Engineering Organized Epithelium Using Nanogrooved Topography in a Gelatin Hydrogel

Home , Hemidesmosome

Engineering organized epithelium using nanogrooved topography in a gelatin hydrogel

John Paul Soleas, BMSc.

A thesis submitted in conformity with the requirements

for the degree of Master of Science

Institute of Medical Science

University of Toronto

Engineering organized epithelium using nanogrooved topography in a gelatin hydrogel

John Paul Soleas, BMSc.

Master of Science, 2012

Institute of Medical Sciences

University of Toronto

Abstract

Tracheal epithelium is organized along two axes: apicobasal, seen through apical ciliogenesis, and planar seen through organized ciliary beating, which moves mucus out of the airway. Diseased patients with affected ciliary motility have serious chronic respiratory infections. The standard method to construct epithelium is through air liquid interface culture which creates apicobasal polarization, not planar organization. Nanogrooved surface topography created in diffusible substrates for use in air liquid interface culture will induce planar organization of the cytoskeleton.

We have created a nanogrooved gelatin device which allows basal nutrient diffusion.

Multiple epithelial cells have been found to align in the direction of the nanogrooves in both sparse and confluent conditions. This device is also congruent with ALI culture as seen through formation of tight junctions and ciliogenesis. Thus, we have created nanogrooved surface topography in a diffusible substrate that induces planar alignment of epithelial cells and cytoskeleton.

Acknowledgements

I would like to thank my mentors, Drs. Alison P. McGuigan and Thomas K. Waddell for their patience, guidance, support, and for their example to always strive to be a better scientist. My future career aspirations as a physician and a scientist are due in no small part to this collaboration and their mentorship. I will forever be grateful to have had the opportunity to train under them.

I am grateful for the friendship and collegial atmosphere of the McGuigan laboratory, the Waddell laboratory group, as well as the entire Latner Thoracic Surgery Research Laboratories. I would like to extend a special thanks to Dr. Siba Haykal for her valuable input, mentorship, and gifts of primary tracheal epithelium, and to Ms. Lily Guo for her valued input, and gift of primary tracheal epithelium.

I thank Dr. Nadeem Moghal’s laboratory for their gift of and discussions on human tracheal epithelial cells.

I thank my program advisory committee members, Drs. Craig Simmons, and Mingyao Liu for their insights and perspectives and critical review of this thesis.

I thank the Canadian Institutes of Health Research and the Natural Science and Engineering Research Council of Canada for the Collaborative Health Research Project grant that made this collaboration possible and to the CIHR Training Program in Regenerative Medicine for their support through a graduate fellowship.

Finally, I thank my family and friends for their unwavering support, love, and humour.

iii

Table of Contents

Abstract ii

Acknowledgements iii

Table of Contents iv

List of Figures and Tables xi

List of Supplementary Figures xii

List of Appendices xiii

List of Abbreviations xiv

Chapter 1 - Literature Review

1.1 Epithelial biology 2

1.1.1 Anatomy 2

1.1.2 Microstructure 3

1.1.3 Broad epithelial functions 4

1.1.4 Airway and respiratory epithelial function 5

1.1.4.1 Mucociliary clearance 8

1.2 Respiratory developmental biology 10

1.3 General epithelial polarization 13

1.4 Apical junctions 16

1.4.1 Tight junctions 16

1.4.1.1 Claudins 17

1.4.1.2 Occludins 18

1.4.2 Adherens junctions 19

1.4.2.1 Cadherins 19

1.5 Hemidesmosomes 20

1.6 Cilia 22

1.6.1 Cilium structure 22

1.6.1.1 Transition zone 24

1.6.2 Ciliogenesis 25

1.6.2.1 FoxJ1 - Master program initiator 25

1.6.2.2 Timing 25

1.6.2.3 Basal body docking and nucleation 26

1.6.2.4 Intraflagellar transport 27

1.6.3 Primary cilium 27

1.6.4 Motile cilium 28

1.6.4.1 Mechanism of motion 28

1.6.4.2 Organization of cilia 29

1.7 Mechanobiology 30

1.8 Summative statement 34

Chapter 2 – Rationale, hypothesis, and aims

2.1 Rationale 36

2.1.1 Primary ciliary dyskinesia 37

2.1.2 How do we organize epithelium? From Soleas et al, 2012 38

2.1.2.1 Chemical signals 40

2.1.2.2 Mechanical signals 45

2.1.3 Study rationale 48

2.2 Hypothesis 49

2.3 Aims 49

Chapter 3 – Device Manufacture

3.1 Introduction 52

3.1.1 Replica moulding rationale 52

3.2 Process rationale 53

3.2.1 Holographic diffraction grating film 53

3.2.2 PDMS 54

3.2.3 Hydrogel choice 54

3.3 Methods 56

3.3.1 Generation of nangrooved PDMS mould 56

3.3.2 Collagen gel creation 56

3.3.3 Gelatin gel creation 57

3.3.4 Creating nanogrooved gel inserts 57

3.3.5 Scanning electron microscopy 58

3.3.6 Effect of scanning electron microscopy preparation on gelatin 59

hydrogel

3.3.7 Statistics 59

3.4 Results 61

3.4.1 Gelatin crosslinking optimization 61

3.4.2 Scanning electron microscopy 62

3.5 Discussion 65

3.5.1 Replication of PDMS 65

3.5.2 Casting gelatin and moulding 65

3.5.3 Cell alignment on the nanogroove gel insert 66

3.5.4 Scanning electron microscopy 66

3.6 Conclusion 67

Chapter 4 – Respiratory epithelium on nanogrooved topography

4.1 Introduction 69

4.2 Rationale 69

4.3 Methods 70

4.3.1 Cell culture 70

4.3.2 Nanogroove PDMS seeding 70

4.3.3 Nanogroove gelatin seeding 71

4.3.4 Air liquid interface culture of BEAS-2B 71

4.3.5 Phase microscopy 71

4.3.6 Fluorescence microscopy 71

4.3.7 Quantification of cellular alignment 72

4.4 Results 73

4.4.1 Cell alignment on nanogrooves 73

4.4.1.1 ARPE19 73

4.4.1.2 IMCD3 77

4.4.1.3 BEAS-2B align on nanogrooves 81

4.4.1.4 Gelatin insert appears congruent with ALI 82 culture 4.5 Discussion 83

4.5.1 Epithelial cell alignment on nanogroove topography 83

vii

4.5.2 Response to topography 86

4.5.3 Substrate stiffness 88

4.5.4 Changing chemistry 90

4.5.5 BEAS-2B polarization on gelatin inserts 90

4.6 Conclusion 91

Chapter 5 –Future directions and conclusions

5.1 Introduction 93

5.2 Rationale 93

5.3 Methods 94

5.3.1 Human tracheal epithelia 94

5.3.2 Nanogroove culture 95

5.3.3 Air liquid interface culture 95

5.3.4 Immunohistochemistry 96

5.3.5 Imaging 97

5.3.6 Quantification of cellular alignment 97

5.37 Quantification of ciliated cells 98

5.4 Results 98

5.4.1 Human tracheal epithelial cells apicobasally polarize in 98

standard ALI culture

5.4.2 Human tracheal epithelial cells apicobasally polarize on gelatin 100

insert ALI culture

5.4.2 Alignment of primary human tracheal epithelial

cell on nanogroove topography 102

viii

5.5 Future directions 103

5.6 Conclusion 105

Chapter 6 - Supplementary Figures

Supplementary Macroscopic image of gelatin gels made from various 107 Figure 1 percentages of gelatin gel

Supplementary Macroscopic image of 5% gelatin gels crosslinked 108 Figure 2 with various concentrations of glutaraldehyde

Supplementary Bronchial epithelial cell line BEAS-2B on collagen 109 Figure 3

Supplementary Scanning electron micrographs of flat substrates 110 Figure 4

Supplementary National Institute of Health 3T3 fibroblasts grown on 111 Figure 5 nanogrooved substrates

Supplementary Murine Inner Medullary Collecting Duct 3 (IMCD3) 112 Figure 6 epithelium on flat PDMS imaged using phase microscopy

Supplementary Murine Inner Medullary Collecting Duct 3 (IMCD3) 113 Figure 7 epithelium on flat PDMS imaged using fluorescent microscopy

Supplementary Human bronchial epithelial cell line (BEAS-2B) 114 Figure 8 sparsely seeded on nanogrooved substrates and imaged using phase microscopy

Supplementary Human bronchial epithelial cell line (BEAS-2B) 115 Figure 9 sparsely seeded on nanogrooved substrates and

imaged using fluorescent microscopy

Supplementary Mean number of ciliated cells in control and gelatin 116 Figure 10 ALI inserts

Chapter 7 – Appendix

Appendix 1 Standard Operating Procedure for isolation of normal 118 human tracheal epithelial cells

Chapter 8 – References 121

List of Figures and tables

Figure 1 Epithelial cell types found within the trachea, bronchioles and alveolus 6

Figure 2 Early morphogenesis of the foregut endoderm 10

Figure 3 The four major epithelial polarity domains are demarcated by various protein 14 complexes

Figure 4 Ciliary structure 23

Figure 5 Examples of the tools of tissue engineering. 40

Figure 6 Specialized exemplar tools of epithelial tissue engineering 42

Figure 7 Generating nanogrooved hydrogels using a PDMS stamp 60

Figure 8 Scanning electron micrographs of nanogrooved substrates 63

Table 1 Contraction of gelatin hydrogel on the macro- and microscale 64

Figure 9 Sparse ARPE19 align on nanogroove topography 74

Figure 10 Confluent ARPE19 align on nanogroove topography 76

Figure 11 IMCD3 do not align morphologically on nanogroove topography 78

Figure 12 Confluent IMCD3 F-actin cytoskeleton aligns on nanogroove topography 80

Figure 13 Confluent normal human bronchial epithelial cell line BEAS-2B grown on 81 nanogrooves have aligned F-actin cytoskeleton

Figure 14 Primary cilia are present on BEAS-2B differentiated on gelatin gels 83

Figure 15 Human tracheal epithelial cells form tight junctions and multiciliated cells 99 during standard ALI culture

Figure 16 Human tracheal epithelial cells form tight junctions and multiciliated cells 101 during ALI culture on a gelatin filter

Figure 17 Human tracheal epithelial cells seeded on nanogrooves do not align 101

List of Supplementary Figures

Supplementary 1 Macroscopic image of gelatin gels made from various 107

percentages of gelatin gel

Supplementary 2 Macroscopic image of 5% gelatin gels crosslinked with various 108 concentrations of glutaraldehyde.

Supplementary 3 BEAS-2B do not align on nanogrooved collagen. 109

Supplementary 4 Scanning electron micrographs of flat substrates 110

Supplementary 5 National Institute of Health 3T3 fibroblasts align on nanogrooved 111 substrates

Supplementary 6 IMCD3 epithelia on flat PDMS do not align morphologically 112

Supplementary 7 IMCD3 epithelia on flat PDMS do not align cytoskeletally 113

Supplementary 8 BEAS-2B sparsely seeded on nanogrooved substrates align 114 morphologically

Supplementary 9 BEAS-2B sparsely seeded on nanogrooved substrates align 115 cytoskeletally Supplementary 10 Mean number of ciliated cells in control and gelatin ALI 116 inserts

xii

List of Appendices

Appendix 1 Standard Operating Procedure for isolation of normal human tracheal 118 epithelial cells – From the Moghal Lab

xiii

List of Abbreviations

AJ Adherens junction ALI Air liquid interface aPKC atypical protein kinase C ATCC American type culture collection BEBM Bronchial epithelial basal media BEGM Bronchial epithelial growth media BM Basement membrane Crb Crumb Dlg Discs large DMEM Dulbecco’s modified eagle medium ECL Extracellular loop ECM Extracellular matrix F-Actin Filamentous actin FAK Focal adhesion kinase GTA Glutaraldehyde HD Hemidesmosomes HTEC Human tracheal epithelial cell IDA Inner dynein arms IF Intermediate filament IFT Intraflagellar transport Lgl Lethal giant larvae MTOC Microtubule organizing centre NEB Neuroepithelial bodies NIH National institute of health ODA Outer dynein arms ODF Outer Dense Fiber PALS1 Protein Associated with Lin Seven 1 PATJ Protein associated with tight junctions PCD Primary ciliary dyskinesia PDMS Poly(dimethlysiloxane) PNEC Pulmonary neuroendocrine cells TJ Tight junction ZO Zona occludens

xiv

Chapter 1| Literature review

1 1.1 Epithelial biology

Epithelial tissues line all tubular organs within the vertebrate body plan1 and form a selective filter between the body and the external environment. The epithelium functions to maintain homeostasis through broadly defined actions such as: ion and fluid secretion, selective trans- epithelial transport, coordinated fluid movement, mucus secretion, and as a physical barrier.2-4

Organ-specific functions of the epithelium are diverse, varied, and determined by the epithelial cell types present.

1.1.1 Anatomy

Epithelial tissues in vivo appear as a sheet of confluent cells that cover the internal surface of a tubular organ, such as the trachea, or the external surface of the body, such as skin.5,6 While mainly functioning as an interface between the body’s internal environment and the outside world, the epithelium also forms a majority of the body’s glandular components. 7

Within the body, distinctive characteristics define epithelial tissues: confluent, avascular, innervated, with regenerative capacities.6,8 They possess highly specialized cell-to-cell9,10 and cell-to-substratum contacts11, which will be discussed later. The substratum on which the epithelial cells reside is secreted by maturing epithelia to form the basal lamina, which, in concert with the more basal reticular fibers, form the basement membrane, which functions as a selectively diffusible filter, while forming the scaffolding critical to epithelial regeneration.12

Finally, the critical defining feature of epithelial tissues which contributes to all other functions and structures are their intrinsic polarity.13 In airway epithelium, apical basal polarity is critical to epithelial function, as the luminal facing cell membrane is very different in composition, structure, and function from the lateral and basal.14 The apical surface is free of cell-to-cell contact, allowing for interaction with the outside world (kidney, lungs, gut), while the

2 lateral is in tight contact with neighbouring cells through tight junctions, adherens junctions, and gap junction proteins which synergistically function to create the selectively permeable barrier of epithelial tissues.15,16 The basal compartment is attached to the basement membrane through various integrins17 where nutrient-waste exchange occurs in close proximity to the underlying vasculature.

1.1.2 Microstructure

The histological anatomy of epithelium is very diverse and complex; the nomenclature of all epithelial tissues is governed by convention, which suggests that all epithelium is given two names: the first indicates the number of cell layers, while the second demarcates cell morphology.6

Based on cell layers, epithelia can be simple having one layer, or can be stratified, having multiple layers. A unique variant known as pseudostratified epithelium occurs in the airway, and will be discussed. Based on cell morphology, epithelia can be squamous and have a large surface area; they could also be cuboidal, or columnar. As with all anatomy, structure is reflected in function and broad statements can be made regarding cell shape and stratification: columnar and cuboidal epithelia are mainly involved in secretion and absorption; squamous cells are used to facilitate simple diffusion and filtration, while stratified epithelium primarily serves in a protective role.

Simple squamous epithelia have a large surface area to volume ratio and thus allow for simple diffusion to quickly occur. This is the structure of epithelium which primarily make up the respiratory zone of the lungs, facilitating the rapid gas exchange within the alveolus.5,6 The simple squamous cells of the respiratory zone are known as Type I pneumocytes. Simple cuboidal epithelium forms the ducts of glands and tubules within the nephron facilitating active

3 transport and osmosis.2,5,6 Simple columnar epithelia are found within the gastrointestinal tract; these tall cells are thicker, and have a larger volume than simple squamous epithelium. The larger volume, while not ideal for simple diffusion, is used to house the transcriptional and translational machinery necessary to manufacture the proteins to facilitate active nutrient transport into the body. 5 Stratified squamous epithelium makes up our epidermis as it is adapted to protection given its thickness and multilayered cellularity. 2

While most of the epithelium above functions in ion and fluid transport or as a simple barrier, epithelial cells also have endocrine and exocrine secretory functions known as glands.

Unicellular examples of exocrine ducts are the goblet and Clara cells found within the lung. 18, 9

The epithelium of conducting airways, generations 0-17 are made up of multiple cell types that form a specialized pseudostratified-columnar epithelium that function in maintaining the airway. Pseudostratified columnar epithelium is a unique form of columnar epithelium. All cells connect directly to the basal lamina; however, not all cells reach the luminal compartment.

This creates a false impression that the epithelia are stratified tissues, which is a false characterization, hence the name pseudostratified. The respiratory zone, by surface area, is made up of squamous epithelium that facilitates gas exchange; there is a small population of simple cuboidal cells which secrete surfactants to decrease surface tension at the air liquid interface.5

1.1.3 Broad Epithelial Functions

Broad airway epithelial functions can be characterized under barrier function and ion transport. Epithelial barrier function is accomplished through a network of tight adherens and gap junctions, in concert with desmosomes. These junctions rivet the lateral aspect of cell membranes creating, through the sheer number of intercellular connections, a fluid barrier that modulates the ‘leakiness’ of the paracellular route of transport.19,20 Thus, to move material across

4 the epithelium, the transcellular route must be followed which allows the epithelium to selectively transport material in and out of the body.

To regulate fluid flow across and near the epithelial barrier, tissues regulate ion transport; water has been found to follow high concentrations of ions. There are two main active transport systems: chloride secretion and sodium absorption. Chloride secretion leads to a buildup of negative charges along the epithelium, which allows for passive diffusion of sodium to balance the charges. This buildup of ions in the extracellular compartment leads to an increase in fluid around the epithelium. 21 The contrary scenario is also true: sodium may be transported into the cell in exchange for potassium, allowing fluid volume to decrease.22 The balance of both these systems creates a homeostatic mechanism to regulate fluid volume. In the airway, this fluid along the lumen is known as periciliary fluid.

1.1.4 Airway and respiratory epithelial function

The respiratory tree is made up of 3 anatomical regions, lined by different combinations of epithelium. The entire respiratory tree is supported by a dense vascular network that functions in nutrient-waste exchange in conductive and proximal areas, and as the site of primary gas exchange in alveolar regions. The vascular and pulmonary trees are mechanically supported by a dense extracellular matrix made of collagens, elastins, and glycoaminoglycans (GAGs).23

The respiratory system can be broken down into two broad areas: conducting and respiratory zones. The conducting zone is responsible for warming, humidifying, and cleaning the air for the respiratory zone, which is the site of gas exchange.9 The conducting zone keeps the airway free of foreign material and bodies through intrinsic defence mechanisms such as metachronic ciliary beating, the secretion of mucus, lysozyme, lactoferrin and secretory IgA18, as well as coughing, and the innate cellular and acquired immune responses.

These zones can further be divided into orders of branching or generations.2, 9 The conducting zone which is composed of the nose, pharynx, larynx, trachea, bronchus, bronchioles and terminal bronchioles comprise generations 0-17, while the respiratory zone of respiratory bronchioles, alveolar ducts and finally sacs comprise generation 17-23. 9 Broadly, conducting and respiratory epithelia are made up of specific and, in some cases, very different cell types

(Figure 1). Two specific epithelial cell types populate the respiratory zone: Type I and Type II pneumocytes. 24 Type I pneumocytes are simple squamous epithelium that are characterized by their large area and small volume, which facilitates quick gas exchange. 25 Type II pneumocytes are large, simple cuboidal epithelium that produce and secrete surfactant. Type II cells are twice as numerous as type I cells. Due to their cuboidal morphology they cover less than 10% of respiratory zone. 24 They are usually found in the corners of alveoli (Figure 1).25

Adapted from 25

Figure 1. Epithelial cell types found within the trachea, bronchioles and alveolus. Within the large, conducting airway (A), ciliated cells and secretory cells are supported by progenitors such as basal cells. Bronchioles have a greater proportion of secretory cells, such as Clara, goblet or neuroendocrine to ciliated epithelium. The alveolus is made up mainly of the squamous airway epithelium cell type 1, and the secretory type 2 cell. Rock and Hogan 2011. Reprinted with permission.

As the conducting epithelium is primarily responsible for cleaning and maintaining the integrity of the airway, the epithelium reflects this function. Throughout the conducting zone multi-ciliated cells are found with approximately 200 – 300 motile cilia that beat in a metachronal fashion. The length and beat frequency of the cilia change according to location, and environment. Broadly speaking, cilia shorten as one moves towards the mouth. 26 An abundant amount of mitochondria produce the necessary energy to drive ciliary beating. These cells are found to be terminally differentiated, as they do not divide under normal circumstances in vivo or in vitro.25

Another secretory cell type of the conducting epithelium is known as the Clara cell. 20

The Clara cell is also known to process xenobiotics through cytochrome P450.27 Histologically, they possesses an apical dome, called the uterodome28, through which it secretes its protein product, known as Clara cell secretory protein (CCSP), Uteroglobin, or Secretoglobin 1A. 29 The specific function of CCSP is not known, however it is reputed to contribute to the Clara cell’s intrinsic xenobiotic elimination activity. Clara cells have been found to have multi-potent, progenitor-like qualities in that after epithelial injury they appear able to generate ciliated cells in mice.30

Neuroendocrine or columnar pulmonary neuroendocrine cells (PNECs) are a rare, basally localized constituent of the conducting airway with progenitor properties.31 Found as single cells within the early generation airways, they cluster together in the intralobar airways and are known as neuroepithelial bodies (NEB).25 They are known to be able to sense stimuli within the airway and calcitonin, bombesin and serotonin which may act as epithelial growth factors. 32

Basal cells are small, flattened, epithelial cells in direct contact with the basal lamina.

They have come to be known as the stem cell of the airway33 and have been found to reside

7 primarily in the rodent trachea33 and throughout the airway in humans34. Their multi-lineage potential has been found to give rise to both ciliated and secretory cell types. 33

Recently, human lung stem cells have been identified that bear the hallmarks of an in vitro/vivo multipotent stem cell, capable of recapitulating epithelial architecture after injury.35

However, criticisms have arisen based on the in vitro expansion required (100 000x) before implantation to repair wounds.36 While successful on the microscale at reconstituting lung architecture this approach has not been proven to recapitulate the macroscopic architecture of the lung.

Another cell type seen primarily within the conducting zone is the mucin secreting goblet cell. These columnar epithelia produce the main proteins and glycoproteins of mucus which traps inhaled foreign material.37 The chemistry of these products confer upon mucus anti- bacterial acidic properties.38 The controlled production of mucus is critical for airway health, mucociliary clearance (reviewed later) and trapping invading bacteria.18 Chemical irritants,39 and immune secretions,40 and airflow 41 can increase mucus secretion. The viscoelastic properties of mucus are due to the mucin proteins, which are essential to mucociliary clearance.42

1.1.4.1 Mucociliary clearance

Mucociliary clearance is a physiological process involving the mechanical clearance of mucus through consistent and coordinated ciliary beating; it is considered the primary innate airway defence. 43 As mucociliary clearance is responsible for the removal of any inhaled particles and substances, its dysfunction can lead to chronic pulmonary symptoms.44 The propulsive force generated by motile cilia moves mucus towards the nasopharynx for elimination. 45 The effectiveness of mucociliary clearance is determined by the composition of mucus, periciliary liquid production, and coordinated ciliary activity. 43

Mucus is a high molecular weight, heavily glycosylated proteinaceous macromolecule that is composed of 1% sodium chloride, 0.5-1% free protein, 0.5-1% mucins; the rest is water. 46

Mucins are 500kDa molecules, bound by hydrogen bonding, hydrophobic interactions, and disulfide bridges.47 The balance of these chemical interactions confers the elasticity required for ciliary propulsion, and the viscosity creates a sticky environment for bacteria to become entangled. 18 Mucus is also made up of proteins with immune functionality, such as lactoferrin, lysozyme, and immunoglobulin A (IgA) which aid in the destruction of bacteria while being cleared from peripheral airways by mucus transport, which may require six hours. 45 Mucins, which perform their physiological role as a tangled mesh of polymer, are made from two genes:

MUC5AC, and MUC5B. 48 This tangled mesh traps inhaled particulate and infectious debris.15

The rate of clearance in mucociliary clearance is dependent on the interaction, effectiveness of beating cilia, and the ciliary beat frequency. 49 The basal mucociliary clearance rate has been found to be approximately 12-15 Hz and the unidirectional movement of mucus results from coordinated ciliary beating or metachrony. 50 Metachrony appears visually as a travelling wave of beating cilia. In actuality, the cilia beat in a highly coordinated fashion that promotes fluid movement unidirectionally.51 To alter the ciliary beat frequency (CBF) the motile cilia have been found to be sensitive to chemical52 and mechanical modalities.53

1.2 Respiratory developmental biology

The trachea and lung epithelium are critical to the initiation and continuation of post natal

life. The development of these highly specialized structures has been under increasingly

intense and elegant biological study.

Adapted from 54

Figure 2. Early morphogenesis of the foregut endoderm. At E9.5 (A) the foregut endoderm is a simple tube, which through mitosis and migration begins to bud into the lungs (B) at E10. At E10.5, the lung has begun to extend posteriorly, while the trachea and esophagus septate (C). By E11.5, the trachea and esophagus have separated and the lung has begun to arborize. Reprinted from Differentiation; research in biological diversity, 74, Que et al., Morphogenesis of the trachea and esophagus: current players and new roles for noggin and Bmps, 422-37, 2006, with permission from Elsevier.

The lung and trachea arises from the ventral foregut endoderm. 55 The endoderm folds ventrally to generate the foregut tube (Fig2A), which through outgrowth and elongation contribute to trachea and lung outgrowth .56 After broncho-pulmonary bifurcation, the dorsal esophageal domain of the foregut begins to separate from the ventral tracheal component E11-12 and days 28-37 in man (Fig2B-D).57 On E9.5 in mice or 28 days in humans, the lung buds appear. 58 Once lung buds have formed, they begin to arborize, and thus increase their surface area. From E10-17.0 the lung undergoes branching morphogenesis, a complex process that creates the entire respiratory tree within the lung.59 From E9.5-16.5, known as the pseudoglandular stage, the arborisation of the lung takes place, from E16.5-17.5 the canalicular stage proceeds as the terminal lung buds begin to narrow, and finally from E18.5 – P5 the lung buds begin to form alveoli. 61 Within humans, alveoli are formed in utero, unlike in the mouse, where alveoli are formed after birth. While the future respiratory tract is created through branching morphogenesis the cell fate of the lung and tracheal progenitors are being determined.

The progenitor cells of the trachea express Sox2, a transcription factor required for epithelial cell development but not found in progenitors destined to become distal lung lineages.60 At

E13.5 within the trachea, Sox2 is expressed in higher quantities in presumptive basal cells60 and by E15.5 – P0, Sox2 is seen in all tracheal and proximal epithelial cells.61 Thus Sox2 expression is found to demarcate the future epithelium through expression in the presumptive epithelial progenitor population: the basal cells. Genetic manipulation of Sox2 has found that the deletion of Sox2 leads to a decrease in the proportion of basal, ciliated, and Clara cells, and an increase in goblet cells while Sox2 overexpression, leads to an increase in neuroendocrine and basal cells.61

The cell fate of tracheal and lung epithelium is established along the proximo-distal axis of the expanding respiratory system as differentiation proceeds in a proximal to distal fashion.55,62 This

11 differentiation is under the regulation of Bmp4 and Wnt signalling, which fates epithelial cells in general. 55

During the pseudoglandular stage, at the emerging tips, there is a proliferative, multipotent progenitor cell which contributes to all lineages, demarcated by the transcription factors Id2 and

Sox9.25 As the respiratory tree elongates and bifurcates the cells left behind begin to differentiate into the future epithelial populations which will constitute the linings of the conducting airway.25

Neuroendocrine cells are among the first cells to appear in the developing conducting airway epithelium at E13.5, ciliated cells follow in the trachea and main bronchi at E14. During the canalicular and saccular stage (E16.5 – P5) the distal lung epithelium, found in the higher respiratory generations (17-23) begin to differentiate, as into Type I and II pneumocytes within the alveoli.

Marked difference between the populations of human and mouse airway epithelium have been found. In mice, only the trachea and main stem bronchi (generations 0 -2) are pseudostratified, whereas the rest of the airway is comprised of columnar or cuboidal epithelium.58 In humans, the entire airway and a majority of the lung is pseudostratified, even at diameters less than 0.5mm.25,58 Only the smallest most distal tubes have simple squamous epithelium which is the most conducive to gas exchange due to a high surface area to volume ratio.63 The composition of both species’ pseudostratified epithelium is similar however, with

70% of the epithelium made up of ciliated, secretory and neuroendocrine clusters and the remaining 30% as basal cells.33,58

1.3 General epithelial polarization

The pseudostratification of the majority of epithelium in humans illustrates the high degree of organization within these cell populations in order to create a functional epithelium.

However, within epithelial cells a great deal of cellular structure and organization is not left to chance. The difference between the membrane seen at the cell apex and basolateral membrane is pronounced in differentiated and mature epithelial cells, suggesting a complex and rigid organizing structure that creates different membrane compartments. This difference in cell compartments is known as polarity. Epithelial polarization is unique because epithelial tissues have highly specialized junctions that are formed as cells polarize and are critical to the function of all epithelial tissues and create separate polarity domains by physical separation.13 There are four well-defined polarity domains in mammalian cells: apical, the luminal facing compartment; junctional, where intercellular adhesions lie; lateral, demarcated apically, by intercellular junctions; and basal, which is in contact with the basement membrane (Figure 3).13,14 Polarity cues are driven by highly conserved signalling clusters.13 There are three main polarity complexes in epithelial cells that are defined by their main protein constituent, the Crumbs,

PAR-3, and Scribble complex.16, 14

Adapted from14

FigureFigure 33.. TheThe 44 majormajor epithelialepithelial polaritypolarity domainsdomains areare demarcateddemarcated byby variousvarious proteinprotein complexes.complexes. TheThe apicalapical (red)(red) compartcompartmentment isis patternedpatterned byby thethe PARPAR--66 complex.complex. TheThe junctionaljunctional (green)(green) complexcomplex isis markedmarked byby PARPAR--33 oror Bazooka.Bazooka. TheThe laterallateral domaindomain isis markedmarked byby thethe inhibitioninhibition ofof thethe PARPAR--66 complexcomplex byby thethe ScribbleScribble complex.complex. FinallyFinally thethe basalbasal membranemembrane isis markedmarked byby thethe variousvarious integriintegrinsns thatthat areare expressedexpressed toto bindbind cellscells toto thethe extracellularextracellular matrix.matrix. ReprintedReprinted fromfrom CurrentCurrent opinionopinion inin cellcell biology,biology, 2323,, St.St. JohnstonJohnston andand SansonSanson,, EEpithelialpithelial polaritypolarity andand morphogenesismorphogenesis,, 540540--6,6, 2011,2011, withwith permissionpermission fromfrom Elsevier.Elsevier. The Crumb (Crb) proteins, Crb 1-3, form a complex with PALS1 (Protein Associated with Lin

Seven 1), PAR-6, aPKC (atypical protein kinase C) and PATJ (Protein associated with tight junctions).14,64 This protein complex defines the apical domain and marks the boundary for the junctional domain, thereby marking the future location of junction formation. Crb appears to induce tight junction formation and the complex as a whole recruits components of tight junction proteins through PATJ. 13 aPKC phosphorylates Crumbs, and other proteins responsible

14 for the critical organization of the apical cytoskeletal.13 A loss of function variant Crb protein leads to a loss of the apical domain in polarizing epithelium, whereas enhanced expression leads to expansion of the apical domain.65 The Crumbs complex is tuned to the activity of Cdc42. 16

When Cdc42 is reduced, the complex of PAR-6, Crb, and aPKC dissociate, which leads to loss of the apical domain.13,66

The PAR-3 complex is made up of PAR3, PAR6, and aPKC. This complex demarcates the junctional domain and limits the expansion of the apical boundary as it is more basally localized than the Crumbs complex. 16,67 The PAR-3 complex has been found to interact directly with components of cell junctions and to regulate Rac activity within cells, which initiates cortical actin ring formation that interacts with developing junctions.13

The Scribble complex in Drosophila is composed of Scribble, lethal giant larvae (Lgl) 1 and 2, Discs large (Dlg), and PAR-114. While proof of direct interaction has yet to be found, mutations in any of the Scribble complex produce similar phenotypes suggesting the proteins work along similar biochemical pathways. 13 The Scribble complex primarily localizes to the lateral membrane, and through interactions with cadherin and actin functions to exclude apical proteins, thus maintaining the lateral domain.14, 16

The genetics of the basal domain have yet to be elucidated. This domain is most likely to be defined by integrin binding to extracellular matrix (ECM) components, which lead to signalling pathways that lead to cytoskeletal rearrangements.68 However, how these pathways function into polarity delineation is currently not known.13,14

A great deal of regulatory cross-talk between polarity complexes has been found.69 The inhibiting activity of Lgl the Crb complex through aPKC inhibition highlights the antagonistic

15 nature of polarity domain specification.13, 14 In particular, aPKC appears to inhibit the Scribble complex through PAR-1 phosphorylation and is in turned repressed through components of the

Scribble complex (Figure 3).14 Another member of the Scribble complex Lgl has been shown to inhibit the activity of aPKC by binding directly to PAR-6/aPKC complex and disrupting apical protein expression along the lateral membrane.73 Thus, the demarcation of epithelial polarity domains appears to be regulated within the three polarity defining complexes.

1.4 Apical junctions

Apical junctions are a defining feature of epithelial monolayers and are found within the junctional domain which is demarcated by the PAR-3 protein.70 Examples of apical junctions found in epithelium are adherens junctions (AJs) and tight junctions (TJs). These conserved junctions are, broadly speaking, organized into three constituents: adhesive proteins embedded within the plasma membrane that facilitate connection of one cell to another along lateral surfaces, adaptor proteins which mediate adhesive proteins connecting to the cytoskeleton of the cell, and finally the cytoskeleton.71

1.4.1 Tight junctions

Tight junctions (TJs) are inter-membrane protein complexes that provide apical cell-to- cell connections that are found in epithelial cells in contact with other epithelial cells. First identified in the 1960s, 72 TJs provide epithelia with structural integrity by creating polarized barriers that are selectively permeable to small molecules and ions, TJs form barriers that seal body cavities, by maintaining the apicobasal polarity of epithelial cells by physically hindering membrane protein diffusion.73 The barrier function of TJs is not absolute, as the variable protein composition can lead to permeability of certain ions and small molecules74, as well as the

16 creation of paracellular pathways that can be altered by the number of TJ strands along the apicobasal membrane axis.75

Located below the apical surface of epithelial monolayers, TJs are composed of a heterogeneous network of strands that form the basis of TJ function. Cytoplasmically, TJs have adaptor proteins known as the zona occludens (ZO-1,2, and 3) along with vinculin which facilitate TJ interactions with the actin cytoskeleton.76 All components of TJs are highly dynamic with regulatory modifications leading to different protein interactions, conformations, and localizations which modulate the adhesive properties of TJs. 73 Some constituents of TJs are claudins and occludins.77

1.4.1.1 Claudins

Twenty-six human claudins have been identified.78 They are 20-27 kDa transmembrane proteins that span the phospholipid bilayer four times, while having cytosolic N- and C-termini.79

They are the main components of TJs80, responsible for structural and paracellular barrier functions. Different combinations of claudins lead to variable ion permeability in the paracellular transport pathway. Claudins have been found to bind homotypically within cells, and across adjacent cells, as well, heterodimerization between certain claudins has been found. 80 The distribution of claudins across various tissues and has been found to vary within certain tissues based on the health of the epithelium.81

Certain structural domains of claudins are of particular interest: extracellular loop 1 (ECL1) is a significant claudin domain80 affecting TJ character through the determination of paracellular charge selectivity; 82 ECL2 is the domain that has been found to bind claudins to the corresponding partner in neighbouring cells;15 finally, the cytoplasmic tail possesses a PDZ-

17 binding motif. PDZ motifs are well-conserved structural domains found in signaling proteins that enables the C-terminus of claudin to interact with cytosolic proteins ZO-1, 2, and 3, indirectly linking claudins to the actin cytoskeleton. 83 This linkage of claudin to the cytoskeleton through various ZO proteins leads to the localization of the entire claudin-ZO complex to the TJ through PDZ-1 signalling of ZO84, the stabilization of TJs, and maintenance of TJ permeability characteristics. 85 This suggests that these associations of claudins, ZO proteins, and the cytoskeleton are critical for the stable integration of claudins into TJ during junction formation.

1.4.1.2 Occludins

Occludins are 65kDa transmembrane proteins that were the first TJ component identified.86

Occludin domains have different functions and regulation: the C-terminus interacts with ZO-1 allowing for the trafficking of occludin to the plasma membrane TJ site.87 Among the extracellular loops, a MARVEL domain (MAL and related proteins for vesicle trafficking and membrane link) is present and is common among junctional proteins involved in membrane binding.73 Occludins have been found to be non-essential to the morphological formation of

TJs.88 However these occludin-deficient tissues display chronic inflammation and poor TJ integrity. Conversely, overexpression has been found to increase transepithelial electrical resistance (TEER). 89 Adding synthetic portions of the extracellular loops, thereby non- competitively binding occludin binding sites, increases epithelial permeability90 which leads to the conclusion that while not essential for TJ formation, occludins are involved in modulating paracellular transport and strengthening cell-to-cell binding.

1.4.2 Adherens junctions

Adherens junctions are located basal to TJs and connect lateral membranes of epithelial cells.71 They are highly dynamic, in vivo and in vitro91, as they are required to release accumulated tension within the cellular monolayer, and to allow for movements of the cell sheet.9 One of the well-known adherens junctions is the zona adherens (ZA) which ultrastructurally appears as a parallel alignment of adjacent cellular membranes over a distance of 200Å with an approximate length of 0.2 -0.5μm. 72 Cytoplasmically, ZAs appear as dense plaques made up of the various cytoskeletal mediator proteins, and the cytoskeleton itself. Actin forms a circumferential belt with myosin that functions as a tension cable that can internally brace a cell and control its shape.92 The primary adhesion molecules in AJs are cadherins.

1.4.2.1 Cadherins

Cadherins were the first family of adhesion molecules found within the AJs. 70 They are a type I single-pass transmembrane glycoprotein that mediate Ca2+-dependent intercellular adhesions through its extracellular domain. 93 The cytoplasmic domain mediates cellular signalling and structural activities through the catenins. 94 In particular, β-catenin has been shown to bind directly to cadherins and α-catenin.95 The presence of α-catenin is critical for actin polymerization near AJs, which is a requirement for cellular adhesion.96 Other members of the catenin family are responsible for modulating levels of cytoplasmic cadherin, and strengthening adhesions by promoting cadherin clustering94 and local actin polymerization.97

Cadherins are ubiquitous along the lateral surface of polarizing epithelium forming homophilic interactions with cadherins on other cells. 98 The adhesiveness of cadherins have been found to

19 be dependent on the presence of extracellular calcium. The Ca2+ binding sites located between extracellular cadherin repeats rigidify the cadherin oligomers when bound.99

Cadherin based AJs are first created through the complexing of the C-terminal domain of cadherin to β-catenin, which then localize to regions of cell-cell contact. 100 Cadherins upon adjacent cells then dimerize, forming the cell-cell adhesions which ‘zipper up’ cadherins into clusters. 101,102 The multiple low affinity interactions are additive and compound to create very tight intercellular adhesions. As clusters of cadherin and β-catenin form, they attract α-catenin which promotes local actin polymerization in parallel bundles, creating the ‘belt’ of cortical actin seen in polarized epithelium.103 ZO-1, a familiar player in junction dynamics, is recruited by the catenins and further reinforces cortical-actin cable formation leading to junctional stabilization.

104,

1.5 Hemidesmosomes

Hemidesmosomes (HD) are adhesive junctions that appear restricted to epithelial cells.105

These junctions function to connect the cytoskeleton through adaptor proteins within the basal component of the cell to confer cell shape and rigidity. They are responsible for maintaining the adherence of epithelial cells to the extracellular matrix.106

Ultrastructurally, HDs are small, electron-dense cytoplasmic plaques which possess intercellular regions that are enriched for cytoskeletal components close to the cytoplasm. 107

Perimembranously, the plaque contains the cytoplasmic tails of the various transmembrane hemidesmosome components that interact and signal into the cell body. 108

HDs are made up of an array of transmembrane, intra-and extra-cellular proteins that enhance the adhesiveness of the junction. α6β4 integrin is the principal factor responsible for the

20 essential connection of the intermediate filaments (IFs) to the ECM and maintaining attachment to the basement membrane (BM).109 The absence of this key integrin results in the absence of the

HD.110 In stratified epithelium, α6β4 is organized in a polarized manner along the basal surface.

111 The β4 integrin component possesses a long cytoplasmic tail, which confers specialized function through binding to plectin which anchors to the intermediate filament cytoskeleton, stabilizes the HD, promotes adhesiveness, and finally binds to laminin-332.112 Finally, α6β4 integrin functions as a nucleating site for the formation of the HD and as a protein backbone in the HD centred signaling cascades both into and out of the cell. Plectin is a 500kDa plakin family member that is a mediator between the HD and cytoplasmic cytoskeletal linker; it can dimerize within itself, interact with IFs, and the multiple domains of the β4 integrin tail. Plectin clusters

α6β4 integrin molecules at the basal cell surface, and finally stabilizes the HD through cytoskeletal connections.109 There are two levels of interaction between plectin and α6β4 integrin: the first level is between the actin binding domain and β4113, not actin; the second level is the plakin domain and a more C-terminal part of the connecting segment of β4. 114 Laminin is the extracellular matrix component that is necessary for firm attachment of basal epithelial cells to BM, and binds to α6β4 integrin. 115

HD assembly and maintenance is dependent of the α6β4 integrin, plectin, and laminin-

322. The β4 tail serves as assembly point for various proteins, including plectin, and extracellular laminin-322. Formation of HD is driven from within: the binding of cytoplasmic tail of β4 integrin is the first step in formation of HDs, and from outside, α6β4 integrin binds to cleaved laminin-332. 116 If laminin-322 is left uncleaved, the cell becomes motile, and is ready to migrate. 117 HDs are dynamic junctions that undergo consistent remodeling which aids in differentiation. 118

1.6 Cilia

A marker of airway epithelial maturation is the formation of motile cilia from the apical aspect of the airway epithelium. Ciliogenesis begins with the formation of a primary or non- motile cilium followed by absorption, followed by the emergence of motile cilia119.

1.6.1 Cilium structure

Cilia are microtubule-based, finger-like membranous projections of the cell membrane

(Figure 4). While cilia are continuous with the apical aspect of the airway epithelial membrane, the ciliary membrane is markedly different from the cell membrane. The ciliary membrane has receptors that make it responsive to signals that modulate ciliary motility such as a purinergic compounds,120 cations121, and cyclic breathing. 122 The internal structure of a cilium is made up of the axoneme, a microtubule based structure that emerges from the ciliary basal body123. The structure of the axoneme is usually found in one of two conserved forms: nine microtubule doublets 124 surrounding a central doublet, 9+2, or nine microtubule doublets

Adapted from125

surrounding no central microtubule

doublets, 9+0 (Figure 4). The 9+2

arrangement is indicative of motile

cilia, while 9+0 is indicative of

primary, non-motile, sensory cilia.

126

The surrounding nine

doublets, or outer doublets, are

made up of one complete A-tubule,

with 13 protofilaments, and an

attached but incomplete B-tubule,

comprising 10 protofilaments made

up of alpha and beta tubulin

monomers. These tubulin Figure 4. Ciliary structure. Cilia are anchored to the cell through a tubulin based structure, the basal body, monomers are modified through which serves as the nucleation site of the axoneme which 127 emerges through the ciliary pocket. The axoneme of the post-translational modification. primary cilium is formed from 9 microtubule doublets, while the motile cilium has an extra doublet in the middle Outer doublet microtubules are of the nine outer doublets. The inner doublet is connected to the outer doublets through radial spokes which provides connected to adjacent neighbours structural support. Dynein arms connect the outer doublets to each other. Reprinted by permission from Macmillan through protein-linkers known as Publishers Ltd: NATURE, 125 2011. nexins. Nexin linkages prevent the

23 movement of the outer doublets with respect to each other.128 Radial spokes, which are complex multi-proteins stabilizers of the axoneme, are attached to the A-tubule through a stalk which is connected to a head, which projects towards the central microtubules129.

The drivers of motile ciliary movement are axonemal dyneins, which are arranged in two rows along the length of the A tubule. 130 Those dynein arms that run closest to the ciliary membrane are known as the Outer Dynein Arms or ODA, while those running closest to the central pair in 9+2 motile cilia are known as the Inner Dynein Arms or IDA. Axonemal dyneins are ATPase driven motors that cause the sliding of adjacent microtubule doublets by interacting with the adjacent B-tubule. 131 As nexin linkages lock the outer doublets with respect to each other, the constraints caused by inter-doublet sliding translate into ciliary beating, beginning from the base of the cilium.132 As the B-tubules terminate distally within the axoneme, the cilium is left with a complete singlet A tubules, along with the central doublet found only in motile cilia. The termination of the doublets, transitioning to singlets, is a conserved feature of motile and primary cilia.133 In multi-ciliated respiratory epithelium, singlets are reinforced by protein linkers which enable more effective transmission of force resulting in better propulsion of the mucus in the trachea. 134

1.6.1.1 Transition zone

The transition zone of the cilium links the axoneme to the basal body, a microtubule organizing centre (MTOC) critical for ciliary assembly.135 The basal body is similar to a centriole as it functions as the nucleation point for the minus ends of the microtubule doublets

The basal body is made up of 9 triplet microtubules, A, B, and C, which anchor the cilium to the cell body. The ciliary axoneme is an extension of these triplets as doublets. The end of the 9

24 triplet microtubules marks the beginning of the transition zone, which extends until the basal plate where the central pair in motile cilia is nucleated.136 Immature cilia have a dense cloud of material surrounding their respective basal bodies which is comprised of proteins that will be used in the construction of cilia.137 There are various structures associated with basal bodies, including centrin fibers, which are calcium binding proteins essential for the replication of basal bodies.138 Also found with basal bodies is the Outer Dense Fiber (ODF) proteins139. ODF2 has been found to be essential for mammalian ciliogenesis, in particular the formation of basal feet, which anchor basal bodies to the apical cytoplasmic microtubule network. 140

1.6.2 Ciliogenesis

1.6.2.1 FoxJ1 – Master programme initiation

The initiators of motile ciliogenesis come from the gene family of FoxJ1 transcription factors141. FoxJ1 is a member of the forkhead transcription factor family, which, when absent from mouse models, leads to a lack of motile cilia,142 suggesting that these transcription factors are critical to the formation of motile cilia.143 The function of FoxJ1 has been found to be in the organization of basal bodies. This organization of basal bodies appears to be through the activation of a Rho GTPase, RhoA, to create the highly organized apical actin mesh characteristic of ciliating epithelia.144 While responsible for creating the docking site for basal bodies, Foxj1 also regulates the gene expression of ciliary components necessary for motility.143

1.6.2.2 Timing

Primary and motile cilia are disassembled before mitosis, as the centrioles which are a part of the basal bodies are needed to organize the microtubules that will participate in the partioning of genetic information and subsequent inheritance by daughter cells. After division,

25 cilia typically form during the G1 or G0 portion of the cell cycle. This is of particular importance for multi-ciliated cells as they serve their physiological function in situ, and are required not to move or change position. Multi-ciliated cells are terminally differentiated, non-motile, and do not undergo mitosis. The timing of ciliary emergence is tightly controlled by the cell cycle and the age of the centriole; older centrioles function as nucleation cites of the axoneme, much sooner than younger, newly created centrioles. 145 Primary and motile cilia are dynamic organelles that can be physically cleaved and removed from the cell body or can be reabsorbed through disassembly.146, 147

1.6.2.3 Basal body docking and nucleation

Basal bodies originate from the mother centriole present from the previous round of mitosis. 148 In multiciliated cells, multiple basal bodies are created that serve as nucleation sites of ciliogenesis, while in primary cilia there is only one basal body.149 As the basal bodies are replicated and mature, they are transported to the cell surface through interactions of the microtubule cytoskeleton with distal appendages, like the basal feet, and ciliary rootlets.139 These interactions which are driven by the planar cell polarity pathway150 guide the basal bodies to dock onto the actin-rich cortex.140 The position and orientation of the newly docked basal body dictates the future alignment of the yet to be formed cilium. Structures associated with the now docked basal body, such as the striated rootlet, are positioned in an orientation that is opposed to the final beat direction.140 Once docking has occurred the ciliary necklace develops, and the ciliary axoneme begins to elongate.151

The basal body serves as the nucleation point of the axonemal microtubules. With the minus gamma tubulin composed end associated with the basal body, αβ tubulin heterodimers

26 with the (+) end oriented towards the emerging ciliary tip polymerize.152 As the axoneme elongates, outer and inner dynein arms attach to the microtubules, which enables the doublets to slide with respect to each other.153 The inner dynein arms control the ciliary waveform during motion,154 and the outer dynein arms control the frequency of ciliary beating through modulation of doublet sliding.155

1.6.2.4 Intraflagellar transport

The mechanisms of moving the cellular construction material necessary to build the cilium is through the bidirectional movement of multi-protein complexes along the axoneme, known as intraflagellar transport (IFT). 156 These multi-protein complexes are known as IFT trains based on their appearance and are found in two distinct varieties: a long variant that is approximately 700nm in length with a periodicity of 40nm and a short variant which is 250nm in length with a periodicity of 16nm. 156 The long variant functions in ciliary tip directed or, anterograde, transport, driven by kinesin-2.156 Short IFT trains are responsible for retrograde, away from the ciliary tip, transport driven by the molecular motor dynein 2.156 These trains are composed of two types of particles, Complex A and B. 157 Complex A is used in retrograde transport and has been found to be necessary for ciliary disassembly, not assembly. 158 While

Complex B is found in anterograde transport and is critical to the assembly and maintenance of primary and motile cilia, the loss of Complex B particles has been found to result in stunted and absent cilia.159

1.6.3 Primary cilium

Initially considered a cell-biological curiosity, then evolutionary dead end, primary cilia have been found to play a role on epithelial cells like kidney collecting duct, bile duct, endocrine

27 pancreas, and the thyroid as well as some mesenchymal tissues such as chondrocytes, fibroblasts, smooth muscle cells, and neurons. 123 Their immotility stems from their lack of a central doublet, the associated structures of ODA and IDA, and radial spokes.132 While immotile, they have been shown to sense physical and biochemical signals.160

1.6.4 Motile cilium

Motile cilia beat cooperatively and in a coordinated fashion to generate fluid flow.

Driving this cooperative movement, motile cilia utilize the two extra microtubules, central pair, in the centre of the axoneme, along with the attached radial spokes and dynein arms. 130 Motile cilia are found along the female reproductive tract to move along ova towards the uterus, on ependymal cells lining the ventricles of the brain, and respiratory epithelium where they function to remove foreign bodies by beating mucus in an oral direction. Motile multi-ciliated cells have rigorously organized microtubule and actin cytoskeletons associated with their cilia at the apex of the cells.140 Ciliated respiratory cells typically possess 200 cilia with uniform dimensions of

6μm and 0.2μm diameter.161 Normal human cilia bend in a wave-like motion, and have differential rates of beating depending on the organ in question and the maturity of the individual.162

1.6.4.1 Mechanism of motion

The function of motile cilia is to generate fluid movement. In the airway, fluid movement that is laminar is most effective in moving mucus out of the airway.163 Based on conserved cellular mechanics, ciliary beating is divided into two phases: a vertical power stroke which moves fluid and a horizontal recovery stroke through the periciliary fluid that primes the motile cilium for the power stroke. The orientation of the power stroke is fixed perpendicular to the

28 central pair orientation. As the dyneins motors that drive ciliary beating are unidirectional motors, there must be two sets of them to drive both phases of ciliary beating, one for power stroke and one for recovery stroke.123

When motile cilia initially begin beating early in their development, their coordination and orientation are driven by genetic cues.164 These genetic cues produce weakly directional flow that functions as part of a positive feedback loop, in that genetics provide a basic orientation of beating which is further amplified by the continuous single direction flow, which serves to further reinforce the direction of beating and coordinate it.165 When groups of cilia become activated to beat in a coordinated fashion, they initiate what is known as a metachronal wave across the epithelium. 166 Metachrony is produced by the beating of each motile cilium in sequence, as opposed to in synchronization, which appears as the propagation of wavelike motion across ciliated airway epithelium.

1.6.4.2 Organization of cilia

Cilia are organized in vivo through fluid flow initiated through planar cell polarity (PCP) protein signalling, which has been proven as necessary and sufficient in determining cilia polarity. 165 Fluid flow works through a positive feedback mechanism where initial polarized beating leads to a stronger signal with more polarized beating. 167Other PCP proteins have not only been found in initiating and maintaining polarized beating, but at the base of cilia.164 Motile cilia and the cytoskeleton have been found directly interacting with basal bodies, the nucleation sites of cilia. 139 Two distinct pools of actin form at the apical surface of multiciliated cells. 140

One pool is at the apex of the multiciliated cell; the second is 0.5um below the cell surface and links the basal body of a cilium and the distal tip of the striated rootlet of the posterior neighbour

29 along the axis of cilia beating. This secondary actin population is key to global ciliary organization as it organizes the distribution of basal bodies. Microtubules inhibition leads to disorganized ciliary beating within a multiciliated cell. Microtubules have been found to directly connect neighbouring basal bodies, with associated structures organized with reference to the direction of beating. Together, actin and microtubules form highly complex apical structures that directly interact and govern the direction of ciliary beating and global coordination.

1.7 Mechanobiology

Cytoskeletal components are integral to epithelial differentiation and polarization. In particular, the organization of multiciliated cells appears to be largely driven by highly organized apical actin and microtubule populations.140 To modify actin and microtubule cytoskeletal components, devices that manipulate the organization of the cytoskeleton in developing epithelium could be used to organize cilia. Such a device would allow for study on the effect that different mechanical forces have on the epithelial cytoskeleton and add to the growing knowledge base of mechanobiology. Mechanobiology is a broad field that studies how cells translate external and internal mechanical forces into biochemical signals that alter cell biology, development, and pathology.

Early work in mechanobiology broadly focused on the apparent relationship between mechanical forces and tissue maturation and remodelling. For example, the earliest work in mechanobiology is from Julius Wolff (1892), a German surgeon who theorized that along lines of mechanical stress, bone structure alignment was produced. In the modern era, cells producing mechanical forces and deforming substrates was first studied in the work of Harris and Stopak

(1980).168 The field has considerably matured since then, and it is widely accepted that cells interact with their substrates and ECM through the generation of mechanical forces, and are

30 instructed by mechanical cues in the same substrates and ECM leading to changes in morphology, cytoskeletal structure, and gene expression. Of great importance to mechanobiology is the question of how mechanical signals are translated from outside the cell in, and from inside the cell out; a process known as mechanotransduction.

Mechanical signals from the exterior environment of the cell are usually found in the

ECM. The various components of the ECM, such as collagens, laminins, and fibronectins bind in a variety of ways and lead to varieties of mechanical properties such as stiffness, and topography. These mechanical signals are translated into the cell and conducted to the cytoskeleton. The cytoskeleton is made up broadly of actin filaments, intermediate filaments, and microtubules.92 As the cytoskeleton provides mechanical support to organelles throughout the cell, these mechanical signals from the external environment influence the entire cell. A major player in interacting and translating these various mechanical factors into the cell from the ECM and adjacent cells are integrins and adherens junctions, respectively.

Integrins are transmembrane heterodimers composed of an extracellular head that interacts with ECM components and two intracellular tails, which interact with cytosolic cellular components critical in mechanotransduction.17,169 Integrins when bound to ECM components are activated and through their cytoplasmic tails bind to focal adhesion kinase (FAK) and paxillin, which binds to α-actinin, which directly binds to filamentous-actin or F-actin. 170 The convergence of these adaptor proteins initiate the creation of nascent adhesions which could ultimately form a specialized attachment site, known as a focal adhesion.92

Focal adhesions begin as nascent connections composed mainly of integrins which, as they mature, are known as focal complexes, and finally mature into focal adhesions. The maturation from a nascent connection to a focal complex and then adhesion is not always necessary for cell

31 attachment, since cell motility and contractility processes require cell-ECM connections to be dynamic and easily moveable. Studies have found that force is required for the maturation of focal adhesions through alteration of focal adhesion complex protein quaternary structure in such a way to reveal alternative hidden or additional protein binding sites for further addition of other adaptor proteins leading to maturation of the nascent complex. 169,171,172 Such forces are applied to these adaptor proteins, such as talin, while under mechanical stress from F-actin stress fibres bound to the developing adhesion complex. As integrins are mechanically strained proteins and connect to the actin cytoskeleton, this strain is transmitted into the cell173 and induces remodeling of the cell architecture.174

As cells adhere they begin interacting with their substrates and migrating over them. For this migration to occur integrins must quickly attach and detach. The region of interest in migrating or motile cells is the actin rich lamellipodia, which constantly extend and retract across the

ECM175, probing the external environment for mechanical and chemical cues, as well as future binding sites.

While integrins are responsible for attachment to the ECM, integrins also translate information to the ECM from the cell through changes in F-actin cytoskeleton tension changes, which can alter substrate mechanical properties. This is known as bidirectional signaling, and leads to the creation of a regulatory loop in which ECM mechanical cues guide alterations in cell architecture; changes in cell architecture promote changes in the ECM, leading to the creation of a feedback loop where cell and ECM consistently alter each other and communicate these changes and alterations through focal adhesions, particularly integrins.176

The actin cytoskeleton participates in transmitting signals out of the cells through focal adhesions by altering the mechanical properties of the actin cytoskeleton. Tension is loaded onto

32 the F-actin cytoskeleton through the dynamic pulling of myosin II heads, which leads to tension in the F-actin fibers, which through their connections to the focal adhesion ultimately exert a dynamic mechanical force on the ECM.

An example of how the mechanical environment affects cell morphology is seen in the substrate stiffness work of Discher and colleagues.177 They have found that stiffer substrates lead to increased focal adhesions resulting in an increased activation of RhoA, which leads to maturation of these focal adhesions and the F-actin cytoskeleton, resulting in cell motility.

Increased substrate stiffness leads to increased cell spreading, contractility, and proliferation178; while on compliant substrates, cells spread less due to decreased focal adhesion formation resulting from decreased integrin activation, which leads to less active RhoA, F-actin formation, and ultimately decreased mechanical forces.68 In fact cells have been found to preferentially migrate from regions of compliant substrates to stiff substrates, a process known as durotaxis. 170

While integrins are responsible for translating mechanical information from the ECM-cell and from the cell-ECM, more apically located on epithelial cells in particular is another form of mechanotransducer, the adherens junction (AJ). While the properties and importance of AJs were documented earlier in this literature review on adherens junctions contributing to epithelial barrier and selective transport functions, they are also critical in mechanotransduction from cell to cell. AJs mature with continuously applied physical forces, and these forces increase leading to orientation changes of the F-actin cytoskeleton, leading to the creation of the actinomyosin belt which has contractile tendencies on epithelial architecture.103 The usual cobblestone phenotype of mature epithelium arises from the interplay between the cortical actinomyosin belt, which drives cells to contract and thereby become more circular, and the AJ they connect to, which essentially pins one cell to another leading to confluent epithelial monolayers. Without

33 remodelling of both junctions and the F-actin cortical belt, mechanical forces would tear the epithelium apart. These morphological conformations are not rigid; epithelial sheets are always remodelling through maturation and migration of monolayers.

1.8 Summative statement

Epithelial tissues line all tubular organs within the body. Epithelial tissues are organized in an organ and tissue-specific matter, where anatomy confers function, and function alters anatomy. A critical function of airway epithelium is mucociliary clearance, in which motile cilia beat in an aligned and coordinated fashion to move mucus and foreign material out of the airway.

This function is critical to maintaining the airway and keeping our respiratory tract clear of foreign material. Once patterned during development, airway epithelium begin to differentiate through polarization of the epithelial cells. This polarization is genetic in origin and causes physical segregation and differential protein trafficking, leading to epithelia with different apical and basal compartments. Airway polarization is usually described in terms of apical junction formation, as well as motile cilia formation. Mechanical stimuli have been shown to induce alteration in airway epithelial differentiation, and could be exploited to promote more physiologically relevant differentiation and polarization.

Chapter 2 | Rationale, hypothesis, and aims

2.1 Rationale Multiciliated, motile epithelial cells are organized along two axes of polarization: an apical – basal axis, and a planar axis. The apical – basal axis is seen through the formation of various apical structures such as tight junctions, adherens junctions, and motile cilia. Basolaterally, structures such as the sodium-potassium ATPase pump, hemidesmosomes, and integrins localize to this compartment. Thus, the apical compartment of the epithelium is made up of different proteins than the basolateral compartment. The planar axis of polarization is seen through the directionality of motile ciliary beating. Cilia beat in metachrony to move mucus out of the airway to be coughed out of the body, or swallowed into the gastrointestinal tract. This beating which moves mucus in one direction is known as mucociliary clearance and is critical to airway and respiratory health. Thus, organized epithelium, in particular, organized ciliary beating is required in vivo for maintenance of the airway epithelium.

There are many examples of diseases with dysfunctional cilia, both primary and motile. A primary ciliary disease such as polycystic kidney disease is associated with defective calcium-ion signalling from the cilium into the cell body, leading to the associated ciliopathy. This cation channel has been found to act as a chemical or mechanical sensor in renal epithelia.179

Ciliopathies of motile cilia involve defects in ciliary beating, leading to disorganized ciliary movements, which in the airway could result in dysfunctional mucociliary clearance. The prototypical disease of dysfunctional or non-functional motile cilia is primary ciliary dyskinesia

(PCD).

2.1.1 Primary ciliary dyskinesia Primary ciliary dyskinesia or PCD, is a genetically heterogeneous disorder that is characterized by primary ciliary dysfunction that leads to, amongst other symptoms, impaired mucociliary clearance in the airway.

PCD is found in 1/16 000 births.180 Generally, the mean age of diagnosis is four years of age.181 Almost all patients present with a chronic, productive cough, recurrent respiratory infections such as rhinosinusitis, otitis media, bronchitis, and pneumonia. 182 In the lower respiratory tract, the chronic and productive cough found in all PCD patients is used as a compensatory mechanism in the absence of mucociliary clearance over short term. 183 Along with chronic cough, pneumonia and bronchitis are key symptoms of PCD in the lower respiratory tract. In addition to the airway, infertility is seen in most male PCD sufferers due to impairments in sperm motility, 184 whereas female fertility issues have been found to be more variable. 185 The current gold standard diagnostics tool used has been the use of transmission electron microscopy on the nasal or bronchial brushings of suspected PCD sufferers.186

PCD was first known as Kartagener’s syndrome after he described his now famous triad of pathologies consisting of situs inversus, chronic sinusitis, and bronchiectasis occurring together in his patients in the 1930’s. 187 Years later, it was noted that patients suffering from

Kartagener’s syndrome had what appeared to be immotile and dysfunctional cilia. Upon further examination of the ultrastructural organization of these non-functional cilia, it was found that they had defects in their ultrastructure. 188 While not immotile, as a result of the defects in ciliary ultrastructure, the cilia were stiff, uncoordinated in their beating, and ineffective in removing foreign bodies from the airway.

The defects associated with PCD patients are apparent in the motile cilia found throughout the body, including the airway, middle ear, and oviduct. This universality of the ciliary defects points towards a generalized disorder of genetic origin.161 Genes that code for proteins involving components of the ciliary axoneme have been found to be mutated or missing. Many patients with PCD have been found to have a complete or partial absence, or inappropriate localization of the outer or inner dynein arms that function as the molecular motors of ciliary motility.189

Further to just the molecular motors of the cilium, radial spoke have been found to be absent or defective in patients with PCD.190 Without radial spokes to guide the organization of the central microtubules pair, cilia have morphological and beat defects.

There is currently no prescribed set standard of care for PCD patients. Most treatments only manage symptoms arising from PCD, in that every effort is made to keep airways clear through vigorous coughing, avoiding cough suppressants, exercise to stimulate deep breathing, and close watch and treatment of bacterial infections that may arise. 191 Once a patient has reached end- stage lung disease, transplantation is one of the only viable interventions that has been shown to be successful.192 Thus, devices that organize epithelium, in particular motile ciliary beating, would be useful to clinicians and would serve as in vitro models of functional airway epithelium.

2.1.2 How do we organize epithelium? From Soleas et al., 2012 193

Tissue engineering (TE) offers a number of strategies to achieve both macroscopic and microscopic cell organization based on the control of chemical and mechanical signals.

Applying TE strategies to organize airway cells into specific and controlled structures will improve the clinical performance of these cells.

The gold standard for the repeatable manufacture of adult airway epithelium in vitro is transwell culture 194. Transwell culture is based on a two compartment culture where primary

38 respiratory epithelial cells are seeded on porous, collagen-coated membranes in liquid culture.

After reaching confluence, liquid from the top compartment is removed, leaving the epithelial sheet exposed to air. This is known as air-liquid-interface (ALI) culture. Over a two week maturation period, epithelial cells form motile-cilia at the apical surface signifying apicobasal polarization. However, transwell culture does not create correctly aligned epithelium with coördinated beating of motile cilia 195. A myriad of TE tools exist to direct cell organization.

These tools, when adapted for epithelial TE, may prove useful for generating more appropriate cell-organization and ciliary alignment in in vitro epithelium.

It is well known that cells are instructed by the materials they grow on, and modify the surfaces they grow on over time. Regulating these instructive signals over time and space is a key challenge of TE. A wide variety of tools have been developed to study the effect of different chemical and mechanical signals on cell behaviour. Most TE tools, however, have been developed for endothelial, muscle, and nerve cells. These cell types do not polarize in an apicobasal fashion and are grown on solid culture substrates. Little work has been reported using these tools to organize epithelium, due to the necessity of special culture conditions required to produces a functional epithelium.193

Airway tissue engineering challenge

In order to apply TE strategies to align structural components of epithelial cells, it is necessary to adapt existing methods for use on the porous membrane of a transwell plate that allows nutrient diffusion to the apical surface of the cells. Here we describe some tools that are currently used in TE, which have the potential to be relevant and useful for engineering epithelium if adapted appropriately. These tools can be classified based on the signal type and method presented. As seen in Figure 5 we will focus on chemical and mechanical signal types.

Chemical signals can be presented in a mobilized or immobilized state while mechanical forces

can be presented in a constant or inducible fashion.

2.1.2.1 Chemical signals

Chemical signals can be immobilized on biomaterial scaffolds in a graded fashion to

guide cell movement and organization (Figure 5A) .196 For example, using an immobilized

From 193

Figure 5. Examples of the tools of tissue engineering. Tools that manipulate the timing and appearance of chemical and mechanical signals offer opportunities to organize and direct the differentiation of developing tissue. Chemical signals can be immobilized, (A) in the form of covalently bonded growth factors that direct cell migration or mobile in a hydrogel, (B) to create a chemotactic signal, through diffusion, for cells to respond to. Mechanical signals can be presented as an inducible force, (C) such as shear flow, to organize cells in the direction parallel to flow or as a constant force, such as substrate stiffness, (D) to modulate cell spreading. From Soleas, et al., 2012.

40 concentration gradient of nerve growth factor and neurotropin-3 on a poly(2-hydroxyethyl- methacrylate) and poly(L-lysine) Moore et al., (2006) were able to guide neurite outgrowth of primary neurons 197. The effect of utilizing two growth factors together was shown to increase the biological response in chick neural cells. This approach has been used to successfully guide the behaviour of fibroblasts198, endothelial cells199, osteocytes,200 and human mesenchymal stem cells201.

Immobilized chemical signals on biomaterials could provide a useful tool for epithelial

TE as multiple growth factors acting together could promote more physiological tissue proliferation, motility, and differentiation in a respiratory epithelial model (Figure 6A). Patterns of immobilized growth factors202 could prove to be of great use in epithelial TE. As well, the difference in epithelial growth and differentiation in wounding experiments could be studied as cells migrated from areas containing immobilized growth factors to areas without growth factors or vice versa.

Another use of immobilized chemical signals on a scaffold is to drive cells down a specific differentiation pathway. A single growth factor on a scaffold to multiple growth factors on solid substrates have been shown to modulate oligodendrocyte differentiation196 and stem cell fate 203. This approach of presenting a chemical signal using a biomaterial to guide differentiation is conceivably useful in epithelial TE as certain differentiation pathways leading to specific lineages could be developed as a model, or a graft of distinct areas of respiratory epithelium.

While the above techniques have been developed for surface culture, the encapsulation of cells within a hydrogel presents an opportunity for cells to be delivered to necessary sites both in vitro and in vivo within a chemically defined 3D environment. Hydrogels can present different

chemical groups and can be bio- or non-degradable over time. Guiding cells using immobilized

chemical signals in defined 3D environments within hydrogels has been seen to have great value

From 193

Figure 6. Specialized exemplar tools of epithelial tissue engineering. Chemical signals can be presented as immobilized growth factors (A) that promote differentiation of airway basal cells to specific cell types in a pattern that is reminiscent of in vivo airway epithelium, or (B) a mobile chemokine gradient of CXCL12 that promotes airway epithelium polarity in the presence of Wnt5a, based on the work of Witze et al., 2008. Mechanical signals can be presented as an inducible force (C) that mimics the transluminal pressure gradient applied to airway epithelium during normal tidal breathing to modulate ciliary beat frequency or as a constant force that organizes epithelial cells cultured on nanogrooved and flat substrates (D). From Soleas, et al., 2012 .

42 in treating retinal degenerative diseases204 and spinal cord injuries 205. While the majority of hydrogels with immobilized signals has been developed for non-epithelial cells, some materials for epithelial applications are already available: to create an oral mucosa equivalent, Kinikoglu and others (2009) developed a co-culture system on a scaffold that presented specific chemical properties 206. Fibroblasts and oral epithelial cells were seeded on this scaffold to create stratified and differentiated epithelium-like oral mucosa. In a refinement of their research,

Kinikoglu and colleagues (2011) used recombinant DNA technology to develop an epithelial TE tool that presented the RGD peptide sequence within a biocompatible polymer that was then electrospun onto elastin and collagen foam, thereby creating a 3D co-culture system of fibroblasts and oral epithelium on scaffolding that presented a static chemical signal to promote specific types of integrin binding 207.While these tools were developed for oral epithelium, their adaptation to air-liquid-interface culture would be a useful application for respiratory epithelial maturation. Lin and colleagues (2006) studied the efficacy of polyglycolic acid (PLGA) as a hydrogel matrix for lung tissue engineering208, while Cortiella and colleagues (2006) did a comparative study of PLGA and Pluronic F-127 (PF-127) hydrogel constructs impregnated with lung cell progenitors209. Both found evidence that suggested that PLGA would be an excellent lung matrix substitute in vitro. The construct was capable of producing specific respiratory epithelial proteins: Clara cell protein 10, and cytokeratins; however, in vivo, these constructs induced potent inflammatory reactions that ablated appropriate epithelial morphogenesis. These results lead to the conclusion that selection of the polymer based on chemistry is very important to creating functional tissue. This shows that while the immobilized chemical signals of PLGA and PF-127 are not ideal for epithelial morphogenesis, a polymer with the correct chemical patterning would facilitate more physiologic respiratory epithelial genesis.

Chemical signals in TE can also be presented to the cell in the form of diffusible and mobile chemical signals released from a material or scaffold (Figure 5B). In a classic example

Richardson and colleagues (2001) developed a polymeric system for dual growth factor delivery that led to differential release kinetics of growth factors and altered the timing of the chemical signals.210 The diffusible chemical signals directed endothelial cell migration to generate vascularized tissues. Single growth factor delivery systems have shown promise in promoting the differentiation and maturation of embryoid bodies211, adipose derived stem cells212, and angiogenesis213. More complex systems of sequential and combinatorial delievery of growth factors on cell-laden scaffolds have been developed for fibroblast culture214. These growth factor delivery systems could be relevant in epithelial TE epithelium as altering the presentation of one or a combination of growth factors; these could be used to discern more elegant and physiologically relevant spatiotemporal effects on epithelial developmental processes, as well as increase cell viability and engraftment in in vivo models. In particular, organization of respiratory epithelium could be controlled by generating gradients of growth factors. Wiltze and colleagues (2008) described using a chemical gradient (CXCL12) to create polarized structures in response to Wnt5a in a melanoma cell line 215. This technique to create polarized structures could conceivably be adapted to create organized-ciliated respiratory epithelium if cells were exposed to a similar gradient of CXCL12 (Figure 6B).

While the effect of growth factors on airway epithelial morphogenesis have not been studied, the manipulation of mucociliary clearance by altering chemical signals present in the maturing epithelium is well documented.53,121 For instance, it is well known that bitter compounds, such as the metabolites of resident bacteria found in cystic fibrosis patients, promote increased mucociliary beating52. Increased calcium and zinc ions accelerate the rate of

44 mucociliary beating 121 as does serotonin in the trachea in an acetylcholine independent pathway.

The chemosensory nature of the epithelium could be exploited through a chemotactic signal embedded in a hydrogel that in a controlled fashion releases the signal, which, that increases mucociliary clearance and promotes a healthier, more clinically relevant epithelium.

2.1.2.2 Mechanical signals

In additional to chemical signals, mechanical signals can be controlled in the cell environment to guide cell behaviour. The mechanical environment sensed by cells can also be modulated by the application of an inducible external force. One of the most common examples of an inducible mechanical force is shear flow to induce cell alignment (Figure 5C). Shear flow has been shown to align cells in the direction of flow and to alter responses to biological signals most clearly in endothelial cells216. The large body of work using shear flow to modulate endothelial cells has looked at how flow induces the organization of endothelial cells in the direction of flow, 216 and modifies the inflammatory response. 217 Examples of shear flow used to modulate epithelium are scant within literature; however, the organized ependymal ciliary beating of rat brain ventricle epithelium in shear flow conditions has been studied167. Applying dynamic shear forces to developing airway epithelium might be very useful to recapitulate physiologic development. In utero fetal breathing movements in amniotic fluid and adult inspiration and expiration of air are both examples of shear flow that could induce the maturation of airway epithelium.

In vivo, there are two inducible mechanical forces exerted on the respiratory epithelium: airflow-induced shear stress and trans-epithelial pressure.53 Button and colleagues have developed two tools to deliver dynamic mechanical forces to the respiratory epithelium: an oscillatory shear stress inducing device which mimics inhalation and expiration mechanical

45 stresses, 218 and a compressive stress device that applies trans-epithelial pressure gradients122,

(Figure 6C). In combination with the findings that mature human airway epithelium is most sensitive to these mechanical stressors122,218 within physiologically relevant boundaries a tissue engineering device can be envisioned that combines these dynamic mechanical forces. This device would induce increases in shear stress and trans-epithelial pressure, thereby increasing ciliary beat frequency and mucociliary clearance. This property could be used to ensure that newly created airway epithelium is kept free of foreign bodies during early morphogenesis.

Based on the anatomy and regenerative potential of the pulmonary system a variety of TE tools available could be utilized to overcome the barriers currently seen in respiratory tissue generation. Tools such as growth factor immobilization and graded morphogen release have shown great promise in epithelial and other model systems which could be rapidly adapted to a respiratory epithelial context. Other tools that manipulate substrate stiffness or topography could be used to promote organized epitheliogenesis by controlling proliferation and differentiation.

Substrate stiffness is a well-studied example of a mechanical signal that is presented in a constant manner (Figure 5D). Substrate stiffness can be utilized to manipulate cell morphology and proliferation. The classic example is the seminal work done by Pelham and Wang in 1997178 where polyacrylamide gels of different stiffness were created to study the effect stiffness has on various cell types. Their work found that fibroblasts cultured on more compliant substrates spread less and became more motile. This model was expanded upon by Discher 177 and further refined to create a high throughput technique to ascertain the appropriate stiffness for specific cell types 219. Examples where substrate stiffness can be exploited to promote specific tissue characteristics are in the heart220 and mammary epithelium221. Substrate stiffness modulation could be used on respiratory epithelium to ascertain and exploit the effect of different stiffness

46 on organization, proliferation and maturation to create a faster growing epithelial sheet that differentiates to a specific mature cell type.

Nanogroove topography

Another aspect of the environment that influence the mechanical environment sensed by the cell is the local surface topography. For example, grooves in substrates can induce organization of cells in the direction of the grooves. Topographic organization of cells has been used to modulate the phenotype of osteoblasts222, cardiomyocytes223, and fibroblasts224.

Nanogrooves specifically have been used to align epithelial cells in the direction of the nanogrooves: MDCK225, 226, human corneal epithelial cells 227, (Figure 6D), and in human mesenchymal stem cells 228.

The early work of Clark et al. (1990) looked at the effect of various grooved topographies on the alignment of three cells types: of particular interest is MDCK, a canine kidney epithelial cell line.225 Starting with a shallow groove of 0.3μm deep and 12μm in pitch, they found that single MDCK rarely aligned in the direction of the grooves. However, on deeper grooves

(1.2μm, and pitch of 12μm) the MDCK aligned while elongating parallel to the direction of the grooves. Manipulation of pitch distance (4 - 24μm) , while holding depth constant (1.1μm) revealed that pitch had little effect on the orientation of the sparse MDCK; while changing depth

(0.2 – 1.9μm) and holding pitch constant (12μm) highlighted the role depth plays in groove topographies as cells on deeper grooves were more aligned in the direction of the grooves. From their work we can conclude that for epithelial cells, increasing the depth of nanogroove topography should increase the alignment of epithelial cells parallel to the grooves. This effect of increasing depth to increase groove guidance has been confirmed in fibroblasts as well.229

Building on the work of Clark et al. (1990) Jin and colleagues (2008) studied the effect of nanogroove substrates on MDCK adhesion, cell cycling, and gene expression.226 They found that

MDCK on nanogrooves adhered better and grew faster than those on flat substrates while their actin cytoskeleton aligned in the direction of the grooves. Cell cycle and gene expression analysis agreed with their findings of faster division of MDCK on nanogrooved substrates by finding that the percentage of cells in S-phase and G2/M phase was higher than cells on smooth substrates and that cyclin D1 mRNA expression was increased on nanogrooved substrates.

Teixeira and colleagues (2003) assessed the alignment and effect of human corneal epithelium on nanogrooved silicon oxide substrates.227 They found that corneal epithelium align and elongate in the direction of the grooves (depth of 600nm and pitch of 400nm). The lamellipodia of these aligned epithelial cells grew out and retracted in the direction of the grooves. In agreement with Clark et al. (1990), Teixeira et al. (2003) found that corneal epithelium was more affected by the depth of the groove more than the pitch.

These data suggest that a variety of epithelial cells should be responsive to nanogroove topography through their morphological and cytoskeletal alignment. The instructive cues of nanogroove topography on epithelium increase with deeper grooves. The effect of altering groove pitch does not readily alter alignment in the direction of the grooves. Further, nanogroove topography appears to alter cellular adhesion, cell cycling, and gene expression, supporting the notion that nanogroove topography could be used in a TE system to organize respiratory epithelium along nanogrooves.

2.1.3 Study rationale

While there are a wide variety of TE tools that we could utilize to organize airway epithelium, we will use a constant mechanical force--that of nanogroove topography, to induce

48 alignment of epithelial tissues. As highlighted above, epithelial tissues have already been found to be responsive to the instructive cues of nanogrooves by aligning and elongating in the direction of the nanogroove topography.

The standard method to construct epithelium in vitro is through transwell filter technology creating an air-liquid-interface (ALI) culture system. While creating ciliated and therefore apicobasal polarized airway epithelium, this epithelium is not organized in a planar fashion as seen through disorganized global ciliary beat alignment. Growing cells on grooved topography is a well-known tissue engineering strategy to align cell morphology and cytoskeleton in artificial tissues. By growing epithelial tissues on nanogrooves we hope to align these tissues in the direction of the nanogrooves and therefore impose planar alignment of these tissues in a morphological and cytoskeletal fashion. To create this planar polarization within ALI culture, we must create these nanogrooves in substrates that will allow for the diffusion of nutrients from the basal compartment. By creating a nanogrooved diffusible substrate, this system should be congruent with ALI culture.

2.2 Hypothesis

It is well known that nanogrooved surface topography induces alignment and elongation of both cellular morphology and cytoskeleton. It is hypothesized that nanogrooved surface topography created in a diffusible substrate for use in air liquid interface culture will induce alignment and elongation of epithelial morphology and cytoskeleton on the microscale in the direction of the grooves.

2.3 Aims

This study aims to (1) create a nanogrooved hydrogel, (2) characterize the morphology and cytoskeletal alignment of epithelial cells on these nanogrooves using phase and light microscopy,

49 and (3) characterize the congruency of this device with air liquid interface culture by assessing apicobasal polarization of airway epithelium.

Chapter 3 | Device manufacture

3.1 Introduction

Problem

To study the alignment of epithelial cells on a nanogrooved substrate congruent to ALI culture we must first find a suitable substrate that allows for the basal diffusion of nutrients and is capable of holding nanogrooved topography in standard tissue culture conditions.

Aims

To create this nanogrooved substrate that allows for basal nutrient diffusion we aimed to (1) find a simple technique that could be used in any standard biology laboratory to create this grooved topography, (2) find a suitable natural hydrogel, (3) create the nanogrooved surface topography on our hydrogel of choice.

3.1.1 Replica moulding rationale

To create an easy-to-use system that did not require the resources or expertise required for soft- or photo- lithography, we sought a replica moulding technique. Replica moulding has been used extensively to create a relief structure of less durable substrates in polymers that are more robust and reusable.230 Replica moulding is used in situations where creating the sought after relief structure multiple times would be inadvisable, too complex, or expensive in the chosen polymer.

Many replica moulding techniques require specialized lithographic techniques to create initial patterned masters. From this master an intermediate mould of PDMS or polyurethane would be created which would be used to mould the polymer of choice. The major drawback of this technique, however, is the initial requirement of specialized engineering and chemical techniques to create the master mould. 231,232

Instead of creating a master of nanogroove topography through laser-etching or soft lithographic techniques which are expensive and require a great deal of expertise, we sought a commonly available, manufactured object that we could mould from. This moulding method would be quicker, less involved, and easily transferable from laboratory to laboratory than standard soft- or photo- lithography. This presented our first drawback of the moulding methodology: reliance on a manufacturer which meant that our options for varying the topography were much more limited than what we could create using lithographic techniques.

The first step in this process was finding a ready-made mould that was accessible and stable for laboratory use.

3.2 Process rationale

To begin creating our nanogrooved hydrogel substrates we required a readily available nanogrooved surface master from which to replicate our topography. While lithographic techniques could have been utilized, these technologies are not commonly available outside of the tissue engineering laboratory. As such, a manufactured nanogrooved master from optics, the holographic diffraction grating film, was used as our topographical master (Figure 7A).

3.2.1 Holographic diffraction grating film

Diffraction gratings are a tool of optics that split light across the electromagnetic spectrum.

One of the most well-known examples of a diffraction grating comes from compact discs.

Diffraction gratings have a nanogrooved surface topography that we exploit as a master for our replica moulding process. We selected a holographic diffraction grating film from Edmund

Optics based on availability and experience using it within our laboratory. Our selection was a clear-polyester-film grating with a pitch of 1 groove per micron. We have calculated the depth of

53 our grooves to approximately 250nm. From the manufacturer, the orientation of the grooves is linear and parallel to the shortest dimension of the film. As the diffraction grating is not very durable and is expensive, we chose to create a polydimethylsiloxane (PDMS) intermediate that would act as an intermediate mould for the hydrogel of choice (Figure 7B).

3.2.2 PDMS

Polydimethylsiloxane (PDMS) is an organosilicon polymer that has wide- ranging applications and is consequently one of the most widely used silicones in soft lithography.232 It has a number of favourable properties that are conducive to creating a reusable nanogroove stamp. Firstly, it is optically clear, allowing us to seed cells directly onto the stamp and use light microscopy to analyse cellular morphometry. Secondly, the stiffness of PDMS can be easily regulated based on the ratio of elastomer to curing agent. Thirdly, PDMS is easy to use and polymerizes under standard laboratory (non-hazardous) conditions.

3.2.3 Hydrogel choice

Since we require an easily mouldable material that allows for basal nutrient diffusion while in transwell culture, the current nanogrooved substrates used for epithelial cells of poly(methyl methacrylate),225 silicon dioxide,227 and polystyrene226 do not represent viable options. Choosing a prospective hydrogel can be a daunting task; factors such as polymerizing temperature, crosslinking, stiffness, surface chemistry, cell adhesion, and degradation may all factor into any one selection. There are many natural and synthetic hydrogel choices available. We limited our search to natural hydrogels as they are found in the basement membrane of epithelium233 diffusible, have been used extensively as coatings in ALI culture of airway epithelial cells

(collagen),234,235 and have been found to be more biocompatible.236 Synthetic hydrogel polymers such as the acrylamides were not used as they require expensive surface chemistry modification

54 for cell adhesion to occur and we wanted to focus on natural hydrogel polymers. As a proof-of- concept study, we limited our search to natural hydrogels that were easily available to our laboratory, that we had experience working with, allowed for basal nutrient diffusion, could be imprinted with grooved topography, and remained stable at tissue culture conditions.

Collagen

Our first candidate hydrogel was PurCol® from Advanced BioMatrix (Cedarlane, Burlington,

Ontario, Canada). Collagen is one of the most widely used natural polymers within the tissue engineering field.236 This bovine collagen solution (3mg/mL) is approximately 97% Type I collagen with the remaining solution being Type III collagen. PurCol® is made from specific bovine that are carefully bred to produce a stable and reproducible collagen mixture which polymerizes at tissue culture conditions and does not revert to a liquid state upon removal from tissue culture conditions. However, batch to batch variations in collagen polymers exist, and the ratios of the different types of collagens have been found, which limit the reproducibility of studies with collagen.

Gelatin

Our second candidate hydrogel available to use was gelatin. Gelatin is a heterogeneous collection of proteins, derived from the breakdown of the triple helix structure of collagen into single strand molecules. There are two types of gelatin which are named according to how they are prepared.

Type A gelatin comes from porcine tissue that is treated with acid and then boiled to extract the collagen proteins. Type B is bovine treated with a base to break down the collagen ultrastructure.

Gelatin readily gels at 4°C and requires protein crosslinking to remain gelled at tissue culture conditions. Type A gelatin has had its mechanical properties extensively studied in the literature allowing us to guess with a high degree of certainty as to the range of gelatin hydrogel stiffness

55 we could conceivably create. 237,238Based on the ease of use and availability of collagen and gelatin, we used these gels in our preliminary experiments of creating a gel that we could imprint with nanotopography.

3.3 Methods

3.3.1 Generation of nangrooved PDMS mould

To create our durable, intermediate mould we fabricated a reusable stamp from the synthetic elastomeric polymer polydimethlysiloxane (PDMS), by replica moulding on a silanized diffraction grating film (Edmund Optics, Barrington, New Jersey, USA) with known nanogroove topography (1um pitch and 250nm depth). The grating surface was silanized with (Tridecafluro-

1,1,2,2, Tetrahydro Octyl 1) – Trichlorosilane and then coated with a mixture of PDMS elastomer and crosslinker at the ratio of 10:1 (Dow Corning Corporation, Midland, Michigan,

USA) (Figure 7A). The coated film was then placed in a vacuum for twenty minutes for degassing and then allowed to cure overnight at 60°C. We were previously aware of the fact that that the diffraction grating grooves were consistent, and always parallel to the shortest dimension which aided us in creating repeatable nanogrooves in a specific direction.

3.3.2 Collagen gel creation

To create the nanogrooved collagen polymer we took 8 parts PurCol® to 1 part Minimum

Essential Medium (MEM) liquid 10x (Invitrogen, Grand Island, NY, USA). The use of MEM

10x was two-fold: first, it provided cell nutrients which would make cell adherence and proliferation better; second, it allowed us to qualitatively assess the acidity of the collagen

(MEM 10x mixture as at acidic concentrations the mixture is a transparent, light yellow, while at more neutral pH conditions the solution is a transparent, light pink). The PurCol® mixture is

56 acidic to keep the collagen constituents from precipitating out of solution; for the collagen to polymerize it must have its acidity reduced. We utilized the sodium bicarbonate buffering system to increase the pH of our collagen-MEM 10x liquid polymer. Drop by drop we added 0.01M sodium bicarbonate (Sigma Aldrich, St. Louis, Missouri, USA) to the acidic collagen-MEM 10x polymer until a colour change was seen from pale yellow to pale pink.

3.3.3 Gelatin gel creation

Gelatin gel was created from type A gelatin (Sigma) that was solubilized in distilled water on a weight per volume scale to give a percentage of gelatin. For example, 5.00 grams of gelatin in 100 millilitres of water would be a 5% gelatin solution. Following the solubilisation of gelatin, the mixture was autoclaved (Tuttnauer Brinkmann 3850E Tabletop Autoclave,

Hauppauge, New York, USA) on a liquid cycle to sterilize the liquid polymer. After sterilization, the gelatin was stored at 4°C until used.

3.3.4 Creating nanogrooved gel inserts

After curing, circular- nanogrooved PDMS stamps were punched out of the elastomer with an 8.0mm biopsy punch (Fray Product Corporation, Buffalo, New York, USA) (Figure 7B).

Next, PurCol® collagen or an autoclaved mixture of 5% (w/v) type A porcine gelatin (Sigma) was cast to a set thickness of 1mm in a 4-well chamber slide (Millipore Ireland, Cork, Ireland) or in a transwell filter (Corning Incorporated, Corning, New York, USA) (Figure 7C).

For collagen gels the nanogrooved PDMS stamp was placed on top of the neutral collagen polymer (Figure 7D). The construct was then allowed to solidify for an hour at 37°C in a tissue culture incubator. After polymerization had occurred, the construct was removed from the incubator and the stamp was carefully removed from the solid collagen polymer (Figure 7E).

Cell specific media was added for at least an hour to ensure that swelling had reached a stable equilibrium. After this swelling period, the substrates were ready for cell seeding.

For gelatin, the nanogrooved PDMS stamp was placed on top of the liquid hydrogel, nanogroove side down, and allowed to set for at least 12 hours at 4°C. After gelation had occurred, the PDMS stamp was carefully removed (Figure 7E). Nanogrooved gelatin substrates were crosslinked with 0.1% (v/v) glutaraldehyde (GTA) made from Type I, 25% stock (Sigma) for 12 hours at 4°C. After the gelatin was crosslinked, the gels were washed to remove the glutaraldehyde: 3 washes of fetal bovine serum (PAA Laboratories Inc., Etobicoke, Ontario,

Canada) were followed by 6 washes with phosphate buffered saline (PBS) (Lonza, Walkersville,

Maryland, USA) with a final wash using cell-appropriate media. All washes were five minutes in lengths. FBS was used as a quenching agent, much like it used in cell passaging protocols to quench the activity of trypsin. In this case we used FBS in an attempt to quench the activity of

GTA. Following the penultimate cell-appropriate media wash, the gelatin was covered with cell- appropriate media overnight at 4°C.

To generate flat inserts we used the same protocol as for the nanogrooved gels, neglecting the step of moulding using the nanogrooved PDMS stamp onto the hydrogel gel surface.

3.3.5 Scanning electron microscopy

Diffraction grating & PDMS

Diffraction grating and PDMS were mounted with carbon colloid paint onto carbon-tape coated stubs and gold coated using a Polaron SC7640 Sputter Coater (Quorum Technologies,

UK) for 45 seconds and then imaged at 5kV in a Hitachi SEM S-3400 (Hitachi High-

Technologies Canada, Inc. Toronto, Ontario, Canada).

Gelatin and collagen gels

Gelatin gels were crosslinked using 0.1% (v/v) type I glutaraldehyde and then dehydrated in an ethanol dilution series of 20%, 40%, 50%, 75%, 90%, 95%, 100% for one hour each and then kept in 100% anhydrous ethanol overnight. After dehydration, the gels were freeze dried using a Labconco FreeZone® 2.5 Litre Freeze Dry System (Labconco, Kansas City, USA) for two hours.

Following the freeze drying process, the gelatin was mounted with carbon colloid paint onto carbon-tape coated stubs and carbon coated using Edwards Coating System E306A,

(Tewksbury, MA. USA) and imaged using a Hitachi S-4500 (Hitachi High-Technologies

Canada, Inc., Toronto ON) at 1.5kV. As we were imaging plastics, silicon polymers, and hydrogels, the accelerating voltage of the SEM was kept below 5kV to avoid charging effects during imaging.

3.3.6 Effect of scanning electron microscopy preparation on gelatin hydrogel

A 1cm block of gelatin was cast, crosslinked, dehydrated in a serial dilution of ethanol and freeze dried. This sample was then compared to the nanogrooved gelatin scanning electron micrograph looking at the change in the length at the macroscale and the change in groove pitch at the microscale. We compared the factor of contraction across each sample run (n=5) knowing at the macroscale our gel was 1cm in length and our grooves began at 1groove/um.

3.3.7 Statistics

To analyse the significance of the change in size of gelatin microgroove topography from native gel to fixed SEM sample, Student's-T test was performed with a P value of significance of less than 0.05.

Figure 7. Generating nanogrooved hydrogels using a PDMS stamp. Liquid Polydimethylsiloxane (PDMS) is cast onto diffraction grating and allowed to cure overnight. Nanogrooved PDMS stamps are cut out using a biopsy punch. Stamps are then placed onto the surface of liquid gelatin. After gelation, stamps are removed and the gels are crosslinked, washed and seeded with cells.

3.4 Results

3.4.1 Gelatin crosslinking optimization

While creating the collagen hydrogels was relatively facile, since collagen polymerizes and is stable at tissue culture conditions gelatin initially, proved more difficult. Gelatin when cast is placed into 4°C to gel and solidify. However, unlike collagen, gelatin does not polymerize and maintain its gelled conformation unless it is kept at 4°C. To overcome this issue the gelatin hydrogels were crosslinked with glutaraldehyde (GTA) (Sigma). From previous work239, GTA is highly toxic to cell culture, therefore we sought out a concentration of both glutaraldehyde and gelatin that would maintain gelatin in a gelled state, and produce the least amount of cytotoxicity. We began a series of experiments testing out various percentages of gelatin crosslinked for six hours overnight with various percentages of GTA. Our early work showed us the while higher percentages of gelatin (10% +) were stiffer, they did not crosslink as well, even at high crosslinker percentage (5% GTA), see Supplementary Figure 1. In the Supplementary

Figure 1, 20%, 15%, 12%, and 10% gelatin were allowed to gel at 4°C overnight and were then crosslinked with 5% GTA overnight at 4°C. Following crosslinking, gels were then placed in standard low glucose media (Invitrogen, Grand Island, NY, USA) and placed in a tissue culture incubator for 48 hours. As can be seen, the higher concentrations of gelatin were not effectively crosslinked at 5% gelatin, even with a high concentration of GTA. We believe this to be due to the increased density of the gelatin monomer which restricted the crosslinking power of GTA to the apical-portion of the stiffer gelatins. We therefore restricted our efforts to gelatin of lower concentration, 5%, and GTA of 0.025%, 0.05%, 0.1%, and 1%. In Supplementary Figure 2, 5% gelatin was allowed to gel overnight at 4°C, and was then crosslinked at 1% GTA, 0.1%, 0.05%, and 0.025% overnight at 4°C. Following crosslinking, gels were then placed in standard low

61 glucose media (Invitrogen) and placed in a tissue culture incubator for 48 hours. As can be seen, the higher concentrations of GTA effectively crosslinked the 5% gelatin, however the lower concentrations did not. Since cell toxicity was a concern of ours we decided to use 5% gelatin with 0.1% GTA at 4°C overnight.

As we had a system that was now viable at tissue culture condition we did not pursue other ranges of gelatin. However, we believe that further altering the concentration of GTA at higher percentages of gelatin could lead to the creation of stiffer gels that are stable at tissue culture conditions. Gelatin of higher concentrations were perhaps fixed only on the most apical region of the gel as increasing amounts of crosslinked monomers hindered the permeation of the entire gel construct with GTA. Perhaps a more dilute GTA, with a much longer period of crosslinking would lead to a viable stiffer gelatin gel. An important consideration is safety with

GTA; while the need to crosslink is important, ideally the lower the concentration of GTA used, the better for the user and for the cells.

3.4.2 Scanning electron microscopy

To characterize the replica moulded PDMS, and the stamped collagen and gelatin, we used scanning electron microscopy (SEM) to visualize the microstructure of these substrates to ascertain if groove transfer from diffraction grating to PDMS to hydrogel construct was possible.

(Figure 8) compares the grooved topography of the diffraction grating, to the replica moulded

PDMS, and the imprinted collagen and gelatin gels. The diffraction grating (Figure 8A) clearly shows the grooved topography, and pitch of 1 groove per micron, which is replicated in relief in the silicon elastomer PDMS (Figure 8B) with less sharpness (to the eye), however, the manufacturer specified pitch of 1 groove per micron is maintained.

As we are able to generate a replica mould of the diffraction grating topography in

PDMS, (which we could use as a more resilient stamp), we next looked at the replication of the

PDMS nanogrooves in the gelatin and collagen hydrogels. While the fibrous structure of collagen is readily seen in the electron micrograph of Figure 8C, the grooved topography seen in the diffraction grating and PDMS is not visible. However, due to the harsh fixation required to image hydrogels using SEM, it is possible that without fixation, the grooves are indeed present in collagen. We predicted that the best test of whether grooved topography was present in our collagen gels would be whether a cell aligned in a similar fashion to collagen as it did on PDMS

Figure 8. Scanning electron micrographs of nanogrooved substrates. Diffraction grating (A) was used as a mould for nanogrooved PDMS (B) which was then used to mould collagen (C) and gelatin hydrogels (D).

63 nanogrooves (Supplementary Figure 3). On 5% gelatin gels, Figure 8D, however, we consistently observed microtopography similar to both the diffraction grating and PDMS.

Therefore, we can generate repeatable and consistent nanogroove topography in 5% gelatin hydrogels as seen through SEM visualization.

Based on the scale of the grooves we see that the gelatin grooves have a shorter pitch than expected. We speculate that this may be a result of the freeze-drying fixation procedure required for SEM preparation. To confirm this was the case we performed macro- and micro-scale analysis of the gelatin gels, looking at how the change at the macroscale of a 1cm block of gelatin compared to the pitch change (Table 1). The mean scale of contraction at the macroscale was 2.3(±0.14) times, while on a microscale, was 2.38(±0.19) times. A Student's t-test, (p<0.05 for significance), indicated a p-value of 0.31. This suggests that the scale of shrinkage is consistent at both macro- and microscales which is consistent with our explanation. This result, further supported our assumption that SEM fixation was the main cause of the change in topographical pitch on gelatin nanogrooves.

Table 1. Contraction of gelatin hydrogel on the macro- and microscale

Contraction on Contraction on Macroscale Microscale Sample 1 2.5 2.7

Sample 2 2.2 2.8 Sample 3 2.6 2.4

Sample 4 2.4 2.3 Sample 5 1.8 1.7

Mean 2.3 2.38 Standard Error 0.14 0.19

of the Mean

To eliminate the effect of fixation we attempted to use environmental scanning electron microscopy (eSEM). The benefits of eSEM are that the substrate does not need to be dehydrated

64 and freeze dried, as the eSEM operated in a partial vacuum at sub-zero temperatures. Ideally, this would eliminate the main driver of our issue in seeing replication of the manufacturer specified pitch. However, the partial vacuum and sub-zero temperatures lead to freeze drying effects as the gel began to freeze and water sublimated, leading to obvious changes in surface topography during imaging.

3.5 Discussion

3.5.1 Replication of PDMS

The creation of a reusable nanogrooved PDMS was necessary to avoid damage to our diffraction grating. While moulding the PDMS was relatively intuitive, minor details such as knowing the direction of the grooves at all times of replication made repeatability much easier.

From the manufacturer we knew that the grooves ran parallel to the shortest direction of the grating. Since only one side of the grating is grooved, we sprayed the grating with 70% ethanol and as the ethanol evaporated on the grooved side, one could see pronounced lines of the ethanol liquid on the grating. After finally moulding, and biopsy punching the grooved out, we took great care to ensure that all grooves faced the same way in storage, and were stamped into gelatin in as similar an orientation as possible.

While PDMS is diffusible to air, it is not permissive to liquid nutrient diffusion and therefore not suitable for use in ALI culture.

3.5.2 Casting gelatin and moulding

With the creation of a PDMS stamp we could then stamp our gelatin hydrogel.

Optimizing this process involved creating thin but manipulatable PDMS stamps, 1-2 millimetres in thickness, by altering the volume of liquid PDMS poured onto the diffraction grating.

Removing the PDMS stamp from the gelatin required straight and smooth splinter forceps that did the minimal amount of damage to the gel surrounding the stamp during removal. After the stamp had been removed and before the GTA was added, the gels were very susceptible to mechanical damage from pipetting.

3.5.3 Cell alignment on the nanogrooved gel insert

After our nanogrooved gelatin was crosslinked, we began seeding our constructs with epithelial cells. Initially, we saw most of our cells as rounded masses or floating within culture.

This led us to believe that traces of GTA left behind after washing were cytotoxic. We began washing our gels more thoroughly by rinsing them with phosphate buffered saline (PBS) on a shaker for one hour and then seeding again, leading to a similar result of dead, or minimally spread cells. Using a technique from cell culture we tried quenching any remaining GTA by washing the gels with fetal bovine serum, and then washing with PBS. Realizing that the amount of time spent washing was not as important as the dilution effect of repeated washes, we increased the amount of FBS washes to 3 for five minutes and subsequent PBS washes to 7 for five minutes followed by leaving the substrates in the media of the cells to be seeded overnight at

4°C. Just before cells were seeded onto the nanogrooves, the media that was present overnight was aspirated, and the cell suspension was then applied. Cells responsive to the nanogrooves aligned within one to two days.

3.5.4 Scanning electron microscopy

Scanning electron microscopy (SEM) of the nanogrooved substrates revealed the successful transfer of grooved topography from diffraction gratings, to PDMS, and to gelatin hydrogels, but not collagen.

Collagen substrates macroscopically, appeared to be a suitable choice for a hydrogel nanogroove substrate. Moulding left an imprint of the PDMS stamp outlined within the collagen.

However, after dehydration, freeze drying, and SEM, the collagen gel appeared as the typical collagen gel, with a fibrillar structure, but no groove topography as seen in PDMS was observed.

This lack of grooves was first thought to arise from the harsh SEM fixation process, however, unfixed, newly stamped collagen gel did not induce alignment of cells that did eventually align on nanogrooved gelatin substrates. This meant that either the grooves were never moulded into the collagen or that the softness of the collagen mitigated the instructive nature of the grooves.

The ratio of collagen protein to the solvent present in PurCol® is 3.0mg/mL; 5% gelatin (w/v) on the other hand is 50mg/mL. The difference between these ratios is a factor 10. This large difference would suggest, that at the percentage used, gelatin produced a much denser, and therefore stiffer gel than collagen. In an attempt to compensate for the difference of gelatin monomer to collagen monomer, we created a series of gelatin mixtures that better approximated the concentration of PurCol® collagen. We found that gelatin below 2% did not gel completely at 4°C. Since we had a gelatin polymer that was stable at tissue culture conditions and was imprinted with the grooves we decided to continue with the gelatin.

3.6 Conclusion

In conclusion, we have created a device that consists of a nanogrooved gelatin hydrogel capable of maintaining surface topography in cell culture conditions. This device is manufactured using replica moulding of PDMS nanogrooves. PDMS nanogrooves are subsequently stamped into gelatin, which is crosslinked and washed.

Chapter 4 – Alignment of epithelium on nanogrooved topography

4.1 Introduction

Problem

Having created and characterized reproducible nanogroove topography in a 5% gelatin hydrogel, we now must verify that this instructive cue guides cell alignment in a variety of epithelial cells and is congruent with air liquid interface culture.

Aims

To test whether our nanogrooved gelatin topography induced alignment of epithelial cells and was congruent with air liquid interface culture, we aimed to (1) assess the morphology of epithelial cells on gelatin nanogrooves, (2) assess the cytoskeletal alignment of epithelial cells seeded on nanogrooved gelatin, and (3) assess device congruency with air liquid interface culture.

4.2 Rationale

Human airway epithelium is lined with motile, multiciliated cells in vivo. When these motile cilia are dysfunctional or immotile it leads to a condition known as primary ciliary dyskinesia

(PCD). The standard method for constructing epithelium in vitro is through transwell filter technology, which creates the air-liquid-interface (ALI) culture system. While this induces apicobasal polarization, the tissue is not organized in a planar orientation as seen through misaligned and uncoordinated ciliary beating, a critical drawback. Thus, a model system which aligns epithelium in a planar fashion would provide a more useful in vitro model of airway epithelium and could be used to study how to organize ciliary beating.

Growing cells on grooved topography is a well-known strategy to align cell morphology and organization in artificial tissues.225,240 However, the substrates currently used to align cells on

69 nanogrooved topography are not appropriate for ALI culture226,227,229, which requires nutrient diffusion from the basal surface.

Having created reproducible nanogroove topography in gelatin, we can now characterize the influence of specific nanogroove substrate topography on the organization of epithelium and assess our devices’ congruency with ALI culture by allowing airway epithelium to undergo apicobasal polarization on a flat gelatin gel.

4.3 Methods

4.3.1 Cell culture

Experiments were done using the human epithelial cell line ARPE-19 (retinal epithelium), murine IMCD3 (inner medullary collecting duct epithelium), and normal human bronchial epithelial cell line (BEAS-2B). All cell lines were from the American Type Culture Collection

(ATCC) (Manassas, Virginia, USA.)

ARPE19 and IMCD3 were maintained in growth medium (Dulbecco’s modified Eagle

Medium [DMEM]/F12 (Invitrogen), 10% fetal bovine serum (PAA Laboratories Inc., Etobicoke,

Ontario, Canada), and 1ug/ml penicillin and streptomycin (Sigma). BEAS-2B were maintained in low glucose DMEM (Invitrogen), supplemented with 10% fetal bovine serum (PAA

Laboratories) and 1ug/ml penicillin and streptomycin (Sigma).

4.3.2 Nanogroove PDMS seeding

Nanogrooved PDMS stamps were sterilized in UV light for 30 minutes and then coated with

10ug/mL fibronectin from bovine plasma (Sigma) overnight at tissue culture conditions. We seeded cell lines at sparse and confluent seeding densities (5000 cells/cm2 and 25 000 cells/cm2) and assessed alignment after 48 hours of culture. (Figure 7F and G).

4.3.3 Nanogroove gelatin seeding

Nanogrooved gelatin was created according to established device manufacture protocol

(Figure 7), after washing with cell specific media, ARPE19, IMCD3 and BEAS-2B we seeded at sparse and confluent seeding densities (5000 cells/cm2 and 25 000 cells/cm2). Cell line cultures were allowed to grow for 48 hours and then assayed for alignment.

4.3.4 Air liquid interface culture of BEAS-2B

Following the protocol for air liquid interface differentiation of Jain et al., (2010),119 we differentiated BEAS-2B on a flat gelatin insert by first creating a 1mm thick flat-gelatin insert in a transwell. We seeded BEAS-2B at our standard confluent seeding density of 25 000 cells/cm2.

After they had reached confluence, we initiated ALI protocol and removed the media from the apical transwell compartment and replaced the basal media with serum free BEAS-2B media

(low glucose DMEM). After 48 hours at ALI, we fixed the samples in preparation for fluorescence microscopy.

4.3.5 Phase microscopy

Cells were imaged using phase microscopy on an Olympus IX81 inverted light microscope

(Olympus Canada Inc., Richmond Hill, Ontario) and imaged using a 20x objective that possessed a long focal length to compensate for the additional thickness of the nanogrooved gelatin.

4.3.6 Fluorescence microscopy

After phase microscopy, cells to be imaged using fluorescent microscopy were fixed with a

4% solution of paraformaldehyde (BioShop, Burlington, Ontario, Canada) for 10 minutes at room temperature and then permeabilized using a 0.1% Triton-X100 (Sigma) for 10 minutes.

Gelatin substrates have non-specific binding sites blocked through covering substrates in a

71 solution of 2.5% bovine serum albumin (Sigma) and 0.1% Tween 20 (BioShop) for 30 minutes at room temperature.

Cytoskeletal staining

Fluorescent imaging of actin was completed by staining F-actin with rhodamine conjugated phalloidin (1:300, Invitrogen). Stained substrates were imaged using a FV1000 Confocal,

Olympus IX81, (Olympus Canada Inc., Richmond Hill, Ontario). Following cytoskeletal staining, substrates were stained with 4',6-diamidino-2-phenylindole (DAPI) to mark cell nuclei

(1: 300, Sigma)

Airway epithelium differentiation staining

Markers of polarization that were stained for were monoclonal mouse anti-Acetylated beta- tubulin for cilia (Sigma). Secondary antibody used was an Alexa 488 goat anti-mouse from

Invitrogen.

4.3.7 Quantification of cellular alignment

To quantify the alignment of plasma membranes and cytoskeletal components, such as actin filaments and microtubules, 20 separate phase contrast or fluorescent images were traced by hand and were analysed using the angle measurement tool of ImageJ (NIH). Within each image, five cells in the four corners was selected, along with one in the centre, was then traced and measured. To be measured, each cell must have had a complete border, that is to say, it was not touching the edges of the image. After the data was compiled, we used the graphing software

Rozetta by Jacek Pazera, to create angular histograms or rose diagrams which clearly visualize the angular information quantified.

4.4 Results

4.4.1 Cell alignment on nanogrooves

4.4.1.1 ARPE19

Since we were interested in assessing the ability of epithelial cells to align in response to microgrooved topography, we conducted experiments with a number of different epithelial cell lines: ARPE19, a human derived retinal pigmented epithelium; IMCD3, a murine derived collecting duct epithelium, and the mesenchymal cell line NIH 3T3 fibroblasts (Supplementary

Figure 5).

Figure 9, shows phase micrographs of ARPE19 cells grown on flat and nanogrooved

PDMS and gelatin at a sparse seeding density. On flat substrates (Figure 9A and C), ARPE19 cells morphologically, do not align in any given direction. This is confirmed through the corresponding angular histograms which show a wide spread of angles derived from the quantification of the cell outlines measured. Between flat PDMS (Figure 9A) and flat gelatin

(Figure 9C) we see that cells on gelatin appear to be more rounded in morphology than those on the PDMS substrate, which appear more spindle-like. We speculate that this difference in morphology is due to the increased stiffness of PDMS over the softer gelatin. In comparison to both flat substrates we see that the ARPE19 grown on nanogrooved PDMS (Figure 9B) are aligned in the direction of the nanogrooves (white double-headed arrow). In the corresponding angular histogram we see that there is a confined spread of cell orientation angles. This shows that sparsely seeded ARPE19 cultured on nanogrooved PDMS are more likely to be oriented in the direction of the grooves. These cells appear to more elongated, and spindle-like than those grown on flat PDMS. ARPE19 grown on nanogrooved gelatin (Figure 9D) also appear elongated in the direction of the grooves as evidenced by the visual orientation of the cells, and

73 by the corresponding angular histogram which shows a dramatic change from the flat gelatin condition, as the angles are more likely to be found within a smaller range.

While sparsely seeded ARPE19 align on PDMS and gelatin nanogrooves, air-liquid

Flat Nanogrooved

PDMS

Gelatin

Figure 9. Sparse ARPE19 align on nanogroove topography. Phase micrographs of sparsely seeded pigmented human retinal epithelial cell line ARPE19 grown on flat (A and C) and nanogrooved (B and D) substrates. Based on the micrographs and corresponding angular histograms it is evident that on flat PDMS (A) and flat gelatin (C) ARPE19 do not align in any given direction. On nanogrooved PDMS (B) we see that ARPE19 cell membranes align in parallel to the direction of the grooves as supported by the corresponding angular histogram. In comparison, ARPE19 grown on nanogrooved gelatin hydrogels (D), also align in the direction of the nanogrooves, as corroborated by the corresponding angular histograms.

74 interface culture of airway epithelium takes place once the monolayer of epithelium has become confluent. Therefore, alignment of confluent epithelial monolayers is necessary if this device is to be used in ALI cultures.

Figure 10 shows phase micrographs of confluent ARPE19 cells grown on flat and nanogrooved PDMS and gelatin. Similar to sparsely seeded ARPE19 on flat substrates (Figure

10A and C), ARPE19 cells morphologically appear not aligned in any given direction. This is confirmed through the corresponding angular histograms which show a wide spread of angles derived from the quantification of the cell outlines measured. Between flat PDMS (Figure 10A) and flat gelatin (Figure 10C) we see a pronounced difference in cell morphology. ARPE19 on flat gelatin are much more rounded than those on flat PDMS; they appear more granulated than their PDMS counterparts; as well, confluent ARPE19 on PDMS appear as spindles and not as typical epithelial cells do. Whereas, those ARPE19 on flat gelatin appear more hexagonal, that is to say, more in the classical epithelial culture morphology. Again this difference in cell morphology is most likely due to the differences in stiffness of PDMS over the softer gelatin. It is interesting to note that ARPE19 appear more epithelial like on softer substrates. In comparison to both flat substrates we see that the ARPE19 grown on nanogrooved PDMS

(Figure 10B) are aligned in the direction of the nanogrooves, (white double-headed arrow). In the corresponding angular histogram we see that there is a narrow distribution of cell orientation angles highlighting that like sparsely seeded ARPE19, confluent ARPE19 on microgrooved

PDMS are oriented in the direction of the grooves in a whole cell manner. These cells appear to more elongated, and spindle like than those grown on flat PDMS. ARPE19 grown on nanogrooved gelatin (Figure 10D) also appear elongated in the direction of the grooves, as evidenced by the very apparent visual orientation of the cells, and by the corresponding angular

75 histogram which shows a dramatic change from the flat gelatin condition, as the angles are found within a smaller range of angles. It is interesting to note that ARPE19 on PDMS in both flat and nanogrooved conditions appear to have similar spindle-like morphology. That is to say that the grooves are instructive to orienting the alignment of the confluent ARPE19 parallel to the

Flat Nanogrooved

PDMS

Gelatin

Figure 10. Confluent ARPE19 align on nanogroove topography. Phase micrographs of confluently seeded pigmented human retinal epithelial cell line ARPE19 grown on flat (A and C) and nanogrooved (B and D) substrates. Based on the micrographs and corresponding angular histograms it is evident that on flat PDMS (A) and flat gelatin (C) ARPE19 do not align in any given direction. However the morphology on flat gelatin does look more rounded and spread out in comparison to the spindle like morphology on flat PDMS. On nanogrooved PDMS (B) we see that ARPE19 cell membranes align in parallel to the direction of the grooves as supported by the corresponding angular histogram. In comparison, ARPE19 grown on nanogrooved gelatin hydrogels (D), also align in the direction of the nanogrooves, as corroborated by the corresponding angular histograms.

76 grooves; the grooves do not appear to alter cell shape from the spindle-like conformation. The grooved topography is clearly capable of inducing alignment of both sparsely and confluent seeded ARPE19.

4.4.1.2 IMCD3

While pigmented retinal epithelium does indeed align in the direction of the grooves, the natural next question was do other types of epithelium, or other cell types, such as mesenchyme align? We found that 3T3 fibroblasts do align in both confluent and sparsely seeded PDMS and gelatin nanogrooves.

Further, we looked at the cell shape and orientation of the murine kidney collecting duct epithelium, IMCD3 at sparse (Figure 11A to C) and confluent (Figure 11D to F) seeding densities on nanogrooved PDMS (Figure 11A and D), and flat (Figure 11B and E) and nanogrooved gelatin (Figure 11C and F). IMCD3 on flat PDMS can be seen in Supplementary

Figure 6. IMCD3 grown on nanogrooved PDMS at sparse seeding densities appears to align in the direction of the grooves in a similar amount of time as sparsely seeded ARPE19. On flat gelatin Figure 14B, we see that visually and through the angular histogram, IMCD3 do not align in any given direction and appear more rounded and bulbous than their counterparts on PDMS.

On nanogrooved gelatin, Figure 11C IMCD3 do not appear to align in the direction of the grooves, which is contrary to the pattern seen in ARPE19. Through analysing the IMCD3 on gelatin nanogrooves it, would appear that when IMCD3 cells, when they are in exclusive contact with the grooved extracellular matrix they are responsive in a whole cell fashion to this instructive cue. On the other hand, if they are in contact with other cells, the grooved topography does not seem to induce whole cell alignment. This conclusion is supported by the difference in the angular histograms of sparsely seeded ARPE19 and IMCD3 on grooved PDMS. In ARPE19,

77 the cells align in the direction of the grooves regardless of whether they are in contact with another cell or completely separate. In IMCD3 cells on the other hand, there appears to exist

Nanogrooved PDMS Flat Gelatin Nanogrooved Gelatin

Sparse

Confluent

Figure 11. IMCD3 do not align morphologically on nanogroove topography. Phase micrographs of murine collecting duct epithelial cell line IMCD3 grown on flat and nanogrooved substrates. Based on the micrographs and corresponding angular histograms it is evident that on sparsely seeded nanogrooved PDMS (A), IMCD3 align in the direction of the nanogrooves. On flat gelatin (B) IMCD3 are more rounded in morphology than or sparsely seeded PDMS and do not align in any given direction as seen in the corresponding angular histogram. Based on the angular histogram of sparse IMCD3 on nanogrooved gelatin (C) we see that they do not align in the direction of the grooves. Morphologically, confluent, IMCD3 on nanogrooved gelatin (D), flat gelatin (D) and nanogrooved gelatin (F) do not align on their respective substrates.

78 distinct cell populations that align in the direction of the grooves, and some that do not align in the direction of the grooves. This can be seen upon close inspection of the angular histogram where in the direction of the grooves, there appears to be cells orienting, however, cells also appear to orient in directions opposed to the grooves. I speculate that IMCD3 that are not in contact with other cells respond to the groove topography by aligning in a whole cell orientation, whereas those that have made contact with other cells do not respond in a whole cell fashion to the groove orientation.

Confluent IMCD3 on nanogrooved PDMS (Figure 11D) and flat or grooved gelatin substrates, Figure 11E and F, did not appear to align in the direction of the grooves, as seen in the micrographs and the corresponding angular histograms. This would seem to support the conclusion that cell to cell contact overrides the signal for cells to align their plasma membranes.

Morphologically, IMCD3 on PDMS and gelatin do not have as pronounced differences as they do in the sparse conditions, yet on PDMS, cell outlines are sharp and easily definable; on gelatin, however, these boundaries are more diffuse and not as sharp. It is evident that IMCD3 do not align at confluent seeding densities like ARPE19 at similar densities.

We next asked if perhaps a component of the cytoskeleton, F-actin aligns in the direction of the grooves. Using confocal microscopy we looked at the alignment of F-actin at sparse

(Supplementary Figure 7) and confluent IMCD3. Since we already know that sparse IMCD3 align as single cells on PDMS and gelatin, we focused on confluent IMCD3. In Figure 12A, confluent IMCD3 seeded on PDMS nanogrooves appeared to have aligned F-actin cytoskeleton, in the direction of the grooves. The actin fibers appear to run parallel to the grooved topography present, as seen in the corresponding angular histogram. It is interesting to note that the actin fibres appear to run across the monolayer with the cytoskeleton of one cell running continuously

79 across the cell-to-cell junctions and into the neighbouring cells, implying some global level of instruction and communication. On flat gelatin, Figure 12B we see that the fine fibers that were present on grooved PDMS are no longer seen regularly. The cytoskeleton and cell membranes do not appear to be organized in any given direction which is supported by the random distribution of angles across the angular histogram. On nanogroove gelatin, Figure 12C, we once again see

Nanogrooved PDMS Flat Gelatin Nanogrooved Gelatin

Confluent

Figure 12. Confluent IMCD3 F-actin cytoskeleton aligns on nanogroove topography. Based on the micrographs and corresponding angular histograms it is evident that confluent IMCD3 actin cytoskeleton aligns on nanogrooved PDMS (A) given the F-actin fibers orientation parallel to the groove orientation. On flat gelatin (B) IMCD3 F-actin is not aligned in any given direction and the cell membrane does not appear aligned, as corroborated by the angular histogram. Based on the angular histogram of confluent IMCD3 on nanogrooved gelatin (C) we see that the actin fibers align in the direction of the grooves. the actin fibers aligning in the direction of the grooves. The actin cytoskeleton appears more robust, in that actin filaments are thicker, and there are fewer fibers formed. As well, the directionality of all fibers is not parallel in the direction of the grooves, which could be a result of the softer gelatin.241

4.4.1.3 BEAS-2B align on nanogrooves

As we have seen that sparse and confluent retinal and collecting duct epithelium, in

addition to fibroblasts, align parallel to the direction of groove topography we next tested

whether

a human airway respiratory cell line, BEAS-2B, aligns in the direction of nanogroove

topography. Sparse and confluent BEAS-2B were grown on nanogrooved PDMS and gelatin and

their alignment was looked at in terms of the cell membrane and cytoskeleton. Since ALI culture

required confluent airway epithelium, sparsely seeded constructs are found in Supplementary

Figure 8 and 9. Confluent BEAS-2B were grown on nanogrooved PDMS (Figure 13A), flat

gelatin (Figure 13B), and nanogroove gelatin (Figure 13C). On nanogrooved PDMS, BEAS-2B

quite obviously align in the direction of the grooves through their cell membranes and actin

cytoskeleton. These observations were quantified using ImageJ and represented in an angular

Nanogrooved PDMS Flat Gelatin Nanogrooved Gelatin

Confluent

Figure 13. Confluent normal human bronchial epithelial cell line BEAS-2B grown on nanogrooves have aligned F-actin cytoskeleton. Based on the micrographs and corresponding angular histograms it is evident that confluent BEAS-2B F-actin cytoskeleton aligns on nanogrooved PDMS (A) given the F-actin fibers orientation parallel to the groove orientation. On flat gelatin (B) BEAS-2B F-actin is not aligned in any given direction and the cell membrane does not appear aligned, as corroborated by the angular histogram. Based on the angular histogram of confluent BEAS-2B on nanogrooved gelatin (C) we see that the actin fibers align in the direction of the grooves.

81 histogram which shows pronounced alignment in the direction of the grooves (white arrow in A).

Much like ARPE19 and sparse IMCD3, aligned BEAS-2B are elongated parallel to the grooves, as well as their cell membranes, and the F-actin cytoskeleton is seen as fibers oriented in the direction of the grooves. It is interesting to note that on flat gelatin, BEAS-2B appear rounded and not oriented in any given direction, both in their cell membranes and their F-actin cytoskeleton, as seen in the corresponding angular histogram; as well, the BEAS-2B grew over each other. On nanogrooved gelatin, confluent BEAS-2B aligned parallel to the nanogrooves with their cell membranes, and with the F-actin cytoskeleton. In comparison to nanogrooved

PDMS, these BEAS-2B were much more spread out, like those grown on flat gelatin. The actin cytoskeleton in particular had multiple fibers per cell aligning in the direction of the grooves.

Unlike confluent IMCD3 in which the actin cytoskeleton aligned across cells, the aligned BEAS-

2B cytoskeleton did not appear to align across cells. This observation could indicate that junctions in BEAS-2B are not as well developed and linked with the cytoskeleton as in IMCD3.

4.4.1.4 Gelatin insert appears congruent with ALI culture

Having seen that BEAS-2B align on gelatin nanogrooves we tested whether our gelatin insert was congruent with ALI culture by seeding BEAS-2B on a flat gelatin insert and assessing ciliogenesis after 2 days of ALI culture. Our observations reveal that after two days of ALI culture, primary cilia form as visualized through fluorescence microscopy of acetylated β-tubulin

(Figure 14), similar to the observations seen in Jain et al., (2010) where primary cilia were seen along a similar timeline.119 This provides preliminary data that our device is congruent with air liquid interface culture, however, through BEAS-2B we do not see motile, multiciliated epithelial cells or tight junction . This result is confirmed in Jain and colleagues paper as well. 119

Figure 14. Primary cilia are present on BEAS-2B differentiated on gelatin gels. Normal human bronchial epithelial cell line BEAS-2B were stained for cilia, (acetylated β-tubulin), green, and for nuclei (4’, 6’ diamidino-2-phenylindole), blue at day 2 of ALI culture.

4.5.1 Epithelial cell alignment on nanogroove topography

Now that we were confident that grooves were present in our PDMS and gelatin hydrogel substrates, we sought to test if the grooves induced planar alignment of epithelial cells. On flat substrates no cells of ARPE19, IMCD3, or BEAS-2B aligned in the direction of the grooves.

This is in keeping with a number of studies on cells, in particular, epithelial cells225, 226,227 where no alignment in cellular morphology was seen in cells of flat substrates. On flat PDMS, sparse and confluent ARPE19 and BEAS-2B appeared more elongated and spindle-like than on flat gelatin. On flat gelatin ARPE19 and BEAS-2B appeared more rounded and lost their spindle-like phenotype. On flat gelatin gels, ARPE19 appear more hexagonal in shape, and have increased granularity. On flat gelatin, sparse IMCD3 appear rounded, without any directionality. It is possible that the differences in cell morphology between flat PDMS and flat gelatin are a result

83 of a wide variety of factors, such as stiffness177,242,241 or surface chemistry 208,209. Further work is required to dissect the exact cause of the change in morphology from PDMS to gelatin.

Using phase microscopy on nanogrooved PDMS and gelatin, sparse and confluent ARPE19 aligned in the direction of the grooves as seen visually and through the corresponding angular histograms. Sparse IMCD3 aligned in the direction grooves as well. In particular, confluent

ARPE19 align quite strikingly on PDMS and gelatin nanogrooves. Sparsely seeded murine collecting duct epithelial cell line IMCD3 align as ARPE19 do on nanogrooved PDMS; their morphology elongates in the direction parallel to the grooves. This alignment is confirmed through the corresponding angular histogram.

Interestingly, and in breaking with ARPE19, IMCD3 do not appear to align on gelatin nanogrooves if in contact with other cells. However, if alone, IMCD3 do show some alignment visually. However, the angular histogram suggests that the majority of IMCD3 on gelatin do not align in the direction of the grooves. Confluent IMCD3 consistently do not respond to nanogroove topography of both PDMS and gelatin, and on flat gelatin. These are reminiscent of the results of Clark et al. (1990), with MDCK on nanogroove topography showing that cell to cell contact altered contact guidance to the grooves.225 Why confluence in MDCK and IMCD3 alters the instructive cue of the grooves is currently unknown; we speculate that it could be the result of an intrinsic organizational program in renal epithelia. Aubin and colleagues (2010) described the alignment and elongation of various cell types within microengineered bars of hydrogel that had varying width. Interestingly, cells that were found to be unresponsive to alignment cues in vitro were found to be unaligned in vivo.243 They suggest that this could be the result of a cell intrinsic program that allows cells to be responsive to alignment and elongation

84 cues. While not providing ample evidence in support of this theory, the authors provide a new area of research into how cell in vivo morphology affects in vitro behaviour.

The morphology of IMCD3 on nanogrooved PDMS was more spindly than on flat or nanogrooved gelatin which appeared more rounded. We again speculate these results are a consequence of differences in substrate stiffness and surface chemistry. While confluent IMCD3 did not align morphologically to the grooves, using fluorescence microscopy, confluent IMCD3

F-actin cytoskeleton did align on nanogrooved PDMS and gelatin. These results are reminiscent of the results seen in Jin et al., (2008) where the F-actin of MDCK cells aligned in the direction of the grooves.226

These two forms of alignment led us to create nomenclature for the two types of alignment we saw, a morphologic or whole cell behaviour that was visualized using phase microscopy, and a cytoskeletal variant found through fluorescence microscopy. As seen in supplemental data, for all cells that we tested, those that align morphologically align with their F-actin cytoskeleton

(ARPE19, 3T3); however, those that align cytoskeletally, particular at confluent conditions

(IMCD3) do not align with their cell membranes. Seeding epithelial cells shows that this grooved topography is capable of inducing planar alignment and elongation of cell morphology and F- actin cytoskeleton in both sparse and confluent conditions.

To our knowledge, airway and respiratory epithelium has never been organized in an aligned and elongated fashion using nanogroove topography. Nanogrooves have been found to organize a wide variety of cells244,245, including retinal epithelial cells227,240 this is known as contact guidance.

4.5.2 Response to topography

Contact guidance of epithelial cells on nanogroove topography has been well characterized on non-diffusible substrates.226,227 Almost all work studying the mechanisms of alignment on groove topography have been done in fibroblasts. Within 20 minutes of fibroblasts being seeded on grooved substrates microtubule alignment has begun to show alignment.246 This alignment of the microtubule cytoskeleton precedes alignment of the fibroblast membrane. It has been found that microtubules organize from the bottom of the cell upwards; implying that the cytoskeleton closest to the grooves aligns first and this alignment cue propagates apically with time. 246 After the alignment of tubulin, actin also aligns in the direction of the grooves within

40 minutes to an hour, as does the cell membrane. Focal adhesions, are the sites of cell adhesions, and function as a bidirectional mechanical transducer of cell-to-ECM and ECM-to- cell signalling information. Thus, the focal adhesion plays a critical role in transmitting the mechanical forces into an adhering cell. This data highlights the integral part the cytoskeleton plays in cellular alignment, as it precedes whole cell alignment, induces it, and finally, the adhesions of aligned cells are guided grooved topography. While certain parallels can be drawn between fibroblasts and epithelial cells, fibroblasts are initially more adherent in vitro than epithelia. This suggests that the timeline outlined by Oakley and colleagues (1993), while perhaps correct in order would occur over a longer period in epithelial cells. Nanogrooves in general have been found to enhance cellular adhesion245, in particular, epithelial cells. 226

In fibroblasts growing on grooves, focal adhesions show changes in distribution. 246 As focal adhesions formed they were distributed in a radial pattern, which after three hours became aligned in the direction of the grooves. Cell adhesions, in the form of focal adhesions, have been found to be impacted by nanotopography as their size can be modified by grooved topography.244

These focal adhesions grow anisotropically, and in the direction of the grooves lead to increases in tension within the adherent cell. 244

As focal adhesion component proteins connect with the actin cytoskeleton, these tensile forces are translated to the actin cytoskeleton through Rho signalling which leads to actin filaments bundling in the direction of the tensile force, or the direction of the grooves. 247 The above theory for how alignment on grooved topography occurs does not take into account observations that cell on nanogrooves migrate in the direction of the grooves. While this may appear self-evident and obvious, the mechanism for this has been the subject of considerable debate.245

The current theory for how cells align while migrating on a substrate involves the highly active, probing filopodia. Filopodia extend and retract search for integrin binding sites.175 On nanogroove topography, filopodia preferentially extend and spread parallel to the direction of the grooves, not orthogonal to their direction. 175 As the filopodia spread anisotropically in the direction of the grooves they bias orientation and migration the cells in the direction of the grooves, through planar cytoskeletal polarization, and ultimately, complete alignment of the cells. This data has been collected in fibroblasts which have many structural differences than epithelial cells, namely, intercellular junctions, which are known to be connected to and stabilized by the actin cytoskeleton. We have found that certain epithelial cell lines, ARPE19 and

BEAS-2B do indeed align in the direction of the grooves, yet IMCD3, do not. This suggests that alignment of the actin cytoskeleton does not necessarily lead to cell membrane polarization and bias in the direction of the grooves. Perhaps the cell to cell contacts of these epithelial cells are more developed and cannot be overcome by the polarized cytoskeleton. To support this

87 hypothesis, BEAS-2B are known to not have well defined junctions meaning that the actin polarization is sufficient to induce whole cell alignment.248

In comparison, IMCD3 do have polarized cytoskeleton on nanogrooves, which form bands across cells but they do not align in the direction of the grooves. IMCD3 have well-formed tight junctions, 249 that we know to be stabilized and organized by the actin cytoskeleton. As the

IMCD3 are receiving instructive cues from the nanogrooved topography, it is possible that TJ homotypically connect, and that the cytoskeleton connecting both dimers aligns in parallel to the planar polarizing signals from the basal cell compartment leading to organization of the actin to appear as one continuous band across the confluent epithelium. In contrast, sparsely seeded

IMCD3 align like ARPE19 on nanogrooved PDMS and gelatin and appear more rounded and without planar organization.

4.5.3 Substrate stiffness

From work within our laboratory we have seen that our gelatin crosslinked with 0.1%

GTA is not as stiff as our PDMS. While no study to our knowledge has looked at how the stiffness of a groove affects the instructive signal from nanogroove cues on epithelium, work has been done in cardiomyocytes. Stiffness and grooved topography was tested using various groove topographies cast in materials of different stiffness. 250 The results suggest that topography and stiffness affect different aspects of cell function. In particular, changes in morphology and orientation were affected by grooved topography, while the contractile function of the nascent cardiomyocytes was coupled with both topography and stiffness. While cardiomyocytes beat in an anisotropic fashion, focal adhesions that are aligned in the direction that best distributes this

88 force are better able to anchor the cell to the underlying ECM and allow for greater junctional formation between cardiac cells. 250

From our own observations we see that there is a difference in cell morphology on stiff and compliant substrates. A great deal of work has been done on the effect of substrate stiffness on cell adhesion. Pelham and Wang (1997), have completed seminal work on the effect of substrate stiffness on cell morphology by changing the stiffness of polyacrylamide gels. This alteration in stiffness led to alterations in cell morphology, and adhesion.178 As in nanogroove topography, the adhesion of cells on stiff or compliant substrates is a critical step in determining overall cell morphology. The adhesion of a cell through integrins, leads to mechanical tension on the cell, through the cytoskeleton. The cytoskeleton responds to this stress through contraction of non-muscle myosin II. The resulting adjustment of the mechanical stresses on the cytoskeleton lead to deformation of either the substrate, meaning it is compliant, or the cell itself deforming, meaning the substrate is quite stiff. As each cell type is different with a different balance and capabilities for sensing and responding to stiffness, what is stiff for one cell may be quite compliant for another. Cells on stiff substrates have more integrin clusters.251 The presence of more integrin clusters allow a cell to distribute a greater amount of force more evenly over a substrate, and these integrin clusters are usually of similar size and distribution on a stiff substrate. 241 On a stiff substrate, fibroblasts polarize; they lengthen and form actin stress fibers in the direction of the developing major axis. The distribution of focal adhesions also changes to conform to the axis of elongation. On compliant substrates fibroblasts produce smaller more numerous focal adhesions which are distributed with a radial orientation. 251 This explains why on flat PDMS, ARPE19 were spindle like: the stiff PDMS substrate created was permissible to organized focal adhesions which lead to changes in cell morphology, thereby becoming more

89 spindle like. On flat gelatin though, more compliant substrate, meant a more radial distribution of focal adhesions, leading to more rounded cells.

It is interesting to note that while focal adhesions, with their mechanosensitive properties play a role in the long term morphology and polarization of fibroblasts, cell orientation in one direction occurs before focal adhesion alignment. Implying that the mechanical forces on a cell change orientation of the whole cell before the cell itself reorganizes cell adhesions. However, for cell membrane stabilization focal adhesions must transition from a radial distribution to an axial one. 251 This transition is most likely driven by the stabilization of those focal adhesions sharing similar orientation to the developing major axis and therefore associated with congruent actin fibers. As this transition occurs on stiff substrates only, then cell adhesions in particular, must be mechanosensitive.

4.5.4 Changing chemistry

While stiffness is one area where PDMS and gelatin nanogrooves differ leading to changes in cell morphology, another factor could be the changing chemical environment. There is evidence that such changes are due to alteration of the chemical signals present on organizing epithelium. Human corneal epithelial cell response has been found to change from parallel to the grooves to orthogonal to the grooves based on the medium in which the epithelial cells are seeded in. 252 However even in media which usually induced orthogonal orientation to the grooves, altering pitch dimensions from 400 to 4000nm cause a switch back to parallel alignment.

4.5.5 BEAS-2B polarization on gelatin inserts

BEAS-2B form primary cilia while undergoing differentiation in ALI culture. While others have seen similar results, it is known that BEAS-2B do not form motile cilia, 119 or well

90 defined and localized apical junctions. 248 Therefore, BEAS-2B, as a cell line, are ineffective to test whether planar alignment or elongation of the cytoskeleton of airway epithelium leads to organized ciliary beating. An effective cell requires the ability to apicobasally polarize, as seen through good apical junction formation, and produce multiciliated cells which have motile cilia.

Our search has led us to begin testing primary tracheal epithelial cells as they are known to produce motile cilia and well defined apical junctions.119,248,253,254

4.6 Conclusions

Epithelial cell types of all variety align morphologically and cytoskeletally on our gelatin nanogroove device. In both sparse and confluent seeding densities, examples of epithelial cells, particularly BEAS-2B align in the direction of the grooves through contact guidance. The alignment of confluent epithelial monolayers is of critical importance as ALI cultures require confluent respiratory epithelia. We tested whether normal human bronchial epithelium on gelatin insert was congruent with ALI culture and found that while BEAS-2B form primary, immotile cilia in ALI culture, they do not form appropriate apical junctions or motile cilia. Therefore

BEAS-2B will not answer the question of whether aligned cytoskeletal components lead to aligned ciliary beating. We require a primary tracheal epithelial cell which does polarize through the formation of apical junctions and motile cilia, thus, allowing us to assess the organization of ciliary beating.

5.1 Introduction

Problem

BEAS-2B epithelium is able to produce primary, immotile cilia on our gelatin transwell insert after 2 days of ALI culture. However, they cannot produce motile cilia, which we require to assess ciliary beat alignment. We looked at primary tracheal epithelial systems to establish

91 which undergo apicobasal polarization in vitro, in our hands. We then tested these polarizing primary airway epithelial cells to assess whether our gelatin insert is congruent with ALI culture by allowing apicobasal polarization. Finally, while we have seen cell line epithelium align in the direction of our nanogroove topography, we asked if primary tracheal epithelial cells align in the direction of our grooves?

Aims

To test whether primary tracheal epithelial cell apicobasally polarize we aimed to (1) assess apical junction and ciliary formation during standard ALI culture, (2) assess apicobasal polarization of human tracheal epithelia on gelatin gels of different thickness, and (3) assess the cytoskeletal alignment of epithelial cells seeded on nanogrooved topography.

5.2 Rationale

From our previous studies PDMS nanogroove stamps were created by moulding PDMS onto diffraction grating nanogrooves (pitch of 1μm and depth of 250nm). After seeding various epithelial cell lines on PDMS and gelatin nanogrooves we have observed that epithelial cells organize on nanogrooved PDMS and gelatin in a cell membrane and cytoskeletal fashion. We have also seen that normal human bronchial epithelial cell line BEAS-2B form primary cilia following 2 days of ALI culture. However, motile cilia are not formed and are required to assess ciliary beat alignment.

We now sought to test whether primary tracheal epithelial cells, which are capable of producing multiciliated cells with motile cilia, polarized in our hands on standard and gelatin insert ALI culture. Airway epithelium lines the entire respiratory tract, comprising the lumen of the nose all the way through to the most distal alveoli where gas exchange occurs. The structure

92 of the epithelium is critical to its function. In particular, multiciliated cells play a critical role in moving mucus out of the respiratory tract; this process is known as mucociliary clearance.

Mucociliary clearance is dependent on the organization of the cilia and how they beat. Cilia must beat in a coordinated, metachronal fashion which moves the mucus consistently out of the airway. Without metachrony, bacteria lodged in mucus would not be removed from the airway and persistent infections would occur.

In this respect, we have defined two axes of organization in the airway epithelium: the apicobasal axis, constituted by a physical difference in the morphology, composition, and structure of the apex from the basal epithelial compartment, and the planar axis, which is seen in the directionality of the beating of ciliated cells propelling mucus out of the airway. The organization of the epithelium along both axes is therefore critical to airway function.

5.3 Methods

5.3.1 Human tracheal epithelia

Human tracheal epithelial cells (HTEC) were a gift from Dr. Nadeem Mohgal’s laboratory. The full protocol for their isolation can be found in Appendix 1. HTEC were maintained in BEGM (Bronchial epithelial growth media) from Lonza clonetics (Walkersville,

Maryland, United States). For their use in polarization studies, we found it best to use HTEC within two passages of plating, as upon visualization, morphological changes occurred from early passages P0-P2, to late passages, >P2. Nanogroove topography studies were done with these higher passage cells.

Bronchial Epithelial Growth Medium (BEGM) from Lonza was added. BEGM is composed of Lonza Bronchial Epithelial Basal Medium (BEBM), and supplemented with the

BEGM SingleQuots Kit Supplement and Growth Factors (Catalog number CC-4175), which is

93 composed of bovine pituitary extract, human recombinant epidermal growth factor, transferrin, triiodothyronine (T3), epinephrine, retinoic acid, hydrocortisone, insulin, and gentamycin. The specific concentrations are not given by the manufacturer. Cells were then counted in a 50:50 mixture of trypan blue (Invitrogen) and HTEC cell suspension to assay the number of dead cells.

Finally cells were plated at approximately 5000 cells/cm2 on PurCol® collagen (Advanced

BioMatrix, Poway, California, USA) coated tissue culture flask coated cell flasks for future passage and use.

5.3.2 Nanogroove culture

Nanogrooved PDMS stamps were sterilized in UV light for 30 minutes and then coated with

10μg/mL PurCol® from Advanced BioMatrix for at least 2 hours in tissue culture conditions..

This bovine collagen product (3mg/mL) is approximately 97% Type I collagen with the remaining solution being Type III collagen.

Nanogrooved gelatin was created according to established device manufacture protocol

(Figure 7), after washing with cell specific media, HTECs were seeded cells at sparse and confluent seeding densities (5000 cells/cm2 and 25 000 cells/cm2). Primary cells were cultured for 72 hours, and then assessed for alignment.

5.3.3 Air liquid interface culture

HTEC were seeded at 50 000 cells/cm2 and grown to confluence on collagen coated transwell inserts in BEGM media. Once confluence was reached upon visual inspection, ALI was initiated by removing apical media and switching the basal compartment to Bronchial Air

Liquid Interface (B-ALI) Differentiation media from Lonza. B-ALI Differentiation media is

94 composed of the following: bovine pituitary extract, human recombinant epidermal growth factor, transferrin, triiodothyronine (T3), epinephrine, retinoic acid, hydrocortisone, insulin, gentamycin, and a proprietary supplement known as B-ALI™ Inducer. While the supplements added to create B-ALI media are identical to BEGM media, B-ALI™ Inducer, and a larger volume of bovine pituitary extract aside, the specific concentrations are not given by the manufacturer. HTEC were maintained at ALI for at least 31 days for full differentiation. After

ALI completion, differentiated epithelium was then prepared for immunocytochemistry.

Gelatin insert air liquid interface culture

To assess the effect of the gelatin insert on ALI culture we created flat gelatin inserts of various thickness (10um, and 500um). After crosslinking with GTA and washing as per our gelatin hydrogel fabrication protocol (Chapter 3), we seeded HTECs on these grooves and allowed the HTEC to polarize under ALI conditions.

5.3.4 Immunocytochemistry

Cell and cytoskeletal orientation samples were fixed with a 4% paraformaldehyde solution (BioShop) for 10 minutes at room temperature, and then permeabilized for 20 minutes with a solution of Triton X-100 (Sigma) in PBS.

Cytoskeletal protein analysis

Fluorescent imaging of actin was completed by staining F-actin with rhodamine conjugated phalloidin (1:300, Invitrogen). Microtubules were stained with a mouse monoclonal antibody to beta tubulin (1:1000, Abcam, ab7792), overnight at 4°C. The secondary stain used was an Alexa 488 goat-anti mouse from Invitrogen (1:250), applied for 45 minutes at room

95 temperature and then removed. Following cytoskeletal staining, substrates were stained with

4',6-diamidino-2-phenylindole (DAPI) to mark cell nuclei (1: 300, Sigma)

Polarization

Markers of polarization that were stained for were monoclonal mouse anti-acetylated beta-tubulin for cilia (Sigma), Mouse anti-ZO-1 unconjugated monoclonal antibody (Invitrogen).

Secondary antibodies used were Alexa 488 goat anti-mouse, and Alexa 546 goat anti-mouse, all from Invitrogen.

5.3.5 Imaging

Stained substrates were imaged using a FV1000 Confocal, Olympus IX81, (Olympus

Canada Inc., Richmond Hill, Ontario) with a 40x/0.8 LUMPlanFI dipping water objective.

5.3.6 Quantification of cellular alignment

To quantify the alignment of plasma membranes and cytoskeletal components, such as actin filaments and microtubules, 20 separate phase contrast or fluorescent images were traced by hand and were analysed using the angle measurement tool of ImageJ (NIH). Within each image, five cells in the four corners were selected, along with one in the centre, was then traced and measured. To be measured, each cell had a complete border, that is to say, it was not touching the edges of the image. After the data was compiled, we used the graphing software Rozetta by

Jacek Pazera, to create angular histograms or rose diagrams which clearly visualize the angular information quantified.

5.3.7 Quantification of ciliated cells

The mean number of ciliated cells was quantified across two samples, three images each of control, 10um and 500um thick gelatin gel HTEC ALI. Within each image, a ciliated cell, seen

96 through the localization of acetylated-tubulin with the nuclear marker DAPI were counted, and then the average was found across the images quantified. After the mean was found, data was compiled into a bar-graph and standard error of the mean was calculated. Since one experimental run occurred statistical analysis could not be validly applied.

5.4 Results

5.4.1 Human tracheal epithelial cells apicobasally polarize in standard ALI culture

While BEAS-2B cell form primary cilia during ALI culture, we require a primary cell that undergoes apical basal polarization as seen through the formation of tight junctions and multiciliated cells. We tested HTECs during standard ALI culture for 31 days and found that these primary tracheal epithelial cells polarize as seen through the formation of tight junctions, visualized through cortically localized zona occludens 1 (ZO-1) staining at day 11 ALI (Figure

15A). At the same time point we see a single patch of cilia, visualized through acetylated β- tubulin (Figure 15B). The corresponding xz-reconstruction of the ciliated epithelium reveals a distinct patch of apically localized acetylated β-tubulin, which we take as evidence of ciliogenesis, a strong marker of apicobasal polarization. At 18 days after the initiation of ALI culture, we see that the tracheal epithelial monolayer has changed in morphology. The ZO-1 in

15C appears less diffuse and more localized to the cellular membrane, and the overall epithelium appears more compact as more cells are present within the same sized area. From day 11 to day

18 ALI, ciliogenesis appears to have continued, with multiple patches of multiciliated cells forming (Figure 15D). The corresponding xz-reconstructions corroborate these results as multiple, apically localized areas of acetylated β-tubulin are apparent. Following the completion of our ALI culture of HTECs at day 31 we see that the epithelial monolayer has compacted even more, with smaller cells as polarization continues. While the ZO-1 staining (Figure 15E)

97 appears more diffuse that our day 18 stain, we see that TJs are still present. Looking at acetylated tubulin (Figure 15F) we see that multiciliated cells are found across the tracheal monolayer in a robust increase in both number of multiciliated cells, and length of cilia (visual inspection) over their day 11 and 18 counterparts. The xz-reconstruction shows apically localized acetylated-tubulin staining in our day 31 sample. Taken together, these results show that we have a primary tracheal epithelial population that is capable of forming an apically-

Figure 15. Human tracheal epithelial cells form tight junctions and multiciliated cells during standard ALI culture. At day 11 ALI (A and B) cortically localized zona occludens 1, ZO-1, (A) staining is seen along with a single patch of acetylated β-tubulin (B), suggesting that tight junctions have formed and ciliogenesis is in progress. At day 18 ALI (C and D) cortically localized ZO-1 (C) staining is seen as are multiple patches of acetylated β-tubulin (D), suggesting that tight junctions are present and the developing epithelium is forming patches of multiciliated cells. On day 31 ALI (E and F) diffuse ZO-1 (E) is seen as are large, prominent areas of multiciliated cells (F) suggesting that apicobasal polarization has occurred.

98 basally polarized epithelium, which will allow us to test whether our gelatin inserts are congruent with ALI culture.

5.4.2 Human tracheal epithelial cells apicobasally polarize on gelatin insert ALI culture

While HTEC polarize with TJ formation and robust multi-ciliogenesis during 31 days of

ALI culture, we require a HTEC that undergo apicobasal polarization on our gelatin device as seen through the formation of TJs and multiciliated cells. Using standard, collagen coated day 31

ALI as a control (Figure 16A and B) we tested HTECs during ALI culture for 31 days while on gelatin of 10um thickness (Figure 16C and D) and 500um thick (Figure 16E and F) and found that these primary tracheal epithelial cells polarize in our hands as seen through the formation of multiciliated cells. In our control, TJ appear in a similar fashion to our previous HTEC ALI with cortically localized ZO-1 staining (Figure 16A) and robust ciliogenesis throughout the epithelium (Figure 16B) that is apically localized by acetylated-tubulin visualization as seen through the corresponding xz-reconstruction. In our 10um thick gelatin ALI experiment we see that while TJ formation has occurred (Figure 16C) similarly to our control in that there is defined cortical localization of ZO-1. Conversely, ciliogenesis (Figure 16D) while occurring, is not as robust across the entire epithelial monolayer. The corresponding xz-reconstruction highlights the apical localization of the motile cilia. The 500um thick gelatin ALI showed diffuse ZO-1 staining that in patches appeared cortically localized (Figure 16E). The ZO-1 staining was also not as sharp as in our 10um thick gelatin or control highlighting that this thickness of gelatin indeed produced less defined TJs. Looking at acetylated-tubulin (Figure

16F) we see that while multiciliated cells are found in the tracheal monolayer they are few and far between and appear shorter in length in comparison to our 10um gelatin. The xz- reconstruction shows apically localized acetylated-tubulin staining pointing to a polarized HTEC

ALI culture. Taken together, these results show that we have a primary tracheal epithelial

population that is capable of forming an apicobasally polarized epithelium on gelatin of 10um

and 500um thickness through the formation of multiciliated cells which will allow us to test

whether aligned cytoskeletal components on gelatin nanogrooves, le+ad to aligned ciliary

beating.

Figure 16. Human tracheal epithelial cells form tight junctions and multiciliated cells during ALI culture on a gelatin filter. In control day 31 ALI (A and B) cortically localized zona occludens 1, ZO-1, (A) staining is seen along with robust patches of acetylated β-tubulin (B), suggesting that apicobasal polarization has occurred through the formation of tight junctions and multiciliated cells. In 10um thick gelatin gel at day 31 ALI (C and D) cortically localized ZO-1 (C) staining is seen as are multiple patches of acetylated β-tubulin (D), suggesting that tight junctions are present with fewer multiciliated cells. In 500um thick gelatin gel at day 31 ALI (E and F) diffuse ZO-1 (E) staining is seen as are sparse patches of acetylated β-tubulin (F), suggesting that multiciliated cells to do form.

100

5.4.3 Alignment of primary human tracheal epithelial cell on nanogroove topography

To see if primary tracheal epithelium respond to our nanogroove topography in a similar

Figure 17. Human tracheal epithelial cells seeded on nanogrooves do not align. Based on the micrographs and corresponding angular histograms it is evident that confluent HTEC on nanogrooved PDMS (A and B) do not align with their F-actin (A) or beta tubulin (B) cytoskeleton in the direction of the grooves (white arrows). On flat gelatin (C and D) HTEC F-actin (C) and beta tubulin (D) are not aligned in any given direction as corroborated by the corresponding angular histogram. Based on the angular histogram of confluent HTEC on nanogrooved gelatin (E and F) we see that the F-actin fibers (E) and tubulin fibers (F) do not align in the direction of the grooves, which is confirmed with the corresponding angular histograms.

101 fashion as the epithelial cell lines we tested, primary human tracheal epithelium (HTEC) were seeded onto collagen coated PDMS nanogrooves (Figure 17A and B), flat gelatin (Figure 17C and D) and onto gelatin nanogrooves (Figure 17E and F). Using confocal microscopy, we studied the orientation of F-actin and β-tubulin, which is a component of microtubules.

Our addition of another cytoskeletal marker was based on literature that suggested that when polarized, primary multiciliated epithelia show polarization apically through cortical actin and through polarized microtubules.139,140 We sought evidence of actin or microtubule alignment in our HTEC cultures. As primary epithelial cells grow slower than cell lines, we allowed primary culture 3 days of culture on nanogrooves before assessing cytoskeletal alignment.

In nanogrooved PDMS (Figure 17A and B), while the actin and microtubule cytoskeleton is sharp and well defined, it is evident that planar alignment and elongation of the cytoskeleton in the direction of the grooves has not occurred. Actin filaments are not parallel to the underlying surface topography while microtubules appeared to be organized in a radial or random fashion within the HTEC. On flat gelatin, we observed that the actin cytoskeleton (Figure 17C) was diffuse, and the fibers were less defined than on nanogrooved PDMS. The corresponding angular histogram showed a range of fiber angle orientation. With regards to microtubules on flat gelatin

(17D), the cytoskeleton appeared diffuse and less defined than on PDMS. On nanogrooved gelatin (Figure 17E and F) we observed a more diffuse arrangement of the F-actin cytoskeleton.

Both actin and microtubules did not appear to organize in the direction of the grooves. Following angular quantification, the corresponding angular histograms showed randomly aligned cytoskeleton seen through the wide distribution of angles. In summary, no planar alignment parallel to the nanogrooves, was evident both visually or through angular quantification in either nanogroove conditions for both actin and tubulin based cytoskeletons.

102

5.5 Future directions

Tracheal epithelium is polarized along two axes: an apicobasal axis, which is evident through difference in structure and function of apical and basal epithelial compartments. In vitro air liquid interface is able to recapitulate apicobasal polarization as seen through apical junction formation and ciliogenesis. The second axis of great importance to tracheal epithelium is the planar axis. This axis is readily understood through the observation of organized and coordinated ciliary beating, which moves foreign bodies and mucus consistently out of the airway; which is critical to mucociliary clearance and maintaining optimal airway health and function. ALI culture does not stimulate planar polarization and thus this epithelium does not function optimally for clinical purposes.

We now have a human primary tracheal epithelial cell, which in our hands can apicobasally polarize, which we have characterized through the formation of multiciliated cells, and tight junctions in standard and gelatin ALI culture. While multi-ciliogenesis does indeed occur in our gelatin ALI inserts there is a positive trend we noticed with increasing the thickness of gelatin: as gelatin thickness increased, the number of ciliated cells decreased. The results of our basic, visual quantification can be seen in Supplementary figure 10. This result allows us now to continue our ongoing experiments to assess the effect of aligned cytoskeletal components on the alignment and organization of motile cilia. To visualize the direction of beating, we will perform a fluorescent bead assay which through live-cell microscopy allows us to track fluorescent beads as they are moved by our motile-ciliated epithelium.167 To assess the effect of the thickness of the gelatin gel a more comprehensive experiment ranging from 10-750um thick gelatin gel will be used during HTEC ALI polarization. These primary cultures require a minimum of 31 days

103 for full apicobasal polarization to occur. While these cultures were differentiating, we began assessing the cytoskeletal alignment of the confluent HTECs on our nanogroove devices.

As seen in Figure 17 HTEC do not align to nanogrooved PDMS or nanogrooved gelatin.

While this lack of alignment is disappointing, the literature to date suggests that altering the mechanical cues of the nanogroove topography should produce a more instructive nanogroove cue and initiate alignment and elongation of the HTEC cytoskeleton in the direction of the grooves. Our first change to the nanogroove topography will be enhancing the depth of the grooves. The is a great body of literature that suggests that increasing depth is a more direct route to induce alignment of the cells on grooved topography.275,293,294,299 Studies have found that while changing pitch somewhat orientation, altering depth has a greater effect.225

While depth may be an overriding signal to force any cell to align in the direction of the grooves, there are other epithelial specific considerations. While fibroblasts are the model cell type for testing contact guidance on grooves they are more simplistic in terms of their polarization. Epithelial cells are generally more complex as apical junctions form and are dynamically regulated by morphogenetic processes.14 Apical junctions are intimately linked and alter the cytoskeleton. These connections are part of a concerted epithelial programme that maintains the architecture of polarization; perhaps this primary tracheal epithelial cell will only form planar aligned cytoskeleton when they are polarized along the apical basal axis.14,140 We speculate that planar alignment of the cytoskeleton in primary epithelial cells may take place during apicobasal polarization as apical junctions are forming. These studies would have to be undertaken during ALI culture.

In the long term a device which organized differentiating and polarizing epithelium through cytoskeletal alignment on nanogroove topography moulded into a diffusible substrate in

104 vitro would create a highly manipulatable device. We could alter stiffness through altering the crosslinking concentration to better replicate the natural stiffness of tracheal basement membrane or replicate a more sclerotic environment. Within our laboratory group we have also considered other forms of topography such mPAD substrates.255 Designed by the laboratory of Chris Chen in Pennsylvania mPAD substrates are an array of elastomeric micropost substrates which we could use to alter the stiffness seen by maturing epithelium without altering the chemical signals presented. Finally, other forms of physical forces outside of topographical features could be utilized: a current project underway is to study the effect of air flow on developing epithelium.

5.6 Conclusions

Current nanogrooved substrates that align cells and cytoskeleton do not allow for nutrient diffusion through substrates moulded with nanogroove topography. We have created a device that consists of a nanogrooved gelatin hydrogel capable of maintaining surface topography in cell culture conditions. This device is manufactured from a PDMS nanogroove intermediate which is subsequently used to mould 5% gelatin. Certain epithelial cells have been found to align morphologically and cytoskeletally on our gelatin nanogroove device, in both sparse and confluent seeding densities. We have found that primary human tracheal epithelia form motile, multiciliated cells after 31 days in standard and gelatin insert ALI culture highlighting that our device is congruent with ALI culture and allowing for future experiments with nanogrooves to assess for ciliary alignment.

105

Chapter 6| Supplementary figures

106

Supplementary Figure 1. Macroscopic image of gelatin gels made from various percentages of gelatin.

Supplementary Figure 1. 20% (A), 15% (B), 12% (C), and 10% (D) gelatin do not remain intact and gelled at tissue culture conditions after crosslinking overnight with 5% glutaraldehyde.

107

Supplementary Figure 2. Macroscopic image of 5% gelatin gels crosslinked with various concentrations of glutaraldehyde.

Supplementary Figure 2. 5% gelatin crosslinked overnight with 1% (A), and 0.1% (B), remain intact after 48 hours of tissue culture conditions. Concentrations of 0.05% (C), and 0.025% (D) do not remain intact and gelled at tissue culture conditions after 48 hours.

108

Supplementary Figure 3. BEAS-2B do not align on nanogrooved collagen.

Supplementary Figure 3. BEAS-2B grown on flat collagen at sparse seeding density (A) do not align. Sparsely (B) and confluently (C) seeded on nanogrooved collagen, BEAS-2B do not align in the direction of the grooves.

109

Supplementary Figure 4 – Scanning electron micrographs of flat substrates

Supplementary Figure 4. Scanning electron micrographs of flat PDMS, gelatin, and collagen. Flat PDMS (A) has no apparent topographical features present, in a similar fashion, flat gelatin (B) has no topographical features. Unstamped collagen (C) shows no repeatable surface topographical features, aside from naturally occurring mesh-like structure.

110

Supplementary Figure 5 – National Institute of Health 3T3 fibroblasts align on nanogrooved substrates.

Supplementary Figure 5. Fluorescent micrographs of NIH 3T3 grown on nanogrooved and flat substrates in sparse and confluent seedings. Sparsely and confluently seeded 3T3 on nanogrooved PDMS (A and B) show alignment in the direction of the grooves, indicated by the arrows, as seen in the corresponding angular histograms. On flat gelatin at sparse (C) and confluent (D) seedings, random cytoskeletal alignment is present, as seen from corresponding angular histograms. However, on nanogroove gelatin hydrogels seeded at sparse (E) and confluent (F) densities, cell orientation parallel to the nanogrooves is seen based on corresponding angular histograms.

111

Supplementary Figure 6 –IMCD3 epithelia on flat PDMS do not align morphologically.

Supplementary Figure 6 – IMCD3 on flat PDMS at sparse and confluent seeding densities. Sparsely seeded IMCD3 on flat PDMS (A) do not show any alignment. Confluent IMCD3 (B) show no alignment on flat PDMS.

112

Supplementary Figure 7 - IMCD3 epithelia on flat PDMS do not align cytoskeletally.

Supplementary Figure 7 – IMCD3 on flat PDMS at sparse and confluent seeding densities. Sparsely seeded IMCD3 on flat PDMS (A) do not show any alignment of F-actin. Confluent IMCD3 (B) show no alignment of F-actin on flat PDMS.

113

Supplementary Figure 8 –BEAS-2B sparsely seeded on nanogrooved substrates align morphologically.

Supplementary Figure 8 – Sparsely seeded BEAS-2B on nanogrooved PDMS (A) aligns in the direction of the grooves, seen in the corresponding angular histogram. On flat gelatin (B), sparsely seeded BEAS-2B do not show any alignment. While on nanogrooved gelatin, BEAS-2B align in the direction of the grooves, seen in the corresponding angular histogram.

114

Supplementary Figure 9 – BEAS-2B sparsely seeded on nanogrooved substrates align cytoskeletally.

Supplementary Figure 9 – Sparse BEAS-2B on nanogrooved PDMS (A) has aligned F-Actin in the direction of the grooves, seen in the corresponding angular histogram. On flat gelatin (B), sparsely seeded BEAS-2B do not show any alignment of the F-actin cytoskeleton. While on nanogrooved gelatin, BEAS-2B F-actin aligns in the direction of the grooves, seen in the corresponding angular histogram.

115

Supplementary Figure 10 – Mean number of ciliated cells in control and gelatin ALI inserts.

Supplementary Figure 10 – In control ALI conditions, the mean number of ciliated cells per confocal micrograph was 100. In 10um thick gelatin, this decreased to 66 ciliated cells per micrograph. In 500um thick gelatin, this decreased to 36 ciliated cells per micrograph.

116

Chapter 7| Appendix

117

Appendix 1 - Standard Operating Procedure for isolation of normal human tracheal epithelial cells – From the Moghal Lab

Personal Protective Equipment (PPE) A. All dissections and cell culture work shall be done in a class II type A/B3 BSL-2 biosafety cabinet. B. Workers shall wear gloves and a lab coat.

Material Fresh tracheal/carinal tissue from lung transplants. Tissue is typically placed in MEM with antibiotics and stored at 4oC for pick-up by us. The surgeons call my cell phone to notify me about a pick-up. We can stock their refrigerator (TMDT Keshavjee Lab, 2nd floor) with up to 6 urine specimen containers containing the storage solution. Typically, the tissue is in only one container. Tissue can be picked up 24 hrs after notification without a significant loss of viability (High pH, e.g. 9.0, also seems to be tolerated by the cells).

Day 1 1. Add antibiotics to 200-250 mL of MEM [Amphotericin B (200X), P/S/G (100X), and gentamycin 1000X)]. Parafilm the MEM bottle after use, as the pH tends to increase with exposure to air. 2. Transfer 25 mL of MEM to a 50 ml tube. 3. Weigh out 4 mg DNAse I. It can blow away very easily, so be careful. Use a blue tip to pipet 1 ml of MEM onto the weigh boat to dissolve the DNAse. Transfer to the 50 ml tube. 4. Weigh out 56 mg of Pronase. Add to the 50 ml tube. Shake to dissolve all powders. It should take a few minutes of shaking and inverting. 5. Filter sterilize the solution in the TC hood into a fresh 50 ml tube using a 30 ml syringe and a 0.22 um filter. Add MEM to bring up the volume to 40 mL and aliquot into 2 x 50 ml tubes (each with 20 mL of MEM solution). 6. Fill a black tray with ice. 7. Place two 10 cm dishes on ice. Pour some MEM into the two dishes. 8. Open a package of sterile forceps and large scissors. Use the covers of the 10 cm dishes to place the instruments on the surface of the Biosafety cabinet. 9. Record the donor number and any other info on the specimen container. 10. Use sterile forceps to place the tissue in one 10 cm dish. 11. Use the scissors to trim away the fatty tissue. 12. Transfer the cleaned tissue to the second 10 cm dish. 13. Use the scissors to cut the tissue longitudinally to expose the inner lumen containing the epithelial cells. Cut the tissue into large “exposed” pieces and wash them with MEM. 14. Split the tissue pieces into the 2 x 50 ml tubes. 15. Incubate at 4oC for 24 hrs (don’t need to agitate the tissue).

Decontamination

1. Transfer pieces of fatty tissue to a pad of paper towels, fold, and dispose of in the Biohazard waste.

118

2. Pour the contaminated MEM solution into a glass beaker with water and commercial bleach to about 5-10% total volume. Soak the instruments in the beaker for at least 20 minutes. 3. Wash the instruments with soap and water, dry, and re-autoclave. 4. Pour the decontaminated solution down the sink. 5. Wipe the Biosafety cabinet surface with 70% ethanol.

Day 2 1. Prepare another 40 mL of digestion solution in MEM. 2. Pour the contents of the tubes into 1 or 2 x 10 cm dishes. Wash the pieces with MEM and transfer the pieces back to the 50 ml tubes. Add fresh digestion solution and incubate for another 24 hrs at 4oC. 3. Transfer the MEM solution containing the cells to new 50 ml tubes. Rinse the 10 cm dish with 10 mL of PBS and combine with the cell suspension. 4. Pellet cells for 10 minutes at 1000 rpm. 5. After verifying the presence of pellets, carefully aspirate the supernatant using a yellow tip at the end of the Pasteur pipette. Leave a little MEM solution at the bottom. 6. Flick the pellet to partially resuspend it. 7. Add 0.5 mL of HeBS to each tube. Use a blue tip to resuspend each pellet. Most of the particulate matter should go into solution. If there are chunks of fatty tissue, remove them with a blue tip. 8. After resuspension, combine the cell suspensions (total volume should be ~1.1 mL). 9. Mix 10 ul of cells with 40 ul of PBS and 10 ul of trypan blue (dilution factor is 6). Wait 5 minutes. 10. Count cells with haemocytometer. Look for beating cilia on some cells. 11. If cells are not too clumped, add 100 ul of FBS to neutralize residual pronase, and pellet for 5 minutes at 1000 rpm. 12. Aspirate as before. Flick to partially resuspend the pellet. 13. Add 500 ul of freezing medium. Resuspend with a blue tip. Add more freezing media so that final cell concentration is close to 1 to 1.3 x 106 cells/ml. 14. Dispense 1 ml/vial. Mark vial with strain number, time point, and date. 15. Freeze at -80oC in either Mr. Frosty or in between 2 Styrofoam racks for no more than 24 hrs.

Day 3 1. Transfer cells from -80oC to liquid nitrogen. Enter information in the excel database on the lab server. 2. Process second digestion. 3. Dispose of the tracheal tissue as on Day 1.

Day 4 1. Transfer cells from second harvest to liquid nitrogen and enter information in the excel database.

Materials Amphotericin B (stored at -20oC, but keep at 4oC after thawing an aliquot)

119

P/S/G (stored at -20oC, but keep at 4oC after thawing an aliquot) Gentamycin (stored at 4oC) MEM (stored at 4oC) DNAse I (DN25, Sigma, stored at -20oC) Pronase (aka Type XIV protease, stored at -20oC) Freezing media (10% FCS, 80% LHC-9, 10% DMSO)(store at -20oC) Autoclaved forceps and scissors (In an emergency, light an alcohol burner in the TC hood, soak the instruments in 95% ethanol, and flame them in the TC hood)

120

Chapter 8| References

121

1 Shubin, N. Your inner fish : a journey into the 3.5-billion-year history of the human body. 1st Vintage Books edn, (Vintage Books, 2009). 2 Jones, C. J. Epithelia : advances in cell physiology and cell culture. (Kluwer Academic, 1990). 3 Birchmeier, W. & Birchmeier, C. Epithelial morphogenesis in development and disease. (Harwood Academic, 1999). 4 Fleming, T. P. Epithelial organization and development. 1st ed edn, (Chapman & Hall, 1992). 5 Johnson, L. R. & Byrne, J. H. Essential medical physiology. 3rd edn, (Elsevier Academic Press, 2003). 6 Marieb, E. N., Mallatt, J. & Wilhelm, P. B. Human anatomy. 5th edn, (Pearson Benhamin Cummings, 2008). 7 Lodish, H. F. Molecular cell biology. 6th edn, (W.H. Freeman, 2008). 8 Warburton, D. et al. Stem/progenitor cells in lung development, injury repair, and regeneration. Proc Am Thorac Soc 5, 703-706, doi:10.1513/pats.200801-012AW (2008). 9 Baum, B. & Georgiou, M. Dynamics of adherens junctions in epithelial establishment, maintenance, and remodeling. J Cell Biol 192, 907-917, doi:10.1083/jcb.201009141 (2011). 10 Sawada, N. et al. Tight junctions and human diseases. Med Electron Microsc 36, 147- 156, doi:10.1007/s00795-003-0219-y (2003). 11 Nievers, M. G., Schaapveld, R. Q. & Sonnenberg, A. Biology and function of hemidesmosomes. Matrix Biol 18, 5-17 (1999). 12 Baranowsky, A. et al. Impaired wound healing in mice lacking the basement membrane protein nidogen 1. Matrix Biol 29, 15-21, doi:10.1016/j.matbio.2009.09.004 (2010). 13 St Johnston, D. & Ahringer, J. Cell polarity in eggs and epithelia: parallels and diversity. Cell 141, 757-774, doi:10.1016/j.cell.2010.05.011 (2010). 14 St Johnston, D. & Sanson, B. Epithelial polarity and morphogenesis. Curr Opin Cell Biol 23, 540-546, doi:10.1016/j.ceb.2011.07.005 (2011). 15 Lal-Nag, M. & Morin, P. J. The claudins. Genome Biol 10, 235, doi:10.1186/gb-2009-10- 8-235 (2009). 16 Shin, K., Fogg, V. C. & Margolis, B. Tight junctions and cell polarity. Annu Rev Cell Dev Biol 22, 207-235, doi:10.1146/annurev.cellbio.22.010305.104219 (2006). 17 Albelda, S. M. & Buck, C. A. Integrins and other cell adhesion molecules. FASEB journal : official publication of the Federation of American Societies for Experimental Biology 4, 2868-2880 (1990). 18 Farmer, S. G. & Hay, D. W. P. The Airway epithelium : physiology, pathophysiology, and pharmacology. (M. Dekker, 1991). 19 Koval, M. Tight junctions, but not too tight: fine control of lung permeability by claudins. Am J Physiol Lung Cell Mol Physiol 297, L217-218, doi:10.1152/ajplung.00196.2009 (2009). 20 Tsukita, S. & Furuse, M. Pores in the wall: claudins constitute tight junction strands containing aqueous pores. J Cell Biol 149, 13-16 (2000). 21 Widdicombe, J. H. & Welsh, M. J. Anion selectivity of the chloride-transport process in dog tracheal epithelium. Am J Physiol 239, C112-117 (1980).

122

22 Nathanson, I., Widdicombe, J. H. & Nadel, J. A. Effects of amphotericin B on ion and fluid movement across dog tracheal epithelium. Journal of applied physiology: respiratory, environmental and exercise physiology 55, 1257-1261 (1983). 23 Dunsmore, S. E. & Rannels, D. E. Extracellular matrix biology in the lung. Am J Physiol 270, L3-27 (1996). 24 Crapo, J. D., Barry, B. E., Gehr, P., Bachofen, M. & Weibel, E. R. Cell number and cell characteristics of the normal human lung. Am Rev Respir Dis 126, 332-337 (1982). 25 Rock, J. R. & Hogan, B. L. Epithelial progenitor cells in lung development, maintenance, repair, and disease. Annu Rev Cell Dev Biol 27, 493-512, doi:10.1146/annurev-cellbio- 100109-104040 (2011). 26 Morgan, K. T., Jiang, X. Z., Patterson, D. L. & Gross, E. A. The nasal mucociliary apparatus. Correlation of structure and function in the rat. Am Rev Respir Dis 130, 275- 281 (1984). 27 Serabjit-Singh, C. J., Wolf, C. R., Philpot, R. M. & Plopper, C. G. Cytochrome p-450: localization in rabbit lung. Science 207, 1469-1470 (1980). 28 Plopper, C. G., Mariassy, A. T. & Hill, L. H. Ultrastructure of the nonciliated bronchiolar epithelial (Clara) cell of mammalian lung: I. A comparison of rabbit, guinea pig, rat, hamster, and mouse. Exp Lung Res 1, 139-154 (1980). 29 Reynolds, S. D., Reynolds, P. R., Pryhuber, G. S., Finder, J. D. & Stripp, B. R. Secretoglobins SCGB3A1 and SCGB3A2 define secretory cell subsets in mouse and human airways. Am J Respir Crit Care Med 166, 1498-1509, doi:10.1164/rccm.200204- 285OC (2002). 30 Evans, M. J., Cabral-Anderson, L. J. & Freeman, G. Role of the Clara cell in renewal of the bronchiolar epithelium. Lab Invest 38, 648-653 (1978). 31 Reynolds, S. D. et al. Conditional clara cell ablation reveals a self-renewing progenitor function of pulmonary neuroendocrine cells. Am J Physiol Lung Cell Mol Physiol 278, L1256-1263 (2000). 32 Shan, L. et al. Bombesin-like peptide receptor gene expression, regulation, and function in fetal murine lung. Am J Physiol Lung Cell Mol Physiol 286, L165-173, doi:10.1152/ajplung.00436.2002 (2004). 33 Rock, J. R. et al. Basal cells as stem cells of the mouse trachea and human airway epithelium. Proc Natl Acad Sci U S A 106, 12771-12775, doi:10.1073/pnas.0906850106 (2009). 34 Nakajima, M. et al. Immunohistochemical and ultrastructural studies of basal cells, Clara cells and bronchiolar cuboidal cells in normal human airways. Pathol Int 48, 944-953 (1998). 35 Kajstura, J. et al. Evidence for human lung stem cells. N Engl J Med 364, 1795-1806, doi:10.1056/NEJMoa1101324 (2011). 36 Chapman, H. A. Toward lung regeneration. N Engl J Med 364, 1867-1868, doi:10.1056/NEJMe1101800 (2011). 37 Carlstedt, I. S., J. K. in Mucus and Mucosa Vol. 109 Ciba Foundation Symposium (ed J. Nugent, O'Connor, M) 157-172 (Pitman Medical, 1984). 38 Lamb, D. & Reid, L. Histochemical types of acidic glycoprotein produced by mucous cells of the tracheobronchial glands in man. J Pathol 98, 213-229, doi:10.1002/path.1710980402 (1969).

123

39 Clark, J. N., Dalbey, W. E. & Stephenson, K. B. Effect of sulfur dioxide on the morphology and mucin biosynthesis by the rat trachea. J Environ Pathol Toxicol 4, 197- 207 (1980). 40 Adler, K. B., Hendley, D. D. & Davis, G. S. Bacteria associated with obstructive pulmonary disease elaborate extracellular products that stimulate mucin secretion by explants of guinea pig airways. Am J Pathol 125, 501-514 (1986). 41 Even-Tzur, N., Kloog, Y., Wolf, M. & Elad, D. Mucus secretion and cytoskeletal modifications in cultured nasal epithelial cells exposed to wall shear stresses. Biophys J 95, 2998-3008, doi:10.1529/biophysj.107.127142 (2008). 42 King, M., Chang, H. K. & Weber, M. E. Resistance of mucus-lined tubes to steady and oscillatory airflow. Journal of applied physiology: respiratory, environmental and exercise physiology 52, 1172-1176 (1982). 43 Knowles, M. R. & Boucher, R. C. Mucus clearance as a primary innate defense mechanism for mammalian airways. J Clin Invest 109, 571-577, doi:10.1172/JCI15217 (2002). 44 Knowles, M. R., Leigh, M. W. & Zariwala, M. A. Cutting edge genetic studies in primary ciliary dyskinesia. Thorax, doi:10.1136/thoraxjnl-2012-201609 (2012). 45 Lund, V. J. Nasal physiology: neurochemical receptors, nasal cycle, and ciliary action. Allergy Asthma Proc 17, 179-184 (1996). 46 Shinogi, J. et al. Quantitative analysis of mucin and lectin in maxillary sinus fluids in patients with acute and chronic sinusitis. Laryngoscope 111, 240-245, doi:10.1097/00005537-200102000-00010 (2001). 47 Silberberg, A. A model for mucus glycoprotein structure. Comparison with submaxillary mucins. Biorheology 25, 799-801 (1988). 48 Verdugo, P., Tam, P. Y. & Butler, J. Conformational structure of respiratory mucus studied by laser correlation spectroscopy. Biorheology 20, 223-230 (1983). 49 Trindade, S. H. et al. Methods for studying mucociliary transport. Braz J Otorhinolaryngol 73, 704-712 (2007). 50 Matsui, H., Randell, S. H., Peretti, S. W., Davis, C. W. & Boucher, R. C. Coordinated clearance of periciliary liquid and mucus from airway surfaces. J Clin Invest 102, 1125- 1131, doi:10.1172/JCI2687 (1998). 51 Antunes, M. B. & Cohen, N. A. Mucociliary clearance--a critical upper airway host defense mechanism and methods of assessment. Curr Opin Allergy Clin Immunol 7, 5-10, doi:10.1097/ACI.0b013e3280114eef (2007). 52 Shah, A. S., Ben-Shahar, Y., Moninger, T. O., Kline, J. N. & Welsh, M. J. Motile cilia of human airway epithelia are chemosensory. Science 325, 1131-1134, doi:10.1126/science.1173869 (2009). 53 Button, B. & Boucher, R. C. Role of mechanical stress in regulating airway surface hydration and mucus clearance rates. Respir Physiol Neurobiol 163, 189-201, doi:10.1016/j.resp.2008.04.020 (2008). 54 Que, J., Choi, M., Ziel, J. W., Klingensmith, J. & Hogan, B. L. Morphogenesis of the trachea and esophagus: current players and new roles for noggin and Bmps. Differentiation 74, 422-437, doi:10.1111/j.1432-0436.2006.00096.x (2006). 55 Cardoso, W. V. & Lu, J. Regulation of early lung morphogenesis: questions, facts and controversies. Development 133, 1611-1624, doi:10.1242/dev.02310 (2006).

124

56 Tam, P. P., Kanai-Azuma, M. & Kanai, Y. Early endoderm development in vertebrates: lineage differentiation and morphogenetic function. Curr Opin Genet Dev 13, 393-400 (2003). 57 Ioannides, A. S. et al. Foregut separation and tracheo-oesophageal malformations: the role of tracheal outgrowth, dorso-ventral patterning and programmed cell death. Dev Biol 337, 351-362, doi:10.1016/j.ydbio.2009.11.005 (2010). 58 Morrisey, E. E. & Hogan, B. L. Preparing for the first breath: genetic and cellular mechanisms in lung development. Dev Cell 18, 8-23, doi:10.1016/j.devcel.2009.12.010 (2010). 59 Metzger, R. J., Klein, O. D., Martin, G. R. & Krasnow, M. A. The branching programme of mouse lung development. Nature 453, 745-750, doi:10.1038/nature07005 (2008). 60 Gontan, C. et al. Sox2 is important for two crucial processes in lung development: branching morphogenesis and epithelial cell differentiation. Dev Biol 317, 296-309, doi:10.1016/j.ydbio.2008.02.035 (2008). 61 Que, J., Luo, X., Schwartz, R. J. & Hogan, B. L. Multiple roles for Sox2 in the developing and adult mouse trachea. Development 136, 1899-1907, doi:10.1242/dev.034629 (2009). 62 Domyan, E. T. & Sun, X. Patterning and plasticity in development of the respiratory lineage. Dev Dyn 240, 477-485, doi:10.1002/dvdy.22504 (2011). 63 Salisbury, J. L., Baron, A. T. & Sanders, M. A. The centrin-based cytoskeleton of Chlamydomonas reinhardtii: distribution in interphase and mitotic cells. J Cell Biol 107, 635-641 (1988). 64 Tepass, U., Tanentzapf, G., Ward, R. & Fehon, R. Epithelial cell polarity and cell junctions in Drosophila. Annu Rev Genet 35, 747-784, doi:10.1146/annurev.genet.35.102401.091415 (2001). 65 Wodarz, A., Hinz, U., Engelbert, M. & Knust, E. Expression of crumbs confers apical character on plasma membrane domains of ectodermal epithelia of Drosophila. Cell 82, 67-76 (1995). 66 Hutterer, A., Betschinger, J., Petronczki, M. & Knoblich, J. A. Sequential roles of Cdc42, Par-6, aPKC, and Lgl in the establishment of epithelial polarity during Drosophila embryogenesis. Dev Cell 6, 845-854, doi:10.1016/j.devcel.2004.05.003 (2004). 67 Horikoshi, Y. et al. Interaction between PAR-3 and the aPKC-PAR-6 complex is indispensable for apical domain development of epithelial cells. J Cell Sci 122, 1595- 1606, doi:10.1242/jcs.043174 (2009). 68 Yu, W. et al. Beta1-integrin orients epithelial polarity via Rac1 and laminin. Mol Biol Cell 16, 433-445, doi:10.1091/mbc.E04-05-0435 (2005). 69 Assemat, E., Bazellieres, E., Pallesi-Pocachard, E., Le Bivic, A. & Massey-Harroche, D. Polarity complex proteins. Biochim Biophys Acta 1778, 614-630, doi:10.1016/j.bbamem.2007.08.029 (2008). 70 Fristrom, D. The cellular basis of epithelial morphogenesis. A review. Tissue Cell 20, 645-690 (1988). 71 Niessen, C. M. & Gottardi, C. J. Molecular components of the adherens junction. Biochim Biophys Acta 1778, 562-571, doi:10.1016/j.bbamem.2007.12.015 (2008). 72 Farquhar, M. G. & Palade, G. E. Junctional complexes in various epithelia. J Cell Biol 17, 375-412 (1963).

125

73 Cummins, P. M. Occludin: one protein, many forms. Mol Cell Biol 32, 242-250, doi:10.1128/MCB.06029-11 (2012). 74 Fromter, E. & Diamond, J. Route of passive ion permeation in epithelia. Nat New Biol 235, 9-13 (1972). 75 Claude, P. & Goodenough, D. A. Fracture faces of zonulae occludentes from "tight" and "leaky" epithelia. J Cell Biol 58, 390-400 (1973). 76 Turner, J. R. 'Putting the squeeze' on the tight junction: understanding cytoskeletal regulation. Semin Cell Dev Biol 11, 301-308, doi:10.1006/scdb.2000.0180 (2000). 77 Aijaz, S., Balda, M. S. & Matter, K. Tight junctions: molecular architecture and function. Int Rev Cytol 248, 261-298, doi:10.1016/S0074-7696(06)48005-0 (2006). 78 Mineta, K. et al. Predicted expansion of the claudin multigene family. FEBS Lett 585, 606-612, doi:10.1016/j.febslet.2011.01.028 (2011). 79 Furuse, M., Sasaki, H., Fujimoto, K. & Tsukita, S. A single gene product, claudin-1 or -2, reconstitutes tight junction strands and recruits occludin in fibroblasts. J Cell Biol 143, 391-401 (1998). 80 Elkouby-Naor, L. & Ben-Yosef, T. Functions of claudin tight junction proteins and their complex interactions in various physiological systems. Int Rev Cell Mol Biol 279, 1-32, doi:10.1016/S1937-6448(10)79001-8 (2010). 81 Overgaard, C. E., Mitchell, L. A. & Koval, M. Roles for claudins in alveolar epithelial barrier function. Ann N Y Acad Sci 1257, 167-174, doi:10.1111/j.1749- 6632.2012.06545.x (2012). 82 Colegio, O. R., Van Itallie, C., Rahner, C. & Anderson, J. M. Claudin extracellular domains determine paracellular charge selectivity and resistance but not tight junction fibril architecture. Am J Physiol Cell Physiol 284, C1346-1354, doi:10.1152/ajpcell.00547.2002 (2003). 83 Ranganathan, R. & Ross, E. M. PDZ domain proteins: scaffolds for signaling complexes. Curr Biol 7, R770-773 (1997). 84 Itoh, M. et al. Direct binding of three tight junction-associated MAGUKs, ZO-1, ZO-2, and ZO-3, with the COOH termini of claudins. J Cell Biol 147, 1351-1363 (1999). 85 Umeda, K. et al. ZO-1 and ZO-2 independently determine where claudins are polymerized in tight-junction strand formation. Cell 126, 741-754, doi:10.1016/j.cell.2006.06.043 (2006). 86 Furuse, M. et al. Occludin: a novel integral membrane protein localizing at tight junctions. J Cell Biol 123, 1777-1788 (1993). 87 Furuse, M. et al. Direct association of occludin with ZO-1 and its possible involvement in the localization of occludin at tight junctions. J Cell Biol 127, 1617-1626 (1994). 88 Saitou, M. et al. Complex phenotype of mice lacking occludin, a component of tight junction strands. Mol Biol Cell 11, 4131-4142 (2000). 89 McCarthy, K. M. et al. Occludin is a functional component of the tight junction. J Cell Sci 109 ( Pt 9), 2287-2298 (1996). 90 Wong, V. & Gumbiner, B. M. A synthetic peptide corresponding to the extracellular domain of occludin perturbs the tight junction permeability barrier. J Cell Biol 136, 399- 409 (1997). 91 Hong, S., Troyanovsky, R. B. & Troyanovsky, S. M. Spontaneous assembly and active disassembly balance adherens junction homeostasis. Proc Natl Acad Sci U S A 107, 3528-3533, doi:10.1073/pnas.0911027107 (2010).

126

92 Lodish, H. F. Molecular cell biology. 5th edn, (W.H. Freeman and Company, 2004). 93 Leckband, D. & Sivasankar, S. Mechanism of homophilic cadherin adhesion. Curr Opin Cell Biol 12, 587-592 (2000). 94 Yap, A. S., Niessen, C. M. & Gumbiner, B. M. The juxtamembrane region of the cadherin cytoplasmic tail supports lateral clustering, adhesive strengthening, and interaction with p120ctn. J Cell Biol 141, 779-789 (1998). 95 Rimm, D. L., Koslov, E. R., Kebriaei, P., Cianci, C. D. & Morrow, J. S. Alpha 1(E)- catenin is an actin-binding and -bundling protein mediating the attachment of F-actin to the membrane adhesion complex. Proc Natl Acad Sci U S A 92, 8813-8817 (1995). 96 Yonemura, S., Wada, Y., Watanabe, T., Nagafuchi, A. & Shibata, M. alpha-Catenin as a tension transducer that induces adherens junction development. Nat Cell Biol 12, 533- 542, doi:10.1038/ncb2055 (2010). 97 Thoreson, M. A. et al. Selective uncoupling of p120(ctn) from E-cadherin disrupts strong adhesion. J Cell Biol 148, 189-202 (2000). 98 Gumbiner, B., Stevenson, B. & Grimaldi, A. The role of the cell adhesion molecule uvomorulin in the formation and maintenance of the epithelial junctional complex. J Cell Biol 107, 1575-1587 (1988). 99 Tsukasaki, Y. et al. Role of multiple bonds between the single cell adhesion molecules, nectin and cadherin, revealed by high sensitive force measurements. J Mol Biol 367, 996- 1006, doi:10.1016/j.jmb.2006.12.022 (2007). 100 Adams, C. L., Nelson, W. J. & Smith, S. J. Quantitative analysis of cadherin-catenin- actin reorganization during development of cell-cell adhesion. J Cell Biol 135, 1899-1911 (1996). 101 Brieher, W. M., Yap, A. S. & Gumbiner, B. M. Lateral dimerization is required for the homophilic binding activity of C-cadherin. J Cell Biol 135, 487-496 (1996). 102 Yap, A. S., Brieher, W. M., Pruschy, M. & Gumbiner, B. M. Lateral clustering of the adhesive ectodomain: a fundamental determinant of cadherin function. Curr Biol 7, 308- 315 (1997). 103 Drees, F., Pokutta, S., Yamada, S., Nelson, W. J. & Weis, W. I. Alpha-catenin is a molecular switch that binds E-cadherin-beta-catenin and regulates actin-filament assembly. Cell 123, 903-915, doi:10.1016/j.cell.2005.09.021 (2005). 104 Itoh, M., Nagafuchi, A., Moroi, S. & Tsukita, S. Involvement of ZO-1 in cadherin-based cell adhesion through its direct binding to alpha catenin and actin filaments. J Cell Biol 138, 181-192 (1997). 105 de Pereda, J. M., Ortega, E., Alonso-Garcia, N., Gomez-Hernandez, M. & Sonnenberg, A. Advances and perspectives of the architecture of hemidesmosomes: lessons from structural biology. Cell Adh Migr 3, 361-364 (2009). 106 Dusek, R. L., Jones, J. C. R., and Green, K. J. in Encyclopedia of Biological Chemistry Vol. 1 569-576 (Elsevier, 2004). 107 Jones, J. C., Hopkinson, S. B. & Goldfinger, L. E. Structure and assembly of hemidesmosomes. Bioessays 20, 488-494, doi:10.1002/(SICI)1521- 1878(199806)20:6<488::AID-BIES7>3.0.CO;2-I (1998). 108 Borradori, L. & Sonnenberg, A. Structure and function of hemidesmosomes: more than simple adhesion complexes. J Invest Dermatol 112, 411-418, doi:10.1046/j.1523- 1747.1999.00546.x (1999).

127

109 Zhang, H. & Labouesse, M. The making of hemidesmosome structures in vivo. Dev Dyn 239, 1465-1476, doi:10.1002/dvdy.22255 (2010). 110 Dowling, J., Yu, Q. C. & Fuchs, E. Beta4 integrin is required for hemidesmosome formation, cell adhesion and cell survival. J Cell Biol 134, 559-572 (1996). 111 Wilhelmsen, K., Litjens, S. H. & Sonnenberg, A. Multiple functions of the integrin alpha6beta4 in epidermal homeostasis and tumorigenesis. Mol Cell Biol 26, 2877-2886, doi:10.1128/MCB.26.8.2877-2886.2006 (2006). 112 Aumailley, M. et al. A simplified laminin nomenclature. Matrix Biol 24, 326-332, doi:10.1016/j.matbio.2005.05.006 (2005). 113 Garcia-Alvarez, B., Bobkov, A., Sonnenberg, A. & de Pereda, J. M. Structural and functional analysis of the actin binding domain of plectin suggests alternative mechanisms for binding to F-actin and integrin beta4. Structure 11, 615-625 (2003). 114 Niessen, C. M., Hulsman, E. H., Oomen, L. C., Kuikman, I. & Sonnenberg, A. A minimal region on the integrin beta4 subunit that is critical to its localization in hemidesmosomes regulates the distribution of HD1/plectin in COS-7 cells. J Cell Sci 110 ( Pt 15), 1705-1716 (1997). 115 de Pereda, J. M., Lillo, M. P. & Sonnenberg, A. Structural basis of the interaction between integrin alpha6beta4 and plectin at the hemidesmosomes. Embo J 28, 1180- 1190, doi:10.1038/emboj.2009.48 (2009). 116 Litjens, S. H., de Pereda, J. M. & Sonnenberg, A. Current insights into the formation and breakdown of hemidesmosomes. Trends Cell Biol 16, 376-383, doi:10.1016/j.tcb.2006.05.004 (2006). 117 Goldfinger, L. E. et al. The alpha3 laminin subunit, alpha6beta4 and alpha3beta1 integrin coordinately regulate wound healing in cultured epithelial cells and in the skin. J Cell Sci 112 ( Pt 16), 2615-2629 (1999). 118 Tsuruta, D., Hopkinson, S. B. & Jones, J. C. Hemidesmosome protein dynamics in live epithelial cells. Cell Motil Cytoskeleton 54, 122-134, doi:10.1002/cm.10089 (2003). 119 Jain, R. et al. Temporal relationship between primary and motile ciliogenesis in airway epithelial cells. Am J Respir Cell Mol Biol 43, 731-739, doi:10.1165/rcmb.2009-0328OC (2010). 120 Lazarowski, E. R. & Boucher, R. C. Purinergic receptors in airway epithelia. Curr Opin Pharmacol 9, 262-267, doi:10.1016/j.coph.2009.02.004 (2009). 121 Woodworth, B. A. et al. Zinc increases ciliary beat frequency in a calcium-dependent manner. Am J Rhinol Allergy 24, 6-10, doi:10.2500/ajra.2010.24.3379 (2010). 122 Button, B., Picher, M. & Boucher, R. C. Differential effects of cyclic and constant stress on ATP release and mucociliary transport by human airway epithelia. J Physiol 580, 577- 592, doi:10.1113/jphysiol.2006.126086 (2007). 123 Satir, P. & Christensen, S. T. Overview of structure and function of mammalian cilia. Annu Rev Physiol 69, 377-400, doi:10.1146/annurev.physiol.69.040705.141236 (2007). 124 Manton, I. Observations with the Electron Microscope on the Cell Structure of the Antheridium and Spermatozoid of Sphagnum. Journal of Experimental Botany 8, 382- 400 (1957). 125 Ishikawa, H. & Marshall, W. F. Ciliogenesis: building the cell's antenna. Nat Rev Mol Cell Biol 12, 222-234, doi:10.1038/nrm3085 (2011).

128

126 Sotelo, J. R. & Trujillo-Cenóz, O. Electron microscope study on the development of ciliary components of the neural epithelium of the chick embryo. Cell Tissue Res 49, 1- 12, doi:10.1007/bf00335059 (1958). 127 Gaertig, J. & Wloga, D. Ciliary tubulin and its post-translational modifications. Curr Top Dev Biol 85, 83-113, doi:10.1016/S0070-2153(08)00804-1 (2008). 128 Boron, W. F. & Boulpaep, E. L. Medical physiology : a cellular and molecular approach. 2nd edn, (Saunders/Elsevier, 2009). 129 Yang, P. et al. Radial spoke proteins of Chlamydomonas flagella. J Cell Sci 119, 1165- 1174, doi:10.1242/jcs.02811 (2006). 130 Afzelius, B. A. & Maunsbach, A. B. Biological ultrastructure research; the first 50 years. Tissue Cell 36, 83-94, doi:10.1016/j.tice.2003.11.001 (2004). 131 Nicastro, D. et al. The molecular architecture of axonemes revealed by cryoelectron tomography. Science 313, 944-948, doi:10.1126/science.1128618 (2006). 132 Fisch, C. & Dupuis-Williams, P. Ultrastructure of cilia and flagella - back to the future! Biol Cell 103, 249-270, doi:10.1042/BC20100139 (2011). 133 Bertelli, E. & Regoli, M. A morphological study of the primary cilia in the rat pancreatic ductal system: ultrastructural features and variability. Acta Anat (Basel) 151, 194-197 (1994). 134 Kubo, A., Yuba-Kubo, A., Tsukita, S. & Amagai, M. Sentan: a novel specific component of the apical structure of vertebrate motile cilia. Mol Biol Cell 19, 5338-5346, doi:10.1091/mbc.E08-07-0691 (2008). 135 Dutcher, S. K. Elucidation of basal body and centriole functions in Chlamydomonas reinhardtii. Traffic 4, 443-451 (2003). 136 Gilula, N. B. & Satir, P. The ciliary necklace. A ciliary membrane specialization. J Cell Biol 53, 494-509 (1972). 137 Doxsey, S. Re-evaluating centrosome function. Nat Rev Mol Cell Biol 2, 688-698, doi:10.1038/35089575 (2001). 138 Salisbury, J. L., Suino, K. M., Busby, R. & Springett, M. Centrin-2 is required for centriole duplication in mammalian cells. Curr Biol 12, 1287-1292 (2002). 139 Kunimoto, K. et al. Coordinated ciliary beating requires Odf2-mediated polarization of basal bodies via basal feet. Cell 148, 189-200, doi:10.1016/j.cell.2011.10.052 (2012). 140 Werner, M. E. et al. Actin and microtubules drive differential aspects of planar cell polarity in multiciliated cells. J Cell Biol 195, 19-26, doi:10.1083/jcb.201106110 (2011). 141 Hackett, B. P. et al. Primary structure of hepatocyte nuclear factor/forkhead homologue 4 and characterization of gene expression in the developing respiratory and reproductive epithelium. Proc Natl Acad Sci U S A 92, 4249-4253 (1995). 142 You, Y. et al. Role of f-box factor foxj1 in differentiation of ciliated airway epithelial cells. Am J Physiol Lung Cell Mol Physiol 286, L650-657, doi:10.1152/ajplung.00170.2003 (2004). 143 Yu, X., Ng, C. P., Habacher, H. & Roy, S. Foxj1 transcription factors are master regulators of the motile ciliogenic program. Nat Genet 40, 1445-1453, doi:10.1038/ng.263 (2008). 144 Pan, J., You, Y., Huang, T. & Brody, S. L. RhoA-mediated apical actin enrichment is required for ciliogenesis and promoted by Foxj1. J Cell Sci 120, 1868-1876, doi:10.1242/jcs.005306 (2007).

129

145 Anderson, C. T. & Stearns, T. Centriole age underlies asynchronous primary cilium growth in mammalian cells. Curr Biol 19, 1498-1502, doi:10.1016/j.cub.2009.07.034 (2009). 146 Parker, J. D. et al. Centrioles are freed from cilia by severing prior to mitosis. Cytoskeleton (Hoboken) 67, 425-430, doi:10.1002/cm.20454 (2010). 147 Piao, T. et al. A microtubule depolymerizing kinesin functions during both flagellar disassembly and flagellar assembly in Chlamydomonas. Proc Natl Acad Sci U S A 106, 4713-4718, doi:10.1073/pnas.0808671106 (2009). 148 Sorokin, S. P. Reconstructions of centriole formation and ciliogenesis in mammalian lungs. J Cell Sci 3, 207-230 (1968). 149 Sorokin, S. Centrioles and the formation of rudimentary cilia by fibroblasts and smooth muscle cells. J Cell Biol 15, 363-377 (1962). 150 Park, T. J., Mitchell, B. J., Abitua, P. B., Kintner, C. & Wallingford, J. B. Dishevelled controls apical docking and planar polarization of basal bodies in ciliated epithelial cells. Nat Genet 40, 871-879, doi:10.1038/ng.104 (2008). 151 Chailley B, B.-U. E., Sandoz D. Structure and assembly of the oviduct ciliary membrane. In Ciliary and Flagellar Membranes. 337–362 (Plenum, 1990). 152 Fawcett, D. W. & Porter, K. R. A study of the fine structure of ciliated epithelia. J Morphol 94, 221-281, doi:10.1002/jmor.1050940202 (1954). 153 Satir, P. Cilia and related microtubular arrays in the eukaryotic cell. 787-817 (Oxford University Press, 1997). 154 Habermacher, G. & Sale, W. S. Regulation of flagellar dynein by phosphorylation of a 138-kD inner arm dynein intermediate chain. J Cell Biol 136, 167-176 (1997). 155 Christensen, S. T. et al. A regulatory light chain of ciliary outer arm dynein in Tetrahymena thermophila. J Biol Chem 276, 20048-20054, doi:10.1074/jbc.M008412200 (2001). 156 Pedersen, L. B. & Rosenbaum, J. L. Intraflagellar transport (IFT) role in ciliary assembly, resorption and signalling. Curr Top Dev Biol 85, 23-61, doi:10.1016/S0070- 2153(08)00802-8 (2008). 157 Cole, D. G. et al. Chlamydomonas kinesin-II-dependent intraflagellar transport (IFT): IFT particles contain proteins required for ciliary assembly in Caenorhabditis elegans sensory neurons. J Cell Biol 141, 993-1008 (1998). 158 Iomini, C., Li, L., Esparza, J. M. & Dutcher, S. K. Retrograde intraflagellar transport mutants identify complex A proteins with multiple genetic interactions in Chlamydomonas reinhardtii. Genetics 183, 885-896, doi:10.1534/genetics.109.101915 (2009). 159 Follit, J. A., Tuft, R. A., Fogarty, K. E. & Pazour, G. J. The intraflagellar transport protein IFT20 is associated with the Golgi complex and is required for cilia assembly. Mol Biol Cell 17, 3781-3792, doi:10.1091/mbc.E06-02-0133 (2006). 160 Singla, V. & Reiter, J. F. The primary cilium as the cell's antenna: signaling at a sensory organelle. Science 313, 629-633, doi:10.1126/science.1124534 (2006). 161 Leigh, M. W. Primary ciliary dyskinesia. Semin Respir Crit Care Med 24, 653-662, doi:10.1055/s-2004-815661 (2003). 162 Green, K. et al. Ventilation inhomogeneity in children with primary ciliary dyskinesia. Thorax 67, 49-53, doi:10.1136/thoraxjnl-2011-200726 (2012). 163 Serway, R. A. Principles of physics. (Saunders College Pub., 1994).

130

164 Mitchell, B. et al. The PCP pathway instructs the planar orientation of ciliated cells in the Xenopus larval skin. Curr Biol 19, 924-929, doi:10.1016/j.cub.2009.04.018 (2009). 165 Mitchell, B., Jacobs, R., Li, J., Chien, S. & Kintner, C. A positive feedback mechanism governs the polarity and motion of motile cilia. Nature 447, 97-101, doi:10.1038/nature05771 (2007). 166 Sanderson, M. J. & Sleigh, M. A. Ciliary activity of cultured rabbit tracheal epithelium: beat pattern and metachrony. J Cell Sci 47, 331-347 (1981). 167 Guirao, B. et al. Coupling between hydrodynamic forces and planar cell polarity orients mammalian motile cilia. Nat Cell Biol 12, 341-350, doi:10.1038/ncb2040 (2010). 168 Harris, A. K., Wild, P. & Stopak, D. Silicone rubber substrata: a new wrinkle in the study of cell locomotion. Science 208, 177-179 (1980). 169 Eyckmans, J., Boudou, T., Yu, X. & Chen, C. S. A hitchhiker's guide to mechanobiology. Dev Cell 21, 35-47, doi:10.1016/j.devcel.2011.06.015 (2011). 170 Lo, C. M., Wang, H. B., Dembo, M. & Wang, Y. L. Cell movement is guided by the rigidity of the substrate. Biophys J 79, 144-152, doi:10.1016/S0006-3495(00)76279-5 (2000). 171 del Rio, A. et al. Stretching single talin rod molecules activates vinculin binding. Science 323, 638-641, doi:10.1126/science.1162912 (2009). 172 Parsons, J. T., Horwitz, A. R. & Schwartz, M. A. Cell adhesion: integrating cytoskeletal dynamics and cellular tension. Nat Rev Mol Cell Biol 11, 633-643, doi:10.1038/nrm2957 (2010). 173 Baker, E. L. & Zaman, M. H. The biomechanical integrin. J Biomech 43, 38-44, doi:10.1016/j.jbiomech.2009.09.007 (2010). 174 Wang, N. et al. Mechanical behavior in living cells consistent with the tensegrity model. Proc Natl Acad Sci U S A 98, 7765-7770, doi:10.1073/pnas.141199598 (2001). 175 Galbraith, C. G., Yamada, K. M. & Galbraith, J. A. Polymerizing actin fibers position integrins primed to probe for adhesion sites. Science 315, 992-995, doi:10.1126/science.1137904 (2007). 176 Nelson, C. M. & Gleghorn, J. P. Sculpting Organs: Mechanical Regulation of Tissue Development. Annu Rev Biomed Eng, doi:10.1146/annurev-bioeng-071811-150043 (2011). 177 Discher, D. E., Janmey, P. & Wang, Y. L. Tissue cells feel and respond to the stiffness of their substrate. Science 310, 1139-1143, doi:10.1126/science.1116995 (2005). 178 Pelham, R. J., Jr. & Wang, Y. Cell locomotion and focal adhesions are regulated by substrate flexibility. Proc Natl Acad Sci U S A 94, 13661-13665 (1997). 179 Nauli, S. M. et al. Polycystins 1 and 2 mediate mechanosensation in the primary cilium of kidney cells. Nat Genet 33, 129-137, doi:10.1038/ng1076 (2003). 180 Torgersen, J. Situs inversus, asymmetry, and twinning. Am J Hum Genet 2, 361-370 (1950). 181 Coren, M. E., Meeks, M., Morrison, I., Buchdahl, R. M. & Bush, A. Primary ciliary dyskinesia: age at diagnosis and symptom history. Acta Paediatr 91, 667-669 (2002). 182 Noone, P. G. et al. Primary ciliary dyskinesia: diagnostic and phenotypic features. Am J Respir Crit Care Med 169, 459-467, doi:10.1164/rccm.200303-365OC (2004). 183 Noone, P. G. et al. Effect of aerosolized uridine-5'-triphosphate on airway clearance with cough in patients with primary ciliary dyskinesia. Am J Respir Crit Care Med 160, 144- 149 (1999).

131

184 Afzelius, B. A. & Eliasson, R. Male and female infertility problems in the immotile-cilia syndrome. Eur J Respir Dis Suppl 127, 144-147 (1983). 185 Halbert, S. A., Patton, D. L., Zarutskie, P. W. & Soules, M. R. Function and structure of cilia in the fallopian tube of an infertile woman with Kartagener's syndrome. Hum Reprod 12, 55-58 (1997). 186 Leigh, M. W. et al. Clinical and genetic aspects of primary ciliary dyskinesia/Kartagener syndrome. Genet Med 11, 473-487 (2009). 187 Kartagener, M. [Not Available]. Schweiz Med Wochenschr 77, 702-706 (1947). 188 Pedersen, H. & Mygind, N. Absence of axonemal arms in nasal mucosa cilia in Kartagener's syndrome. Nature 262, 494-495 (1976). 189 Fliegauf, M. et al. Mislocalization of DNAH5 and DNAH9 in respiratory cells from patients with primary ciliary dyskinesia. Am J Respir Crit Care Med 171, 1343-1349, doi:10.1164/rccm.200411-1583OC (2005). 190 Sturgess, J. M., Chao, J., Wong, J., Aspin, N. & Turner, J. A. Cilia with defective radial spokes: a cause of human respiratory disease. N Engl J Med 300, 53-56, doi:10.1056/NEJM197901113000201 (1979). 191 Flume, P. A. et al. Cystic fibrosis pulmonary guidelines: chronic medications for maintenance of lung health. Am J Respir Crit Care Med 176, 957-969, doi:10.1164/rccm.200705-664OC (2007). 192 Date, H. et al. Living-donor lobar lung transplantation for primary ciliary dyskinesia. Ann Thorac Surg 71, 2008-2009 (2001). 193 Soleas, J. P., Paz, A., Marcus, P., McGuigan, A. P., Waddell, T.K. Engineering Airway Epithelium. Journal of Biomedicine and Biotechnology 2012, doi:10.1155/2012/982971 (2012). 194 Crespin, S. et al. Approaches to study differentiation and repair of human airway epithelial cells. Methods in molecular biology 742, 173-185, doi:10.1007/978-1-61779- 120-8_10 (2011). 195 Koike, T. et al. Cultured epithelial grafting using human amniotic membrane: The potential for using human amniotic epithelial cells as a cultured oral epithelium sheet. Arch Oral Biol, doi:10.1016/j.archoralbio.2011.04.009 (2011). 196 Aizawa, Y., Leipzig, N., Zahir, T. & Shoichet, M. The effect of immobilized platelet derived growth factor AA on neural stem/progenitor cell differentiation on cell-adhesive hydrogels. Biomaterials 29, 4676-4683, doi:10.1016/j.biomaterials.2008.08.018 (2008). 197 Moore, K., MacSween, M. & Shoichet, M. Immobilized concentration gradients of neurotrophic factors guide neurite outgrowth of primary neurons in macroporous scaffolds. Tissue Eng 12, 267-278, doi:10.1089/ten.2006.12.267 (2006). 198 Turturro, M. V. & Papavasiliou, G. Generation of Mechanical and Biofunctional Gradients in PEG Diacrylate Hydrogels by Perfusion-Based Frontal Photopolymerization. J Biomater Sci Polym Ed, doi:10.1163/092050611X566450 (2011). 199 Leslie-Barbick, J. E., Moon, J. J. & West, J. L. Covalently-immobilized vascular endothelial growth factor promotes endothelial cell tubulogenesis in poly(ethylene glycol) diacrylate hydrogels. J Biomater Sci Polym Ed 20, 1763-1779, doi:10.1163/156856208X386381 (2009). 200 Subramani, K. & Birch, M. A. Fabrication of poly(ethylene glycol) hydrogel micropatterns with osteoinductive growth factors and evaluation of the effects on

132

osteoblast activity and function. Biomed Mater 1, 144-154, doi:10.1088/1748- 6041/1/3/009 (2006). 201 Re'em, T., Tsur-Gang, O. & Cohen, S. The effect of immobilized RGD peptide in macroporous alginate scaffolds on TGFbeta1-induced chondrogenesis of human mesenchymal stem cells. Biomaterials 31, 6746-6755, doi:10.1016/j.biomaterials.2010.05.025 (2010). 202 Yu, L. M., Miller, F. D. & Shoichet, M. S. The use of immobilized neurotrophins to support neuron survival and guide nerve fiber growth in compartmentalized chambers. Biomaterials 31, 6987-6999, doi:10.1016/j.biomaterials.2010.05.070 (2010). 203 Pompe, T., Salchert, K., Alberti, K., Zandstra, P. & Werner, C. Immobilization of growth factors on solid supports for the modulation of stem cell fate. Nat Protoc 5, 1042-1050, doi:10.1038/nprot.2010.70 (2010). 204 Ballios, B. G., Cooke, M. J., van der Kooy, D. & Shoichet, M. S. A hydrogel-based stem cell delivery system to treat retinal degenerative diseases. Biomaterials 31, 2555-2564, doi:10.1016/j.biomaterials.2009.12.004 (2010). 205 Kim, H., Tator, C. H. & Shoichet, M. S. Chitosan implants in the rat spinal cord: Biocompatibility and biodegradation. J Biomed Mater Res A 97, 395-404, doi:10.1002/jbm.a.33070 (2011). 206 Kinikoglu, B. et al. Reconstruction of a full-thickness collagen-based human oral mucosal equivalent. Biomaterials 30, 6418-6425, doi:10.1016/j.biomaterials.2009.08.010 (2009). 207 Kinikoglu, B., Rodriguez-Cabello, J. C., Damour, O. & Hasirci, V. The influence of elastin-like recombinant polymer on the self-renewing potential of a 3D tissue equivalent derived from human lamina propria fibroblasts and oral epithelial cells. Biomaterials 32, 5756-5764, doi:10.1016/j.biomaterials.2011.04.054 (2011). 208 Lin, Y. M., Boccaccini, A. R., Polak, J. M., Bishop, A. E. & Maquet, V. Biocompatibility of poly-DL-lactic acid (PDLLA) for lung tissue engineering. J Biomater Appl 21, 109- 118, doi:10.1177/0885328206057952 (2006). 209 Cortiella, J. et al. Tissue-engineered lung: an in vivo and in vitro comparison of polyglycolic acid and pluronic F-127 hydrogel/somatic lung progenitor cell constructs to support tissue growth. Tissue Eng 12, 1213-1225, doi:10.1089/ten.2006.12.1213 (2006). 210 Richardson, T. P., Peters, M. C., Ennett, A. B. & Mooney, D. J. Polymeric system for dual growth factor delivery. Nat Biotechnol 19, 1029-1034, doi:10.1038/nbt1101-1029 (2001). 211 Carpenedo, R. L. et al. Homogeneous and organized differentiation within embryoid bodies induced by microsphere-mediated delivery of small molecules. Biomaterials 30, 2507-2515, doi:10.1016/j.biomaterials.2009.01.007 (2009). 212 Bhang, S. H. et al. Locally delivered growth factor enhances the angiogenic efficacy of adipose-derived stromal cells transplanted to ischemic limbs. Stem Cells 27, 1976-1986, doi:10.1002/stem.115 (2009). 213 Yang, F. et al. Gene delivery to human adult and embryonic cell-derived stem cells using biodegradable nanoparticulate polymeric vectors. Gene Ther 16, 533-546, doi:10.1038/gt.2008.182 (2009). 214 Luo, Y., Kobler, J. B., Zeitels, S. M. & Langer, R. Effects of growth factors on extracellular matrix production by vocal fold fibroblasts in 3-dimensional culture. Tissue Eng 12, 3365-3374, doi:10.1089/ten.2006.12.3365 (2006).

133

215 Witze, E. S., Litman, E. S., Argast, G. M., Moon, R. T. & Ahn, N. G. Wnt5a control of cell polarity and directional movement by polarized redistribution of adhesion receptors. Science 320, 365-369, doi:10.1126/science.1151250 (2008). 216 Noria, S. et al. Assembly and reorientation of stress fibers drives morphological changes to endothelial cells exposed to shear stress. Am J Pathol 164, 1211-1223, doi:10.1016/S0002-9440(10)63209-9 (2004). 217 Sheikh, S., Rainger, G. E., Gale, Z., Rahman, M. & Nash, G. B. Exposure to fluid shear stress modulates the ability of endothelial cells to recruit neutrophils in response to tumor necrosis factor-alpha: a basis for local variations in vascular sensitivity to inflammation. Blood 102, 2828-2834, doi:10.1182/blood-2003-01-0080 (2003). 218 Tarran, R. et al. Normal and cystic fibrosis airway surface liquid homeostasis. The effects of phasic shear stress and viral infections. J Biol Chem 280, 35751-35759, doi:10.1074/jbc.M505832200 (2005). 219 Mih, J. D. et al. A multiwell platform for studying stiffness-dependent cell biology. PLoS One 6, e19929, doi:10.1371/journal.pone.0019929 (2011). 220 Bhana, B. et al. Influence of substrate stiffness on the phenotype of heart cells. Biotechnology and bioengineering 105, 1148-1160, doi:10.1002/bit.22647 (2010). 221 Miroshnikova, Y. A. et al. Engineering strategies to recapitulate epithelial morphogenesis within synthetic three-dimensional extracellular matrix with tunable mechanical properties. Phys Biol 8, 026013, doi:10.1088/1478-3975/8/2/026013 (2011). 222 Biggs, M. J., Richards, R. G., Gadegaard, N., Wilkinson, C. D. & Dalby, M. J. The effects of nanoscale pits on primary human osteoblast adhesion formation and cellular spreading. J Mater Sci Mater Med 18, 399-404, doi:10.1007/s10856-006-0705-6 (2007). 223 Kim, D. H. et al. Modulation of adhesion and growth of cardiac myocytes by surface nanotopography. Conference proceedings : ... Annual International Conference of the IEEE Engineering in Medicine and Biology Society. IEEE Engineering in Medicine and Biology Society. Conference 4, 4091-4094, doi:10.1109/IEMBS.2005.1615362 (2005). 224 Curtis, A. S. et al. Cells react to nanoscale order and symmetry in their surroundings. IEEE Trans Nanobioscience 3, 61-65 (2004). 225 Clark, P., Connolly, P., Curtis, A. S., Dow, J. A. & Wilkinson, C. D. Topographical control of cell behaviour: II. Multiple grooved substrata. Development 108, 635-644 (1990). 226 Jin, C. Y. et al. Nanoscale surface topography enhances cell adhesion and gene expression of madine darby canine kidney cells. J Mater Sci Mater Med 19, 2215-2222, doi:10.1007/s10856-007-3323-z (2008). 227 Teixeira, A. I., Abrams, G. A., Bertics, P. J., Murphy, C. J. & Nealey, P. F. Epithelial contact guidance on well-defined micro- and nanostructured substrates. J Cell Sci 116, 1881-1892, doi:10.1242/jcs.00383 (2003). 228 Yim, E. K., Darling, E. M., Kulangara, K., Guilak, F. & Leong, K. W. Nanotopography- induced changes in focal adhesions, cytoskeletal organization, and mechanical properties of human mesenchymal stem cells. Biomaterials 31, 1299-1306, doi:10.1016/j.biomaterials.2009.10.037 (2010). 229 Loesberg, W. A. et al. The threshold at which substrate nanogroove dimensions may influence fibroblast alignment and adhesion. Biomaterials 28, 3944-3951, doi:10.1016/j.biomaterials.2007.05.030 (2007).

134

230 Xia, Y. et al. Complex Optical Surfaces Formed by Replica Molding Against Elastomeric Masters. Science 273, 347-349 (1996). 231 Liu, Y. et al. Electrospun fiber template for replica molding of microtopographical neural growth guidance. Small 8, 676-681, doi:10.1002/smll.201101199 (2012). 232 Folch, A., Ayon, A., Hurtado, O., Schmidt, M. A. & Toner, M. Molding of deep polydimethylsiloxane microstructures for microfluidics and biological applications. J Biomech Eng 121, 28-34 (1999). 233 Paulsson, M. Basement membrane proteins: structure, assembly, and cellular interactions. Crit Rev Biochem Mol Biol 27, 93-127, doi:10.3109/10409239209082560 (1992). 234 You, Y., Richer, E. J., Huang, T. & Brody, S. L. Growth and differentiation of mouse tracheal epithelial cells: selection of a proliferative population. Am J Physiol Lung Cell Mol Physiol 283, L1315-1321, doi:10.1152/ajplung.00169.2002 (2002). 235 Coleman, D. L., Tuet, I. K. & Widdicombe, J. H. Electrical properties of dog tracheal epithelial cells grown in monolayer culture. Am J Physiol 246, C355-359 (1984). 236 Lee, K. Y. & Mooney, D. J. Hydrogels for tissue engineering. Chem Rev 101, 1869-1879 (2001). 237 Bigi, A., Cojazzi, G., Panzavolta, S., Rubini, K. & Roveri, N. Mechanical and thermal properties of gelatin films at different degrees of glutaraldehyde crosslinking. Biomaterials 22, 763-768 (2001). 238 Martinez-Diaz, G. J., Nelson, D., Crone, W. C. & Kao, W. Y. J. Mechanical and chemical analysis of gelatin-based hydrogel degradation. Macromol Chem Physic 204, 1898-1908, doi:DOI 10.1002/macp.200350042 (2003). 239 McGuigan, A. P. & Sefton, M. V. Modular tissue engineering: fabrication of a gelatin- based construct. J Tissue Eng Regen Med 1, 136-145, doi:10.1002/term.14 (2007). 240 Dalton, B. A. et al. Modulation of epithelial tissue and cell migration by microgrooves. J Biomed Mater Res 56, 195-207 (2001). 241 Prager-Khoutorsky, M. et al. Fibroblast polarization is a matrix-rigidity-dependent process controlled by focal adhesion mechanosensing. Nat Cell Biol 13, 1457-1465, doi:10.1038/ncb2370 (2011). 242 Yeung, T. et al. Effects of substrate stiffness on cell morphology, cytoskeletal structure, and adhesion. Cell Motil Cytoskeleton 60, 24-34, doi:10.1002/cm.20041 (2005). 243 Aubin, H. et al. Directed 3D cell alignment and elongation in microengineered hydrogels. Biomaterials 31, 6941-6951, doi:10.1016/j.biomaterials.2010.05.056 (2010). 244 McNamara, L. E. et al. Nanotopographical control of stem cell differentiation. J Tissue Eng 2010, 120623, doi:10.4061/2010/120623 (2010). 245 Bettinger, C. J., Langer, R. & Borenstein, J. T. Engineering substrate topography at the micro- and nanoscale to control cell function. Angew Chem Int Ed Engl 48, 5406-5415, doi:10.1002/anie.200805179 (2009). 246 Oakley, C. & Brunette, D. M. The sequence of alignment of microtubules, focal contacts and actin filaments in fibroblasts spreading on smooth and grooved titanium substrata. J Cell Sci 106 ( Pt 1), 343-354 (1993). 247 Burridge, K. & Chrzanowska-Wodnicka, M. Focal adhesions, contractility, and signaling. Annu Rev Cell Dev Biol 12, 463-518, doi:10.1146/annurev.cellbio.12.1.463 (1996). 248 Stewart, C. E., Torr, E. E., Mohd Jamili, N. H., Bosquillon, C. & Sayers, I. Evaluation of differentiated human bronchial epithelial cell culture systems for asthma research. J Allergy (Cairo) 2012, 943982, doi:10.1155/2012/943982 (2012).

135

249 Rauchman, M. I., Nigam, S. K., Delpire, E. & Gullans, S. R. An osmotically tolerant inner medullary collecting duct cell line from an SV40 transgenic mouse. Am J Physiol 265, F416-424 (1993). 250 Wang, P. Y., Yu, J., Lin, J. H. & Tsai, W. B. Modulation of alignment, elongation and contraction of cardiomyocytes through a combination of nanotopography and rigidity of substrates. Acta Biomater 7, 3285-3293, doi:10.1016/j.actbio.2011.05.021 (2011). 251 Wang, C. C., Jamal, L. & Janes, K. A. Normal morphogenesis of epithelial tissues and progression of epithelial tumors. Wiley Interdiscip Rev Syst Biol Med 4, 51-78, doi:10.1002/wsbm.159 (2012). 252 Teixeira, A. I. et al. The effect of environmental factors on the response of human corneal epithelial cells to nanoscale substrate topography. Biomaterials 27, 3945-3954, doi:10.1016/j.biomaterials.2006.01.044 (2006). 253 Khoufache, K. et al. Primary in vitro culture of porcine tracheal epithelial cells in an air- liquid interface as a model to study airway epithelium and Aspergillus fumigatus interactions. Med Mycol 48, 1049-1055, doi:10.3109/13693786.2010.496119 (2010). 254 Widdicombe, J. H., Sachs, L. A., Morrow, J. L. & Finkbeiner, W. E. Expansion of cultures of human tracheal epithelium with maintenance of differentiated structure and function. BioTechniques 39, 249-255 (2005). 255 Yang, M. T., Fu, J., Wang, Y. K., Desai, R. A. & Chen, C. S. Assaying stem cell mechanobiology on microfabricated elastomeric substrates with geometrically modulated rigidity. Nat Protoc 6, 187-213, doi:10.1038/nprot.2010.189 (2011).

136