Mehmet Nizamoglu

The Extracellular Matrix to Bind Them All: Interactions Between ECM and Cells in Lung Fibrosis and Influences Thereof Mehmet Nizamoglu

The studies presented in this thesis were performed within an industrial collaborative PhD program between University Medical Center Groningen (UMCG) and Boehringer Ingelheim. The printing of this thesis was financially supported by the University of Groningen, Graduate School of Medical Sciences, the Dutch Society for Matrix Biology (NVMB) and Boehringer Ingelheim. ISBN (printed): 978-94-6483-345-4 Cover design: Albano Tosato Layout: Wouter Aalberts | Print: Ridderprint | © Mehmet Nizamoglu, 2023 All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted in any form or by any means without the permission of the author, or when applicable, from the editorial publishers.

The Extracellular Matrix to Bind Them All: Interactions Between ECM and Cells in Lung Fibrosis and Influences Thereof PhD thesis to obtain the degree of PhD at the University of Groningen on the authority of the Rector Magnificus Prof. J.M.A. Scherpen and in accordance with the decision by the College of Deans. This thesis will be defended in public on Wednesday 27 September 2023 at 14.30 hours by Mehmet Nizamoğlu born on 29 June 1992 in Kadıköy, Istanbul, Turkey

Supervisors Prof. J.K. Burgess Prof. H.I. Heijink Prof. B.N. Melgert Assessment Committee Prof. R.A. Bank Prof. G. Westergren-Thorsson Prof. M. Kolb

Paranymphs Mugdha M. Joglekar Roderick H. J. de Hilster

“The head may err, but never the blood.” Nakajima Atsushi, in Light, Wind and Dreams

TABLE OF CONTENTS Chapter 1: General Introduction 9 Chapter 2: The multi-faceted extracellular matrix: unlocking its secrets for understanding the perpetuation of lung fibrosis 17 Chapter 3: Abnormal collagen structure resulting from lack of contribution of collagen type XIV in lungs of patients with idiopathic pulmonary fibrosis 57 Chapter 4: Innovative 3D models for understanding mechanisms underlying lung diseases: powerful tools for translational research 71 Chapter 5: Dysregulated cross talk between alveolar epithelial cells and stromal cells in IPF reduces epithelial regenerative capacity 111 Chapter 6: 3D lung models – 3D extracellular matrix models 127 Chapter 7: Current possibilities and future opportunities with threedimensional lung ECM-derived hydrogels 165 Chapter 8: An in vitro model of fibrosis using crosslinked native extracellular matrix-derived hydrogels to modulate biomechanics without changing composition 183 Chapter 9: Fibroblast remodeling of extracellular matrix is directed by the fibrotic nature of the three-dimensional microenvironment 217 Chapter 10: General discussion and future perspectives 289 Appendices: English summary 306 Nederlandse samenvatting 309 Acknowledgements 313 Curriculum vitae 323 List of publications 324

CHAPTER 1 General Introduction

10 Chapter 1 Lung fibrosis encompasses several different pulmonary diseases, all of which result in fibrotic scar tissue accumulation in the lung interstitium, the region between the alveolar epithelium and blood vessels [1]. The most common type is idiopathic pulmonary fibrosis (IPF) which has an unknown etiology and is mainly characterized by progressive fibrosis and a poor prognosis [2]. The survival rate for patients with IPF is very low: a recent meta-analysis reported an overall median survival of 3.2 years [3]. Despite the fact that IPF is the most common type of lung fibrosis, it is still classified as a rare disease with a prevalence of 3-45 in 100,000 [4]. The current school of thought suggests that the IPF originates from repeating (micro)injuries in the alveolar epithelium of lung tissue, followed by an aberrant wound repair response involving the recruitment of the (myo)fibroblasts to the injured area [5]. Unfortunately, to date, there is no permanent cure for IPF except for lung transplantation [6]. The only two available therapeutics, Nintedanib and Pirfenidone, merely slow the progression of fibrosis, increase the median survival rate but cannot reverse or cure established disease [7, 8]. One of the main reasons for a lack of new treatment strategies is the lack of thorough knowledge of the mechanisms underlying disease progression, as the fibrotic process is already, usually, at a very late stage at the time of diagnosis. Another roadblock is the lack of adequate laboratory and animal models to investigate the complex disease mechanisms of IPF [9]. The formation of aberrant extracellular matrix (ECM), deposited as scar tissue, is a key disease mechanism not only for IPF but for all fibrotic lung diseases [10]. Forming the immediate environment around the cells, the ECM is composed of a plethora of proteins, proteoglycans and glycosaminoglycans [11]. ECM provides a mechanical scaffold for the resident cells to attach; moreover, it provides bioactive cues through its own composition, as well as the growth factors that are stored in it [12]. Lung ECM is of vital importance for lung function through providing structural support and elasticity [13]. In fibrotic lung diseases, including IPF, the native ECM is disrupted with respect to its composition, mechanics and organization [14]. These collective changes during fibrosis result in a stiff and rigid scaffold instead of a soft and elastic network [15, 16]. In the last decade, our thoughts on how ECM plays a role in biological processes, has evolved drastically from an inert scaffold towards a bioactive and instructive network. Parallel to these revelations, our understanding on the realms of how ECM might be involved in the progression of a fibrotic response in lung diseases has also improved. We now know that the presence of fibrotic ECM alone, in vitro, guides cells towards a more pro-fibrotic state in a positive feedback loop, possibly resulting in generation of more fibrotic ECM [17-19]. Together with aberrant composition, disorganized fibers and altered mechanics, abnormal cells

11 General Introduction in fibrotic ECM mark the four neighboring regions with complex relationships with the realms of fibrotic ECM (Figure 1). “One Ring to rule them all, One Ring to find them, One Ring to bring them all, and in the darkness bind them.” – J.R.R. Tolkien. However, the details of the overruling capacity of ECM among the interactions between the ECM and the resident cells are still uncharted. By investigating and characterizing the interactions between the ECM and cells, which are already imprinted by the fibrotic ECM in vivo, we can broaden our horizons on how the pro-fibrotic cycle continues towards further activation of cells or disruption of composition, organization and mechanics of ECM. Figure 1: The landscape and influence of the Realms of Extracellular Matrix (ECM). Aberrant composition, disorganized fibers, altered biomechanics and abnormal cells are important areas that regulate the interactions between the (fibrotic) ECM and resident cells are shaped during the course of the fibrotic response. 1

12 Chapter 1 AIM AND OUTLINE OF THIS THESIS The overarching aim of this thesis was to investigate and characterize the interactions between the fibrotic ECM and cells, to elucidate how these contribute to subsequent pro-fibrotic reactions. In Chapter 2, I outlined the changes in the composition, mechanics and organization of ECM during lung fibrosis and summarize the recent advances regarding how these changes might pave the way to a better understanding of fibrotic responses. In Chapter 3, I reported how collagen type XIV, an ECM protein, is involved in fibrotic lungs. While its role in the organization of fibrillar collagens was previously demonstrated, the status of collagen type XIV in IPF was not described prior. In this study, I investigated how the relative proportion of collagen type XIV protein is different in lungs of patients with IPF at the whole tissue level, as Ill as in specific tissue compartments. In Chapter 4, I critically summarized the cutting-edge technologies used to mimic lung microenvironments in three-dimensional (3D) in vitro conditions. After reviewing different in vitro tools, and their advantages and disadvantages, I discussed the challenges associated with such models. Moreover, I provided a detailed overview of the recent translational applications for each model and characterizations of such models. In Chapter 5, I investigated regenerative responses of cells isolated from the lungs of IPF patients and compared them to cells isolated from non-IPF lungs using an organoid model system. I hypothesized that supportive cells from the stromal niche/ microenvironment imprinted by a fibrotic ECM carry over their “lessons” and result in a dysregulated regenerative response of epithelial cells. I compared the number and size of organoids developed from unfractionated cells obtained from non-IPF lungs to IPF lungs. In addition, I isolated epithelial cells from these unfractionated cell populations to test their regenerative capacity without the influence of additional cells. In Chapter 6, I described the state-of-the-art status of 3D in vitro models used to mimic lung ECM. I explained the advantages of 3D over two-dimensional (2D) culture systems and exemplified the most commonly used materials to create a 3D in vitro lung microenvironment.

13 General Introduction In Chapter 7, I provided a perspective on recent developments in lung ECM-derived hydrogels. After a brief review of the innovative science utilizing such hydrogels, I deliberated on the path towards developing new technologies based on ECM-derived hydrogels. In Chapter 8, I harnessed the potential of lung ECM-derived hydrogels by artificially introducing fiber crosslinking. I hypothesized that applying additional fiber crosslinking would result in increased stiffness, and in turn would trigger cellular changes that are also observed during the fibrosis process. I investigated the changes in hydrogel mechanics and fiber organization, and examined cellular responses with respect to fibroblast activation. By creating an in vitro model that represents changes in the mechanical properties of fibrotic ECM alone, I succeeded in separating the mechanical influence of fibrotic ECM from the biomechanical influence on cells. In Chapter 9, I examined ECM and cell interactions using both IPF and non-IPF human lung ECM-derived hydrogels and human primary lung fibroblasts. I hypothesized that the origin of microenvironment would overrule the origin of the fibroblasts, and that IPF matrix can drive fibrotic responses in fibroblasts from normal lungs. I tested this hypothesis by comparing the influence of the microenvironment with the influence of the cell origin by combining IPF or non-IPF ECM-derived hydrogels with IPF or nonIPF fibroblasts, ultimately resulting in a combinatorial comparison. I characterized changes in collagen amount and collagen fiber organization, glycosaminoglycan content as well as mechanical properties of hydrogels with and without fibroblasts to compare the instructiveness of non-IPF and IPF microenvironments. In Chapter 10, the outcomes of the individual chapters of this thesis are integrated and discussed. The future perspectives of our findings with respect to interactions between ECM and cells in IPF are also included in this discussion. In the context of cellular functions, mechanical forces and fibrillar organization, this section concludes with how ECM brings them all, binds them all, and ultimately, rules them all. 1

14 Chapter 1 REFERENCES 1. Wijsenbeek M, Suzuki A, Maher TM. Interstitial lung diseases. Lancet 2022: 400(10354): 769-786. 2. Khor YH, Ng Y, Barnes H, Goh NSL, McDonald CF, Holland AE. Prognosis of idiopathic pulmonary fibrosis without anti-fibrotic therapy: a systematic review. Eur Respir Rev 2020: 29(157): 190158. 3. Ryerson CJ, Kolb M. The increasing mortality of idiopathic pulmonary fibrosis: fact or fallacy? Eur Respir J 2018: 51(1): 1702420. 4. Podolanczuk AJ, Thomson CC, Remy-Jardin M, Richeldi L, Martinez FJ, Kolb M, Raghu G. Idiopathic Pulmonary Fibrosis: State of the Art for 2023. Eur Respir J 2023: 2200957. 5. Martinez FJ, Collard HR, Pardo A, Raghu G, Richeldi L, Selman M, Swigris JJ, Taniguchi H, Wells AU. Idiopathic pulmonary fibrosis. Nat Rev Dis Primers 2017: 3(1): 17074. 6. Lederer DJ, Martinez FJ. Idiopathic Pulmonary Fibrosis. N Engl J Med 2018: 378(19): 18111823. 7. Guenther A, Krauss E, Tello S, Wagner J, Paul B, Kuhn S, Maurer O, Heinemann S, Costabel U, Barbero MAN, Muller V, Bonniaud P, Vancheri C, Wells A, Vasakova M, Pesci A, Sofia M, Klepetko W, Seeger W, Drakopanagiotakis F, Crestani B. The European IPF registry (eurIPFreg): baseline characteristics and survival of patients with idiopathic pulmonary fibrosis. Respir Res 2018: 19(1): 141. 8. Behr J, Prasse A, Wirtz H, Koschel D, Pittrow D, Held M, Klotsche J, Andreas S, Claussen M, Grohe C, Wilkens H, Hagmeyer L, Skowasch D, Meyer JF, Kirschner J, Glaser S, Kahn N, Welte T, Neurohr C, Schwaiblmair M, Bahmer T, Oqueka T, Frankenberger M, Kreuter M. Survival and course of lung function in the presence or absence of antifibrotic treatment in patients with idiopathic pulmonary fibrosis: long-term results of the INSIGHTS-IPF registry. Eur Respir J 2020: 56(2): 1902279. 9. Moss BJ, Ryter SW, Rosas IO. Pathogenic Mechanisms Underlying Idiopathic Pulmonary Fibrosis. Annu Rev Pathol 2022: 17(1): 515-546. 10. Upagupta C, Shimbori C, Alsilmi R, Kolb M. Matrix abnormalities in pulmonary fibrosis. Eur Respir Rev 2018: 27(148): 180033. 11. Burgstaller G, Oehrle B, Gerckens M, White ES, Schiller HB, Eickelberg O. The instructive extracellular matrix of the lung: basic composition and alterations in chronic lung disease. Eur Respir J 2017: 50(1). 12. Theocharis AD, Skandalis SS, Gialeli C, Karamanos NK. Extracellular matrix structure. Adv Drug Deliv Rev 2016: 97: 4-27. 13. Burgess JK, Harmsen MC. Chronic lung diseases: entangled in extracellular matrix. Eur Respir Rev 2022: 31(163): 210202. 14. Herrera J, Henke CA, Bitterman PB. Extracellular matrix as a driver of progressive fibrosis. J Clin Invest 2018: 128(1): 45-53.

15 General Introduction 15. Booth AJ, Hadley R, Cornett AM, Dreffs AA, Matthes SA, Tsui JL, Weiss K, Horowitz JC, Fiore VF, Barker TH, Moore BB, Martinez FJ, Niklason LE, White ES. Acellular normal and fibrotic human lung matrices as a culture system for in vitro investigation. Am J Respir Crit Care Med 2012: 186(9): 866-876. 16. de Hilster RHJ, Sharma PK, Jonker MR, White ES, Gercama EA, Roobeek M, Timens W, Harmsen MC, Hylkema MN, Burgess JK. Human lung extracellular matrix hydrogels resemble the stiffness and viscoelasticity of native lung tissue. Am J Physiol Lung Cell Mol Physiol 2020: 318(4): L698-L704. 17. Parker MW, Rossi D, Peterson M, Smith K, Sikstrom K, White ES, Connett JE, Henke CA, Larsson O, Bitterman PB. Fibrotic extracellular matrix activates a profibrotic positive feedback loop. J Clin Invest 2014: 124(4): 1622-1635. 18. Philp CJ, Siebeke I, Clements D, Miller S, Habgood A, John AE, Navaratnam V, Hubbard RB, Jenkins G, Johnson SR. Extracellular Matrix Cross-Linking Enhances Fibroblast Growth and Protects against Matrix Proteolysis in Lung Fibrosis. Am J Respir Cell Mol Biol 2018: 58(5): 594-603. 19. Liu F, Lagares D, Choi KM, Stopfer L, Marinkovic A, Vrbanac V, Probst CK, Hiemer SE, Sisson TH, Horowitz JC, Rosas IO, Fredenburgh LE, Feghali-Bostwick C, Varelas X, Tager AM, Tschumperlin DJ. Mechanosignaling through YAP and TAZ drives fibroblast activation and fibrosis. Am J Physiol Lung Cell Mol Physiol 2015: 308(4): L344-357. 1

CHAPTER 2 The multi-faceted extracellular matrix: unlocking its secrets for understanding the perpetuation of lung fibrosis Mehmet Nizamoglu & Janette K. Burgess Reproduced with permission from Springer Nature from the publication in Current Tissue Microenvironment Reports: Nizamoglu, M., & Burgess, J. K. (2021). The multi-faceted extracellular matrix: Unlocking its secrets for understanding the perpetuation of lung fibrosis. Current Tissue Microenvironment Reports, 2(4), 53–71.

18 Chapter 2 ABSTRACT Purpose of review: Lung fibrosis is currently thought to stem from an aberrant wound healing response after recurring (micro) injuries in the lung epithelium, together with disrupted crosstalk between epithelial and stromal cells. An important factor in lung fibrosis is the abnormal deposition of extracellular matrix (ECM). In this review, we extend the view of ECM to summarize how aberrant structural organization and degradation of ECM contributes to (perpetuation of) lung fibrosis. Recent findings: Fibrotic changes in ECM including altered composition, such as increased collagens, coupled with mechanical properties, such as increased stiffness or abnormal fiber crosslinking, promote pro-fibrotic responses in cells in this microenvironment. Similarly, changes in matrix degrading enzymes and release of degradation products from ECM proteins also perpetuate cellular fibrotic responses. Summary: In lung fibrosis, irreversible ECM structure, organization and architectural alterations drive a perpetuating fibrotic response. Targeting strategies abrogating the abnormal ECM or ECM-degrading enzymes accompanied by prognostic and/or diagnostic approaches based on ECM fragments may provide novel alternatives to current therapeutic approaches for lung fibrosis.

19 The multi-faceted extracellular matrix: unlocking its secrets for understanding the perpetuation of lung fibrosis INTRODUCTION Lung fibrosis is a common characteristic of the heterogeneous group of interstitial lung diseases (ILDs). Of these the most common is idiopathic pulmonary fibrosis (IPF), which is a chronic, progressive lung disease, with a very poor survival rate (median: 3-5 years) [1]. Currently, there is no cure for IPF, other than lung transplantation, and while there are two therapeutic agents, pirfenidone and nintedanib, that can slow the disease progression, these therapies are not effective in all patients and have adverse, sometimes severe side effects [2]. Lung fibrosis is currently thought to result from an aberrant wound healing response following recurrent microinjuries to the alveolar epithelium, augmented by aberrant cross-talk between the fibroblasts and epithelial cells resulting in an excessive and abnormal deposition of extracellular matrix (ECM) proteins [3-5]. Under normal physiological conditions, ECM is composed of a multitude of different proteins, glycosaminoglycans (GAGs), and glycoproteins (collagen types I, III, IV and VI, fibronectin, laminin, periostin, and hyaluronic acid are a few examples), forming a dynamic network that provides support to the cells embedded within it [6, 7]. In addition to its structural support function, ECM is a bioactive component of the tissue and it provides cues to all cells to influence/instruct their behavior. In fibrosis, deposition of several different ECM proteins such as collagens and fibronectin is increased, while others are decreased, changing the biochemical composition of the tissue [8]. As a natural consequence of the changes in the protein composition and organization, the biomechanical properties of fibrotic lung tissues are also altered: fibrotic lungs are stiffer, have a greater degree of collagen crosslinking and altered topography [9-11]. This catalogue of changes was previously thought to only be the result of the fibrotic process within the tissue; however, a plethora of recent studies have illustrated the changes in the ECM are an emerging contributor to the disease progression process itself, influencing different cell types and cellular mechanisms [12-17]. Moreover, with the advances in the single-cell RNA sequencing methods, lung resident cell populations are shown to have great heterogeneity in lung fibrosis, compared to healthy lungs, which would also impact the diversity of ECM changes in fibrosis [18-21]. Interestingly, it has recently been suggested (in the context of embryonic development) that each cell type expresses its own unique ECM gene profile (indicative of the production of an individual ECM protein profile) that becomes more refined as the cells differentiate towards end-stage cells such as fibroblasts [22]. This finding implies the importance of the ECM microenvironment, which is disrupted in fibrosis, for the maintenance of a homeostatic status in tissue. However, the detailed mechanisms regarding how altered properties of ECM affect cellular responses or contribute to the cellular heterogeneity present in fibrosis and the consequent influence upon the disease outcome are yet to be investigated completely. 2

20 Chapter 2 In this review, we summarize the varied aspects of the contribution of ECM in lung fibrosis and how ECM influences cellular responses. First, we focus on updates for understanding how changes in ECM composition, coupled with altered mechanical properties, impact cellular responses. Then, we look beyond the ECM scaffold to illustrate how ECM degradation and the released bioactive ECM fragments play a role in lung fibrosis. Finally, we reflect on how targeting (changes in) ECM can be leveraged to provide new avenues for managing lung fibrosis. ECM CHANGES IN FIBROSIS AND THEIR FUNCTIONAL CONSEQUENCES Composition and Crosslinking In pulmonary fibrosis, changes in the quantities of ECM proteins have been extensively described [6, 9, 23] : including, but not limited to, increased collagen types I and III, fibronectin, periostin, and hyaluronic acid. One of the most important pieces of evidence illustrating how fibrotic ECM induces fibrotic responses in fibroblasts, as a result of the feedback in two dimensional (2D) cell culture models, was described by Parker et al [14]. In concert, primary lung fibroblasts cultured on scaffolds made with stacked sections of decellularized IPF lung were shown to produce a protein output that mirrored the fibrotic matrix composition compared to the fibroblasts cultured on scaffolds made with control lung tissue [24]. By comparing the decellularized fibrotic and alveolar tissue-derived sections of mouse ex vivo lung tissue scaffolds, the fibrotic microenvironment was found to decrease the spontaneous movement speed of immortalized mouse fibroblasts, compared to healthy mouse tissue [13]. The effect of the microenvironment was shown to also influence responses in other cells: Monocyte-derived macrophages in the fibroblastic-foci were found to perpetuate the fibrotic response, suggesting the fibrotic microenvironmental cues were guiding these cellular responses [25]. Similarly, pericytes were also shown to have higher gene and protein expression of α smooth muscle actin (α-SMA) when cultured on decellularized IPF lung samples compared to decellularized control lung samples [26]. Interestingly, culturing alveolar epithelial cells on IPF lung derived decellularized matrices was found to protect alveolar epithelial cells from transforming growth factor β (TGF-β) induced apoptosis, while additionally strengthening the profibrotic response of IPF lung-derived decellularized matrix-seeded fibroblasts to TGF-β via engagement of integrin α2β1, compared with cells seeded on non-disease control lung derived decellularized matrices [27]. These studies collectively show the influence of the fibrotic ECM on different cells, illustrating the different responses of the cells to the changing microenvironment in lung fibrosis.

21 The multi-faceted extracellular matrix: unlocking its secrets for understanding the perpetuation of lung fibrosis Along with the changes in the biochemical distribution of the ECM proteins in fibrosis, post-translational modifications of these proteins are also altered. Collagen protein synthesis starts within the rough endoplasmic reticulum, with post-translational modifications adding hydroxyl groups to proline and lysine residues (Figure 1) [28]. Individual collagen molecules come together within the Golgi to form the triple helical structure, forming the procollagen molecule. This trimer then is secreted from the Golgi into the extracellular space, where its pro-collagen ends at both the C- and N-terminals are cleaved to generate the mature collagen molecule. The collagen molecules self-assemble to begin forming fibrils before lysyl oxidases (LOX), LOX-like enzymes (LOXLs) and transglutaminases (TGs) actively crosslink the triple helices to each other, forming the collagen fibers [28]. Increased expression and amount of LOXL1 and LOXL2 was reported in IPF lung tissue compared with non-disease control lung tissue [11]. In concert, fibrotic fibroblasts were found to have higher expression of TG2 compared with healthy fibroblasts in vitro [29]. Crosslinking of the collagen fibers by LOX/LOXL has also been shown to promote the TGF-β induced stiffening of the microenvironment [11]. ECM deposited by from IPF-lung derived fibroblasts increased the expression of LOXL3 and TG2; and in turn, the increased crosslinking of this ECM was demonstrated to boost fibroblast proliferation and adhesion [29]. These data together suggest that the increased collagen crosslinking, and dysregulation of the crosslinking enzyme amounts in pulmonary fibrosis could contribute to the positive feedback loop which Parker et al. first described [14]. Figure 1: Schematic illustration of synthesis, secretion and crosslinking of collagen fibrils. RER: Rough endoplasmic reticulum; LOX: lysyl oxidase; LOXL: LOX-like; TG: transglutaminase. 2

22 Chapter 2 Stiffness and Viscoelasticity Changes in the biomechanics of fibrotic lung tissue result directly from the altered and abnormal distribution and modification of the ECM proteins in lung fibrosis. Lung ECM has a viscoelastic nature that can dissipate the stress applied to it via various sources, such as mechanical forces changing with breathing in and out [10]. Among many biomechanical parameters, stiffness of the tissue is strongly associated with lung fibrosis and has been well-documented: native IPF lung samples were shown to have higher stiffness than control lung samples (1.96 ± 0.13 kPA vs 16.52 ± 2.25 kPA) and this difference remained similar also in decellularized lung samples (IPF lung sample: 7.34 ± 0.6 kPa, control lung sample: 1.6 ± 0.08 kPA) [9, 10]. Stiffness, like many other mechanotransducers, induces the Hippo pathway through yes-associated protein (YAP) – PDZ-binding motif (TAZ) signaling, resulting in the perpetuation of fibrosis [30]. Several in vitro models have been developed to assess the effect of stiffness on lung cells: fibroblasts cultured on 2D hydrogels with higher stiffness were shown to migrate faster, along with a greater cell spread area, compared to the fibroblasts cultured on hydrogels with lower (more physiological-like) stiffness [31]. Similarly, a stiffer 2D culture environment was shown to increase fibroblast activation via chromatin remodeling compared to softer surfaces, accompanied by increased nuclear volume in these fibroblasts [32]. Higher stiffness of fibronectin coated polyacrylamide hydrogels was shown to decrease fibroblast activating protein expression while increasing the cell spreading area and αSMA expression in murine lung fibroblasts compared with softer hydrogels; on the other hand, changes in the stiffness of collagen type I coated polyacrylamide hydrogels did not change the cellular response [33]. Likewise, comparing the effect of different stiffness values of polyacrylamide hydrogels and incorporation of solubilized matrix from healthy or IPF lungs on pericytes seeded on these hydrogels showed that increased cell area and higher expression of αSMA resulted from the increase in the stiffness of the hydrogel rather than the ECM composition [26]. Interestingly, a study by Matera et al. suggested opposing effects of stiffness on lung fibroblasts in 2D and 3D cultures: higher stiffness of 2D cultures promoted myofibroblast differentiation while stiffer 3D cultures limited the differentiation of the lung fibroblasts [12]. Lastly, blocking the YAP – TAZ pathway of mechanotransduction in IPF lung-derived fibroblasts resulted in decreased expression of ECM proteins while ECM-degradation enzyme gene expression levels increased compared to the untreated IPF fibroblasts [34]. All of these studies together indicate different effects of stiffness on cells. It is highly possible that the combination of altered composition and increased stiffness induces different cellular responses in different cells. More investigation on separating the contribution of altered composition and stiffness of the fibrotic microenvironment

23 The multi-faceted extracellular matrix: unlocking its secrets for understanding the perpetuation of lung fibrosis could improve our understanding and help identify novel therapeutic approaches for targeting the progression of lung fibrosis. Another emerging parameter among the biomechanical properties of ECM is the viscoelasticity, which is the ability to dissipate an applied stress through time [35]. The importance of the viscoelasticity of ECM both in healthy and diseased conditions has been recently reviewed elsewhere [36]. Similar to many other tissues and organs, lung ECM has viscoelastic properties and the loss of viscoelastic relaxation in fibrotic tissues has recently been established by our group [10]. The implications of (the loss of) viscoelasticity on cellular function in vivo is yet to be clarified; however, it is known that changes in the viscoelasticity of the microenvironment can affect cell migration, proliferation and ECM deposition by the cells [36]. All of these cellular functions are recognized as being altered in lung fibrosis, so now the challenge lies in separating the individual contributions of these different mechanical stimuli to the perpetuation of the fibrotic response. Promisingly, a recent study revealed the possibility of modifying stiffness and viscoelasticity independently of each other [37]. Developing advanced in vitro culture systems will further our understanding of viscoelasticity and its contribution to the progression of the lung fibrosis. Topography Topography of the ECM influences many cellular responses including migration and proliferation, as recently reviewed by Ouellette et al. [38]. While the altered composition alone could influence the topography of the ECM in lung fibrosis, the increased crosslinking and abnormal alignment of the fibers in the ECM are two other important factors changing the topography. In lung fibrosis, the topography of the ECM is drastically altered (Figure 2), due to the increased mature and organized collagen content, compared to the healthy lungs [11]. 2

24 Chapter 2 Figure 2: Representative scanning electron microscopy (SEM) images of decellularized lung parenchyma from non-disease control donors (upper row) and IPF patients (lower row). One of the mechanisms by which abnormal topography plays a role in lung fibrosis is the altered microstructures in the protein organization. In a study by Seo et al., comparison of adipose-derived stromal cells seeded on collagen type I networks with thin fibers and low pore size to cells seeded on networks with thick fibers and bigger pore size revealed that changing the microstructure increased differentiation of these stromal cells to myofibroblasts ~1.5X [39]. Along with fiber thickness and pore size, fiber alignment is an important parameter in ECM topography. Increased migration speed of primary lung fibroblasts seeded on collagen type I -methacrylated gelatin hydrogels was observed in highly aligned network samples, compared to hydrogels with less aligned networks [40]. In another study, increasing fiber density independent of the stiffness resulted in higher surface area of seeded dermal fibroblasts in 3D in vitro culture [41]. While collagen type I hydrogels with different stiffness values were used to test the effect of microstructure in the study by Seo et al., it is difficult to conclude the stiffness-independent contribution of the microstructure. As the abovementioned changes (stiffness, viscoelasticity and topography) in the fibrotic ECM occur simultaneously during fibrosis, more studies using advanced biomaterials are required to examine the individual contributions of such properties to the perpetuation of the fibrotic response.

25 The multi-faceted extracellular matrix: unlocking its secrets for understanding the perpetuation of lung fibrosis Storage of growth factors in the ECM As the non-cellular part of the tissue microenvironment, the ECM serves as a storage depot for many different growth factors and other soluble proteins, many of which are important for regulating the fibrotic response. Within the ECM, it is predominantly the glycosaminoglycans (GAGs) that serve as a reservoir for growth factors in the extracellular space. The negatively charged residues within the GAGs provide multiple binding sites for the positively charged amino acids within many growth factors. Through these binding interactions the growth factors are bound to the ECM, protecting them from degradation and conserving them until they are required for local signaling [42, 43]. The changing protein content and organization in fibrotic lung diseases is likely to alter the presence, amount, and the availability of the factors stored within ECM. Among these ECM proteins, fibronectin can bind many growth factors and soluble proteins, including latent TGF-β-binding protein-1 (LTBP1) [44]. Increased amounts of fibronectin and LTBP-1 in fibrosis could lead to a greater storage capacity of ECM for TGF-β. Activation of the stored TGF-β can be directed via mechanical stimuli due to prestress on the ECM, applied by cells or decreased viscoelastic relaxation of the ECM itself [45]. TGF-β activation is also regulated by mechanisms driven by other proteins including fibulin-1, an ECM glycoprotein, which is also found in greater amounts in the lung tissues of IPF patients [46, 47]. The binding of growth factors important for lung development and repair, including TGF-β, fibroblast growth factor (FGF) 1 and FGF2 and hepatocyte growth factor (HGF), and their interaction with their relevant receptors are dependent on the sulfation state of GAGs such as heparan sulfate, dermatan sulfate and chondroitin sulfate [42, 48]. Using hydrogels established using decellularized lung ECM (which is devoid of most GAGs) combined with heparan sulfate, dermatan sulfate or chondroitin sulfate with either TGF-β, FGF2 or HGF, Uhl and colleagues recently illustrated that matrix associated growth factor-dependent and -independent GAG effects in parallel with GAG-dependent and -independent matrix-associated growth factor effects are important for regulating cellular responses in lung in vitro models [48]. There is an increase in heparan sulfate, chondroitin sulfate, dermatan sulfate and hyaluronan in IPF lungs compared to controls [49], suggesting a greater capacity for anchoring important growth factors for regulating reparative or fibrotic processes in these tissues. Analyses of the sulfation state of the GAGs in the IPF tissues found that the highly sulfated GAGs were located predominantly in the regions of interaction between the fibrotic and less fibrotic tissues, potentially indicating a central role for the GAGs in providing growth factors for promoting the high fibrotic activity within these regions. 2

26 Chapter 2 Table 1: Summary of recent studies illustrating changes related to ECM in fibrosis and their impacts on cell responses. Category Changes in ECM Cellular Response Reference ECM Crosslinking Pan-inhibition of LO activity Lower rates of TGF-β induced collagen remodeling and ECM stiffness in human primary lung fibroblasts seeded on decellularized cell-derived ECM [11] Increased ECM crosslinking by TG2 Higher rates of proliferation in human primary lung fibroblasts seeded on decellularized IPF cell-derived ECM [29] Knock out of LOXL3 by siRNA Lower rates of fibroblast-tomyofibroblast differentiation in normal human lung fibroblasts [50] Deficiency of LOXL1 Lower expressions of TGF-β, collagen type I and αSMA in TGF-β overexpression mice model of fibrosis at day 35 compared to wild type mice of the same model [51] Stiffness Increased stiffness in 2D polyacrylamide hydrogels Faster migration and higher spread area normal human lung fibroblasts [31] Increased stiffness in 2D collagen type I-coated culture higher rate of activation and higher nuclear volume in murine primary lung fibroblasts [32] Increased stiffness of fibronectin coated polyacrylamide hydrogels Higher αSMA expression in murine lung fibroblasts [33] Increased stiffness of polyacrylamide hydrogels functionalized with solubilized matrix from lungs Higher cell area and higher αSMA expression in human primary microvascular pericytes [26] Increased stiffness in 2D dextranbased hydrogels functionalized with MMP-cut sites Higher myofibroblast differentiation in normal human lung fibroblasts [12] Increased stiffness in 3D dextranbased hydrogels functionalized with MMP-cut sites Lower myofibroblast differentiation in normal human lung fibroblasts [12] Topography Thicker collagen fibers in 3D collagen type I hydrogels Higher myofibroblast differentiation in human adipose derived stromal cells [39] Increased alignment of collagen type I -methacrylated gelatin hydrogel networks Increased migration speed in human primary lung fibroblasts in 2D [40] Increasing fiber density in 3D dextran-based hydrogels functionalized with MMP-cut sites Higher myofibroblast differentiation, cell spread area, YAP translocation and proliferation rate in normal human lung fibroblasts [12] LO: Lysyl oxidase, ECM: Extracellular matrix, TG: Transglutaminase, LOXL: Lysyl oxidase-like, TGF-β: Transforming growth factor β, αSMA: α Smooth muscle actin, MMP: Matrix metalloproteinase, YAP: Yes-associated protein

27 The multi-faceted extracellular matrix: unlocking its secrets for understanding the perpetuation of lung fibrosis It is not unlikely that the ECM in fibrotic lung disease would have enhanced storage capacity for bioactive factors as a result of the increased amount of the abovementioned ECM proteins, among others, and the activation and release of these factors from the ECM would be boosted by the biomechanical changes in the tissues in lung fibrosis. Further investigations regarding the ECM-stored bioactive factors are necessary for improving our understanding of the contribution of the repository function of the ECM to the progression of lung fibrosis. Collective impact of the altered ECM scaffold in fibrotic lung disease All in all, the altered ECM in fibrosis generates diverse influences which impact cellular phenotypes; as summarized in Table 1. While the biochemical changes in the fibrotic microenvironment have been demonstrated with proteomics analyses via mass spectrometry [6], the accompanying biomechanical changes, such as increase in stiffness or loss of viscoelastic relaxation [10], require further investigation. While these changes could be simply the result of the altered biochemical composition, there are other emerging contributing factors such as collagen crosslinking that require further clarification. With the new developments in the field of biomaterials, advanced in vitro culture systems will be generated to mimic the specific biomechanical properties of the fibrotic microenvironment. Such systems will further improve our understanding of how the biomechanical properties of the ECM, either individually or collectively, contribute to the perpetuation of fibrotic disease in the lung. Eventually, such knowledge should illuminate how such properties could be targeted via therapeutic intervention for treatment of lung fibrosis. BEYOND THE ECM SCAFFOLD The role of ECM degrading enzymes and their regulators The ECM is a dynamic microenvironment that is constantly being remodeled as elements are degraded and newly deposited during normal tissue maintenance and particularly under conditions of disease pathogenesis. While all cell types synthesize, secrete and orchestrate deposition of ECM (including epithelial cells, mesenchymal cells, endothelial cells and immune cells), the fibroblasts are recognized as the major ECM producing cell type in fibrotic tissues. Leukocytes and macrophages, but also mesenchymal and epithelial cells produce enzymes that regulate the degradation of the ECM. The most well recognized group are the matrix metalloproteinases (MMPS), but also serine or cysteine proteases have a role in maintaining a healthy homeostasis within the ECM [52, 53]. Activity of the enzymes that degrade the ECM is tightly balanced by endogenous inhibitors (tissue inhibitors of MMPs (TIMPs)), serpins or cystatins [54-56]. 2

28 Chapter 2 In fibrotic lung disease, an easy assumption would be that there would be reduced levels of matrix degradative enzymes, in particular MMPs, as this is where the greatest amount of research has focused, as an explanation as to why there is increased ECM deposition. However, multiple studies report increased levels of several MMPs associated with fibrotic lung disease, reviewed in [53, 57, 58]. While this appears paradoxical, it is important to realize that in addition to degrading ECM proteins, the range of substrates MMPs can process and activate includes cell receptors, chemokines and growth factors [59]. Through the generation of chemotactic gradients or activation of specific proinflammatory or profibrotic factors, in cooperation with disruption of basement membranes and other physical barriers within the tissue, MMPs can influence the influx of inflammatory cells into the site of tissue injury, which in turn then contribute to the development / perpetuation of fibrosis [60, 61]. A number of MMPs have been specifically linked to fibrosis in the lungs, summarized in Table 2. A selection of MMPs associated with fibrotic lung disease, particularly IPF, are highlighted herein. MMP1 MMP1, considered a “classic” collagenase, cleaves interstitial collagens including collagen type I and III. MMP1 protein levels are increased in bronchial alveolar lavage and gene and protein expression levels are increased lung tissue from IPF, compared to non-fibrotic, patients [63, 69, 79, 96]. A single nucleotide polymorphism, within the AP-1 binding domain of the MMP1 promoter (which increases transcription of MMP1) is observed more frequently in patients with IPF who smoke, than those who do not [97]. In a murine system, MMP1 enhanced cellular migration, increased wound closure rate, and protected cells from apoptosis. Increased MMP1 in alveolar epithelial cells repressed mitochondrial respiration, reduced the production of reactive oxygen species (both total and mitochondrial) and also, under normoxic conditions, increased expression of hypoxia-inducible factor-1α (HIF-1α) [98]. The fact that MMP1 was upregulated via increased HIF-1α induction under hypoxic conditions in the alveolar epithelial cells suggests a role for MMP1 in bidirectional cross talk regulating alveolar epithelial cell functions in fibrosis. Intriguingly, MMP1 has recently been reported to be part of a set of signature genes illustrating the link between IPF and lung cancer. MMP1 was suggested to be a promising candidate gene driving significant expression changes through the transition from healthy tissue to IPF and non-small cell carcinoma [99]. MMP1 is primarily located in reactive bronchial epithelial cells, hyperplastic type 2 pneumocytes in honeycomb cysts and in alveolar macrophages, with little to no expression being observed in interstitial mesenchymal cells, suggesting that the localization of the increased MMP1 in fibrotic lung tissue does not facilitate the degradation of the fibrotic deposits [69].

29 The multi-faceted extracellular matrix: unlocking its secrets for understanding the perpetuation of lung fibrosis Table 2: Different types of MMPs, their origins, involvement in lung fibrosis and possible mechanisms of action. Substrates for cleavage sourced primarily from [62]. MMP Source in vivo Localization in lung tissue Profile in lung fibrosis Substrates for Cleavage Possible mechanism of action References MMP1 Collagenase 1 Bronchial and AECs Macrophages Reactive bronchial epithelial cells, hyperplastic type 2 pneumocytes in honeycomb cysts and in alveolar macrophages ↑ in plasma, serum & BAL from IPF, ↑ gene and protein in IPF lung tissue ECM substrates: Collagen types I, II, III, VII and X; gelatins; aggrecan; link protein; entactin; tenascin; perlecan Non-ECM substrates α2-M; α1-PI; α1antichymotrypsin; IGFBP-2, 3, 5; proIL-1β; CTGF Regulates AEC migration, wound closure and resistance to apoptosis. Possible bidirectional cross talk (via HIF1α) regulating AEC functions in fibrosis [63-67] MMP2 Gelatinase A Bronchial and AECs, Fibroblasts and fibrocytes Subepithelial myofibroblasts foci, close to basement membrane disruption. In ECM surrounding fibroblast foci. AECs and basal bronchiolar epithelial cells ↑ in plasma and BAL from IPF ↑ protein in IPF lung tissue ECM substrates Gelatins; collagen types IV, V, VII, X and XI; Ln; Fn; elastin; aggrecan; link protein Non-ECM substrates ProTGF-β; FGF receptor I; MCP-3; IGFBP-5; proIL-1β; galectin-3; plasminogen Postulated disrupted integrity of sub-epithelial/ endothelial basement membrane resulting in infiltration of factors and interstitial cells to the alveolar space promoting fibro-proliferative response [67-71] MMP3 Stromelysin 1 Bronchial and AECs, Alveolar Macrophages, and fibroblasts Regions of bronchiolization close to aberrant ECM deposits Weaker expression in lymphoid aggregates ↑ in plasma and BAL from IPF ↑ gene and protein in IPF lung tissue ECM substrates Aggrecan; decorin; gelatins; Fn; Ln; collagen types III, IV, IX and X; tenascin; link protein; perlecan Non-ECM substrates lamin, casein, IGFBP-3; proIL-1β; HB-EGF; CTGF; E-cadherin; α1antichymotrypsin; α1-PI; α2-M; plasminogen; uPA; proMMP-1, 7, 8, 9, 13 Regulating activation of TGF-β through release from latency associated peptide and latent TGF-β binding protein 1 Induction of CTGF [67, 72-77] 2

30 Chapter 2 Table 2: Different types of MMPs, their origins, involvement in lung fibrosis and possible mechanisms of action. Substrates for cleavage sourced primarily from [62]. (continued) MMP Source in vivo Localization in lung tissue Profile in lung fibrosis Substrates for Cleavage Possible mechanism of action References MMP7 Matrilysin Bronchial epithelial cells, aberrantly activated AECs, mononuclear phagocytes and fibrocytes Aberrant activated AECs and in bronchiolar epithelial cells ↑ in plasma and serum from IPF ↑ gene and protein in IPF lung tissue ECM substrates Aggrecan; gelatins; Fn; Ln; elastin; entactin; collagen type IV; tenascin; decorin; link protein, Non-ECM substrates osteopontin, β4 integrin, E-cadherin, syndecan, FasL, plasminogen, Proα-defensin; proTNFα; CTGF; HB-EGF; RANKL; IGFBP-3; plasminogen Regulating neutrophil transepithelial influx, via the shedding of syndecani-CXCL1 complexes, Protecting the fibroblast from undergoing apoptosis vis removal of FasL from their surfaces [63, 65-68, 72, 78-82] MMP8 Collagenase 2 Bronchial epithelial cells, neutrophils, macrophages Bronchial epithelial cells in regions of moderately severe and severe fibrosis, macrophages Type II AECs are positive for MMP-8 in control lung tissue but there is minimal or no staining in these cells in regions of moderately severe and severe fibrosis in IPF lungs ↑ in plasma and BAL from IPF ↑ gene and protein in lung homogenates and tissue from IPF ECM substrates Collagen types I, II and III; gelatins; aggrecan; link protein Non-ECM substrates α1-PI Promotes fibrocyte migration [58, 66-68, 72, 83, 84] MMP9 Gelatinase B AECs, neutrophils, macrophages, fibrocytes and fibroblasts in fibroblastic foci Metaplastic AECs, alveolar and interstitial macrophages and fibroblasts in fibrotic foci ↑ in plasma from IPF ↑ gene and protein in lung tissue from IPF ECM substrates Gelatins; collagen types III, IV and V; aggrecan; elastin; entactin; link protein, vitronectin; N-telopeptide of collagen type I Non-ECM substrates ProTGF-β; IL-2 receptor α; Kit-L; IGFBP-3; proIL-1β; ICAM-1; α1-PI; galectin-3; plasminogen Fibroblasts in healthy lung tissue express Thy-1 (a glycophosphatidylinositollinked glycoprotein) which suppresses MMP9 expression. IPF fibroblasts lack Thy-1 leading to expression of MMP9, which correlates with regions of active fibrogenesis, suggesting a role for MMP9 in induction of fibrosis. Activation of TGF-β1 in the extracellular space [67-69, 72, 79, 85]

31 The multi-faceted extracellular matrix: unlocking its secrets for understanding the perpetuation of lung fibrosis Table 2: Different types of MMPs, their origins, involvement in lung fibrosis and possible mechanisms of action. Substrates for cleavage sourced primarily from [62]. (continued) MMP Source in vivo Localization in lung tissue Profile in lung fibrosis Substrates for Cleavage Possible mechanism of action References MMP10 Stromelysin 2 Bronchial and AECs and alveolar macrophages AECs, macrophages, and peripheral bronchiolar epithelial cells ↑ in serum and BAL from IPF ↑ protein in lung tissue from IPF ECM substrates Aggrecan; Fn; Ln; collagen types III, IV and V; link protein, gelatin, Non-ECM substrates lamin, casein, Pro-1, 8, 10 Possible role in regulation of macrophage migration and polarization driving fibrotic response [58, 86-88] MMP12 Macrophage elastase macrophages, and lung stromal cells Not available ↑ in plasma and BAL from IPF ↑ fragment from collagen type IV released after MMP12 degradation in IPF ECM substrates Elastin; aggrecan; Fn; collagen type IV; osteonectin; Ln; nidogen Non-ECM substrates Plasminogen; apolipoprotein(a) Postulated disrupted integrity of sub-epithelial/ endothelial basement membrane resulting in infiltration of factors and interstitial cells to the alveolar space promoting fibro-proliferative response [67, 89, 90] MMP13 Collagenase 3 Bronchial and AECs, alveolar macrophages and fibroblasts Bronchial and AECs and alveolar macrophages ↑ in plasma from IPF ↑ protein in lung homogenates from IPF ECM substrates Collagen types I, II, III, IV, IX, X and XIV; aggrecan; Fn; tenascin; osteonectin; Ln; Perlecan Non-ECM substrates CTGF; ProTGF-β; MCP-3; α1antichymotrypsin Not known as yet [67, 91] MMP14 MT1-MMP AECs, alveolar macrophages and endothelial cells. AECs and alveolar macrophages ↑ gene and protein in lung tissue from IPF ECM substrates Collagen types I, II and III; gelatins; aggrecan; Fn; Ln; fibrin; Ln-5 Non-ECM substrates ProMMP-2; proMMP-13; CD44; MCP-3; tissue, transglutaminase Facilitates fibroblast migration through disruption of ECM barriers [69, 70, 92] 2

32 Chapter 2 Table 2: Different types of MMPs, their origins, involvement in lung fibrosis and possible mechanisms of action. Substrates for cleavage sourced primarily from [62]. (continued) MMP Source in vivo Localization in lung tissue Profile in lung fibrosis Substrates for Cleavage Possible mechanism of action References MMP15 MT2-MMP AECs and endothelial cells AECs ↑ gene and protein in lung tissue from IPF ECM substrates Fn; tenascin; nidogen; aggrecan; perlecan; Ln ProMMP-2; tissue transglutaminase Not known as yet [70] MMP16 MT3-MMP AECs and fibroblasts AECs and fibroblasts in fibroblastic foci ↑ gene and protein in lung tissue from IPF ECM substrates Collagen type III; Fn; gelatin Non-ECM substrates ProMMP-2; tissue transglutaminase Not known as yet [70] MMP19 Monocytes, macrophages, fibroblasts, and endothelial cells Hyperplastic AECs overlying fibrotic areas ↑ gene in lung tissue from IPF ECM substrates Collagen type IV; gelatin; Fn; tenascin; aggrecan; COMP; Ln; nidogen Non-ECM substrates IGFBP-3 Stimulates epithelial cell wound healing and migration Promotes fibroblast migration, proliferation and ECM component synthesis [93] MMP28 Epilysin bronchial and AECs Bronchial and AECs ↑ in serum from IPF ↑ gene and protein in lung tissue from IPF ECM substrates Unknown Non-ECM substrates Casein Promoting M2 macrophage programming [64, 94, 95] AEC – alveolar epithelial cells; TGF – transforming growth factor; α2-M, α2-macroglobulin; α1-PI, α1-proteinase inhibitor; COMP, cartilage oligomeric matrix protein; CTGF, connective tissue growth factor; Fas-L, Fas ligand; FGF, fibroblast growth factor; Fn, fibronectin; HB-EGF, heparin-binding epidermal growth factor like growth factor; IGFBP, insulin-like growth factor binding protein; ICAM-1, inter-cellular adhesion molecule 1; Kit-L, kit ligand; Ln, laminin; MCP-3, monocyte chemotactic protein-3; MMP, matrix metalloproteinases; MT-MMP, membrane-type MMP; PG, proteoglycan; proIL-1β, pro interleukin-1β ; Pro, proteinase type; proTNF-α, pro tumor necrosis factor-α; proTGF-β, pro transforming growth factor β; ProMMP, latent MMP; RASI-1, rheumatoid arthritis synovium inflamed-1; RANKL, receptor activator for nuclear factor κB ligand; uPA, urokinase plasminogen activator