and KGF to heparin (a heavily-sulfated HS) was performed by surface plasmon resonance using an immobilized heparin biochip and BIAcore 3000 analyzer

(Biacore AB, Uppsala, Sweden), also as previously described (23). Size and sulfation necessary

79 for heparin to bind HGF and KGF were determined by competition of heparin oligosaccharides and selectively de-sulfated heparin for HGF and KGF preventing their binding to the immobilized heparin biochip. These data were normalized to binding of HGF and KGF to the biochip in the absence of any soluble heparin (100% binding to heparin biochip). The reverse percentage of HGF/KGF-heparin biochip binding was calculated and graphed to demonstrate the percentage of HGF/KGF binding to soluble heparin oligosaccharides and selectively de-sulfated heparin (23).

Lung/Alveolar Permeability and Inflammation Quantification

Lung/alveolar permeability and inflammation were measured as described in Chapter V.

Statistical Analyses

Statistical analyses were performed as described in Chapter V. Significance by t-test and

Mann-Whitney test are noted by *, and by Mann-Whitney alone by #.

Results

Alveolar Heparan Sulfate is Relatively Un-Sulfated in Mice and Patients with Direct Lung Injury

As the functions of HS are largely dependent on both their overall and/or specific sequences of sulfation, we measured the disaccharide sulfation patterns of HS in BAL fluid from mice treated with intratracheal LPS and in HME fluid from ARDS patients. Alveolar HS from mice treated with intratracheal LPS was relatively un-sulfated, with 92% of HS disaccharides being un-sulfated (0S) 12 hours after LPS treatment (Figure VII-1A). Alveolar HS from ARDS patients was similarly un-sulfated (median of 79.5% 0S) indicating that our findings in mice were translationally relevant (Figure VII-1B). Sulfation of alveolar HS modestly increased through the timecourse of intratracheal LPS-induced lung injury, as alveolar un-sulfated HS decreased to 85% 4 days after intratracheal LPS instillation. This was concurrent with an

80 increase in NS2S HS (from 1% to 9% at 12 hours and 4 days, respectively, after intratacheal

LPS), the same sulfation pattern that comprised the majority of alveolar HS believed to line the airspace epithelium during homeostasis (i.e. alveolar HS sulfation after intratracheal heparinase

I/III treatment (Figure V-1E)).

Figure VII-1: Alveolar Heparan Sulfate is Heavily Un-Sulfated in Mice During Intratracheal LPS-Induced Lung Injury and in ARDS Patients. A) Alveolar HS sulfation 12 hours, 2 days, and 4 days after intratracheal LPS. B) HME HS sulfation in ARDS patients. Modified from Haeger, et al, 2018 (1).

Heparan Sulfate Sulfation and Size Characteristics Necessary for Binding to Epithelial

Reparative Growth Factors

While long (≥ 20 saccharides in length), soluble alveolar HS is heavily un-sulfated during lung injury, and therefore likely lacks the sulfation necessary to bind growth factors involved in alveolar epithelial repair. Despite this, we sought to determine the size and sulfation of HS necessary to bind HGF and KGF, growth factors implicated in alveolar epithelial repair, to determine if these growth factors have unexpected affinity to specific (i.e. unsulfated) HS sulfation patterns. Using surface plasmon resonance, we quantified binding of HGF and KGF to soluble heparin (heavily sulfated HS) of different length and sulfation patterns. We determined that heparin of a length between 4-10 saccharides that contains N-sulfation is necessary for KGF

81 binding, as heparin tetrasaccharides did not bind KGF, but decasaccharides did, and only 35% of

KGF bound to N-de-sulfated heparin (Figure VII-2A-B). KGF still bound nearly 100% of heparin oligosaccharides that were 2-O- and 6-O-de-sulfated, indicating that 2-O and 6-O sulfation of HS are not necessary to bind KGF.

Binding of heparin to HGF was dependent on slightly different structural properties.

While there was slight binding of heparin decasaccharides to HGF (13% HGF binding), heparin

18 saccharides in length exhibited nearly complete HGF binding (Figure VII-2C). N-sulfation was also necessary for binding of heparin to HGF, as only 26% of HGF bound to N-de-sulfated heparin (Figure VII-2D). 6-O sulfation of heparin was also relatively important for HGF binding (58% KGF binding to 6-O-desulfated heparin); however, 2-O sulfation was not necessary. Based upon these findings, while alveolar HS is long enough to bind both KGF and

HGF (≥ 20 saccharides in length), it contains very little N- and 6-O sulfation and therefore likely does not contain the sulfation necessary to bind KGF or HGF.

Exogenous Heparin Octosaccharides or Full-Length Heparan Sulfate do not Affect Repair After

Intratracheal LPS-Induced Lung Injury

The lack of sulfation of alveolar HS after intratracheal LPS differs greatly from the N- and 2-O sulfate enriched alveolar HS after intratracheal heparinase I/III (Chapter I). While the reasons why and mechanisms by which alveolar HS sulfation differs between these two models are unknown, we aimed to determine the effect of heavily sulfated exogenous HS/heparin on recovery from lung injury to determine whether the lack of sulfation of alveolar HS during lung injury may aid in facilitating, or oppositely may contribute a delay in, alveolar epithelial repair.

On day 4 after intratracheal LPS we treated mice with intratracheal heparin octasaccharides and full-length heparan sulfate, two heparin/HS structures that differ in size and sulfation pattern, and

82 euthanized mice the following day (day 5 after LPS) (Figure VII-3A). Based upon the previously described surface plasmon resonance experiments, the heparin octasaccharides and full-length HS both contain the sulfation necessary to bind HGF and KGF, while the heparin octasaccharides are not long enough to bind HGF and possibly KGF. We detected no change in

BAL protein or BAL neutrophils in mice treated with either intratracheal heparin octasaccharides or full-length HS in comparison to PBS, indicating that heavily sulfated HS/heparin of moderate and full length do not affect these measures of alveolar permeability or inflammation (Figure

VII-3B and C) despite their ability to bind HGF and KGF.

Figure VII-2: Size and Sulfation of Heparin Necessary to Bind Keratinocyte Growth Factor and Hepatocyte Growth Factor. Using surface plasmon resonance competition assays binding of KGF and HGF to A and C) different length heparin oligosaccharides and B and D) selectively de-sulfated heparin were quantified.

Figure VII-3: Intratracheal Heparin Octasaccharides and Full Length Heparan Sulfate do not Affect Alveolar Permeability or Inflammation on Day 5 After Intratracheal LPS. A) Four days after intratracheal 3mg/kg LPS instillation mice were treated with 6µg intratracheal heparin oligosaccharides and full length HS. On day 5 after LPS instillation (one day after heparin/HS instillation), B) BAL protein and C) BAL total cells were measured.

Alveolar Chondroitin Sulfate is 4-O Sulfated in Mice and Patients with Direct Lung Injury

In contrast to HS, alveolar CS from mice treated with intratracheal LPS was comparatively heavily sulfated (Figure VII-4A). Twelve hours after intratracheal LPS instillation, alveolar CS was enriched in 4-O sulfation (69% of CS disaccharides being 4S).

HME fluid CS from ARDS patients exhibited a similar sulfation pattern and was similarly enriched in 4-O sulfation (median of 78%), again confirming the translational validity of our mouse model (Figure VII-4B). Through the timecourse of intratracheal LPS-induced lung injury, alveolar CS sulfation decreased. A reduction in 4-O sulfation from 69% to 43% and dual

4-O and 6-O sulfation (4S6S) from 9% to 3% occurred concurrently with an increase in un- sulfated CS (from 22 to 53%) from 12 hours to 4 days after intratracheal LPS. Although we

84 were able to determine that alveolar CS is heavily 4-O sulfated after intratracheal LPS, limitations with the mass spectrometry technique used does not allow for the discrimination between IdoA and GlcA, and as such, whether the alveolar 4-O sulfated CS is CS-A enriched or

CS-B (DS) enriched remains unknown.

Figure VII-4: Alveolar Chondroitin Sulfate is Enriched in 4-O Sulfation in Mice During Intratracheal LPS-Induced Lung Injury and in ARDS Patients. A) Alveolar CS sulfation 12 hours, 2 days, and 4 days after intratracheal LPS. B) HME CS sulfation in ARDS patients. Modified from Haeger, et al, 2018 (1).

Intratracheal Protamine does not Affect the Severity of or Recovery from Lung Injury

As alveolar CS is heavily 4-O sulfated after LPS instillation, we hypothesized that the soluble alveolar CS may be biologically active and aimed to determine its effect on the severity of and recovery from intratracheal LPS-mediated lung injury. To inhibit 4-O sulfated CS, we utilized protamine, a positively charged protein that non-specifically binds negatively charged molecules. Although 4-O sulfated CS comprises the majority of all alveolar HS and CS sulfation, suggesting that any effect of protamine on GAGs is largely due to its neutralization of

CS 4-O sulfation, we cannot exclude that protamine may have an effect on other negatively charged molecules (i.e. extracellular DNA or RNA). However, as no specific inhibitor of 4-O sulfated CS exists, we utilized protamine to gain insight into any potential effect 4-O sulfated CS

85 may have on lung injury severity or recovery from lung injury.

As protamine binds sulfated GAGs and is heavily positively charged, intratracheal administration of high dose protamine induces lung injury (data not shown), likely by inducing charge-mediated plasma membrane disruption and neutralizing GAGs within the epithelial glycocalyx. Therefore, before treating mice with intratracheal protamine after LPS, we first determined a dose of protamine that does not cause lung injury itself (Figure VII-5A).

Treatment with 10µg intratracheal protamine on two consecutive days (day 0 and day 1) did not increase lung permeability or inflammation in naïve mice, and therefore, was the dose chosen for additional experiments (Figure VII-5B and C).

To determine the effect of protamine on the severity of lung injury we treated mice with

10µg intratracheal protamine both 6 hours and 1 day following intratracheal LPS instillation and euthanized mice the following day (2 days after intratracheal LPS) (Figure VII-5D). We observed no difference in BAL protein or total cells in mice treated with intratracheal protamine, in comparison to PBS as control (Figure VII-5E and F). To determine the effect of protamine on recovery from lung injury, we performed a similar experiment and treated mice with 10µg intratracheal protamine both 3 days and 4 days following intratracheal LPS instillation and euthanized the mice on day 5 of the experiment, when resolution of alveolar inflammation and permeability is occurring (Figure VII-5G). We again observed no difference in BAL protein or total cells in mice treated with intratracheal protamine, in comparison to PBS as control (Figure

VII-5E and F). While these data indicate that 4-O sulfated CS does not affect the severity of or recovery from intratracheal LPS-mediated lung injury, we are unable to confirm that the dose of protamine used effectively inhibits CS binding to other mediators in vivo. In addition, any effect of CS on lung injury or repair after injury may be independent of 4-O sulfation.

Exogenous CS-A Induces Mild Inflammation and Enhances Intratracheal LPS-Induced Lung

Injury

Given the limitations of using protamine to inhibit 4-O sulfated CS, we performed complementary experiments to determine whether supplementation with exogenous 4-O sulfated

CS (CS-A) affects the severity of lung injury. We treated mice with 50µg/g intratracheal CS-A both 6 hours and 1 day after intratracheal LPS instillation (Figure VII-6A). In contrast to the lack of effect of protamine on LPS-induced lung injury, supplementation with exogenous CS-A enhanced alveolar permeability and inflammation (total cells and neutrophils) on day 2 after LPS instillation, indicating that CS-A may induce alveolar inflammation or enhance LPS-induced inflammation (Figure VII-6B-D).

To determine whether CS-A alone induces inflammation, thereby acting as a DAMP or chemokine, we treated mice once with the same supraphysiologic dose used in the previous experiment (50µg/g), and a dose similar to the total amount of lavagable CS after intratracheal

LPS instillation (10µg) (Figure VII-6E). Mice treated with supraphysiologic CS-A (50µg/g) exhibited a modest increase in BAL total cells (significance by Mann-Whitney) but no increase in BAL protein, in comparison to mice treated with PBS as control (Figure VII-6F and G). In contrast, there was no difference in BAL protein or total cells in mice treated with 10µg CS-A, in comparison to PBS (Figure VII-6F and G). These findings indicate that high concentrations of

CS-A may serve as a mild chemokine, inducing modest inflammation itself; however, whether enhancement of LPS-induced inflammation by supraphysiologic CS-A is due to modulation of other inflammatory mediators and whether physiologic concentrations of CS-A also modulate

L PS-induced inflammation remain unknown.

Figure VII-5: Intratracheal Protamine does not Affect Alveolar Permeability or Inflammation on Day 2 or Day 5 After Intratracheal LPS. A) In determining a dose of protamine the does not induce alveolar hyper-permeability or inflammation, mice were treated with 10µg intratracheal protamine on day 0 and day 1. After euthanasia on day 2, B) BAL protein and C) BAL total cells were measured. D) To determine the effect of protamine on lung injury severity, mice were treated with 10µg intratracheal protamine 6 hours and again 1 day after instillation of 3mg/kg intratracheal LPS. After euthanasia on day 2, E) BAL protein and F) BAL total cells were measured. G) To determine the effect of protamine on recovery from lung injury, mice were treated with 10µg intratracheal protamine on day 3 and day 4 after instillation of 3mg/kg intratracheal LPS. After euthanasia on day 5, H) BAL protein and I) BAL total cells were measured.

Figure VII-6: High Dose Intratracheal Chondroitin Sulfate-A Enhances LPS-Induced Permeability and Inflammation and Induces Modest Inflammation in the Absence of a Prior Stimulation. A) To determine the effect of 4-O sulfated CS on lung injury, mice were supplemented with 50µg/g CS-A 6 hours and again 1 day after 3mg/kg intratracheal LPS. After euthanasia on day 2, B) BAL protein, C) BAL total cells, and D) BAL neutrophils were measured. E) To determine if intratracheal CS-A alone, at physiologic or supraphysiologic doses, induces inflammation and increased permeability, mice were treated with 10µg or 50µg/g CS-A. After euthanasia 24 hours later, F) BAL protein and G) BAL total cells were measured.

Summary of Results

• Shed alveolar HS is heavily un-sulfated during direct lung injury in mice and in ARDS

patients.

• Exogenous sulfated HS, that is able to bind HGF and KGF, may not affect recovery from

lung injury.

• Shed alveolar CS is extensively 4-O sulfated during direct lung injury in mice and in

ARDS patients.

• Alveolar CS-A (a type of 4-O sulfated CS), may induce alveolar inflammation alone, and

enhances LPS-induced lung inflammation and permeability.

Discussion

In this chapter we described the sulfation pattern and potential effects of soluble alveolar

HS and CS on alveolar injury and subsequent repair during intratracheal LPS-induced lung injury. We determined that during direct lung injury in mice and ARDS in human patients, soluble alveolar HS is heavily un-sulfated, while alveolar CS is enriched in 4-O sulfation. As much of the functions of HS and CS arise from their sulfation, these findings indicate that alveolar HS may have little biologic activity whereas alveolar CS may affect the severity of or recovery from lung injury.

As alveolar HS sulfation differed after intratracheal heparinase I/III (NS2S HS) versus

LPS (0S HS), we aimed to determine whether the lack of sulfation of alveolar HS after intratracheal LPS was beneficial, or oppositely delayed alveolar repair. However, treatment with heavily sulfated heparin octasaccharides and full-length HS did not affect alveolar permeability nor inflammation at a timepoint during active alveolar repair. While we cannot exclude that a different concentration, duration, or time of heparin/HS treatment may provide a different result,

90 these data indicate that heavily sulfated HS does not affect direct lung injury or subsequent repair. As such, a lack of sulfation of alveolar HS after intratracheal LPS may be neither beneficial nor harmful for alveolar repair following direct lung injury.

In contrast to the absence of an effect of alveolar HS, we determined that supraphysiologic concentrations of alveolar 4-O sulfated CS (CS-A) can induce mild alveolar inflammation and enhance LPS-induced inflammation. Despite these findings, whether the alveolar 4-O sulfated CS present after intratracheal LPS is composed of CS-A and/or CS-B units and any effect of the endogenous 4-O sulfated CS on LPS-induced inflammation remains unknown. To better determine the effect of endogenous alveolar 4-O sulfated CS on LPS- induced lung injury, removal of alveolar CS 4-O sulfation by intratracheal administration of a 4-

O endosulfatase or depletion of CS 4-O sulfation by genetic knockout of CS a 4-O sulfotransferase could be utilized. Recent identification and production of a recombinant marine bacterial CS 4-O endosulfatase makes removal of alveolar CS 4-O sulfation feasible; however, as genetic knockdown of only one of the three CS 4-O sulfotransferases (C4ST1-3) and one CS-

B/DS 4-O sulfotransferase exist (C4ST-1 gene trap), and it is unknown which CS/DS 4-O sulfotransferase mediates initial 4-O sulfation of the lung injury-induced alveolar CS, depletion of CS 4-O sulfation may prove difficult (51, 177).

CHAPTER VIII

CONCLUSIONS AND FUTURE DIRECTIONS9

Perspective

The alveolar epithelial glycocalyx is a negatively charged carbohydrate-rich layer lining the epithelial cell surface. Despite its discovery over 50 years ago, little is known about its composition, function, and fate during injury and disease (75). In contrast, the pulmonary endothelial glycocalyx is known to be enriched in HS, which critically contributes to the function of the endothelial glycocalyx to maintain endothelial barrier integrity and inhibit endothelial- leukocyte adhesion. Furthermore, during indirect lung injury heparanase mediates endothelial glycocalyx-HS degradation thereby increasing endothelial permeability and facilitating neutrophil-endothelial adhesion (23, 24). While lung epithelial cells are known to express the

HS/CS proteoglycans syndecan-1 and syndecan-4, and shedding of epithelial syndecan-1 and syndecan-4 occurs in response to pro-inflammatory stimuli in vitro and in vivo (76, 135), the composition and function of GAGs on the alveolar epithelial cell surface is poorly understood.

As such, in this thesis we aimed to identify the presence and function of HS and CS within the healthy adult alveolar epithelium, and determine their fate and function during direct lung injury and subsequent alveolar epithelial repair.

Summary of Findings

In Chapters V-VII we determined that alveolar epithelial HS is N- and 2-O sulfated and contributes to the alveolar epithelial barrier to protein, that alveolar shedding of HS (partially

MMP-mediated) and CS occurs and correlates with permeability in a mouse model of direct lung injury and in ARDS patients, and that shed alveolar HS is un-sulfated and likely does not affect

9Portions of this chapter were previously published in the American journal of respiratory cell and molecular biology. Reprinted with permission of the American Thoracic Society. Copyright © 2018 American Thoracic Society. (1, 2)

92 lung injury or subsequent repair, while shed alveolar CS is heavily 4-O sulfated and may contribute to alveolar inflammation during direct lung injury (Figure VIII-1). A brief detailed summary of the findings from each chapter (Chapters V-VII) are described below.

Figure VIII-1: Thesis Findings. Modified from Haeger, et al, 2016 (2).

Chapter V

In Chapter V, we used immunofluorescence to identify the expression of HS and CS within the alveolar epithelium. By intratracheal instillation of selective HS or CS lyases, we determined that alveolar epithelial HS, but not CS, contributes to the alveolar epithelial barrier to protein. We determined that this increase in alveolar permeability was not mediated by HS fragments generated by heparinase I/III, and that epithelial HS degradation does not alter the expression of tight junction proteins ZO-1 and claudin-18. Furthermore, we discovered that the alveolar epithelial glycocalyx HS is enriched in N- and 2-O sulfation, which provides additional insight into what proteins alveolar epithelial glycocalyx-HS may bind.

Chapter VI

In Chapter VI, we determined the fate of alveolar epithelial HS and CS during intratracheal LPS-induced direct lung injury. We detected an increase in alveolar HS and CS in mice treated with intratracheal LPS and in patients mechanically ventilated for ARDS. As doxycycline partially reduced alveolar shedding of HS and syndecan-1 after intratracheal LPS,

MMPs likely contribute to the HS-bound syndecan-1 shedding during direct lung injury. An increase in MMP-2 and MMP-9 protein and activity were observed after intratracheal LPS; however, knockout of MMP-9 did not protect against, but rather nearly significantly exacerbated,

HS and syndecan-1 shedding. MMP-9 mice exhibited a trend towards increased MMP-2 protein, indicating that MMP-2 may instead cleave syndecan-1 and its bound HS.

Concentrations of both alveolar HS and CS correlated with alveolar permeability in mice treated with intratracheal LPS and in ARDS patients. While partial reduction of HS shedding by doxycycline did not significantly inhibit alveolar permeability, complete abrogation of HS shedding may lead to a reduction in alveolar permeability. A trend towards increased alveolar

HS in ARDS patients with direct versus indirect lung injury was detected, indicating that HME fluid (alveolar) HS may serve as a biomarker for ARDS due to direct lung injury.

Chapter VII

In Chapter VII we demonstrated that the alveolar HS shed during direct lung injury is un- sulfated, while shed alveolar CS is enriched in 4-O sulfation. Given its lack of sulfation, alveolar

HS likely does not modulate inflammation or damage during direct lung injury, or subsequent repair. As soluble HS can bind and sequester growth factor ligands from their receptors, we hypothesized that the absence of sulfation within shed HS enables a permissive alveolar microenvironment for growth factor signaling, potentially imparting evolutionary benefit.

However, intratracheal administration of heavily sulfated full-length HS did not affect recovery from lung injury, despite its ability to bind KGF and HGF.

In contrast to the lack of activity by HS, alveolar 4-O sulfated CS may modulate inflammation during lung injury. Supraphysiologic doses of intratracheal 4-O sulfated CS (CS-

A) induced mild alveolar inflammation itself and enhanced LPS-induced alveolar inflammation.

However, treatment with intratracheal protamine to inhibit sulfated CS did not affect alveolar inflammation following intratracheal LPS, suggesting that CS may modulate LPS-induced inflammation in a manner unaffected by 4-O sulfation (or negative charge). This observation requires future investigation.

Future Directions

Mechanism by which Epithelial Heparan Sulfate Contributes to Alveolar Barrier

Function

In Chapter V, we demonstrated that epithelial HS contributes to the alveolar barrier to protein as epithelial HS degradation increased alveolar permeability. While we determined that heparinase I/III-generated fragments are not responsible for the observed increase in alveolar permeability, and that there were no changes in ZO-1 or claudin-18 protein expression after heparinase I/III instillation, we were unable to affirmatively determine the mechanism by which epithelial HS contributes to the alveolar barrier. As described in Chapter I, HS, in a sulfate sequence-independent manner, can form a rigid gel-like layer that opposes the flux of proteins, acting as a molecular sieve (2). Unfortunately, it is nearly impossible to prove that alveolar epithelial HS, independent of its interactions with or effects on other mediators, is a structural component of the alveolar epithelial barrier. Although cells express HS and CS proteoglycans in vitro, the thickness of their glycocalyx is dramatically reduced in comparison to the thickness of

95 the in vivo glycocalyx (178). As such, the composition and function of the epithelial glycocalyx in vitro may greatly differ from the epithelial glycocalyx in vivo, and, therefore, any studies to determine the mechanism by which HS contributes to the alveolar epithelial barrier in vitro may invalid or inaccurate. While in vivo neutralization of the negative charge of sulfated alveolar epithelial HS is feasible by intratracheal instillation of a HS-binding positively charged protein

(i.e. protamine), there are no proteins that specifically bind HS. Furthermore, reduction of the negative charge of HS by removal of 6-O sulfation can be accomplished by treatment with recombinant 6-O endosulfatases; however, alveolar epithelial HS is enriched in N- and 2-O sulfation, and no N- or 2-O endosulfatases are known to exist (179).

Alternatively, alveolar epithelial HS may form and contribute to the robustness of the aqueous hypophase that lies beneath the layer of pulmonary surfactant and lines the luminal surface of the alveolar epithelium (similar to the importance of pulmonary endothelial HS to formation of the ESL). It is thought that the aqueous hypophase is critical for surfactant metabolism and function, which is further evidenced by mathematical modeling studies that show decreased aqueous hypophase thickness increases the time necessary for surfactant spreading and equilibrium formation (180, 181). Furthermore, measurement of the aqueous hypophase thickness is feasible and has been done in normal rat lungs using low-temperature scanning electron microscopy (182). As such, low-temperature scanning electron microscopy could be performed on lungs from mice treated with intratracheal heparinase I/III to determine if epithelial HS degradation decreases aqueous hypophase thickness. Lung compliance and quantification of BAL surfactant composition could also be performed to determine if intratracheal heparinase I/III decreases lung compliance and alters the composition of alveolar surfactant.

Differences in Sulfation of Alveolar Heparan Sulfate

In Chapter V, we performed mass spectrometry analyses of HS fragments liberated after intratracheal heparinase I/III, demonstrating that alveolar epithelial HS is enriched in N- and 2-O sulfation. In contrast, after intratracheal LPS instillation (Chapter VI), we detected heavily un- sulfated HS within the alveolar airspace. As intratracheal LPS induces alveolar epithelial, basement membrane, and ECM damage, a portion of HS within the alveolar airspace after intratracheal LPS may originate from additional sources of HS shedding other than the alveolar epithelium. This is evidenced by an increase in the average maximal alveolar HS detected after intratracheal LPS (924 ng/ml (Figure VI-1C)) in comparison to intratracheal heparinase I/III instillation (597 ng/ml (Figure V-1C)), indicating that approximately 35% of alveolar HS may originate from the ECM and basement membrane, assuming equal release of alveolar epithelial

HS between the two models. Despite additional sources of alveolar HS, the sulfation of alveolar

HS in mice treated with LPS (92% 0S 12 hours after instillation (Figure VII-1A)) is still markedly more reduced than if the 35% of alveolar HS derived from ECM and basement membrane after LPS were completely un-sulfated.

Given the discrepancy in observed versus calculated alveolar HS sulfation after intratracheal LPS, active de-sulfation, remodeling, or sulfation-specific loss of HS likely occurs during LPS-induced lung injury. While extracellular 6-O endosulfatases that remove HS 6-O sulfation exist, no 2-O or N- endosulfatases have been described (179). As epithelial HS is thought to be heavily N- and 2-O sulfated, according to our findings (Figure VI-1C), nearly complete de-sulfation of alveolar HS during LPS-induced lung injury is unlikely to be responsible by a 6-O endosulfatase alone.

Alternatively, selective internalization or binding of heavily sulfated alveolar HS to inflammatory or epithelial cells within the alveolus, thereby enriching the alveolar airspace with lavagable un-sulfated HS, could occur. As sulfation is necessary for HS to bind other mediators, selective binding of sulfated HS to cell-surface proteins, and subsequent internalization, is theoretically feasible. To determine if heavily-sulfated HS is bound to the surfaces of cells within the alveolar airspace after intratracheal LPS, a BAL with a strong ionic solvent could be performed to release ionically bound HS from alveolar cell surfaces. Furthermore, while it would be expected that similarly-sulfated HS would bind to the surfaces of alveolar cells to the same extent after intratracheal heparinase I/III as after LPS instillation, HS fragments generated after intratracheal heparinase I/III are likely short oligosaccharides and may not be long enough to bind to cell-surface proteins.

Evolutionary Benefit of Redundant Heparan Sulfate Sheddases during Lung Injury

While MMP-2 may mediate the majority of alveolar HS shedding during intratracheal

LPS-induced lung injury, the susceptibilities of HS, CS, syndecan-1, and syndecan-4 shedding to the broad-spectrum MMP inhibitor, doxycycline, all varied. These findings suggest that shedding of different HS/CS proteoglycans may be mediated by different sheddases, but also that multiple redundant HS/CS proteoglycan sheddases may exist. The existence of redundant mechanisms for epithelial shedding of HS and CS suggests an evolutionary beneficial of such shedding during lung infection and injury. Many viruses and bacteria use HS as a receptor to adhere and infect host cells, including Staphylococcus aureus and Streptococcus pneumoniae, common causes of hospital and community acquired pneumonia (16, 183-185). Additionally, soluble HS can act as a decoy receptor to bind and inhibit the adherence of S. aureus and S. pneumoniae to lung epithelial cells. Thus, epithelial HS shedding during infection may be

98 beneficial to inhibit the propagation of infection by both depleting bacterial receptors on host cells and by increasing soluble HS decoy receptors. This evolutionary importance would support the presence of redundant, insult-specific pulmonary epithelial sheddases. To determine whether alveolar epithelial HS serves as a host receptor for pathogenic bacteria, intratracheal instillation of heparinase I/III prior to administration of intratracheal bacteria could be performed and quantification of lung bacterial burden could be measured. Furthermore, to determine if alveolar

HS serves as a decoy receptor for bacteria, the lung bacterial burden of mice co-administered with intratracheal HS and bacteria could be performed and compared to the lung bacterial burden of mice treated with intratracheal instillation of bacteria alone.

Alveolar Heparan Sulfate and Chondroitin Sulfate as Biomarkers of Direct Lung Injury

In Chapter VI, we detected alveolar shedding of HS, CS, syndecan-1 and syndecan-4 in mice after intratracheal LPS instillation and in ARDS patients. We also detected a modest trend towards increased alveolar HS in patients with ARDS due to direct versus indirect lung injury, indicating that HME HS could potentially be used as a biomarker for direct lung injury-mediated

ARDS. While GAG mass spectrometry is expensive and time consuming, inexpensive and rapid identification of sulfated GAGs can be performed using colorimetric dimethylmethylene blue

(DMMB) detection. However, as alveolar HS is heavily un-sulfated in ARDS patients, and

GAG sulfation is necessary for DMMB detection, DMMB detection may not be a useful HS biomarker assay. To determine if DMMB detection may be a useful biomarker assay, we calculated the total sulfated HS and CS, as measured in Chapter VI, in HME fluid from patients with ARDS due to direct lung injury (pneumonia/aspiration) versus indirect lung injury (sepsis).

We detected a weak, if any, trend towards increased HME total sulfated HS and CS in patients with direct versus indirect lung injury (Figure VIII-2). Given the limited number of HME

samples, especially from patients with sepsis-induced ARDS, quantification of HME total

sulfated HS and CS in more patients is necessary to conclusively determine whether DMMB

detection may be a feasible biomarker assay to distinguish ARDS driven by direct versus indirect

lung injury.

Figure VIII-2: HME Total Sulfated Heparan Sulfate and Chondroitin Sulfate in ARDS Patients. HME total sulfated HS and CS in ARDS patients with pneumonia/aspiration versus sepsis.

Determination of Alveolar Chondroitin Sulfate Subunit Composition and Function

In Chapter VII, we determined that alveolar CS is heavily 4-O sulfated in mice treated

with intratracheal LPS and in ARDS patients. Additionally, we discovered that supraphysiologic

concentrations of CS-A induces mild inflammation on its own and enhances LPS-induced

inflammation. While these findings indicate that, during direct lung injury, alveolar CS-A may

potentiate inflammation, the percentage of alveolar 4-O sulfated CS that is comprised of CS-A

versus CS-B/DS units remains unknown. As such, the function of alveolar CS based upon our

findings from treatment with exogenous CS-A cannot be firmly concluded. To determine the

relative abundance of IdoA- versus GlcA-containing CS, thereby approximating the amount of

alveolar CS-B versus CS-A, fragment sizes of alveolar CS after degradation with bacterial

chondroitinases that selectively cleave IdoA-GalNAc or GlcA-GalNAc bonds can be measured,

as has been described previously (186, 187). Once the relative amount of alveolar CS-A and CS-

100

B are determined, additional experiments can be performed targeting CS-A/CS-B biosynthesis and/or sulfation to determine the function of endogenous alveolar CS on lung injury.

101

REFERENCES

1. Haeger SM, Liu X, Han X, McNeil JB, Oshima K, McMurtry SA, Yang Y, Ouyang Y, Zhang F, Nozik-Grayck E, Zemans RL, Tuder RM, Bastarache JA, Linhardt RJ, Schmidt EP. Epithelial Heparan Sulfate Contributes to Alveolar Barrier Function and is Shed During Lung Injury. Am J Respir Cell Mol Biol 2018.

2. Haeger SM, Yang Y, Schmidt EP. Heparan Sulfate in the Developing, Healthy, and Injured Lung. Am J Respir Cell Mol Biol 2016; 55: 5-11.

3. Esko JD, Kimata K, Lindahl U. Proteoglycans and Sulfated Glycosaminoglycans. In: nd, Varki A, Cummings RD, Esko JD, Freeze HH, Stanley P, Bertozzi CR, Hart GW, Etzler ME, editors. Essentials of Glycobiology. Cold Spring Harbor (NY); 2009.

4. Sasarman F, Maftei C, Campeau PM, Brunel-Guitton C, Mitchell GA, Allard P. Biosynthesis of glycosaminoglycans: associated disorders and biochemical tests. J Inherit Metab Dis 2016; 39: 173-188.

5. Ogawa H, Hatano S, Sugiura N, Nagai N, Sato T, Shimizu K, Kimata K, Narimatsu H, Watanabe H. Chondroitin sulfate synthase-2 is necessary for chain extension of chondroitin sulfate but not critical for skeletal development. PLoS One 2012; 7: e43806.

6. Freeze HH, Elbein AD. Glycosylation Precursors. In: Varki A, Cummings RD, Esko JD, Freeze HH, Stanley P, Bertozzi CR, Hart GW, Etzler ME, editors. Essentials of Glycobiology, 2nd ed. Cold Spring Harbor (NY); 2009.

7. Klaassen CD, Boles JW. Sulfation and sulfotransferases 5: the importance of 3'- phosphoadenosine 5'-phosphosulfate (PAPS) in the regulation of sulfation. FASEB J 1997; 11: 404-418.

8. Prydz K. Determinants of Glycosaminoglycan (GAG) Structure. Biomolecules 2015; 5: 2003- 2022.

9. Sarrazin S, Lamanna WC, Esko JD. Heparan sulfate proteoglycans. Cold Spring Harb Perspect Biol 2011; 3.

10. Beeson JG, Chai W, Rogerson SJ, Lawson AM, Brown GV. Inhibition of binding of malaria- infected erythrocytes by a tetradecasaccharide fraction from chondroitin sulfate A. Infect Immun 1998; 66: 3397-3402.

102

11. Rogers CJ, Clark PM, Tully SE, Abrol R, Garcia KC, Goddard WA, 3rd, Hsieh-Wilson LC. Elucidating glycosaminoglycan-protein-protein interactions using carbohydrate microarray and computational approaches. Proceedings of the National Academy of Sciences of the United States of America 2011; 108: 9747-9752.

12. Shipp EL, Hsieh-Wilson LC. Profiling the sulfation specificities of glycosaminoglycan interactions with growth factors and chemotactic proteins using microarrays. Chem Biol 2007; 14: 195-208.

13. Luo J, Kato M, Wang H, Bernfield M, Bischoff J. Heparan sulfate and chondroitin sulfate proteoglycans inhibit E-selectin binding to endothelial cells. J Cell Biochem 2001; 80: 522-531.

14. Gill S, Wight TN, Frevert CW. Proteoglycans: key regulators of pulmonary inflammation and the innate immune response to lung infection. Anat Rec (Hoboken) 2010; 293: 968- 981.

15. Asada M, Shinomiya M, Suzuki M, Honda E, Sugimoto R, Ikekita M, Imamura T. Glycosaminoglycan affinity of the complete fibroblast growth factor family. Biochimica et biophysica acta 2009; 1790: 40-48.

16. Bernfield M, Gotte M, Park PW, Reizes O, Fitzgerald ML, Lincecum J, Zako M. Functions of cell surface heparan sulfate proteoglycans. Annu Rev Biochem 1999; 68: 729-777.

17. Herndon ME, Stipp CS, Lander AD. Interactions of neural glycosaminoglycans and proteoglycans with protein ligands: assessment of selectivity, heterogeneity and the participation of core proteins in binding. Glycobiology 1999; 9: 143-155.

18. Kuschert GS, Coulin F, Power CA, Proudfoot AE, Hubbard RE, Hoogewerf AJ, Wells TN. Glycosaminoglycans interact selectively with chemokines and modulate receptor binding and cellular responses. Biochemistry 1999; 38: 12959-12968.

19. Tully SE, Rawat M, Hsieh-Wilson LC. Discovery of a TNF-alpha antagonist using chondroitin sulfate microarrays. J Am Chem Soc 2006; 128: 7740-7741.

20. Volpi N. Dermatan sulfate: Recent structural and activity data. Carbohyd Polym 2010; 82: 233-239.

103

21. Levy-Adam F, Feld S, Cohen-Kaplan V, Shteingauz A, Gross M, Arvatz G, Naroditsky I, Ilan N, Doweck I, Vlodavsky I. Heparanase 2 interacts with heparan sulfate with high affinity and inhibits heparanase activity. J Biol Chem 2010; 285: 28010-28019.

22. Peterson SB, Liu J. Multi-faceted substrate specificity of heparanase. Matrix Biol 2013; 32: 223-227.

23. Yang Y, Haeger SM, Suflita MA, Zhang F, Dailey KL, Colbert JF, Ford JA, Picon MA, Stearman RS, Lin L, Liu X, Han X, Linhardt RJ, Schmidt EP. Fibroblast Growth Factor Signaling Mediates Pulmonary Endothelial Glycocalyx Reconstitution. Am J Respir Cell Mol Biol 2017; 56: 727-737.

24. Schmidt EP, Yang Y, Janssen WJ, Gandjeva A, Perez MJ, Barthel L, Zemans RL, Bowman JC, Koyanagi DE, Yunt ZX, Smith LP, Cheng SS, Overdier KH, Thompson KR, Geraci MW, Douglas IS, Pearse DB, Tuder RM. The pulmonary endothelial glycocalyx regulates neutrophil adhesion and lung injury during experimental sepsis. Nat Med 2012; 18: 1217-1223.

25. Goodall KJ, Poon IK, Phipps S, Hulett MD. Soluble heparan sulfate fragments generated by heparanase trigger the release of pro-inflammatory cytokines through TLR-4. PLoS One 2014; 9: e109596.

26. Kaneiwa T, Mizumoto S, Sugahara K, Yamada S. Identification of human hyaluronidase-4 as a novel chondroitin sulfate hydrolase that preferentially cleaves the galactosaminidic linkage in the trisulfated tetrasaccharide sequence. Glycobiology 2010; 20: 300-309.

27. Honda T, Kaneiwa T, Mizumoto S, Sugahara K, Yamada S. Hyaluronidases Have Strong Hydrolytic Activity toward Chondroitin 4-Sulfate Comparable to that for Hyaluronan. Biomolecules 2012; 2: 549-563.

28. Ra HJ, Parks WC. Control of matrix metalloproteinase catalytic activity. Matrix Biol 2007; 26: 587-596.

29. Nagase H, Visse R, Murphy G. Structure and function of matrix metalloproteinases and TIMPs. Cardiovasc Res 2006; 69: 562-573.

30. Arpino V, Brock M, Gill SE. The role of TIMPs in regulation of extracellular matrix proteolysis. Matrix Biol 2015; 44-46: 247-254.

104

31. Manon-Jensen T, Itoh Y, Couchman JR. Proteoglycans in health and disease: the multiple roles of syndecan shedding. FEBS J 2010; 277: 3876-3889.

32. Longo GM, Xiong W, Greiner TC, Zhao Y, Fiotti N, Baxter BT. Matrix metalloproteinases 2 and 9 work in concert to produce aortic aneurysms. J Clin Invest 2002; 110: 625-632.

33. Edwards DR, Handsley MM, Pennington CJ. The ADAM metalloproteinases. Mol Aspects Med 2008; 29: 258-289.

34. Apte SS. A disintegrin-like and metalloprotease (reprolysin type) with thrombospondin type 1 motifs: the ADAMTS family. Int J Biochem Cell Biol 2004; 36: 981-985.

35. White JM. ADAMs: modulators of cell-cell and cell-matrix interactions. Curr Opin Cell Biol 2003; 15: 598-606.

36. Stanton H, Melrose J, Little CB, Fosang AJ. Proteoglycan degradation by the ADAMTS family of proteinases. Biochim Biophys Acta 2011; 1812: 1616-1629.

37. Moseley R, Waddington RJ, Embery G. Degradation of glycosaminoglycans by reactive oxygen species derived from stimulated polymorphonuclear leukocytes. Biochim Biophys Acta 1997; 1362: 221-231.

38. Panasyuk A, Frati E, Ribault D, Mitrovic D. Effect of reactive oxygen species on the biosynthesis and structure of newly synthesized proteoglycans. Free Radic Biol Med 1994; 16: 157-167.

39. Bayir H. Reactive oxygen species. Crit Care Med 2005; 33: S498-501.

40. Moseley R, Waddington R, Evans P, Halliwell B, Embery G. The chemical modification of glycosaminoglycan structure by oxygen-derived species in vitro. Biochim Biophys Acta 1995; 1244: 245-252.

41. Kliment CR, Englert JM, Gochuico BR, Yu G, Kaminski N, Rosas I, Oury TD. Oxidative stress alters syndecan-1 distribution in lungs with pulmonary fibrosis. J Biol Chem 2009; 284: 3537-3545.

105

42. Kliment CR, Tobolewski JM, Manni ML, Tan RJ, Enghild J, Oury TD. Extracellular superoxide dismutase protects against matrix degradation of heparan sulfate in the lung. Antioxid Redox Signal 2008; 10: 261-268.

43. Horwitz AL, Crystal RC. Content and synthesis of glycosaminoglycans in the developing lung. J Clin Invest 1975; 56: 1312-1318.

44. Schmid K, Grundboeck-Jusco J, Kimura A, Tschopp FA, Zollinger R, Binette JP, Lewis W, Hayashi S. The distribution of the glycosaminoglycans in the anatomic components of the lung and the changes in concentration of these macromolecules during development and aging. Biochim Biophys Acta 1982; 716: 178-187.

45. Vaccaro CA, Brody JS. Ultrastructural localization and characterization of proteoglycans in the pulmonary alveolus. Am Rev Respir Dis 1979; 120: 901-910.

46. Kim CW, Goldberger OA, Gallo RL, Bernfield M. Members of the syndecan family of heparan sulfate proteoglycans are expressed in distinct cell-, tissue-, and development- specific patterns. Mol Biol Cell 1994; 5: 797-805.

47. Filmus J, Selleck SB. Glypicans: proteoglycans with a surprise. J Clin Invest 2001; 108: 497- 501.

48. Li Y, Miao L, Cai H, Ding J, Xiao Y, Yang J, Zhang D. The overexpression of glypican-5 promotes cancer cell migration and is associated with shorter overall survival in non- small cell lung cancer. Oncol Lett 2013; 6: 1565-1572.

49. Yu X, Li Y, Chen SW, Shi Y, Xu F. Differential expression of glypican-3 (GPC3) in lung squamous cell carcinoma and lung adenocarcinoma and its clinical significance. Genet Mol Res 2015; 14: 10185-10192.

50. Wilson DG, Phamluong K, Lin WY, Barck K, Carano RA, Diehl L, Peterson AS, Martin F, Solloway MJ. Chondroitin sulfate synthase 1 (Chsy1) is required for bone development and digit patterning. Dev Biol 2012; 363: 413-425.

51. Kluppel M, Wight TN, Chan C, Hinek A, Wrana JL. Maintenance of chondroitin sulfation balance by chondroitin-4-sulfotransferase 1 is required for chondrocyte development and growth factor signaling during cartilage morphogenesis. Development 2005; 132: 3989- 4003.

106

52. Weibel ER, Gomez DM. Architecture of the human lung. Use of quantitative methods establishes fundamental relations between size and number of lung structures. Science 1962; 137: 577-585.

53. Rackley CR, Stripp BR. Building and maintaining the epithelium of the lung. J Clin Invest 2012; 122: 2724-2730.

54. Murray JF. The structure and function of the lung. Int J Tuberc Lung Dis 2010; 14: 391-396.

55. Hsia CC, Hyde DM, Weibel ER. Lung Structure and the Intrinsic Challenges of Gas Exchange. Compr Physiol 2016; 6: 827-895.

56. Weibel ER. Lung morphometry: the link between structure and function. Cell Tissue Res 2017; 367: 413-426.

57. Townsley MI. Structure and composition of pulmonary arteries, capillaries, and veins. Compr Physiol 2012; 2: 675-709.

58. Gebb S, Stevens T. On lung endothelial cell heterogeneity. Microvasc Res 2004; 68: 1-12.

59. Yang Y, Schmidt EP. The endothelial glycocalyx: an important regulator of the pulmonary vascular barrier. Tissue Barriers 2013; 1.

60. Dull RO, Cluff M, Kingston J, Hill D, Chen H, Hoehne S, Malleske DT, Kaur R. Lung heparan sulfates modulate K(fc) during increased vascular pressure: evidence for glycocalyx-mediated mechanotransduction. Am J Physiol Lung Cell Mol Physiol 2012; 302: L816-828.

61. Dull RO, Mecham I, McJames S. Heparan sulfates mediate pressure-induced increase in lung endothelial hydraulic conductivity via nitric oxide/reactive oxygen species. Am J Physiol Lung Cell Mol Physiol 2007; 292: L1452-1458.

62. Gao L, Lipowsky HH. Composition of the endothelial glycocalyx and its relation to its thickness and diffusion of small solutes. Microvasc Res 2010; 80: 394-401.

63. Pahakis MY, Kosky JR, Dull RO, Tarbell JM. The role of endothelial glycocalyx components in mechanotransduction of fluid shear stress. Biochem Biophys Res Commun 2007; 355: 228-233.

107

64. Wang L, Fuster M, Sriramarao P, Esko JD. Endothelial heparan sulfate deficiency impairs L- selectin- and chemokine-mediated neutrophil trafficking during inflammatory responses. Nat Immunol 2005; 6: 902-910.

65. Gorin AB, Stewart PA. Differential permeability of endothelial and epithelial barriers to albumin flux. J Appl Physiol Respir Environ Exerc Physiol 1979; 47: 1315-1324.

66. Niessen CM. Tight junctions/adherens junctions: basic structure and function. J Invest Dermatol 2007; 127: 2525-2532.

67. Georas SN, Rezaee F. Epithelial barrier function: at the front line of asthma immunology and allergic airway inflammation. J Allergy Clin Immunol 2014; 134: 509-520.

68. Dai CY, Dai GF, Sun Y, Wang YL. Loss of p120 catenin aggravates alveolar edema of ventilation induced lung injury. Chin Med J (Engl) 2013; 126: 2918-2922.

69. Rezaee F, Georas SN. Breaking barriers. New insights into airway epithelial barrier function in health and disease. Am J Respir Cell Mol Biol 2014; 50: 857-869.

70. Krause G, Winkler L, Mueller SL, Haseloff RF, Piontek J, Blasig IE. Structure and function of claudins. Biochim Biophys Acta 2008; 1778: 631-645.

71. Schlingmann B, Molina SA, Koval M. Claudins: Gatekeepers of lung epithelial function. Semin Cell Dev Biol 2015; 42: 47-57.

72. Overgaard CE, Mitchell LA, Koval M. Roles for claudins in alveolar epithelial barrier function. Ann N Y Acad Sci 2012; 1257: 167-174.

73. Sartori C, Matthay MA. Alveolar epithelial fluid transport in acute lung injury: new insights. Eur Respir J 2002; 20: 1299-1313.

74. Flodby P, Kim YH, Beard LL, Gao D, Ji Y, Kage H, Liebler JM, Minoo P, Kim KJ, Borok Z, Crandall ED. Knockout Mice Reveal a Major Role for Alveolar Epithelial Type I Cells in Alveolar Fluid Clearance. Am J Respir Cell Mol Biol 2016; 55: 395-406.

75. Brooks RE. Ruthenium red stainable surface layer on lung alveolar cells; electron microscopic interpretation. Stain Technol 1969; 44: 173-177.

108

76. Pruessmeyer J, Martin C, Hess FM, Schwarz N, Schmidt S, Kogel T, Hoettecke N, Schmidt B, Sechi A, Uhlig S, Ludwig A. A disintegrin and metalloproteinase 17 (ADAM17) mediates inflammation-induced shedding of syndecan-1 and -4 by lung epithelial cells. J Biol Chem 2010; 285: 555-564.

77. Fehrenbach H. Alveolar epithelial type II cell: defender of the alveolus revisited. Respir Res 2001; 2: 33-46.

78. Mason RJ. Biology of alveolar type II cells. Respirology 2006; 11 Suppl: S12-15.

79. Goerke J. Pulmonary surfactant: functions and molecular composition. Biochimica et biophysica acta 1998; 1408: 79-89.

80. Veldhuizen R, Nag K, Orgeig S, Possmayer F. The role of lipids in pulmonary surfactant. Biochim Biophys Acta 1998; 1408: 90-108.

81. Pastva AM, Wright JR, Williams KL. Immunomodulatory roles of surfactant proteins A and D: implications in lung disease. Proc Am Thorac Soc 2007; 4: 252-257.

82. Matthay MA, Clerici C, Saumon G. Invited review: Active fluid clearance from the distal air spaces of the lung. J Appl Physiol (1985) 2002; 93: 1533-1541.

83. Jain R, Barkauskas CE, Takeda N, Bowie EJ, Aghajanian H, Wang Q, Padmanabhan A, Manderfield LJ, Gupta M, Li D, Li L, Trivedi CM, Hogan BL, Epstein JA. Plasticity of Hopx(+) type I alveolar cells to regenerate type II cells in the lung. Nat Commun 2015; 6: 6727.

84. Barkauskas CE, Cronce MJ, Rackley CR, Bowie EJ, Keene DR, Stripp BR, Randell SH, Noble PW, Hogan BL. Type 2 alveolar cells are stem cells in adult lung. J Clin Invest 2013; 123: 3025-3036.

85. Giangreco A, Reynolds SD, Stripp BR. Terminal bronchioles harbor a unique airway stem cell population that localizes to the bronchoalveolar duct junction. Am J Pathol 2002; 161: 173-182.

86. Ware LB, Matthay MA. Keratinocyte and hepatocyte growth factors in the lung: roles in lung development, inflammation, and repair. Am J Physiol Lung Cell Mol Physiol 2002; 282: L924-940.

109

87. Crosby LM, Waters CM. Epithelial repair mechanisms in the lung. Am J Physiol Lung Cell Mol Physiol 2010; 298: L715-731.

88. Tong L, Bi J, Zhu X, Wang G, Liu J, Rong L, Wang Q, Xu N, Zhong M, Zhu D, Song Y, Bai C. Keratinocyte growth factor-2 is protective in lipopolysaccharide-induced acute lung injury in rats. Respir Physiol Neurobiol 2014; 201: 7-14.

89. Lee JW, Fang X, Krasnodembskaya A, Howard JP, Matthay MA. Concise review: Mesenchymal stem cells for acute lung injury: role of paracrine soluble factors. Stem Cells 2011; 29: 913-919.

90. Maniscalco WM, Campbell MH. Alveolar type II cells synthesize hydrophobic cell- associated proteoglycans with multiple core proteins. Am J Physiol 1992; 263: L348-356.

91. Maniscalco WM, Campbell MH. Transforming growth factor-beta induces a chondroitin sulfate/dermatan sulfate proteoglycan in alveolar type II cells. Am J Physiol 1994; 266: L672-680.

92. Zhang H, Newman DR, Sannes PL. HSULF-1 inhibits ERK and AKT signaling and decreases cell viability in vitro in human lung epithelial cells. Respir Res 2012; 13: 69.

93. Leiner KA, Newman D, Li CM, Walsh E, Khosla J, Sannes PL. Heparin and fibroblast growth factors affect surfactant protein gene expression in type II cells. Am J Respir Cell Mol Biol 2006; 35: 611-618.

94. Force ADT, Ranieri VM, Rubenfeld GD, Thompson BT, Ferguson ND, Caldwell E, Fan E, Camporota L, Slutsky AS. Acute respiratory distress syndrome: the Berlin Definition. JAMA 2012; 307: 2526-2533.

95. Ashbaugh DG, Bigelow DB, Petty TL, Levine BE. Acute respiratory distress in adults. Lancet 1967; 2: 319-323.

96. Blank R, Napolitano LM. Epidemiology of ARDS and ALI. Crit Care Clin 2011; 27: 439- 458.

97. Calfee CS, Delucchi K, Parsons PE, Thompson BT, Ware LB, Matthay MA, Network NA. Subphenotypes in acute respiratory distress syndrome: latent class analysis of data from two randomised controlled trials. Lancet Respir Med 2014; 2: 611-620.

110

98. Meyer NJ, Christie JD. Genetic heterogeneity and risk of acute respiratory distress syndrome. Semin Respir Crit Care Med 2013; 34: 459-474.

99. Shaver CM, Bastarache JA. Clinical and biological heterogeneity in acute respiratory distress syndrome: direct versus indirect lung injury. Clin Chest Med 2014; 35: 639-653.

100. Meyer NJ, Daye ZJ, Rushefski M, Aplenc R, Lanken PN, Shashaty MG, Christie JD, Feng R. SNP-set analysis replicates acute lung injury genetic risk factors. BMC Med Genet 2012; 13: 52.

101. Matthay MA, Zemans RL. The acute respiratory distress syndrome: pathogenesis and treatment. Annu Rev Pathol 2011; 6: 147-163.

102. Thompson BT, Chambers RC, Liu KD. Acute Respiratory Distress Syndrome. N Engl J Med 2017; 377: 1904-1905.

103. Sarkar M, Niranjan N, Banyal PK. Mechanisms of hypoxemia. Lung India 2017; 34: 47-60.

104. Sylvester JT, Shimoda LA, Aaronson PI, Ward JP. Hypoxic pulmonary vasoconstriction. Physiol Rev 2012; 92: 367-520.

105. Burnham EL, Janssen WJ, Riches DW, Moss M, Downey GP. The fibroproliferative response in acute respiratory distress syndrome: mechanisms and clinical significance. Eur Respir J 2014; 43: 276-285.

106. Abraham E. Neutrophils and acute lung injury. Crit Care Med 2003; 31: S195-199.

107. Rittirsch D, Flierl MA, Day DE, Nadeau BA, McGuire SR, Hoesel LM, Ipaktchi K, Zetoune FS, Sarma JV, Leng L, Huber-Lang MS, Neff TA, Bucala R, Ward PA. Acute lung injury induced by lipopolysaccharide is independent of complement activation. Journal of immunology 2008; 180: 7664-7672.

108. Steinberg KP, Milberg JA, Martin TR, Maunder RJ, Cockrill BA, Hudson LD. Evolution of bronchoalveolar cell populations in the adult respiratory distress syndrome. Am J Respir Crit Care Med 1994; 150: 113-122.

109. Lee WL, Downey GP. Neutrophil activation and acute lung injury. Curr Opin Crit Care 2001; 7: 1-7.

111

110. Perl M, Lomas-Neira J, Venet F, Chung CS, Ayala A. Pathogenesis of indirect (secondary) acute lung injury. Expert Rev Respir Med 2011; 5: 115-126.

111. Orfanos SE, Mavrommati I, Korovesi I, Roussos C. Pulmonary endothelium in acute lung injury: from basic science to the critically ill. Intensive Care Med 2004; 30: 1702-1714.

112. Reutershan J, Ley K. Bench-to-bedside review: acute respiratory distress syndrome - how neutrophils migrate into the lung. Crit Care 2004; 8: 453-461.

113. Lomas-Neira J, Venet F, Chung CS, Thakkar R, Heffernan D, Ayala A. Neutrophil- endothelial interactions mediate angiopoietin-2-associated pulmonary endothelial cell dysfunction in indirect acute lung injury in mice. American journal of respiratory cell and molecular biology 2014; 50: 193-200.

114. Blondonnet R, Constantin JM, Sapin V, Jabaudon M. A Pathophysiologic Approach to Biomarkers in Acute Respiratory Distress Syndrome. Dis Markers 2016; 2016: 3501373.

115. Grommes J, Soehnlein O. Contribution of neutrophils to acute lung injury. Molecular medicine 2011; 17: 293-307.

116. Murphy LS, Wickersham N, McNeil JB, Shaver CM, May AK, Bastarache JA, Ware LB. Endothelial glycocalyx degradation is more severe in patients with non-pulmonary sepsis compared to pulmonary sepsis and associates with risk of ARDS and other organ dysfunction. Ann Intensive Care 2017; 7: 102.

117. Schmidt EP, Li G, Li L, Fu L, Yang Y, Overdier KH, Douglas IS, Linhardt RJ. The circulating glycosaminoglycan signature of respiratory failure in critically ill adults. J Biol Chem 2014; 289: 8194-8202.

118. Freeman CG, Parish CR, Knox KJ, Blackmore JL, Lobov SA, King DW, Senden TJ, Stephens RW. The accumulation of circulating histones on heparan sulphate in the capillary glycocalyx of the lungs. Biomaterials 2013; 34: 5670-5676.

119. Li L, Ling Y, Huang M, Yin T, Gou SM, Zhan NY, Xiong JX, Wu HS, Yang ZY, Wang CY. Heparin inhibits the inflammatory response induced by LPS and HMGB1 by blocking the binding of HMGB1 to the surface of macrophages. Cytokine 2015; 72: 36- 42.

112

120. Li X, Liu Y, Wang L, Li Z, Ma X. Unfractionated heparin attenuates LPS-induced IL-8 secretion via PI3K/Akt/NF-kappaB signaling pathway in human endothelial cells. Immunobiology 2015; 220: 399-405.

121. Wildhagen KC, Garcia de Frutos P, Reutelingsperger CP, Schrijver R, Areste C, Ortega- Gomez A, Deckers NM, Hemker HC, Soehnlein O, Nicolaes GA. Nonanticoagulant heparin prevents histone-mediated cytotoxicity in vitro and improves survival in sepsis. Blood 2014; 123: 1098-1101.

122. Beck-Schimmer B, Schwendener R, Pasch T, Reyes L, Booy C, Schimmer RC. Alveolar macrophages regulate neutrophil recruitment in endotoxin-induced lung injury. Respir Res 2005; 6: 61.

123. Skerrett SJ, Liggitt HD, Hajjar AM, Ernst RK, Miller SI, Wilson CB. Respiratory epithelial cells regulate lung inflammation in response to inhaled endotoxin. Am J Physiol Lung Cell Mol Physiol 2004; 287: L143-152.

124. Zemans RL, Colgan SP, Downey GP. Transepithelial migration of neutrophils: mechanisms and implications for acute lung injury. Am J Respir Cell Mol Biol 2009; 40: 519-535.

125. Aschner Y, Zemans RL, Yamashita CM, Downey GP. Matrix metalloproteinases and protein tyrosine kinases: potential novel targets in acute lung injury and ARDS. Chest 2014; 146: 1081-1091.

126. Frank JA. Claudins and alveolar epithelial barrier function in the lung. Annals of the New York Academy of Sciences 2012; 1257: 175-183.

127. Coyne CB, Vanhook MK, Gambling TM, Carson JL, Boucher RC, Johnson LG. Regulation of airway tight junctions by proinflammatory cytokines. Molecular biology of the cell 2002; 13: 3218-3234.

128. Looi K, Troy NM, Garratt LW, Iosifidis T, Bosco A, Buckley AG, Ling KM, Martinovich KM, Kicic-Starcevich E, Shaw NC, Sutanto EN, Zosky GR, Rigby PJ, Larcombe AN, Knight DA, Kicic A, Stick SM. Effect of human rhinovirus infection on airway epithelium tight junction protein disassembly and transepithelial permeability. Exp Lung Res 2016: 1-16.

129. Ingenito EP, Mora R, Cullivan M, Marzan Y, Haley K, Mark L, Sonna LA. Decreased surfactant protein-B expression and surfactant dysfunction in a murine model of acute lung injury. American journal of respiratory cell and molecular biology 2001; 25: 35-44.

113

130. Schmidt R, Markart P, Ruppert C, Wygrecka M, Kuchenbuch T, Walmrath D, Seeger W, Guenther A. Time-dependent changes in pulmonary surfactant function and composition in acute respiratory distress syndrome due to pneumonia or aspiration. Respiratory research 2007; 8: 55.

131. Dushianthan A, Cusack R, Goss V, Postle AD, Grocott MP. Clinical review: Exogenous surfactant therapy for acute lung injury/acute respiratory distress syndrome--where do we go from here? Crit Care 2012; 16: 238.

132. Ware LB, Matthay MA. Alveolar fluid clearance is impaired in the majority of patients with acute lung injury and the acute respiratory distress syndrome. Am J Respir Crit Care Med 2001; 163: 1376-1383.

133. Berthiaume Y, Matthay MA. Alveolar edema fluid clearance and acute lung injury. Respir Physiol Neurobiol 2007; 159: 350-359.

134. Mutlu GM, Machado-Aranda D, Norton JE, Bellmeyer A, Urich D, Zhou R, Dean DA. Electroporation-mediated gene transfer of the Na+,K+ -ATPase rescues endotoxin- induced lung injury. Am J Respir Crit Care Med 2007; 176: 582-590.

135. Li Q, Park PW, Wilson CL, Parks WC. Matrilysin shedding of syndecan-1 regulates chemokine mobilization and transepithelial efflux of neutrophils in acute lung injury. Cell 2002; 111: 635-646.

136. Matthay MA, Ware LB, Zimmerman GA. The acute respiratory distress syndrome. J Clin Invest 2012; 122: 2731-2740.

137. Rubenfeld GD. Confronting the frustrations of negative clinical trials in acute respiratory distress syndrome. Ann Am Thorac Soc 2015; 12 Suppl 1: S58-63.

138. Acute Respiratory Distress Syndrome N, Brower RG, Matthay MA, Morris A, Schoenfeld D, Thompson BT, Wheeler A. Ventilation with lower tidal volumes as compared with traditional tidal volumes for acute lung injury and the acute respiratory distress syndrome. N Engl J Med 2000; 342: 1301-1308.

139. Briel M, Meade M, Mercat A, Brower RG, Talmor D, Walter SD, Slutsky AS, Pullenayegum E, Zhou Q, Cook D, Brochard L, Richard JC, Lamontagne F, Bhatnagar N, Stewart TE, Guyatt G. Higher vs lower positive end-expiratory pressure in patients with acute lung injury and acute respiratory distress syndrome: systematic review and meta- analysis. JAMA 2010; 303: 865-873.

114

140. Sud S, Friedrich JO, Adhikari NK, Taccone P, Mancebo J, Polli F, Latini R, Pesenti A, Curley MA, Fernandez R, Chan MC, Beuret P, Voggenreiter G, Sud M, Tognoni G, Gattinoni L, Guerin C. Effect of prone positioning during mechanical ventilation on mortality among patients with acute respiratory distress syndrome: a systematic review and meta-analysis. CMAJ 2014; 186: E381-390.

141. National Heart L, Blood Institute Acute Respiratory Distress Syndrome Clinical Trials N, Wiedemann HP, Wheeler AP, Bernard GR, Thompson BT, Hayden D, deBoisblanc B, Connors AF, Jr., Hite RD, Harabin AL. Comparison of two fluid-management strategies in acute lung injury. N Engl J Med 2006; 354: 2564-2575.

142. Lai M, Lampert IA, Lewis PD. The influence of fixation on staining of glycosaminoglycans in glial cells. Histochemistry 1975; 41: 275-279.

143. Rambourg A, Neutra M, Leblond CP. Presence of a "cell coat" rich in carbohydrate at the surface of cells in the rat. Anat Rec 1966; 154: 41-71.

144. Bignon J, Jaubert F, Jaurand MC. Plasma protein immunocytochemistry and polysaccharide cytochemistry at the surface of alveolar and endothelial cells in the rat lung. J Histochem Cytochem 1976; 24: 1076-1084.

145. Sun X, Li L, Overdier KH, Ammons LA, Douglas IS, Burlew CC, Zhang F, Schmidt EP, Chi L, Linhardt RJ. Analysis of Total Human Urinary Glycosaminoglycan Disaccharides by Liquid Chromatography-Tandem Mass Spectrometry. Anal Chem 2015; 87: 6220- 6227.

146. Dargaville PA, South M, Vervaart P, McDougall PN. Validity of markers of dilution in small volume lung lavage. Am J Respir Crit Care Med 1999; 160: 778-784.

147. van den Born J, Salmivirta K, Henttinen T, Ostman N, Ishimaru T, Miyaura S, Yoshida K, Salmivirta M. Novel heparan sulfate structures revealed by monoclonal antibodies. J Biol Chem 2005; 280: 20516-20523.

148. Brennan TV, Lin L, Huang X, Cardona DM, Li Z, Dredge K, Chao NJ, Yang Y. Heparan sulfate, an endogenous TLR4 agonist, promotes acute GVHD after allogeneic stem cell transplantation. Blood 2012; 120: 2899-2908.

149. Beaudet JM, Weyers A, Solakyildirim K, Yang B, Takieddin M, Mousa S, Zhang F, Linhardt RJ. Impact of autoclave sterilization on the activity and structure of formulated heparin. Journal of pharmaceutical sciences 2011; 100: 3396-3404.

115

150. Coulson-Thomas VJ, Chang SH, Yeh LK, Coulson-Thomas YM, Yamaguchi Y, Esko J, Liu CY, Kao W. Loss of corneal epithelial heparan sulfate leads to corneal degeneration and impaired wound healing. Invest Ophthalmol Vis Sci 2015; 56: 3004-3014.

151. Mensah SA, Cheng MJ, Homayoni H, Plouffe BD, Coury AJ, Ebong EE. Regeneration of glycocalyx by heparan sulfate and sphingosine 1-phosphate restores inter-endothelial communication. PLoS One 2017; 12: e0186116.

152. Lindahl U, Kusche-Gullberg M, Kjellen L. Regulated diversity of heparan sulfate. J Biol Chem 1998; 273: 24979-24982.

153. Saumon G, Soler P, Martet G. Effect of polycations on barrier and transport properties of alveolar epithelium in situ. Am J Physiol 1995; 269: L185-194.

154. Qing Q, Zhang S, Chen Y, Li R, Mao H, Chen Q. High glucose-induced intestinal epithelial barrier damage is aggravated by syndecan-1 destruction and heparanase overexpression. J Cell Mol Med 2015; 19: 1366-1374.

155. Bode L, Salvestrini C, Park PW, Li JP, Esko JD, Yamaguchi Y, Murch S, Freeze HH. Heparan sulfate and syndecan-1 are essential in maintaining murine and human intestinal epithelial barrier function. J Clin Invest 2008; 118: 229-238.

156. Lilly JD, Parsons CL. Bladder surface glycosaminoglycans is a human epithelial permeability barrier. Surg Gynecol Obstet 1990; 171: 493-496.

157. Woodcock TE, Woodcock TM. Revised Starling equation and the glycocalyx model of transvascular fluid exchange: an improved paradigm for prescribing intravenous fluid therapy. Br J Anaesth 2012; 108: 384-394.

158. Lygizos MI, Yang Y, Altmann CJ, Okamura K, Hernando AA, Perez MJ, Smith LP, Koyanagi DE, Gandjeva A, Bhargava R, Tuder RM, Faubel S, Schmidt EP. Heparanase mediates renal dysfunction during early sepsis in mice. Physiol Rep 2013; 1: e00153.

159. Folz RJ, Abushamaa AM, Suliman HB. Extracellular superoxide dismutase in the airways of transgenic mice reduces inflammation and attenuates lung toxicity following hyperoxia. J Clin Invest 1999; 103: 1055-1066.

116

160. Nozik-Grayck E, Suliman HB, Majka S, Albietz J, Van Rheen Z, Roush K, Stenmark KR. Lung EC-SOD overexpression attenuates hypoxic induction of Egr-1 and chronic hypoxic pulmonary vascular remodeling. Am J Physiol Lung Cell Mol Physiol 2008; 295: L422-430.

161. Finigan JH, Faress JA, Wilkinson E, Mishra RS, Nethery DE, Wyler D, Shatat M, Ware LB, Matthay MA, Mason R, Silver RF, Kern JA. Neuregulin-1-human epidermal receptor-2 signaling is a central regulator of pulmonary epithelial permeability and acute lung injury. J Biol Chem 2011; 286: 10660-10670.

162. Ng HH, Narasaraju T, Phoon MC, Sim MK, Seet JE, Chow VT. Doxycycline treatment attenuates acute lung injury in mice infected with virulent influenza H3N2 virus: involvement of matrix metalloproteinases. Exp Mol Pathol 2012; 92: 287-295.

163. Ferguson ND, Fan E, Camporota L, Antonelli M, Anzueto A, Beale R, Brochard L, Brower R, Esteban A, Gattinoni L, Rhodes A, Slutsky AS, Vincent JL, Rubenfeld GD, Thompson BT, Ranieri VM. The Berlin definition of ARDS: an expanded rationale, justification, and supplementary material. Intensive Care Med 2012; 38: 1573-1582.

164. al-Hakim A, Linhardt RJ. Electrophoresis and detection of nanogram quantities of exogenous and endogenous glycosaminoglycans in biological fluids. Appl Theor Electrophor 1991; 1: 305-312.

165. McNeil JB, Shaver CM, Kerchberger VE, Russell DW, Grove BS, Warren MA, Wickersham NE, Ware LB, McDonald WH, Bastarache JA. Novel Method for Non- invasive Sampling of the Distal Airspace in Acute Respiratory Distress Syndrome. Am J Respir Crit Care Med 2017.

166. Ducharme A, Frantz S, Aikawa M, Rabkin E, Lindsey M, Rohde LE, Schoen FJ, Kelly RA, Werb Z, Libby P, Lee RT. Targeted deletion of matrix metalloproteinase-9 attenuates left ventricular enlargement and collagen accumulation after experimental myocardial infarction. J Clin Invest 2000; 106: 55-62.

167. Kheradmand F, Rishi K, Werb Z. Signaling through the EGF receptor controls lung morphogenesis in part by regulating MT1-MMP-mediated activation of gelatinase A/MMP2. J Cell Sci 2002; 115: 839-848.

168. Oshima K, Haeger SM, Hippensteel JA, Herson PS, Schmidt EP. More than a biomarker: the systemic consequences of heparan sulfate fragments released during endothelial surface layer degradation (2017 Grover Conference Series). Pulm Circ 2018; 8: 2045893217745786.

117

169. Deepa SS, Umehara Y, Higashiyama S, Itoh N, Sugahara K. Specific molecular interactions of oversulfated chondroitin sulfate E with various heparin-binding growth factors. Implications as a physiological binding partner in the brain and other tissues. The Journal of biological chemistry 2002; 277: 43707-43716.

170. Campo GM, Avenoso A, Campo S, D'Ascola A, Traina P, Calatroni A. Chondroitin-4- sulphate inhibits NF-kB translocation and caspase activation in collagen-induced arthritis in mice. Osteoarthritis Cartilage 2008; 16: 1474-1483.

171. Iovu M, Dumais G, du Souich P. Anti-inflammatory activity of chondroitin sulfate. Osteoarthritis Cartilage 2008; 16 Suppl 3: S14-18.

172. Vallieres M, du Souich P. Modulation of inflammation by chondroitin sulfate. Osteoarthritis Cartilage 2010; 18 Suppl 1: S1-6.

173. Rachmilewitz J, Tykocinski ML. Differential effects of chondroitin sulfates A and B on monocyte and B-cell activation: evidence for B-cell activation via a CD44-dependent pathway. Blood 1998; 92: 223-229.

174. Wang JY, Roehrl MH. Glycosaminoglycans are a potential cause of rheumatoid arthritis. Proc Natl Acad Sci U S A 2002; 99: 14362-14367.

175. Zhang W, Sun F, Niu H, Wang Q, Duan J. Mechanistic insights into cellular immunity of chondroitin sulfate A and its zwitterionic N-deacetylated derivatives. Carbohydr Polym 2015; 123: 331-338.

176. Zhou J, Nagarkatti P, Zhong Y, Nagarkatti M. Immune modulation by chondroitin sulfate and its degraded disaccharide product in the development of an experimental model of multiple sclerosis. J Neuroimmunol 2010; 223: 55-64.

177. Wang W, Han W, Cai X, Zheng X, Sugahara K, Li F. Cloning and characterization of a novel chondroitin sulfate/dermatan sulfate 4-O-endosulfatase from a marine bacterium. J Biol Chem 2015; 290: 7823-7832.

178. Potter DR, Damiano ER. The hydrodynamically relevant endothelial cell glycocalyx observed in vivo is absent in vitro. Circulation research 2008; 102: 770-776.

118

179. Singer MS, Phillips JJ, Lemjabbar-Alaoui H, Wang YQ, Wu J, Goldman R, Rosen SD. SULF2, a heparan sulfate endosulfatase, is present in the blood of healthy individuals and increases in cirrhosis. Clin Chim Acta 2015; 440: 72-78.

180. Siebert TA, Rugonyi S. Influence of liquid-layer thickness on pulmonary surfactant spreading and collapse. Biophys J 2008; 95: 4549-4559.

181. Zuo YY, Veldhuizen RA, Neumann AW, Petersen NO, Possmayer F. Current perspectives in pulmonary surfactant--inhibition, enhancement and evaluation. Biochim Biophys Acta 2008; 1778: 1947-1977.

182. Bastacky J, Lee CY, Goerke J, Koushafar H, Yager D, Kenaga L, Speed TP, Chen Y, Clements JA. Alveolar lining layer is thin and continuous: low-temperature scanning electron microscopy of rat lung. J Appl Physiol (1985) 1995; 79: 1615-1628.

183. Barbier F, Andremont A, Wolff M, Bouadma L. Hospital-acquired pneumonia and ventilator-associated pneumonia: recent advances in epidemiology and management. Curr Opin Pulm Med 2013; 19: 216-228.

184. Ishiguro T, Takayanagi N, Yamaguchi S, Yamakawa H, Nakamoto K, Takaku Y, Miyahara Y, Kagiyama N, Kurashima K, Yanagisawa T, Sugita Y. Etiology and factors contributing to the severity and mortality of community-acquired pneumonia. Intern Med 2013; 52: 317-324.

185. Rajas O, Quiros LM, Ortega M, Vazquez-Espinosa E, Merayo-Lloves J, Vazquez F, Garcia B. Glycosaminoglycans are involved in bacterial adherence to lung cells. BMC Infect Dis 2017; 17: 319.

186. Karamanos NK, Vanky P, Syrokou A, Hjerpe A. Identity of dermatan and chondroitin sequences in dermatan sulfate chains determined by using fragmentation with chondroitinases and ion-pair high-performance liquid chromatography. Anal Biochem 1995; 225: 220-230.

187. Pacheco B, Maccarana M, Malmstrom A. Dermatan 4-O-sulfotransferase 1 is pivotal in the formation of iduronic acid blocks in dermatan sulfate. Glycobiology 2009; 19: 1197- 1203.

188. Ito Y, Ahmad A, Kewley E, Mason RJ. Hypoxia-inducible factor regulates expression of surfactant protein in alveolar type II cells in vitro. American journal of respiratory cell and molecular biology 2011; 45: 938-945.

119

189. Ashikari-Hada S, Habuchi H, Sugaya N, Kobayashi T, Kimata K. Specific inhibition of FGF-2 signaling with 2-O-sulfated octasaccharides of heparan sulfate. Glycobiology 2009; 19: 644-654.

190. Fannon M, Forsten KE, Nugent MA. Potentiation and inhibition of bFGF binding by heparin: a model for regulation of cellular response. Biochemistry 2000; 39: 1434-1445.

191. Goetz R, Mohammadi M. Exploring mechanisms of FGF signalling through the lens of structural biology. Nat Rev Mol Cell Biol 2013; 14: 166-180.

192. Ito Y, Correll K, Schiel JA, Finigan JH, Prekeris R, Mason RJ. Lung fibroblasts accelerate wound closure in human alveolar epithelial cells through hepatocyte growth factor/c-Met signaling. Am J Physiol Lung Cell Mol Physiol 2014; 307: L94-105.

120

APPENDIX A

HEPARAN SULFATE MODULATES HEPATOCYTE GROWTH FACTOR

SIGNALING IN ALVEOLAR TYPE II CELLS

Introduction

In Chapter VII we aimed to determine the effect of the shed alveolar HS and CS on lung injury and repair. While investigating whether the shed alveolar HS may affect lung repair, we identified that N- and 6-O sulfated HS longer than 10 saccharides binds HGF, but that the shed alveolar HS, while long, was nearly entirely un-sulfated (92% 0S (Figure VII-1A)). Given these findings, we concluded that it was unlikely for the shed alveolar HS to bind and affect alveolar

HGF in vivo. While this near absence of sulfation of shed HS may create an alveolar microenvironment permissive to repair (given the absence of soluble HS capable of sequestering growth factors away from their receptors (168)), intratracheal instillation of full-length soluble sulfated HS did not in fact alter alveolar permeability or inflammation during lung injury resolution in vivo. In conjunction with these in vivo findings discussed in Chapter VII, we performed additional in vitro studies to determine whether soluble HS affects HGF signaling in a mouse ATII-like cell line (MLE-12 cells) or in rat primary ATII cells.

Objectives

• Determine if soluble HS affects HGF signaling in alveolar epithelial cells in vitro.

• Determine if the presence/absence of alveolar epithelial cell-surface HS alters the effect

of soluble HS on HGF signaling in vitro.

121

Materials and Methods

Materials

Human recombinant hepatocyte growth factor (HGF) was purchased from Sigma. Full- length HS was purchased from Celsus Laboratories, and HS oligosaccharides, provided by Dr.

Robert Linhardt (Rensselaer Polytechnic University, Troy, NY, USA), were created by size purification of digested full-length HS. MLE-12 cells were provided by Dr. Rachel Zemans

(University of Michigan, Ann Arbor, MI, USA).

Animals

Experiments were approved by the University of Colorado Institutional Animal Care and

Use Committee and conducted in accordance with the National Institutes of Health guidelines.

Sprague Dawley rats purchased from Harlan (Indianapolis, IN) were used to isolate primary

ATII cells.

Alveolar Type II Cell Isolation, Culture, and Treatment

Primary ATII cells were isolated from the lungs of Sprague Dawley rats. Lungs were digested using pancreatic porcine elastase (Worthington Biochemical Corporation; Lakewood,

NJ, USA), minced, and subjected to density gradients to isolate an ATII cell-enriched population, as previously described (188). Isolated ATII cells were cultured in Dulbecco’s

Modified Eagle Media (DMEM) containing 10% fetal bovine serum (FBS). Twenty-four hours after isolation, ATII cells were serum starved in 1% FBS for 24 hours and were then treated with

30ng/ml HGF and 1µg/ml full-length HS for 30 minutes. The treated cells were then lysed for protein.

MLE-12 Cell Culture, Treatment, and Scratch Closure

122

MLE-12 cells were cultured in DMEM media containing 10% FBS. Once cells reached approximately 75% confluence, cells were serum starved in DMEM media containing 1% FBS for 8 hours. For experiments investigating the effect of HGF and HS on ERK phosphorylation, following serum starvation, cells were treated with 10ng/ml HGF and 1µg/ml HS for 30 minutes and then lysed for protein. In experiments in which MLE-12 cells were pre-treated with heparinase I/III, cells were treated with 2U/ml enzymatically active or heat-inactivated (prepared as described in Chapter V) heparinase I/III during serum starvation prior to HGF and HS treatment. For experiments determining the effect of HGF on scratch closure, cells were grown to 100% confluence, and following serum starvation, a scratch was performed with a 10µl pipette tip. Cells were then rinsed once with DMEM and treated with 10ng/ml HGF. Images were taken immediately and 18 and 24 hours after scratch induction. Scratch width was measured using ImageJ and percent scratch closure, relative to initial scratch width, was calculated.

Cell-Surface Heparan Sulfate Staining

Cells were rinsed with cold PBS and staining for HS (HS 10E4 antibody) was performed as described in Chapter V; however, all staining procedures were performed at 4°C to inhibit endocytosis of the HS 10E4 and/or secondary antibody.

Western Blotting

5x Laemmli buffer was added to equal amounts of extracted protein and heated at 100°C for 10 minutes. The protein extracts were then loaded into Criterion Tris-HCl gels (BioRad) followed by gel electrophoresis and transfer to PVDF membrane. The membranes were blocked, stained for p-ERK (phosphorylated ERK), total ERK, and -actin (Cell Signaling Technology;

Danvers, MA, USA), and developed and imaged as described in Chapter V. Western blot

123 pERK/total ERK band intensities from each experiment were normalized to the band intensity from cells treated with HGF alone, which was set to 100%.

Statistical Analyses

Western blot band intensities were analyzed for statistical significance using a Wilcoxon

Signed Rank Test and one-sample t-test to determine a difference in a group’s band intensity average from 100 (normalized band intensity from HGF-treated cells). Results were considered statistically significant if p<0.05 (significance by t-test only noted by *).

Results

We utilized downstream ERK phosphorylation as an index of epithelial HGF signaling in

MLE-12 cells and rat primary ATII cells. To determine the effect of HS fragments on HGF signaling, we treated cells with HGF and HS (full-length and oligosaccharides of different lengths) or PBS as control. In MLE-12 cells, we detected increased ERK phosphorylation (by t- test only) in cells treated with HGF alone (in comparison to PBS), confirming the utility of ERK phosphorylation as a downstream readout of HGF signaling (Figure A-1A). HGF-induced ERK phosphorylation was somewhat attenuated in cells treated simultaneously with HS. Furthermore, the degree of ERK phosphorylation inhibition by HS appeared to increase as HS length increased. In primary rat ATII cells, we similarly observed a trend towards decreased ERK phosphorylation in cells treated simultaneously with HGF and HS, in comparison to cells treated with HGF alone (Figure A-1B). While limited, these studies indicate that HS, in a length- dependent manner, may inhibit HGF signaling in naïve ATII cells, likely by binding HGF and sequestering it away from its receptor (in a manner similar to what has been shown previously with other growth factors (189)).

124

Figure A-1: Heparan Sulfate may inhibit Hepatocyte Growth Factor Signaling in ATII Cells. A) pERK, total ERK, and -actin expression in MLE-12 cells treated with 10ng/ml HGF and 1µg/ml HS of varied length for 30 minutes. pERK/total ERK band intensities in all groups, normalized to HGF (100%), were graphed. B) pERK, total ERK, and -actin expression in rat primary ATII cells treated with 30ng/ml HGF and 1µg/ml full-length HS for 30 minutes. pERK/total ERK band intensities in all groups, normalized to HGF (100%), were graphed.

While soluble HS can bind and sequester growth factor ligands from their cell-surface

receptors in naïve cells, it is thought that sequestration of growth factor ligands by HS is partly

due to competition with cell-surface HS for growth factor ligand binding. As such, removal of

cell-surface HS can transform soluble HS from an inhibitor of growth factor signaling into an

activator of signaling (Figure A-2A) (189, 190). Given that epithelial shedding of HS occurs

125 during lung injury (Chapter VI), we aimed to determine if soluble HS activates HGF signaling in

MLE-12 cells devoid of cell-surface HS. To remove MLE-12 cell-surface HS, we treated cells with recombinant enzymatically active heparinase I/III or heat-inactivated heparinase I/III as control, and confirmed cell-surface HS loss by performing surface staining for HS (HS 10E4 antibody) in live, unfixed MLE-12 cells at 4°C (Figure A-2B).

After determining a concentration of heparinase I/III that removes MLE-12 cell-surface

HS, we then performed an experiment similar to that performed in Figure A-1B in MLE-12 cells pre-treated with enzymatically active heparinase I/III, or heat-inactivated heparinase I/III as control. During serum starvation, we pre-treated MLE-12 cells with heparinase I/III, and then again treated cells with HGF and HS for 30 minutes and extracted protein for ERK phosphorylation detection by western blot. In MLE-12 cells pre-treated with heat-inactivated heparinase I/III before HGF and/or HS treatment, we detected a similar pattern of ERK phosphorylation as in naïve (no heparinase pre-treatment) MLE-12 cells treated with HGF and

HS (Figure A-2C). After pre-treatment with enzymatically active heparinase I/III, cells treated with HGF and HS disaccharides, octasaccharides, and full-length HS exhibited a trend towards decreased ERK phosphorylation in a length-dependent manner as observed previously, in comparison to cells treated with HGF alone. In contrast, heparinase I/III pre-treated MLE-12 cells treated with HGF and HS decasaccharides did not exhibit a trend towards decreased ERK phosphorylation, in comparison to heparinase I/III pre-treated cells treated with HGF alone.

These data indicate that soluble HS of certain lengths may have the ability to both activate or inhibit ATII cell HGF signaling depending on the presence or absence of ATII cell-surface HS.

126

Figure A-2: Heparan Sulfate may not Inhibit Hepatocyte Growth Factor Signaling in ATII Cells Devoid of Cell-Surface Heparan Sulfate. A) Soluble HS can inhibit growth factor signaling cells that contain cell-surface HS, but facilitate growth factor signaling in cells devoid of cell-surface HS. B) MLE-12 cells were treated with 2U/ml enzymatically active or heat- inactivated HS and stained for cell surface HS (HS 10E4 - green) and DAPI (blue). pERK, total ERK, and -actin expression in MLE-12 cells treated with 10ng/ml HGF and 1µg/ml HS of varied length for 30 minutes following pre-treatment with C) 2U/ml heat-inactivated heparinase I/III or D) enzymatically active heparinase I/III. pERK/total ERK band intensities in all groups, normalized to HGF (100%), were graphed. Panel A modified from Goetz et al, 2013 (191).

127

While we demonstrated that soluble HS may modulate ATII cell HGF-induced ERK phosphorylation, we next aimed to determine whether HS also affects functional measures of

HGF signaling, including scratch closure. Using MLE-12 cells, we first aimed to confirm that

HGF accelerates scratch closure as has been shown previously (192). Following serum starvation and scratch induction, we treated MLE-12 cells with HGF, or PBS as control, and monitored scratch closure over time. We observed no difference in scratch closure in cells treated with HGF, in comparison to PBS as control (Figure A-3). As MLE-12 cells are a transformed cell line, we hypothesized that MLE-12 cell growth and migration may be independent of HGF signaling. As such, we aimed to perform these experiments in primary rat

ATII cells; however, due to technical difficulties these experiments could not be performed.

Given our in vivo findings that full-length HS did not affect resolution of lung injury, and the technical difficulties experienced, further studies were not performed.

128

Figure A-3: Hepatocyte Growth Factor may not Accelerate MLE-12 Scratch Closure. MLE-12 cells were grown to confluence and a scratch was induced. After scratch induction, cells were treated with 10ng/ml HGF, or PBS as control, and scratch closure was measured at 0, 18, and 24 hours. Percent scratch closure over time was graphed. Summary of Results

• HGF induces ERK phosphorylation in MLE-12 cells.

• HS, of varied lengths, may inhibit HGF signaling in naïve MLE-12 cells and primary

ATII cells.

• In MLE-12 cells devoid of cell-surface HS, soluble HS decasccharides may not inhibit

HGF signaling, while HS disaccharides, octasaccharides, and full-length HS inhibit HGF

signaling (similar to their effect on naïve MLE-12 cells).

129

Discussion

In this appendix we determined that soluble HS of varied lengths (from disaccharides to full-length) likely inhibits HGF signaling in naïve ATII cells in a HS length-dependent potency.

Inhibition of HGF signaling by HS likely occurs due to competition with cell-surface HS for

HGF, leading to a reduction in HGF binding to cell-surface HS and its cell-surface receptor, cMET. In contrast, in ATII cells devoid of cell-surface HS (like what occurs during lung injury), soluble HS of distinct lengths may no longer inhibit, and indeed facilitate, HGF signaling.

We hypothesize that soluble HS long enough to bind HGF, but too short to bind both HGF and cMET, may inhibit HGF signaling independent of the presence/absence of cell-surface HS, as binding of HGF to these short HS oligosaccharides would prevent HGF binding to cMET. In contrast, soluble HS of intermediate lengths that can bind one HGF ligand and one cMET receptor may inhibit HGF signaling in naïve cells (via competition with cell-surface HS for

HGF), but may facilitate HGF signaling in the absence of competitive cell-surface HS.

Furthermore, we hypothesize that soluble HS of long lengths (i.e. full-length HS) that can bind several HGF ligands, may act as a HGF sink independent of the presence/absence of cell-surface

HS, as even in the absence of competitive cell-surface HS, long HS can likely only facilitate binding of one HGF ligand to one cMET receptor despite its binding to multiple HGF ligands.

The hypotheses offered in this discussion are extremely speculative and, additionally, may only be valid during scenarios in which the HGF:HS concentration ratio is not large.

During states of HGF excess, soluble HS would not have the ability to bind all/most of HGF, thereby allowing HGF to access cMET despite the presence of soluble HS. Given that an effect of HS on HGF signaling is dependent on the concentration, length, and sulfation of both soluble and cell-surface HS, in addition to the concentration of HGF and cMET, these hypotheses are

130 difficult to test in vitro and are even more complex to study in vivo. While, in chapter VII, we did not detect an effect of intratracheally instilled HS on the resolution from lung injury, this may be due to the concentrations of HGF and HS present within the alveolar airspace; however, proving this would likely be difficult.

131