Th2 T Cells Are Required for Biomaterial-Mediated Functional Muscle Regeneration

TH2 T CELLS ARE REQUIRED FOR BIOMATERIAL-MEDIATED FUNCTIONAL MUSCLE REGENERATION

by Kaitlyn Noelle Sadtler

A dissertation submitted to Johns Hopkins University in conformity with the requirements for the degree of Doctor of Philosophy

Baltimore, Maryland March, 2016

Abstract

Regenerative medicine therapies that primarily target stem cells have achieved limited success. An alternative strategy is to focus on immune cells, the first responders to traumatic wounds, which can interact directly with biomaterial scaffolds. Here, we investigate how biomaterial scaffolds shape the immune microenvironment in non- traumatic subcutaneous and traumatic muscle wounds (VML) and ultimately impact tissue regeneration. A diverse population of immune cells is recruited into scaffolds and the surrounding area, including macrophages, dendritic cells, T lymphocytes and B lymphocytes. The scaffolds induced a pro-regenerative type-2 response, which, in the

VML is characterized by an mTOR/Rictor-dependent Th2 pathway and IL-4-dependent macrophage polarization, critical for functional muscle regeneration. Targeting the adaptive components of the immune system during the process of biomaterials design may support the development of future therapies that efficiently control immune balance in tissues, ultimately stimulating tissue repair.

Thesis Readers

Jennifer Elisseeff, Ph.D.

Department of Biomedical Engineering, Johns Hopkins University

Drew Pardoll, M.D. Ph.D.

Sidney Kimmel Comprehensive Cancer Center, Johns Hopkins University School of

Medicine

Acknowledgements

The past 3.5 years have been some of the most exciting, stressful, rewarding years of my life. Throughout that time, I have had the chance to work with amazing individuals and meet friends I will have for the rest of my life.

Professionally, I would like to thank Ken Estrellas, a fellow graduate student in the Elisseeff lab that helped me immensely with the muscle studies. He was always willing to lend a hand with dissections, sample preparation, histology and manuscript editing. This included meeting me in lab at 5:00am to start processing samples that I had to have at the core facility by noon, even bringing me in breakfast one of the mornings after I slept at lab. Ken – I cannot thank you enough, you have some of the greatest motivation and the best heart I have seen in science, and wish you the best in your career and life. Additionally, I would like to thank both Vince Beachley and Matt Wolf, who were in the lab as postdocs. Vince, you were my mentor during my rotation even though you had no idea what I was doing with all the immunology, you were always there to lend a helping hand and tell me to relax a bit when I tried to get everything done in a short period of time. Matt, after Vince left, you got to deal with the hyper-productive grad student with a heavy dose of immunology. You helped out more than I could have hoped, your knowledge of the field and biomaterials was very helpful as I waded through the biomaterial and engineering literature as a molecular biologist, and have become a great friend who is always willing to take a red pen to anything I write and provide amazing feedback. For about a year, right when things started picking up, I worked with an undergraduate student, Brian Allen, who has since moved on to grad school. Brian, thanks for all of your help in sample prep and PCR. Couldn’t have finished these experiments as quickly without your helping hand. Best of luck in your career, you will do great things. Franck Housseau, from the Pardoll lab, thank you very much for all of your

iii guidance in immunology, and being open to the new project and field as it developed, your guidance and willingness to answer my questions is greatly appreciated.

Additionally, I would like to thank the members of my thesis committee (Dr.

Andrew Ewald, Dr. Jordan Green, Dr. Jonathan Schneck) that guided me through my project and helped me with paper planning and my plans after graduate school. Dr. Drew

Pardoll, my committee chair, I greatly appreciate the collaboration with your lab and all of the insight you had from your years of immunology experience, and your excitement and enthusiasm for the project. Finally, my PI, Dr. Jennifer Elisseeff. Jennifer, I couldn’t have asked for a better advisor, you challenged me and gave me the resources to grow as a scientist. The opportunities I had in your lab were amazing, from the research at Hopkins and collaborations stretching multiple differences, to conferences taking me to exciting places and meeting some of the biggest names in the field. I cannot thank you enough.

Personally, I have met many amazing people at Hopkins. A non-exhaustive list of friends I’d like to thank for being there throughout the years includes Rebecca Tweedell,

Donna Dang, Meredith Stone, Jackie Pham, Melissa Bowman, Emily Bergbower, Iris

Chen, Alex Mims, Nina Hosmane, Karen Cravero, Katie and Kyle Bruner, Ashley Cook,

Kathleen Cunningham, and Heather Jacobs. I’d especially like to thank Rebecca

Tweedell who was my roommate and classmate for the first two years of grad school.

We worked together through first year classes, and sitting together while studying for our qualifying exams for weeks in silent solidarity. You are one of the best friends I have made during grad school and always there when experiments failed or things got too stressful, and on the other side you were always there for celebratory pie or froyo when we got things right.

From the minute I was born, my family has always supported me in my endeavors. My brother, Sean, thanks for being there for Sam and I when we were growing up, and being an amazing friend and brother. Sam, my sister and partner in

iv adventure. Life is always interesting when you are around, you are always there to take a phone call, and up for an adventure at any time. I cannot wait for our future explorations around the world, you are amazing, the first doctor in our family and the best sister I could’ve asked for. Mom and Dad, I couldn’t have done any of this without you. Thank you for supporting me through the years and always promoting me to do my best, even when I didn’t think I could achieve it. Thank you for making me stubborn and driven, and willing to help others along the way. You always told me I could achieve whatever I set my mind to and, well, here is the result of that. This is for you two.

Table of Contents

Abstract ...... ii Acknowledgements ...... iii List of Abbreviations ...... viii List of Tables ...... x List of Figures ...... xi Chapter 1: Introduction ...... 1 Introduction ...... 2 Tissue-Based Class Control...... 2 Immune Privileged Sites and Clonal Ignorance ...... 4 Development and The Immune System ...... 5 Shared Signaling Molecules: Immune Response and Developmental Biology ...... 5 Functional Healing: Tissue Engineering’s Approach to Development...... 7 Polymer Scaffolds and the Surrounding Tissue Environment ...... 9 Polyethylene glycol (PEG) Hydrogels ...... 10 Foreign Body Response (FBR) Microenvironment...... 11 Synthetics Coated with Extracellular Matrix (ECM) Scaffolds...... 11 Decellularized Extracellular Matrix: A Model and Scaffold ...... 12 Small Intestinal Submucosa (SIS) ...... 12 Urinary Bladder Matrix (UBM) ...... 13 Acellular Adipose Tissue (AAT)...... 13 As an Immunomodulatory Reagent and Tissue Model...... 15 Biochemical and Physical Alterations of Scaffolds...... 16 Physical Alteration of ECM & Synthetics ...... 17 Cyclodextrin and RGD modification of Polymeric Scaffolds ...... 18 Growth Factor Modification and Modulation of Microenvironment ...... 18 Summary ...... 19 Chapter 2: Methods...... 24 Tissue ECM Preparation ...... 25 Cell Culture ...... 26 RT-PCR ...... 26 Subcutaneous ECM Implantation ...... 28 Volumetric Muscle Loss Surgery ...... 28 Histology ...... 29 Flow Cytometry ...... 30 Treadmill Testing for Muscle Function ...... 31 Chapter 3: Differential effects of tissue-ECM source on cell function ...... 33 Tissue matrix arrays for high-throughput screening and systems analysis of cell function ...... 35 Abstract ...... 35 Introduction ...... 36 Methods ...... 37 Results ...... 57 Discussion ...... 66

The Scaffold Immune Microenvironment: Biomaterial-Mediated Immune Polarization in Traumatic and Non-Traumatic Applications ...... 99 Abstract ...... 99 Introduction ...... 100 Methods ...... 102 Results ...... 107 Discussion ...... 111 Composition and immune environment of urinary bladder matrix scaffolds ...... 124 Abstract ...... 124 Introduction ...... 125 Materials and Methods ...... 126 Results ...... 131 Discussion ...... 135 Chapter 4: The Role of the Adaptive Immune System In Formation of A Pro- Regenerative Scaffold Immune Microenvironment ...... 143 Developing a Pro-Regenerative Biomaterial Scaffold Microenvironment Requires T helper 2 cells ...... 145 Abstract ...... 145 Methods ...... 145 Main Text ...... 154 Chapter 6: Conclusions and Future Directions ...... 186 Conclusions and Future Directions ...... 187 Curriculum Vitae ...... 197 Permission Letters to Reprint or Use Copyrighted Material ...... 201

vii

List of Abbreviations

AAT acellular adipose tissue ARG1 arginase-1 CCL chemokine (C-C motif) ligand CCR chemokine (C-C motif) receptor CD cluster of differentiation DAMP damage-associated molecular pattern ECM extracellular matrix FAPs fibro/adipogenic progenitors FBGCs foreign body giant cells FBR foreign body response FN fibronectin HA hyaluronic acid hDAM human adipose tissue-derived extracellular matrix IFNg interferon gamma IL interleukin iNOS inducible nitric oxide synthase OA osteoarthritis PCL poly(capro-lactone) PEG poly(ethylene glycol) Retnla resistin-like alpha RGD arginine-glycine-aspartic acid ROS reactive oxygen species SAM scaffold-associated macrophage SIM scaffold immune microenvironment SIS small intestinal submucosa TBI traumatic brain injury TGFb transforming growth factor beta TLR toll-like receptor

viii

TNFa tumor necrosis factor alpha UBM urinary bladder matrix VEGF vascular endothelial growth factor VML volumetric muscle loss

List of Tables

Table 1.1. Immune Polarizations and Their Effector Functions ...... 21

Table 1.2. Role of Immune Populations in Inflammation and Tissue Regeneration ...... 22

Table 1.3 Common Materials used in Tissue Engineering ...... 23

Supplementary Table 2.1: One-way ANOVA results for cancer cells and macrophages on 2D tissue arrays...... 95

Supplemental Table 2.2: Results of correlation analysis between ECM protein quantity and in vitro assay outcomes...... 97

Supplemental Table 2.3: Osteogenesis gene ontology enrichment analysis results for tissue ECMs...... 98

Supplementary Table 5.1. Wilcoxon Rank Sum Test Results on sorted CD3+ T cells . 184

Supplementary Table 5.2. Wilcoxon Rank Sum and Linear Regression Test Results on sorted F4/80+ macrophages in WT and Rag1-/- mice...... 185

List of Figures

Figure 2.1: 2D and 3D tissue and organ model ECM arrays...... 70

Figure 2.2: Tissue ECM protein composition and analysis...... 71

Figure 2.3: Stem cell–tissue ECM interactions in 2D and 3D arrays...... 72

Figure 2.4: Cancer cell and macrophage interactions with tissue microarrays...... 74

Figure 2.5: Systems biology analysis of tissue proteomic composition and in vitro function...... 76

Supplementary Figure 2.1: Characterization of tissue ECM spots...... 78

Supplementary Figure 2.2: Tissue array cell-seeding method optimization...... 80

Supplementary Figure 2.3: Optimization of cell-tissue ECM for formation of breast cancer–tissue ECM 3D spheroids...... 82

Supplementary Figure 2.4: Array spots generated from diseased tissues...... 84

Supplementary Figure 2.5: Fidelity of tissue processing, array spotting and proteomic analysis...... 85

Supplementary Figure 2.6: Characterization and protein composition of tissue arrays prepared with multiple processing methods...... 87

Supplementary Figure 2.7: Comparison of macrophage morphology on arrays prepared with different processing methods...... 88

Supplementary Figure 2.8: Cellular differentiation and viability in 3D spheroids...... 90

Supplementary Figure 2.9: Attachment of bone marrow–derived macrophages to different tissue substrates on array...... 92

Supplementary Figure 2.10: In vitro and in vivo macrophage polarization in response to tissue ECM...... 93

Figure 3.1: Subcutaneous injection of particulate ECM scaffolds induces encapsulation cellular infiltration...... 117

Figure 3.2: Immune response to scaffolds is dominated by CD11b+ Macrophages communicating with CD3+ T cells...... 118

Figure 3.3: Immune infiltrate profile depends on tissue source and time post-injection 119

Figure 3.4: Myeloid subtypes defined by F4/80, CD11c, CD206 and CD86 expression.

...... 120

Figure 3.5: T cell profile is dominated by CD4+ T cells ...... 121

Figure 3.6: Detailed profile of myeloid cells in a scaffold-treated volumetric muscle wound ...... 122

Figure 3.7 Myeloid profile is dependent upon scaffold and presence of adaptive immune cells ...... 123

Figure 4.1: UBM materials characterization...... 137

Figure 4.2: Particulate UBM induces alterations in macrophage phenotype in vitro. ... 138

Figure 4.3: The scaffold immune microenvironment of UBM...... 139

Figure 4.4: UBM promotes an M2-macrophage phenotype that matures over time..... 140

Figure 4.5: UBM-treated muscle wounds recruit a diverse immune cell repertoire...... 141

Figure 4.6: UBM induces a systemic IL-4 upregulation correlated with local antigen- presenting M2-macrophages...... 142

Figure 5.1: Biomaterial scaffolds induce a Th2 response in volumetric muscle wounds

...... 161

Figure 5.2: M(IL-4) pro-regenerative myeloid polarization induced by scaffolds is Th2- dependent...... 163

Figure 5.3: Systemic immune homeostasis is modified by application of biomaterial scaffolds...... 165

Figure 5.4: Th2/M(IL-4) responses to biomaterial-treated muscle wound promote functional tissue regeneration ...... 166

Supplementary Figure 5.1: Materials characterization and selection...... 168

xii

Supplementary Figure 5.2: Cell recruitment to muscle injury...... 170

+ Supplementary Figure 5.3: FoxP3 Treg populations at 1 and 3 weeks post-operation. 171

Supplementary Figure 5.4: Data spread of gene expression profiling of CD3+ cells sorted from 1 week post surgery muscle defects...... 172

Supplementary Figure 5.5: M2/M(IL4) Gene expression in scaffold-treated muscle wounds...... 173

Supplementary Figure 5.6: Myeloid polarization in WT, Rag1-/- and Cd4-/- mice...... 174

Supplementary Figure 5.7: Adoptive Transfer of CD4+ T cells into Rag1-/- mice...... 176

Supplementary Figure 5.8: Data spread of gene expression profiling of cells sorted from

1 week post surgery muscle defects...... 177

Supplementary Figure 5.9: Gene ontology analysis of adaptive immune dependent gene expression changes in SIM F4/80+ macrophages associated with wound healing and tissue regeneration...... 178

Supplementary Figure 5.10: Gene expression in draining lymph nodes at 1 and 3 weeks post-operation...... 179

Supplementary Figure 5.11: Computed Tomography imaging reveals irregular muscle density in Rag1-/- mice...... 180

Supplementary Figure 5.12: Quadriceps muscle at 3 weeks post-operation in WT and

Rag1-/- mice...... 181

Supplementary Figure 5.13: Collagen and adipose-related gene expression increases in

Rag1-/- mice...... 182

Supplementary Figure 5.14: T cell participation in muscle regeneration and fibro/adipogenic lineage commitment...... 183

xiii

Chapter 1: Introduction

This introduction has been reprinted with permission from Springer International

Publishing License #3823780882688

Sadtler, Kaitlyn, et al. "Integrating Tissue Microenvironment with Scaffold Design to

Promote Immune-Mediated Regeneration." Biomaterials in Regenerative Medicine and the Immune System. Springer International Publishing, 2015. 35-51.

Introduction

As the field of biomedical engineering further develops the ability to engineer diverse organs, the need to understand the role of the tissue microenvironment in host immune responses to the engineered tissues or scaffolds has surfaced. It has been appreciated that there are varying responses to pathogens dependent upon the location in which that pathogen is encountered. This principle also applies to biomaterials such as synthetics and tissue-derived scaffolds made from extracellular matrix (ECM) when used in regenerative medicine. We will begin our discussion on immune responses to these materials by first considering the ability of the host immune system to cater its response appropriately to the tissue microenvironment, and how that environment in turn alters the response.

Tissue-Based Class Control

Depending upon the location in which the host receives a danger signal, the immune system can alter the magnitude and polarization of that response. This signal can arise from specific sources, such as invading pathogens, or from damage associated molecular patterns (DAMPs) created upon tissue damage (1). After receiving this signal, cells present in the vicinity will integrate the signals from their location as well

2 as from the type of danger that is present and respond appropriately. This is the basis of a theory known as tissue based class control (2).

Three commonly defined immune polarizations are TH1, TH2, and TH17, referring to the T Cells involved in the class of response. Broadly, TH1 and TH2 immune polarizations are also known as type-1 and type-2 responses. There is a fourth type, Treg that will also be discussed in this chapter. Major cytokines and cells that are associated with these phenotypes are shown in Table 1. These type-1 and type-2 immune polarizations also involve differential activation of myeloid cells such as macrophages, neutrophils and eosinophils. It is important to note that there is a large continuum in immune polarization, where individual cells rarely fit directly into these categorizations, and intermediate phenotypes are also observed (3). One example is TH3 responses induced in situations of oral tolerance (4). Oral tolerance is when an antigen is given orally and in the gut induces a response that has several hallmarks of both the TH2 response, Interleukin-4 (IL-4) and Interleukin-5 (IL-5) production, as well as the Treg response, Interleukin-10 (IL-10) and transforming growth factor beta (TGF-β), inducing a tolerance to that antigen. Engineers and immunologists exploited this phenomenon of oral tolerance to treat individuals that suffer from allergic reactions, such as peanut allergies (5). In this example, patients are orally exposed to low doses of antigen that slowly tolerize their immune reaction, reducing or preventing the strong eosinophilic and

IgE induced allergen hypersensitivity response.

There are other factors that help direct the antigen-experienced cells to the correct location for their response. Such factors include expression of adhesion proteins on the immune cell surface that are specific for proteins or carbohydrates present in the target location. Cells that have received an antigen signal from the intestine upregulate

x α4β7 integrin that binds to specific sialyl-Lewis acids present on intestinal epithelia (6).

When these immune phenotypes are presented in the wrong location, they could result

3 in severe inflammation and damage of delicate tissues, or in an incomplete clearance of an infection. In addition to cell adhesion molecules, chemokines (CCL or CXCL) and chemokine receptors (CCR or CXCR) are another set of ligands and receptors important for tissue-specific homing of lymphocytes (7). Chemokines, such as CCL25, are expressed by small intestinal epithelial cells, endothelial cells and other associated cells, which bind receptors on CCR9+ T Cells, lymphocytes that are primed for small intestinal homing. In another example, α4β1+CLA+ (cutaneous lymphocyte antigen, a glycoprotein)

CCR4+CCR10+ T Cells bind carbohydrates, known as selectins, CCL17 and CCL27, which are expressed by cutaneous cells. These ligands are differentially expressed in homeostatic and inflamed tissue, demonstrating how an alteration in the immune status of tissues changes the recruitment of different effector cells. Chemokines can also serve to recruit specific cell types to an area such as neutrophils and eosinophils (8).

For tissue engineers, the intricacies of the host immune response complicate the goal of immune acceptance and remodeling by introducing variables dependent upon the host tissue in which the engineered tissue is designed to integrate. If we can leverage the polarization tendencies of varying tissues, and the optimal immune response for regeneration, we will be able to create a more rational scaffold design that incorporates immunological instructions to stimulate or guide tissue repair.

Immune Privileged Sites and Clonal Ignorance

Integration of signals received from both the challenge encountered, as well as the location in which this challenge occurs, ultimately determines the phenotype of the immune response. In certain tissues, there is another method to avoid excessive immune-mediated tissue damage, a phenotype known as clonal ignorance. Here, antigens from organs that remain sequestered from the rest of the body, these “immune privileged” organs, such as brain, eye and testes, are ignored by the host immune

4 system (9). At immune privileged sites, these antigens are not “tolerized” but the host does not mount a response to them. Misregulation of this system can result in various autoimmune pathologies, such as multiple sclerosis.

If we are better able to understand the intricacies of immune responses to biomaterials as a function of their tissue microenvironment, we will be able to design enhanced scaffolds and better predict the host response to these regenerative grafts.

Through manipulation of the biomaterials-response system, we could also create an appropriate immune response to help guide proper regeneration and formation of a functional replacement tissue. A delicate balance between promoting immune recruitment while avoiding excessive inflammation will have to be reached to optimize the remodeling of these tissues. This immune modulation could be obtained through modification of scaffolds, changing the biomaterial as a function of the tissue to be regenerated, while driving the proper inflammatory response and immune environment.

Development and The Immune System

As tissue engineers drive the growth of new tissue development to replace a non-functional tissue or organ, we can use insights from naturally occurring regeneration and development to guide scaffold design. We will first look at immunology as it pertains to developmental biology, including several signaling molecules that have different roles in both systems, and then look at direct evidence of immune system involvement in natural and engineered tissue regeneration.

Shared Signaling Molecules: Immune Response and Developmental Biology

One of the most interesting evolutionary adaptations is the use of the “Toll” and

“TLR” signaling pathways for dorsal-ventral patterning in Drosophila melanogaster and innate immune receptor signaling in vertebrates. In Drosophila the Toll receptor binds a

5 protein called Spätzle, which activates a downstream cascade of proteins, resulting in degradation of Cactus, and nuclear translocation of Dorsal inducing a ventral cell fate.

This pathway is also important in antifungal response in adult Drosophila (9). TLR’s also induce the NFκB signaling pathway, which results in degradation of IκBα, and nuclear translocation of p65/RelA, activating the expression of inflammatory genes (9). In this situation, Toll is orthologous to TLR, and Dorsal is orthologous to p65/RelA. At the evolutionary level, the immune system and tissue development are tightly integrated, shown in this adaptation of a signaling pathway for both embryonic development, and immune response.

Looking closer at the dynamics within a single organism, we can see functions for various proteins both in development and the immune response. The TGF-β family of proteins has many roles in immunology, including the induction of regulatory responses along with IL-10, and has been associated with the alternative polarization of macrophages that do not produce an inflammatory phenotype (10). The TGF-β family of proteins is also important in differentiation between the endoderm, mesoderm, and ectoderm germ layers in various organisms (11). In amphibians, TGF-β-like signals in the organizer (a cellular structure in embryos) specify the dorsal mesoderm. In birds, signaling by proteins in this family is important in primitive streak formation, a structure in which there is epithelial-to-mesenchymal transition of mesodermal precursors. Another important cytokine, IL-4, promotes type 2 immunity and alternative activation of macrophages, and is also associated with allergy and hypersensitivity reactions. IL-4 has also been implicated as a driving force in myotube formation during muscle development (12). Developing myotubes secrete IL-4, recruiting endogenous progenitor cells and inducing fusion with existing myoblasts to promote mature myotube formation.

Interestingly, one of the largest inhibitory signals of muscle growth comes from myostatin, which is also in the TGF-β family of proteins. The overlap of signaling

6 molecules that are important in both embryonic development and immune response suggests that a varying immune microenvironment during development will alter the ultimate phenotype of that tissue. This association between development and the immune system also has implications for regenerative medicine. Just as researchers have attempted to harness elements of normal developmental biology pathways to enhance new tissue development and repair, the immune system may also be critical to consider when engineering strategies for regeneration.

Scientists studying natural limb reconstruction in amphibians have directly implicated the role of the immune system in regeneration (13). After severing the limb of an axolotl, the limb bud was injected with liposomes, containing clodronate, a cytotoxic compound that is released into the cytoplasm of phagocytic cells, such as macrophages, after uptake of liposomes. Upon local depletion of macrophages, limb bud formation and proper limb regeneration was inhibited. This study demonstrated that local macrophages were necessary for proper limb bud formation and limb regeneration in axolotls. Without immune interference, axolotls are capable of regenerating a normal limb about 2.5 months post-amputation.

Through the use of techniques implemented by developmental biologists and integration of their findings on immune-mediated tissue development, engineers could better mimic the local environment in natural tissue regeneration. Alterations of the immune microenvironment of engineered tissues could promote or inhibit differentiation tendencies of various progenitor cells dependent upon the signals they receive from resident immune cells.

Functional Healing: Tissue Engineering’s Approach to Development

One major goal of regenerative medicine is to create a functional replacement for missing or damaged tissue. Since the development of new tissue has many corollaries

7 with embryonic development, we can assume that the immune system and associated signaling molecules play a pivotal role in repair and regeneration. Several studies on the role of the immune system in tissue engineering, and a new field of

“immunoengineering” have appeared in recent years (14, 15). The goal of immunoengineering is to create an immune response through biological or pharmacological intervention that will yield a desired outcome, such as decreased inflammation or targeted destruction of cancer cells. Working in the field of macrophage biology, several groups have characterized various polarizations of macrophage response that can alter the outcome of regeneration; more specifically, studying the M1-

M2 axis of macrophage polarization (16, 17). Others have discussed the possibility of manipulating the macrophage response to biomaterials to induce a better outcome of host immune response (18). In their view, macrophages are the driving force behind both positive and negative aspects of the host response to biomaterials. Since macrophages are responsible for both inflammation and scarring, they could potentially be a good target for manipulation and improving integration of new tissues with the surrounding environment. A plethora of heterogeneity exists in these polarizations; varying scaffolds, both natural and synthetic, will elicit varying responses.

Macrophage biology has been implicated in several aspects of development and regeneration (19, 20). Eosinophils, another immune cell type associated with type-2 polarization, are important mediators of both muscle and liver regeneration. As mentioned previously, IL-4 signaling is required for myotube fusion and formation of a mature muscle fiber. Through various mouse models that eliminated IL-4 and IL-13 signaling, Heredia et. al. showed that IL-4 secreted by eosinophils was required for differentiation of satellite cells (muscle stem cells) through signals received from fibro/adipogenic progenitors (FAPs) resident in muscle tissue (21). Furthermore, IL-4 prevented adipogenic differentiation of these precursors and therefore prevented

8 accumulation of ectopic adipocytes in muscle. These type-2 signals also promote clearance of necrotic debris resulting from muscle trauma or degeneration. In this study, a cascade of signals was outlined starting with a type-2 innate immune response signaling to FAPs through IL-4 and IL-13, which then promoted satellite cell differentiation into regenerated muscle. Type-2 signals were generated by eosinophils in cardiotoxin-mediated muscle damage. IL-4 and IL-13 signaling was not required in satellite cells for induction of regeneration.

Eosinophils and type-2 immunity have also been associated with liver regeneration in partial hepatectomy and toxin-mediated liver damage models (22).

Through the use of an eosinophil knockout (ΔdblGATA) mouse model and an IL-4 signaling knockout (IL-4Rα-/-) model, Goh et. al. concluded that both eosinophil and IL-4 signaling were required for liver regeneration. In this case, selective inhibition of IL-4 signaling in hepatocytes decreases proliferation measured by Ki67 (a cellular proliferation marker) and BrdU incorporation.

Polymer Scaffolds and the Surrounding Tissue Environment

Even relatively inert polymeric scaffolds are capable of eliciting varying types of immune responses. Evaluation of these responses is important for the selection of biomaterials for use in drug delivery or as a scaffold, and screening methods are being developed to provide insight into proper material design (23). Outside of the typical foreign body response (FBR), host immune polarization is dependent upon the type of biomaterial and location of the synthetic scaffold implant. This includes a varying degree of immune cell activation, inflammation and subsequent scaffold degradation rates dependent upon the tissue environment (24). With one compound capable of eliciting different responses, it is important that we understand these responses and better cater our implants to the immune microenvironment that it will be subjected to.

Polyethylene glycol (PEG) Hydrogels

Synthetic polymers are used in many locations in the body due to the flexibility of their design and low immunogenicity. One polymer that is well established in tissue engineering is poly(ethylene glycol) (PEG). Studies conducted by Hillel et al. (2011) on a

PEG-based soft tissue replacement showed a variance in immune response dependent upon location of injection. Implants placed in the dorsal subcutaneous area of rats showed a decreased inflammatory response compared to those adjacent to adipose tissue in preliminary clinical trials (25). This observation lead to further investigation of the location-regulated immune response through implantation of PEG hydrogels in varying locations of Sprague-Dawley rats (26). These rats received implants subcutaneously, intraperitoneally and postlaterally in the thigh near a large adipose depot. After one week of incubation the implants were excised to evaluate the early immune response. Subcutaneous implants showed the least inflammatory cell recruitments, followed by intraperitoneal implants, with adipose inducing the largest response as reported previously. Cells of the immune system such as macrophages can produce reactive oxygen species (ROS) that can lead to hydrolysis of PEG hydrogels.

M1 macrophages express iNOS (inducible nitric oxide synthase, NOS2), which will catalyze the production of nitric oxide (a reactive oxygen species) whose primary function is to defend against bacteria. Neutrophils and other immune cells are also capable of producing ROS, leading to altered implant degradation kinetics. Not only will the immune response directly affect the degradation and lifetime of biodegradable synthetics, it will also produce different signals that cells residing in that scaffold will receive, depending on the biomaterial’s tissue location.

Foreign Body Response (FBR) Microenvironment

Within the context of a foreign body response (FBR) to materials such as PEG, there are both systemic and local effects of the immune response (27-29). Locally, the biomaterial or implant is surrounded by a dense fibrous capsule and often induces the fusion of macrophages to form foreign body giant cells (FBGCs), signaling poor graft integration. Proximal to the implant the FBR is dominated by type-2 cytokines that help induce collagen deposition and wall off the foreign body. Bacterial infection at the site of implantation can yield an inflammatory type-1 immune response, causing damaging release of free radicals that increase degradation of the biomaterial. Immune cells that interact with foreign bodies often have a mixed phenotype, showing signs of both collagen deposition and scar formation (Arginase-1, IL-4) characteristic of a type-2 response, as well as inflammatory mediators (Interferon-γ (IFN-γ), Tumor necrosis factor

α (TNF-α), iNOS) characteristic of a type-1 response.

Synthetics Coated with Extracellular Matrix (ECM) Scaffolds

To promote the acceptance of polymer-based scaffolds, researchers have coated synthetic mesh with a decellularized urinary bladder matrix (UBM) to delay the inflammatory reaction associated with a foreign body response to polymer scaffolds (30,

31). In these studies, polypropylene surgical mesh was coated with UBM or dermal ECM that had been digested with pepsin. An abdominal wall injury study showed that proximity of macrophages to the synthetic mesh was correlated with a variation in their polarization. Macrophages that were prevented from interacting with the synthetic material and instead remained on the ECM coating showed a more M2-like polarization, whereas those in direct contact with the mesh were more inflammatory, showing an M1- like phenotype. This coating thereby moderated the host-implant response and allowed for a decrease in inflammation at the implant location.

Decellularized Extracellular Matrix: A Model and Scaffold

One promising material for use in regenerative medicine is decellularized extracellular matrix, or ECM. ECM-based scaffolds are derived from native tissue that has been thoroughly cleansed of cellular material, leaving behind large structural proteins and associated growth factors and small molecules (32-34). Decellularization is achieved by treating the tissue with a variety of solvents meant to disrupt cell membranes and remove cells and nucleic acids. These solvents include, but are not limited to, acids, detergents, and enzyme solutions. Current ECM scaffolds that are commercially available and approved for human clinical use include Acellular Dermis

(CollamendTM), Urinary Bladder Matrix (Matristem®), and Small Intestinal Submucosa,

(OASIS® Wound Matrix). Concerns have been raised about the efficiency of decellularization and possible xenogeneic effects, but clinically tissue-derived ECM appears to be a promising scaffold with specific immunomodulatory properties that promote regeneration (35).

Small Intestinal Submucosa (SIS)

The Badylak group at University of Pittsburgh are pioneers in the field of decellularized ECM and have studied the properties of porcine small intestinal submucosa (SIS). SIS has been evaluated in many animal models such as murine and canine, and also used in human tendon repair as well as abdominal cavity defects and other applications. Currently they are investigating the use of various pepsin-digested

ECM scaffolds as hydrogels for regeneration (36). ECM hydrogels have been tested with success in skeletal muscle regeneration after acute large-scale trauma (37, 38). Their research has also investigated the polarization of macrophages along the M1-M2 axis and its implications in graft acceptance and remodeling versus inflammation and

12 rejection. They have associated the alternative (M2) state of macrophage activation with the use of SIS as opposed to the inflammatory (M1) activation state. SIS contains several growth factors that are known to be important in tissue regeneration, including transforming growth factor beta (TGF-β) and vascular endothelial growth factor (VEGF), which are key regulators in wound repair and vascularization. The matrix has also been shown to include fibroblast growth factor-2 and extracellular matrix proteins such as glycosaminoglycans and proteoglycans, though its structure is predominated by collagens. SIS is also available commercially through various companies.

Urinary Bladder Matrix (UBM)

Decellularized urinary bladder (UBM) has been made commercially available through the company ACell marketed under the product name Matristem. UBM is used in both sheet and particulate forms as a scaffold for regeneration and wound healing.

Much like SIS and dermal ECM products, UBM in its sheet form is used in abdominal wall repair and skin lesion repair. The particulate form was tested in topical applications to aide in complex tissue regeneration. UBM hydrogels that are formed by pepsin digestion of ECM are under evaluation for the treatment of traumatic brain injury (TBI) and tested as an undigested gel as well as a lyophilized powder (39). UBM has also been combined with synthetics through incorporation via electrospinning (40). These scaffolds represent a hybrid between typical synthetic scaffolds and biologically derived

ECM, offering some of the signals and ECM macromolecules to promote healing and regeneration.

Acellular Adipose Tissue (AAT)

Though largely composed of lipid triglycerides for energy storage, white adipose tissue also has an associated connective tissue that can be isolated. This acellular

13 tissue is obtained by extrusion of lipids, followed by standard decellularization procedures such as acid and detergent treatments. AAT is currently being investigated in its promise as a soft-tissue replacement as it forms a thick paste when hydrated (41).

In a porcine model, both human and porcine AAT were injected subcutaneously and then excised to gauge the integration with host tissue as well as the inflammatory response. In most cases, there were a large number of cell-dense aggregates that appeared on the periphery of the xenogeneic human-derived ECM, while fewer aggregates formed around the allogeneic implant. These cellular aggregates stained densely for the macrophage marker CD11b in both xenogeneic and allogeneic implants.

After a 2-week incubation of porcine AAT in rats, de novo adipose tissue was observed in the soft tissue implants. Vascularization of subcutaneous grafts was observed in an athymic mouse model as early as 1-week post implantation.

AAT has also been tested as a model for breast cancer progression and metastasis in response to various drugs used to treat these malignancies (42). Breast cancer cells were cultured in human adipose tissue-derived extracellular matrix (hDAM) in the presence of Doxorubicin and Lapatinib to measure their behavior by migration, proliferation, and cell shape. This model is stated to be more tissue-mimetic than current ex vivo cancer models such as three-dimensional Matrigel culture and conventional 2- dimensional cell culture. Inhibition of EGFR signaling (a target effect of Lapatinib) was increased in hDAM cultures compared to a 2D system, whereas the effect of

Doxorubicin was decreased in hDAM compared to 2D assays. The increased ability to study the tumor microenvironment will allow for an easier and more effective technique for engineering cancer therapies.

As an Immunomodulatory Reagent and Tissue Model

The role of the tissue microenvironment in host responses to pathogens or injury has been an exciting area of research in the field of immunology. It is well accepted that the surrounding tissues in which that response is occurring largely dictate the polarization of that response (2, 43). This is especially relevant in immune privileged organs such as the brain and eye, where an inflammatory response would be catastrophic to the function of a tissue. These effects are caused not only by the cellular component of an organ, but also the extracellular matrix and associated small proteins, sugars, and other molecules. Decellularized ECM is a promising tool for bioengineers as a scaffold, and has recently demonstrated utility as an immunomodulatory reagent. A more thorough understanding of the ability of ECM to induce these polarizations, as well as an in-depth characterization of the responses of cells to varying tissue ECM, have yet to be demonstrated. With further understanding of these interactions, we could leverage the polarization tendencies of various ECM scaffolds with a synthetic ability to promote a desired immune phenotype in regenerative medicine and disease models.

Macrophage polarization in response to a scaffold directly affects the outcome of that scaffold being either rejected or remodeled (16, 18). In general, the field has concluded that an abundance of M1 macrophages, expressing iNOS as well as a host of inflammatory cytokines, TNFα, interleukin 1 beta (IL-1β) and interleukin 6 (IL-6) among others, signals poor scaffold integration with surrounding tissue. Conversely, M2 macrophages expressing arginase 1 (ARG1), Fizz1, and cytokines such as IL-10 are important in suppressing inflammation and healing damaged tissue. Previous research has verified the ability of decellularized small intestinal submucosa (SIS) to promote an

M2 macrophage phenotype in vivo (16, 44, 45). Using SIS as both a scaffold and powder for wound healing has shown the promise of ECM as a powerful immunomodulatory reagent.

Further studies would need to involve the characterization of this response as a whole, with a deeper understanding of the intricacies of immune polarization. There are some pitfalls using a purely M2-promoting scaffold, including the essence of the type-2 phenotype: collagen deposition. In tissue engineering our goal is to create a functional replacement tissue, which is hindered by the deposition of large collagenous scarring.

Pathogenic type-2 immunity and arginase expression, which leads to collagen deposition, is present in fibrotic diseases such as pulmonary fibrosis (46) and tracheal stenosis (47). As previously mentioned, type-2 cytokines are present in the foreign body response (FBR) to synthetic materials, which yields fibrotic encapsulation of engineered devices. In order for a scaffold to create an optimal immune microenvironment, we would need it to promote a balance between the M1 and M2 phenotypes. This balance would prevent any damaging inflammation caused by a strong M1 reaction, while avoiding a large M2-derived scarring response rendering the tissue non-functional. Through the use of decellularized ECM scaffolds, researchers have demonstrated the ability of varying tissues to exert polarization influence on the M1/M2 axis of macrophage activation.

These polarizations are dependent upon both the ECM tissue source, as well as whether this source is xenogeneic or allogeneic in nature.

Biochemical and Physical Alterations of Scaffolds

Through the integration of the physical and chemical properties of scaffolds, it is possible to modify biomaterials to better promote the desired cellular outcome (48, 49).

Physical modification of scaffolds includes the use of particle and fiber-based polymers, as well as particle and sheet-based ECM scaffolds. These materials can be further modified biochemically with proteins and small molecules that promote cell polarization to a desired lineage or allow for a more conducive environment for progenitor cells to grow and differentiate.

Physical Alteration of ECM & Synthetics

Decellularized ECM has been used in sheet and particulate forms, depending upon the desired application. ECM sheets are typically created out of a thin section of tissue that has been decellularized, maintaining its overall architecture. This sheet can be fenestrated, with openings introduced to its structure, promoting vascularization and cell migration. Particulate ECM is typically tissue that has been decellularized and lyophilized prior to milling, either through a liquid-nitrogen cryogenic mill or through a knife mill. The latter produces larger particle sizes and does not require the ECM to be lyophilized. Cryo-milling involves cooling lyophilized ECM with liquid nitrogen to make it very brittle, then milling it with a magnetic rod. This yields a material with a size range in the micrometer scale. It is possible to further decrease particle size by hydraulic shearing to create nanometer-sized ECM.

Recently, researchers have shown that different physical properties of synthetic scaffolds can alter the behavior of cells that come in contact with that material. Density and torsion of polycaprolactone (PCL)-based scaffolds have been changed to modify the polarization of immune cells such as macrophages (50). The authors claim that differing cyclic strain on human peripheral blood mononuclear cells (PBMCs) induces the presence of M2-polarization, promoting the deposition of new extracellular matrix. It is important to note that in all cases, strained and unstrained, CCR7+ M1 macrophages were present in addition to the presence or absence of CD206+ M2 macrophages. The majority of cells were M1 phenotype at day 1, however under intermediate (7%) torsional strain the population of M2 cells increased over time.

Cyclodextrin and RGD modification of Polymeric Scaffolds

Through the use of PEG-DA (polyethylene glycol diacrylate) and PEG-MA

(polyethylene glycol monoacrylate) modified Matrigel hydrogels, Beck et al. demonstrated a varying cellular response to modified physical and adhesive properties of these hydrogels (51). PEG-DA is able to crosslink and will increase hydrogel stiffness, while PEG-MA is not able to crosslink and therefore limits hydrogel stiffness. With increasing concentrations of PEG-DA, Beck et. al. observed altered growth and differentiation of mammary epithelial organoids in vitro. With greater scaffold stiffness, they observed less differentiation and formation of primitive branching. They also tested the effect of epithelial cell dissemination from PEG-Matrigel constructs by varying the concentration of adhesion motifs. These motifs were added through the use of a cyclodextrin ring molecule with covalently attached RGD (arginine-glycine-aspartic acid), which is a peptide sequence that is recognized adhesion proteins known as integrins.

With the addition of these extracellular adhesion signals, the authors showed that cell dissemination was inhibited by stiffness of the hydrogel, and induced by chemically added adhesion signals.

Growth Factor Modification and Modulation of Microenvironment

Through manipulation of growth factors, researchers have been able to administer exogenous signaling proteins to an area of tissue damage to promote proper repair (52). In this example, an engineered fibronectin (FN) fragment was modified with factor XIIIa. The segment of factor XIIIa that was introduced to the recombinant FN fragment polymerizes fibrin and is cross-linked to the scaffold. The modified FN fragment displays both integrin and growth factor (GF) binding domains. When applied exogenously, therapeutic growth factors will associate with the GF domain of the FN fragment, which has been linked to the host ECM through factor XIIIa cross-linking as

18 described previously. This localized the GF near integrin-binding domains of the recombinant protein, bringing cells into close contact with the growth factor and allowing signaling to induce proliferation and differentiation. Growth factors tested with the binding peptide included vascular endothelial growth factor-A (VEGF-A), platelet-derived growth factor-BB (PDGF-BB) and bone morphogenetic protein-2 (BMP-2). Combinations of these growth factors were combined to promote skin and bone repair.

Application of these peptide-manipulations has also been tested in osteoarthritis

(OA) treatment. Modification of a hyaluronic acid (HA) binding peptide allowed researchers to extend the lifetime of HA injected into the synovium of the knee joint in an

OA model (53). Using a biochemical approach, in several platforms, it is possible to localize signaling and structural proteins to a location of interest, which could be exploited for use in immunomodulation by synthetically creating an optimal immune microenvironment.

Summary

The tissue microenvironment plays a large role in the host response to biomaterials. This phenomenon has been observed with other immunological challenges such as pathogens that elicit different responses dependent upon the location in which the host encounters that pathogen. Current research has shown a variance in the extent of the inflammatory response to a scaffold as well as its degradation kinetics when applied in different tissue locations. These altered immunologic reactions also occur with biologically derived scaffolds such as ECM. Small intestinal submucosa (SIS), Urinary

Bladder Matrix (UBM) and Acellular Adipose Tissue (AAT) are all examples of ECM- derived scaffold that have been used successfully in the clinic and are under evaluation for their precise role in wound healing and regeneration of new tissue. ECM can act both as a regenerative scaffold and a bioengineering tool to mimic the tissue environment,

19 promoting certain immune phenotypes that will provide a conducive setting for regeneration. These tendencies for certain immune polarizations will vary depending upon the tissue source of the ECM, much like how the immune response to a scaffold changes depending upon the location within a cellular tissue. Through physical and chemical manipulation of the biomaterial with reference to its location of implantation, researchers will be able to create an optimal setting for functional healing and generate a more effective replacement tissue.

Table 1.1. Immune Polarizations and Their Effector Functions Table 1: Immune Polarizations and Their Effector Functions

Cytokines & Antibody Main Major Cells Pathogenesis Chemokines Class Tissues/Targets CTLs, NK Cells, Local: Damaging IFN-γ, TNF-α, IL-1β, Skin, intracelluar T 1 macrophages expressing IgG inflammation, cell death H IL-6, IL-12 pathogens iNOS, dendritic cells Systemic: Cachexia Local: Fibrosis, Scar Eosinophils, basophils, deposition IL-4, IL-5, IL-13, IL- Mucosal surfaces, T 2 macrophages expressing IgE Systemic: Asthma/Allergy, H 10 helminths, allergy ARG1, B Cells diarrhea, promotes cancer survival IL-6, TGF-β, IL-17, Skin, extracellular T 17 Neutrophils IgG Autoimmunity H IL-22, IL-23 pathogens Local: Antigenic ignorance All organs, T IL-10, TGF-β T Cells, Macrophages IgA Systemic: Can promote reg immunosuppression cancer survival

Table 1.2. Role of Immune Populations in Inflammation and Tissue Regeneration Table 2: Role of Immune Population in Inflammation and Tissue Regeneration

Immune Function Constructive Phenotypes Destructive Phenotypes

Inducing helpful inflammation to M1 macrophages express Phagocytic cells of the innate clear pathogens and recruit iNOS which can lead to faster immune system that are capable of adaptive immune cells. Deposition scaffold degradation, over- Macrophages clearing debris and secreting of collagenous scars during wound active inflammation, M2 inflammatory messengers as well as healing and required for limb macrophages can induce over- antigen presentation regeneration in axolotls. scarring and fibrosis

Includes neutrophils, basophils, eosinophils. Phagocytic and Recruitment of other cells to the chemotactic cells of the innate site of injury or infection. Increased PMN infiltration & PMN's immune system that are involved in Polarizing to secrete cytokines cell death can lead to exudate clearing debris, activation and such as IL-4 to help induce wound accumulation. polarization of adaptive immune healing responses. system.

Major mediators of cellular adaptive Secretion of cytokines that help Overactive immune response immunity. Help polarize immune polarize immune response, act as can induce inflammation, T Cells response and subsets are responsible regulators to prevent an overactive heightened regulatory for killing of target cells. response. response

TTableable 3: 1.3 CommonCommon Materials Materials us useded in Tinis sTissueue Engine Engineeringering

Material Applications Commercial Products drug delivery, PLGA poly(lactic-co-glycolic acid) nanofiber scaffolds, Vicryl nanoparticles Synthetics PCL polycaprolactone nanofiber scaffolds Capronor dermal filler, Dacron, Bio-Anchor, PLLA poly-L-lactide nanofiber scaffolds DEXON

SurgiMend®, Alloderm, ACD Acellular dermis hernial repair, skin lesion repair, FlexHD® abdominal cavity defect repair, Oasis® Wound Matrix, ECECMM SIS Small intestinal submucosa volumetric muscle CorMatrix®, DynaMatrix® loss, burns, multi- UBM Urinary bladder matrix purpose degradable Matristem® scaffolds

Chapter 2: Methods

Below are subsets of important methods that are used in multiple manuscripts. The methods for each specific section are listed within the text of that section.

Tissue ECM Preparation

Porcine derived tissues (Wagner Meats, Mt. Airy MD) were processed following a standard protocol. Samples were formulated into a paste through the use of a knife-mill processor (Retsch, Germany) with particle sizes no larger than 5 mm2 and rinsed thoroughly with running distilled water until blood was cleared from samples. Bone samples were pre-treated for decalcification by incubation in 10% formic acid (Sigma) for

3 days, which was verified by a colorimetric calcium test (STANBIO Laboratory). Tissues were then incubated in 3.0% peracetic acid (Sigma) on a shaker at 37oC for 4 hours, with a change to fresh acid solution after 1 hour. pH was adjusted to 7 with thorough water and PBS rinsing, and tested after solution was freshly changed and tissue rested for 20 minutes. Samples were washed once more with distilled water then transferred to a 1% Triton-X100 (Sigma) + 2 mM sodium EDTA (Sigma) solution on a stir plate at 400 rpm, room temperature for 3 days, changing the solution daily. Tissues were rinsed thoroughly with distilled water until no bubbles formed from detergent upon agitation.

Finally, processed tissues were incubated in 600 U/ml DNase I (Roche Diagnostics) + 10 mM MgCl2 (J. T. Baker) + 10% Antifungal-Antimycotic (Gibco®) for 24 hours. Tissues were rinsed thoroughly with distilled water, then frozen at -80oC and lyophilized for 3 days. The dry sample was turned into a particulate form using a SPEX SamplePrep

Freezer/Mill (SPEX CertiPrep). ECM powder was stored between -20oC and -80oC and

UV sterilized prior to use. Collagen from bovine tendon (Sigma) was cryomilled using the

SPEX SamplePrep Freezer/Mill to form a particulate similar to the whole tissue ECM samples.

Cell Culture

Murine immortalized bone marrow macrophages (iBMM, Michele de Palma, Ecole

Polytechnique Federale Lausanne) were cultured as per developer’s protocol in IMDM

(Gibco®) media containing 20% FBS (Hyclone, GE Healthcare Life Sciences), L-

Glutamine (Gibco®), PenStrep (Life Technologies), and 50ng/ml M-CSF (Recombinant

Mouse, BioLegend). iBMM’s were cultured on plates coated with decellularized ECM powder for 24 hours in growth media.

RT-PCR

In vitro samples: After 24 hours cells were harvested in TRIzol (Life Technologies) and

RNA was extracted using a combination of TRIzol and RNeasy Mini (Qiagen) column- based isolations. cDNA was synthesized through the use of SuperScript Reverse

Transcriptase III (Life Technologies) as per manufacturer’s instructions. RT-PCR was conducted on an Applied Biosystems Real Time PCR Machine using SYBR Green (Life

Technologies) as a reporter and the following primers: B2m forward CTC GGT GAC

CCT GGT CTT TC, B2m reverse GGA TTT CAA TGT GAG GCG GG; Tnfa forward GTC

CAT TCC TGA GTT CTG, Tnfa reverse GAA AGG TCT GAA GGT AGG; Il1b forward

GTA TGG GCT GGA CTG TTT C, Il1b reverse GCT GTC TGC TCA TTC ACG; Inos forward GAC GAG ACG GAT AGG CAG AG, Inos reverse GTG GGG TTG TTG CTG

AAC TT; Arg1 forward CAG AAG AAT GGA AGA GTC AG, Arg1 reverse CAG ATA TGC

AGG GAG TCA CC; Retnla forward CTT TCC TGA GAT TCT GCC CCA G, Retnla reverse CAC AAG CAC ACC CAG TAG CA; Ccl2 forward GCT CAG CCA GAT GCA

GTT AAC, Ccl2 reverse CTC TCT CTT GAG CTT GGT GAC (Integrated DNA

Technologies).

In vivo samples: In vivo inguinal and axillary/brachial lymph node samples from volumetric muscle loss (VML) studies were homogenized in TRIzol and RNA was extracted using a combination of TRIzol and RNeasy Mini (Qiagen) column-based isolations. cDNA was synthesized through the use of SuperScript Reverse Transcriptase

III (Life Technologies) as per manufacturer’s instructions. RT-PCR was conducted on an

Applied Biosystems Real Time PCR Machine using SYBR Green (Life Technologies) as a reporter and the following primers: B2m forward CTC GGT GAC CCT GGT CTT TC,

B2m reverse GGA TTT CAA TGT GAG GCG GG; Tnfα forward GTC CAT TCC TGA

GTT CTG, Tnfα reverse GAA AGG TCT GAA GGT AGG; Il1β forward GTA TGG GCT

GGA CTG TTT C, Il1β reverse GCT GTC TGC TCA TTC ACG; Retnla forward CTT TCC

TGA GAT TCT GCC CCA G, Retnla reverse CAC AAG CAC ACC CAG TAG CA; Ifnγ forward TCA AGT GGC ATA GAT GTG GAA, Ifnγ reverse TGA GGT AGA AAG AGA

TAA TCT GG; Il4 forward ACA GGA GAA GGG ACG CCA T, Il4 reverse ACC TTG GAA

GCC CTA CAG A. Whole muscle samples were processed similarly to lymph nodes to isolate RNA and produce cDNA. Primers used included those previously described and:

Arg1 forward CAG AAG AAT GGA AGA GTC AG, Arg1 reverse CAG ATA TGC AGG

GAG TCA CC; Col1a1 forward CTG GCG GTT CAG GTC CAA T, Col1a1 reverse TTC

CAG GCA ATC CAC GAG C; Fabp4 forward TCA CCT GGA AGA CAG CTC CT, Fabp4 reverse AAT CCC CAT TTA CGC TGA TG; AdipoQ forward TCC TGG AGA GAA GGG

AGA GAA AG, AdipoQ reverse TCA GCT CCT GTC ATT CCA ACA T; Lep forward TTC

ACA CAC GCA GTC GGT AT, Lep reverse ACA TTT TGG GAA GGC AGG CT; Actb forward ATG TGG ATC AGC AAG CAG GA, Actb reverse AAG GGT GTA AAA CGC

AGC TCA (Integrated DNA Technologies). F4/80+ and CD3+ cells from volumetric muscle wounds were sorted directly into RNA lysis buffer; RLT buffer (Qiagen) + β- mercaptoethanol (Sigma). RNA was isolated using an RNeasy Micro Kit (Qiagen) with carrier RNA and on-column DNase treatment. cDNA synthesis was performed with a

High Capacity cDNA Reverse Transcription Kit (Applied Biosystems). Isolated RNA underwent preamplification prior to plating in custom 96-well TaqMan® Array Fast Plates

(Life Technologies) and gene expression was detected on an Applied Biosystems

StepOne Real-Time PCR System.

Subcutaneous ECM Implantation

ECM was hydrated with 1X PBS to form implants of 0.3 mg/ml. 0.2 ml of the ECM paste was injected twice subcutaneously into female 6.5 week old wild type C57BL/6 mice

(Charles River Laboratories) both proximal and distal locations. Injection sites were sterilized with ethanol and Povidone-Iodine to prevent any inflammation altering the response to the scaffold. Implants were dissected with surrounding tissue including or excluding skin and processed for histology and FACS, respectively. All animal procedures were done in accordance with the Johns Hopkins University ACUC guidelines.

Volumetric Muscle Loss Surgery

Six- to eight-week-old female wild type C57BL/6 (Charles River), B6.129S7-

Rag1tm1Mom/J, BALB/c-Il4ratm1Sz/J, or B6.129S2-Cd4tm1Mak/J (Jackson Laboratories) mice were anesthetized with 4.0% isoflurane and maintained under 2.5% isoflurane. Hair was removed from the lower extremities with an electric razor (Oster). After ethanol sterilization of the surrounding skin, a 1.5-cm incision was created between the knee and hip joint to access the quadriceps femoris muscle. Through the use of surgical scissors, a 3 mm x 3 mm deep defect was created in the quadriceps femoris muscle group. The resulting bilateral defects were filled with 0.05 cc of a 250 - 350 mg/ml biomaterial scaffold (UV-sterilized ECM (manufactured in house) or Collagen (Sigma)) or 0.05 cc of a vehicle (saline) control. Mice were under anesthesia for 10 – 15 minutes during

28 surgical preparation and procedure before return to cage and monitored until ambulatory. Directly after surgery, mice were given subcutaneous carprofen (Rimadyl®,

Zoetis) at 5 mg/kg for pain relief and were maintained on Uniprim® antibiotic feed (275 ppm Trimethoprim and 1365 ppm Sulfadiazine, Harlan Laboratories) until the end of study to prevent opportunistic infections. After 1 (7 days), 3 (24 days) and 6 (42 days) weeks, the mice were sacrificed and their entire quadriceps femoris muscle was removed by cutting from the knee joint along the femur to the hip joint. Both inguinal and axillary/brachial lymph nodes and whole muscle samples for RNA isolation were flash frozen in liquid nitrogen and stored at -80oC until RNA extraction. All animal procedures in this study were conducted in accordance with an approved Johns Hopkins University

IACUC protocol.

Histology

SubQ Samples: Implants were fixed overnight in 10% formalin prior to dehydration and paraffin embedding. 5 μm sections were rehydrated then prepared for immunohistochemistry (IHC) or direct staining with hematoxylin and eosin (Sigma-

Aldrich). IHC samples were treated with a citrate antigen retrieval buffer, 10 mM sodium citrate (J.T. Baker) + 0.05% Tween20 (Sigma-Aldrich) at pH 6, for 30 minutes in a vegetable steamer. Sections were stained with primary antibodies against CD11b, CD3,

Neutrophil Elastase, or CD11c (AbCam) overnight at 4oC then visualized using

SuperPictureTMPolymer Detection Kit, HRP-DAB (Life Technologies). These samples were then counterstained with hematoxylin (Sigma-Aldrich).

Volumetric Muscle Loss Lymph Nodes and Muscle Tissue: Inguinal lymph nodes were harvested and fixed in formalin overnight before dehydration and paraffin embedding, microtome sectioning, then histological examination via hematoxylin and eosin staining.

Muscle samples were prepared as fresh-frozen samples for cryosectioning by flash freezing in isopentane after mounting in Tragacanth gum (Sigma Life Science). A

Microm HM 550 cryostat (Fisher Scientific) was used to obtain 10 μm cryosections from

5-7 different regions of each muscle roughly 300 μm apart. Sections were stained with a

Hematoxylin and Eosin protocol (Sigma Aldrich) or with a Modified Masson’s Trichrome protocol.

Flow Cytometry

In vitro: In vitro samples were harvested using Accutase (Life Technologies) and washed with cold 1X PBS. Then, cells were stained with the following antibody panel: F4/80 PE-

Cy7 (BioLegend), CD11b Pacific Blue (BioLegend), CD11c APC-Cy7 (BD Biosciences),

CD86 AlexaFluor700 (BioLegend), MHCII (I-A/I-E) AlexaFluor488 (BioLegend), CD206

APC (BioLegend) and LIVE/DEAD® Fixable Aqua Dead Cell Stain Kit (Life

Technologies). Samples were fixed using the BD Cytofix/CytopermTM kit (BD

Biosciences), and run on BD LSRII Cell Analyzer, data was analyzed using FlowJo Flow

Cytometry Analysis Software (Treestar). M1/M2 polarization levels were determined by mean fluorescence intensity of CD86 and CD206 in LIVE/DEAD® Fixable Aqua Dead

Cell Stain -F4/80+CD11c- cells.

In vivo: Subcutaneous ECM implants and VML samples were harvested then finely diced using a scalpel in 1XPBS. Resultant material was digested for 45 minutes at 37oC in 5 mg/ml Liberase TL (Roche Diagnostics) + 0.2 mg/ml DNase I (Roche Diagnostics) in serum-free RPMI (Gibco). Digest was filtered through a 100 μm cell strainer then washed twice with 1XPBS + 0.5 mM EDTA and once with 1XPBS. Cells were resuspended in 5 ml 1XPBS and carefully layered atop 5 ml Lympholyte-M (Cedarlane), then spun for 20 minutes at 1200 xg. Cellular interphase was washed twice with 1XPBS

30 then transferred to a 96-well plate for antibody staining. Isolated cells were stained with the following antibody panel: Viability Aqua (Life Technologies), CD19 BrilliantViolet 421

(BioLegend), CD3 AlexaFluor 488 (BioLegend), CD34 PerCP-Cy5.5 (BioLegend),

CD11c APC-Cy7 (BD Biosciences), F4/80 PE-Cy7 (Biolegend), CD86 AlexaFluor700

(BioLegend), CD206 APC (BioLegend). After staining cells were fixed and analyzed as per previously described. Viability Aqua negative (live) cells were evaluated based off of percent population of T Cells (CD3+), B Cells (CD19+), Dendritic Cells (CD11c+),

Progenitor Cells (CD34+) and Macrophages (F4/80+). As with in vitro studies, macrophages were further analyzed for polarization by mean fluorescence intensity of

F4/80+ cells in CD86 AlexaFluor700 and CD206 APC channels. All analysis was performed in FlowJo Flow Cytometry Analysis Software (Treestar). T cell analysis was performed using the following panel: CD3 AlexaFluor488 (Biolegend), CD4 PE/Cy7

(Biolegend), CD8 AlexaFluor700 (Biolegend), FoxP3 Pacific Blue (Biolegend) and

Fixable Viability Dye eFluor780 (eBioscience). Myeloid compartment analysis in the volumetric muscle wound at 1 week post-injury was done with the following antibody panel: Fixable Viability Dye eFluor®780 (eBioscience), F4/80 PE-Cy7 (BioLegend),

CD11b AlexaFluor700 (BioLegend), CD11c APC (BioLegend), Ly6C Per/CP-Cy5.5

(BioLegend), Ly6G PacificBlue (BioLegend), CD86 BrilliantViolet510 (BioLegend),

CD206 PE (BioLegend), MHCII I-A/I-E AlexaFluor488 (BioLegend).

Treadmill Testing for Muscle Function

48 hours prior to testing mice were trained on treadmill apparatus running at 5 m/min and increased by 1 m/min every minute for a total of 5 minutes. Mice were run to exhaustion starting at 5 m/min and increased by 1 m/min every minute. Exhaustion was defined as when the mouse stayed on the pulsed shock grid for a continuous 30

31 seconds (Treat NMD: Brunelli et al 2007, Denti et al 2006). Animals were tested at least

48 hours prior to harvesting for analysis via FACS, PCR, or histology.

Chapter 3: Differential effects of tissue-ECM source on cell function

Summary

The tissue microenvironment has profound effects on the immune system’s response to a foreign body such as a pathogen or a biomaterial implant. The following research focuses on how this microenvironment changes a host response to regenerative scaffolds used in tissue engineering, and how we can manipulate that immune environment to promote tissue growth and regeneration. Through the use of decellularized extracellular matrix (ECM) scaffolds, we are able to investigate how immune cells interact with these matrices and how the interplay between the scaffold and cells create the host response. This response is dependent upon the tissue from which the ECM scaffold is derived, as well as the location within the host that the scaffold is placed. In this manner, our ECM scaffolds can be viewed as both a biomaterial in a pre-existing tissue microenvironment (implantation site) as well as an engineered environment (tissue or organ from which the ECM is derived) that alter the immune response.

Tissue matrix arrays for high-throughput screening and systems analysis of cell

function

This work has been published and is reprinted here with permission from Nature

Methods.

Beachley, Vince Z., Matthew T. Wolf, Kaitlyn Sadtler, Srikanth S. Manda, Heather

Jacobs, Michael R. Blatchley, Joel S. Bader, Akhilesh Pandey, Drew Pardoll, and

Jennifer H. Elisseeff. "Tissue matrix arrays for high-throughput screening and systems analysis of cell function." Nature Methods (2015).

Abstract

Cell and protein arrays have demonstrated remarkable utility in the high- throughput evaluation of biological responses; however, they lack the complexity of native tissue and organs. Here we spotted tissue extracellular matrix (ECM) particles as two-dimensional (2D) arrays or incorporated them with cells to generate three- dimensional (3D) cell-matrix microtissue arrays. We then investigated the responses of human stem, cancer and immune cells to tissue ECM arrays originating from 11 different tissues. We validated the 2D and 3D arrays as representative of the in vivo microenvironment by means of quantitative analysis of tissue-specific cellular responses, including matrix production, adhesion and proliferation, and morphological changes after culture. The biological outputs correlated with tissue proteomics, and network analysis identified several proteins linked to cell function. Our methodology enables broad screening of ECMs to connect tissue-specific composition with biological activity, providing a new resource for biomaterials research and further understanding of regeneration and disease mechanisms

Introduction

Tissues and organs in the body are composed of cells and their surrounding

ECM, generated by self-assembly and cellular processing (54). Tissue specificity is determined by the unique composition of the tissue—from hundreds of different biomolecules—and supramolecular structures that interact physically, chemically and biologically with cells to regulate cellular-level functions (55-58). Ongoing research continues to elucidate how the structural and compositional properties of the ECM influence resident cells (59, 60). Despite the use of tissue-derived materials in the clinic, detailed mechanistic information on how tissue ECMs directly influence cell behavior or repair processes is largely unavailable.

Strategies based on arraying DNA, RNA, single ECM proteins or synthetic biomaterials allow high-throughput in vitro screening of cellular functions and biological outputs on diverse substrates (61-65). For example, 2D microarray libraries of synthetic polymers have been used to delineate the optimal scaffold composition for lineage- specific stem cell differentiation (61, 62). ECM proteins have been integrated with synthetic hydrogels to identify combinations that stimulate stem cell osteogenesis in three dimensions (66-68). Individual and combinatorial screening of purified proteins in microarray formats has suggested mechanisms of cell-protein interactions (63) and identified candidate cell-protein interactions that correlate with cancer metastasis (64).

All of these previous arrays started with simple building blocks, such as polymers or proteins that can be tested in a combinatorial manner. However, cells in the body exist in tissues and organs with a complex ECM that includes hundreds of different molecules ranging from nanometer-sized fibrils to micrometer-sized units that can modulate cell behavior (69).

Tissue ECMs have been used for regenerative medicine and wound healing in

36 humans (37, 70, 71), typically to match 'like with like' (72-75)—for instance, stem cells cultured on liver ECM to create new liver tissue (73). However, broader screening of tissue ECM properties may elucidate more general biological functions and novel therapeutic entities. In order for the understanding and use of tissue-derived biomaterials to advance, high-throughput screening tools are needed to probe variability in ECM composition and complex cell-matrix interactions in vitro. This would allow the intricate mechanisms of cell-material response and repair processes to be teased apart to explain how these materials can be used to influence cell behavior in vivo.

Here we developed 2D and 3D tissue ECM arrays for screening biological responses to tissue-specific scaffold microenvironments. We processed tissues to remove soluble tissue components and then mechanically fragmented the matrix to create tissue ECM microparticles that retained the proteomic complexity of the natural

ECM in a medium compatible with array fabrication (spotting). We characterized a range of cell-matrix interactions at cellular and functional levels, including mineralization, cell adhesion and proliferation, gene expression, and changes in cell morphology, using stem cells, cancer cells and macrophages. We then correlated tissue ECM array outputs with proteomic composition to build networks that highlighted candidate proteins responsible for tissue-specific differences in cell function.

Methods

Tissue processing.

Porcine tissues were harvested from 6-month-old market-weight pigs weighing approximately 100 kg (Wagner's Meats, Mt. Airy, Maryland, USA). Mouse tissues were harvested from normal mammary glands of 6–10-week-old female wild-type C57BL/6 mice and early-stage (diameter: ~9 mm) and late-stage (diameter: ~16 mm) tumors from

MMTV-PyMT mutant mice (Jackson Laboratories). Skeletal muscle was harvested from

37 the quadriceps of 26–32-week-old female Dmdmdx-5Cv mutant mice (Jackson

Laboratories) and wild-type control mice. Whole tissues and organs were cut into pieces approximately 100 mm3 in size and rinsed several times with phosphate-buffered saline

(PBS). Bone tissue required an additional decalcification preparation in 10% formic acid for 18 h at room temperature, and fat was mechanically pressed to reduce the lipid content before processing. Unless otherwise noted, tissue was processed by incubation with three different solutions, with thorough washing in PBS between each step24: (1)

3% peracetic acid for 3 h at 37 °C, (2) 1% Triton X-100 containing 2 mM EDTA for 18 h at 37 °C and (3) 600 U/mL DNase (Sigma) containing 10 mM MgCl2 for 18 h at 37 °C.

After the final treatment, the tissue was washed thoroughly with PBS followed by distilled water and then lyophilized.

For experiments comparing multiple processing protocols, the following modifications were used: fresh—tissues were not treated with acid, detergent or DNase; mild—tissues were incubated in (1) 3% peracetic acid for 10 min at 37 °C and (2) 600

U/mL DNase containing 10 mM MgCl2 for 18 h at 37 °C; moderate—tissues were incubated in (1) 3% peracetic acid for 3 h at 37 °C, (2) 1% Triton X-100 containing 2 mM

EDTA for 18 h at 37 °C and (3) 600 U/mL DNase containing 10 mM MgCl2 for 18 h at 37

°C; harsh—tissues were incubated in (1) 3% peracetic acid for 3 h at 37 °C, (2) 1%

Triton X-100 containing 2 mM EDTA for 18 h at 37 °C, (3) 4% SDS for 16 h at 37 °C and

(4) 600 U/mL DNase containing 10 mM MgCl2 for 18 h at 37 °C; and digest—ECM particles treated using the moderate protocol were further digested in a 1 mg/ml solution of porcine pepsin in 0.01 NHCl for 72 h.

Tissue particle fabrication.

Lyophilized processed tissue was cryogenically pulverized in a cryomill (SPEX

6770, SPEX SamplePrep) at −195 °C under liquid nitrogen. Approximately 300–500 mg of sample was processed in each batch. Cryomill settings were 8–15 1-min cycles at 10 cycles s−1 with 3-min cooling periods between runs. The resulting powder was suspended in distilled water or DMEM at 10 mg/mL and sonicated with a probe sonicator

(GE 130PB, Cole-Parmer) at an output power of 10–15 W two times for 30 s in an ice bath. Water suspensions were centrifuged at 4,000 r.p.m. for 10 min and resuspended in fresh deionized (DI) water. Sonication was repeated, and the suspension was filtered through a 40-μm cell sieve. The final concentrations of solutions were determined from the mass of lyophilized aliquots.

Microarray-chip fabrication.

Glass coverslips (22 × 60 mm) were cleaned and functionalized with methacrylate groups using a silane reaction, as previously described10. Acrylamide was mixed with bis-acrylamide and dissolved in DI water at concentrations of 10.55% and

0.55% (wt/vol). A photoinitiator solution of Irgacure (I2959) dissolved in methanol at 200 mg/mL was added to the acrylamide solution at a concentration of 10% (vol/vol). A 20-μL drop of working solution was pipetted onto the functionalized 22 × 60–mm coverslip, and an untreated 22 × 50–mm glass slide was carefully placed on top of the liquid to form a thin layer estimated to be 36 μm thick. The solution was polymerized with UV light (~2 mW/cm2) for 10 min, and the 22 × 50–mm coverslip was removed after incubation in DI water for 30 min. Gel-coated slides were soaked in DI water overnight and dried on a hot plate at 40 °C for 45 min.

Silicon gaskets with arrays of 3-mm-diameter wells (CWCS-50R, Grace Bio-Labs) were placed on the dry gel-coated slides with 40 wells in full contact. Collagen (C7661,

Sigma-Aldrich) dissolved at 0.25 mg/mL in 0.1 M acetic acid was pipetted (9 μL) in each chamber and allowed to dry overnight. Next, 10 μL of tissue particle suspension was spotted in each of the collagen-coated wells. The working tissue particle concentration was 2 mg/mL for suspensions of liver, lung, spleen and small intestine and 3 mg/mL for all other tissue types. The selected concentration was determined visually as the amount required to form a complete monolayer spot without visualization of the underlying acrylamide gel surface. Mixed cardiac and lung ECM solutions were combined at ratios of 1:3, 3:1 and 1:1 (vol/vol) and spotted at a total concentration of 3 mg/ml to form composite ECM substrates. Spotted chips were left to dry overnight in a cell-culture hood at room temperature, and the gaskets were removed. Chips were sterilized in a culture hood with UV light for 30 min on each side.

Tissue spot variability.

The reproducibility of the spotting techniques was validated by histochemical staining and proteomic analysis. Dried substrates spotted from ECM particle suspensions of 11 different tissues and collagen controls were stained using Masson's trichrome. Three replicates of each tissue type were visually compared for color content and distribution. We quantitatively compared the overall staining variation between tissues using a hue filter in ImageJ to threshold the percentage of each total spot area contained in the hue ranges 151–159, 209–229, 237–255, 188–238 and 227–234. The s.d. and coefficient of variation were calculated for each tissue (n = 3 spots) at each hue range.

Protein-size characterization.

We compared the effects of different processing methods on protein size for lung and cardiac tissues by boiling samples in Laemmli buffer and running them on an SDS-

PAGE gel at 120 V for 1.5 h. Changes in protein size after pepsin digestion were compared for tissue ECM particles from bladder, cartilage, lung, spleen and purified collagen. Undigested tissue was either prepared in Laemmli buffer as described above or digested in 1 mg/mL porcine pepsin (Sigma) in 0.01 M HCl. After 72 h of agitation in pepsin solution, the pH was neutralized and a sample was run on an SDS-PAGE gel to allow comparison of changes in protein size distribution relative to that of control undigested tissue particles.

Proteomics analysis.

Tissue samples were reduced with dithiothreitol, alkylated with iodoacetamide and digested in solution with trypsin (Promega) at 37 °C overnight. Digested peptides were acidified and dried. The peptides were reconstituted in 40 μL of 2% acetonitrile,

0.1% formic acid, and 6 μL were injected (15% of the total volume). Protein identification by liquid chromatography–tandem mass spectrometry (LC-MS/MS) analysis of peptides was performed using a Q-Exactive interfaced with a Thermo Easy-nLC 1000 system

(Thermo Scientific) or a Velos Orbitrap (Thermo Scientific) interfaced with a

NanoACQUITY UPLC system (Waters). Peptides were fractionated by reversed-phase

HPLC on a 75 μm × 12 cm column with a 15-μm emitter tip (New Objective, Woburn,

MA) packed in-house with Magic C18AQ (5 μm, 120 Å, Michrom Bioresources) using a

0–90% acetonitrile, 0.1% formic acid gradient over 90 min at 300 nL/min. Eluting peptides were sprayed directly into the Q-Exactive or Velos at 2.0 kV. Q-Exactive survey scans were acquired from 350–1800 m/z with up to 15 peptide masses (precursor ions) individually isolated with a 2.0-Da window and fragmented (MS/MS) using a collision

41 energy of 27 and 30-s dynamic exclusion. Precursor and fragment ions were analyzed at

70,000 and 17,500 resolution, respectively. Velos survey scans were acquired at 350–

1,800 m/z with up to eight peptide masses (precursor ions) individually isolated with a

1.9-Da window and fragmented (MS/MS) using a collision energy of 35 and 30-s dynamic exclusion. Precursor and fragment ions were analyzed at 30,000 and 15,000 resolution, respectively.

The mass spectrometry–derived data were searched against a combined human and porcine RefSeq protein database (version 65 with common contaminants added) using the SEQUEST HT search algorithm through Proteome Discoverer (version

1.4.1.14, Thermo Scientific). Search parameters included a maximum of one missed trypsin cleavage, cysteine carbamidomethylation as a fixed modification, and methionine oxidation as a variable modification. The precursor mass tolerance was 20 ppm, the fragment mass tolerance was 0.05 Da, and the maximum peptide length was specified as seven amino acids. Peptides that passed the 1% false discovery rate threshold were used for protein identification. Protein inference was based on rules of parsimony as employed by Proteome Discoverer software. The mass spectrometry proteomics data have been deposited to the ProteomeXchange Consortium (76) via the PRIDE partner repository with the data set identifier PXD002571. We obtained gene-level measurements by summing the PSMs for all proteins corresponding to a given gene. We normalized the summed spectral counts in each experiment by dividing by the total acquired tandem mass spectra as previously described6. Small intestine spectra from each experiment were compared for repeatability and averaged before normalization.

The identified genes were then categorized as ECM (subclassified as collagen, proteoglycan or glycoprotein), ECM-associated (subclassified as ECM-affiliated protein,

ECM regulator or secreted factor) or non-ECM as previously described (77). The proteomic composition of processed tissues was compared to that of the native tissues

42 from a recent draft map of the human proteome6 (online resource available at http://www.humanproteomemap.org/). Proteins found in both processed tissues and the corresponding native tissues were normalized between 0 and 1 and summarized in separate heat maps for ECM, ECM-associated and non-ECM components using Gene-E software (http://www.broadinstitute.org/cancer/software/GENE-E/). Hierarchical clustering for each protein was performed across all processed and native tissues in each heat map. The total abundance of ECM and non-ECM components in each tissue was determined from the sum of the normalized PSMs for all proteins in each category.

The variation in proteomic composition between replicate spots in a fabricated microarray was evaluated using three separate aliquots of a representative 3 mg/ml small intestine particle suspension. The total normalized PSMs and the corresponding number of identified proteins in ECM and non-ECM categories were compared for each run. Protein-identification consistency was evaluated for rare proteins (identified by one

PSM) and more abundant proteins (identified by two or more PSMs) across the three runs and defined as follows: No Overlap, identified in one of three samples; 2 of 3, identified in two of three samples; or 3 of 3, identified in all samples.

Cell-seeding methods.

The uniformity of cell distribution on 2D culture plates can be enhanced with a confined chamber. To limit variation in the initial cell-seeding density on each tissue array spot, we designed custom confined flow chambers. The flow chambers were made from polydimethylsiloxane (PDMS) such that the depth of media over the microarray was approximately 1 mm. The tissue microarrays were placed on a tissue culture plate (one well, rectangular; Nunc), and their edges were sealed with Parafilm to stabilize them and prevent media from flowing under the tissue microarray chip. The PDMS flow chamber was pressed over the array to seal the chamber, and cell suspension was slowly injected

43 through the inlet. The locations of tissue spot replicates were randomized in the microarrays so that each type was evenly distributed across rows and columns to minimize systematic error associated with flow patterns during cell seeding

(Supplementary Fig. 2a).

We performed a cell adhesion assay on control microarrays to quantitatively investigate the experimental error associated with cell seeding and washing procedures.

Control microarrays were fabricated in which all spots were collagen controls. Breast cancer cells were seeded at 15,000 cells/cm2 in serum-free media and allowed 1.5 h to attach. The PDMS flow chambers were removed, and the microarrays were washed four times in PBS, incubated in media with 10% FBS for 30 min, stained with calcein AM (3

μg/ml) for 30 min, and imaged. The total number of cells per spot was counted with

ImageJ. The assay was run in duplicate, and the total error was quantified as the coefficient of variation for all spots in the microarray. We analyzed the systematic error by quantifying the average number of cells per spot in each row and each column. The coefficient of variation between rows or columns was calculated for each experiment.

We calculated the final coefficient of variation in rows or columns by averaging the row and column coefficients of variation of duplicate tissue microarray chips.

3D microtissue optimization.

We screened different ratios of cells to tissue ECM weight to optimize parameters for the formation of breast cancer tissue ECM 3D spheroids. Cells were suspended at 500,000 cells/mL and added to tissue ECM suspended at 4–0.25 mg/ml concentrations at a 1:1 (vol/vol) ratio. 40 μL of the cell-tissue suspension was pipetted into a 96-well hanging drop culture plate (InSphero) to form a hanging drop. Seven different tissue particle types (bone, cardiac, kidney, liver, lung, spleen and collagen) were mixed at final concentrations of 0, 0.125, 0.25, 0.5, 1 and 2 mg/ml with a final

44 concentration of 250,000 breast cancer cells/ml in hanging drops. We imaged spheroids with a 2.5× objective on a Zeiss inverted microscope in bright field on days 1, 2, 4 and 6 to capture spheroid-formation kinetics over time. The relative cell viability and/or proliferation at day 6 was compared between all groups with alamarBlue assay (Life

Technologies).

Spheroid microarray fabrication.

We formed the microtissue array by first preparing an array of 1-mm wells in a plastic mold. The wells were filled with water and cleared of air bubbles. Cell-tissue particle spheroids were transferred from a GravityTRAP (InSphero AG) culture dish into the wells of the plastic mold. After tissues had settled to the bottom of the wells, the mold was infiltrated with a 2% agarose solution in water at 70 °C and allowed to cool and form a gel. The agarose diffused into the water-filled chambers and surrounded the microtissues. After cooling, the agarose block was removed, dehydrated and infiltrated with paraffin similarly to the previously described procedure (78). Dehydration was performed with graded ethanol solutions (100 mL: 30%, 50%, 70%, 80%, 95% × 2,

100% × 2) applied for 3 h each, after which 100% ethanol was applied overnight.

Ethanol solutions were cleared with HistoClear II (100 mL) three times for 2 h and once overnight, and the samples were infiltrated with paraffin (100 mL, 60 °C, four times for 2 h) and cast in paraffin.

Macrophage cell culture and morphological analysis.

BMDM progenitors were isolated from the femurs of wild-type C57BL/6 mice and differentiated toward a macrophage phenotype in a mixture of 80% DMEM-F12

(supplemented with 10% FBS and 1% penicillin-streptomycin) and 20% fibroblast (L929) conditioned media (10% FBS, 100 mM L-glutamine and 1% penicillin-streptomycin).

Macrophage progenitors were differentiated for 7 d, with media changed at day 4. The resulting BMDMs were suspended in M1 or M2 macrophage polarization media (2.5 ×

105 cells in 1.5 mL) on the tissue microarrays and the surrounding acrylamide. The M1 polarization medium contained 200 ng/mL lipopolysaccharide (055:B5, Sigma) and 20 ng/mL interferon-γ (PeproTech), and M2 polarization medium contained 20 ng/mL IL-4

(Peprotech). Cells were incubated for 24 h in their respective cytokine environments before being fixed, stained for actin and imaged and analyzed with the Cellomics platform. BMDMs were suspended in unsupplemented macrophage-differentiation media for seeding on tissue spots with different degrees of processing.

Cancer cell culture and adhesion analysis.

Human skin cancer cell line A375 (ATCC, CRL-1619) and human lung cancer cell line A549 (ATCC, CCL-185) were cultured in growth media. Human breast cancer cell line Hs 578T (ATCC, HTB-126) was cultured in growth media supplemented with

0.01 mg/mL bovine insulin. Melanoma-derived mouse cancer cell lines (B16-F0 (ATCC,

CRL-6322), B16-F10 (ATCC, CRL-6475) and B16-KY8 (ref. (79)) and L929 fibroblasts were cultured in growth media. We verified cell line identity by referencing the

International Cell Line Authentication Committee database of cross-contaminated or misidentified cell lines. Cell lines were not tested for mycoplasma contamination.

Cancer cells and L929 fibroblasts were seeded in serum-free media at a final concentration of 15,000 cells/cm2 and allowed to attach for 2 h. Cells were washed gently four times with PBS, incubated in media with 10% serum for 45 min and then washed again with PBS. Cells were stained with calcein AM at 3 μL/mL from 1 mg/mL stock for 30 min at 37 °C and imaged at 50× magnification. The total number of cells per spot was counted with ImageJ. Two chips with different tissue spot patterns

(Supplementary Fig. 2a) were analyzed for each cell type for a total of six spots of each

46 tissue type (n = 6). We quantified cell adhesion for each tissue by normalizing to the solubilized collagen–coated control spots.

Cancer cell proliferation analysis.

Chips were placed in one-well rectangular culture dishes (Nunc) surrounded by

Parafilm and seeded with labeled breast cancer cells via the PDMS gasket method.

Cells were pre-stained in a T-75 flask with CellTracker Green dye CMFDA (5- chloromethylfluorescein diacetate, Invitrogen) according to the manufacturer's instructions. A suspension of dyed cells was seeded onto the chips at a concentration of

5,000 cells/cm2 in 10% FBS media in a confined flow chamber. Cells were allowed to attach overnight. The next day the confined flow chamber was removed and the media was changed. Cells were imaged at 50× magnification on day 1 and day 2 after seeding.

By day 3 the tracker dye's signal had become too weak for segmentation imaging. At day 7 cells were incubated with calcein AM at 3 μg/ml for 30 min at 37 °C and imaged, and the number of viable cells per spot was quantified using ImageJ. Autofluorescence of the liver ECM made it impossible to resolve cell staining at day 2, so these data were not available. Cells formed an overconfluent monolayer and peeled off of the soluble collagen control spot at day 7, so these data also were not available. The cell number over time was used in an exponential growth model to estimate the doubling rate when cells were seeded on different tissue spots.

Final cell number = Initial cell number x 2kt

Cell attachment was confirmed at day 1, and a growth-lag phase was assumed between day 1 and 2. Thus t represents a 5-d time span, the day 2 and day 7 cell counts represent the initial and final cell counts, respectively, and the doubling rate is calculated as k.

Viability analysis of 3D cancer cell and tissue ECM microtissues was conducted with the alamarBlue assay. Relative cell viability was compared between all groups by repeated sampling from wells incubated with alamarBlue reagent (Life Technologies) at days 1, 3 and 6 (n = 3).

hASC culture and analysis.

hASCs were isolated as previously described (80) and passaged at 90% confluence in growth media (GM) with media changes every 3 d. hASC differentiation was induced with osteogenic differentiation media (DMEM, 10% FBS, 1% penicillin- streptomycin, 100 nM dexamethasone, 50 μM ascorbic-acid-2-phosphate, 10 mM β- glycerophosphate).

2D microarray. For culture on microarray chips, hASCs were seeded on tissue microarrays (Nunc) at 6,000 cells/cm2 as described above. hASCs were cultured on the microarrays for 5 d in GM to reach confluence, after which the media was replaced with either osteogenic media or GM and the cells were cultured for an additional 6 d before being fixed and stained with Alizarin Red.

3D microtissues. Microtissues were formed by self-assembly of cells and tissue particles into spheroids in hanging drop culture (81). Tissue particle suspensions were diluted to

0.8 mg/mL in serum-free DMEM culture media and sterilized with UV light (~1.5 mW/cm2 for 20 min). hASCs were suspended in GM at 850,000 cells/mL with a 1:1 mixture of

ECM particle suspension for microtissue formation as described above. We changed the microtissue medium with GM every 2 d by replacing half of the hanging drop volume with fresh media twice. Microtissues were cultured for 6 d in GM and then transferred to 96- well GravityTRAP (InSphero) plates with control GM or osteogenic media with media

48 changes every 3 d. After 11 d of culture, microtissues were fixed, embedded in well molds as described above, sectioned and stained with Masson's trichrome or Alizarin

Red.

Macrophage and hASC PCR.

Changes in mouse macrophage and hASC gene expression after culture with tissue ECM were determined in 2D and 3D microenvironments, respectively.

2D macrophage microarrays. Bone, cardiac, liver, lung and spleen tissues were processed using a modified procedure. Tissues were mechanically fragmented using a knife-mill processor (Retsch, Germany) into particles with sizes no larger than 10 mm 3 and rinsed with thoroughly distilled water until blood was fully cleared from the samples.

Bone samples were decalcified by incubation in 10% formic acid (Sigma) for 3 d and verified by a colorimetric calcium test (STANBIO Laboratory). All tissues were then incubated in 3.0% peracetic acid (Sigma) with agitation at 37 °C for 4 h and moved to fresh solution 1 h after the end of the incubation period. The pH was adjusted to 7 by thorough, extensive rinsing with water and PBS. Samples were transferred to a 1%

Triton X-100 (Sigma) + 2 mM sodium EDTA (Sigma) solution and incubated with agitation at room temperature for 3 d, with the solution replaced by fresh solution daily.

Tissues were rinsed thoroughly with distilled water and incubated in 600 U/ml DNase I

(Roche Diagnostics) + 10 mM MgCl2 (Sigma) + 10% antifungal-antimycotic (Gibco) at

37 °C for 24 h. Tissues were rinsed thoroughly with distilled water, frozen at −80 °C and lyophilized. Tissue particles were prepared from lyophilized samples as described above. Tissue particle suspensions in distilled water were adjusted to 4–5 mg/ml for coating of six-well plates with 1 ml of solution followed by air-drying. After drying, plates were sterilized under UV light for 1 h and then rinsed to remove any nonadhered

49 particles before being seeded with macrophages. Macrophages were cultured as described above in M1, M2 or nonpolarizing M0 media conditions for 24 h (n = 3).

Macrophage RNA was extracted for PCR analysis using TRIzol reagent (Life

Technologies), and RNA was purified using RNeasy Mini columns (Qiagen). cDNA was synthesized through the use of SuperScript Reverse Transcriptase III (Life

Technologies) per the manufacturer's instructions. We carried out RT-PCR on an

Applied Biosystems Real Time PCR Machine using SYBR Green (Life Technologies) as a reporter. Macrophage polarization was evaluated on the basis of the expression of genes associated with M1 polarization (Il1b, Nos2 and Tnf) and M2 polarization (Arg1,

Retnla (Fizz1) and Il10). Expression was calculated relative to nonpolarized M0 macrophages cultured in uncoated control wells using B2m as the reference gene via the 2−ΔΔCt method and the following primers: B2m forward, CTC GGT GAC CCT GGT

CTT TC; B2m reverse, GGA TTT CAA TGT GAG GCG GG; Tnf forward, GTC CAT TCC

TGA GTT CTG; Tnf reverse, GAA AGG TCT GAA GGT AGG; Il1b forward, GTA TGG

GCT GGA CTG TTT C; Il1b reverse, GCT GTC TGC TCA TTC ACG; Nos2 forward,

GAC GAG ACG GAT AGG CAG AG; Nos2 reverse, GTG GGG TTG TTG CTG AAC TT;

Arg1 forward, CAG AAG AAT GGA AGA GTC AG; Arg1 reverse, CAG ATA TGC AGG

GAG TCA CC; Retnla forward, CTT TCC TGA GAT TCT GCC CCA G; Retnla reverse,

CAC AAG CAC ACC CAG TAG CA; Il10 forward, TCT CAC CCA GGG AAT TCA AA;

Il10 reverse, AAG TGA TGC CCC AGG CA (Integrated DNA Technologies). The fold change in expression for each polarizing media condition was used to create a heat map with hierarchical clustering by tissue ECM coating.

3D hASC microtissues. Osteogenic differentiation was determined in bone and lung microtissues using quantitative RT-PCR. Microtissues were formed as described above for 7 d in GM followed by an additional 14 d in osteogenic media. A total of three

50 microspheres per treatment group were pooled and homogenized for RNA extraction using TRIzol reagent per the manufacturer's instructions. RNA was then reverse- transcribed to cDNA using Superscript III reverse transcriptase (Life Technologies) according to the manufacturer's instructions. Real-time PCR was carried out with a

StepOnePlus Real-time PCR system (Applied Biosystems). We calculated Bglap gene expression at day 21 for each tissue relative to that at day 7 using Gapdh as the reference gene via the 2−ΔΔCt method and the following primers: Bglap forward, CCT

CAC ACT CCT CGC CCT AT; Bglap reverse, CTT GGA CAC AAA GGC TGC AC;

Gapdh forward, AGG AGC GAG ATC CCT CCA AA; Gapdh reverse, AAA TGA GCC

CCA GCC TTC TC.

Histology.

Microarray chips were fixed in 4% paraformaldehyde before staining procedures.

3D microtissues were fixed in paraformaldehyde for 1 h before being embedded in microarrays as described above. Paraffin blocks containing microtissue arrays were sectioned at 5 μm and stained as described below.

Actin and nuclear staining. Cells were permeabilized in 1% Triton X-100 and stained for actin with Alexa Fluor 647–conjugated phalloidin (Life Technologies) at a 1:25 dilution in

0.1% BSA (Sigma) in PBS for 30 min at 37 °C. Nuclei were stained with DAPI (100 ng/mL, Life Technologies) for 30 min.

In vitro macrophage immunofluorescence. Macrophages seeded on tissue arrays in M1 or M2 polarizing cytokines as described above were evaluated for expression of the M2 and M1 markers Arg-1 and iNOS. After 24 h, macrophages were fixed for 10 min in 4% paraformaldehyde and permeabilized with 0.1% Triton X-100 in Tris-buffered saline

(TBS) for 10 min. Nonspecific protein interactions were blocked with 1% BSA, 2% horse serum in 0.05% Tween-20. Tissue microarrays were then incubated overnight at 4 °C with primary antibodies to Arg-1 (rabbit polyclonal, diluted 1:100; GTX109242, GeneTex) and iNOS (mouse monoclonal (4E5), diluted 1:300; ab129372, Abcam) diluted in 1%

BSA in TBS. Arrays were washed and probed with secondary FITC-conjugated goat anti-rabbit IgG (diluted 1:250; 111-095-003, Jackson ImmunoResearch) and Alexa Fluor

568–conjugated goat anti-mouse IgG (diluted 1:250; A-11004, Invitrogen) in 1% BSA for

1 h. ECM autofluorescence was then blocked by treatment with 0.1% Sudan Black B in

70% ethanol for 20 min. Arrays were washed, coverslipped and imaged. We corrected variations in background autofluorescence by subtracting the mean background intensity of acellular regions of each tissue array spot using ImageJ software.

Subcutaneous tissue ECM implantation and immunofluorescence. Animal experiments were conducted in accordance with guidelines set by the Johns Hopkins University

Animal Care and Use Committee. Bone and collagen tissue ECM particulate was hydrated with saline (100 mg dry wt/0.2 ml) and injected subcutaneously on the dorsum of 6–8-week-old female C57BL/6 mice. The mice were killed after 1 week, and implants were explanted, fixed in formalin and embedded in paraffin for sectioning. Sections were then deparaffinized, rehydrated and immunolabeled for the pan-macrophage marker

F4/80 and the M1 marker iNOS. Antigen retrieval was conducted in citrate (10 mM, pH

6) for 30 min in a vegetable steamer, and samples were then rinsed with TBS in 0.05%

Tween-20. Sections were blocked with 1% BSA, 2% goat serum and 0.05% Tween-20 for 1 h. Sections were then incubated overnight at 4 °C with primary antibodies to F4/80

(rat monoclonal (BM8), diluted 1:100; ab16911, Abcam) and iNOS (mouse monoclonal

(4E5), diluted 1:200; ab129372, Abcam) in blocking solution. Sections were washed and probed with FITC-conjugated goat anti-rabbit IgG (diluted 1:250; 111-095-003, Jackson

ImmunoResearch) and Alexa Fluor 594–conjugated goat anti-rat IgG (diluted 1:250; A-

11007, Invitrogen) secondary antibodies in blocking solution for 1.5 h. ECM autofluorescence was then blocked by treatment with 0.1% Sudan Black B in 70% ethanol for 20 min. Sections were counterstained with DAPI for 5 min, washed, coverslipped and imaged.

Alizarin Red staining. Staining of calcified matrix was performed on microarray chips with

Alizarin Red (40 mM, pH 4.1) for 25 min. Chips were rinsed with water briefly three times and then a fourth time for 5 min before rapid dehydration and clearing in acetone, acetone:xylene (50:50) and xylene, followed by coverslipping. Sectioned slides containing microtissue arrays were deparaffinized; rehydrated; stained with Alizarin Red for 5 min; washed; dehydrated and cleared rapidly with acetone, 1:1 acetone-xylene, xylene; and coverslipped.

Electron microscopy.

Dry tissue microarrays were placed in a desiccator overnight to dry completely, attached to aluminum stubs by carbon sticky tabs (Ted Pella) and coated with 20 nm of

AuPd with a Denton Vacuum Desk III sputter coater (Denton Vacuum). Stubs were viewed and digital images were captured at a 60° tilt on a LEO 1530 field emission scanning electron microscope operating at 1 kV.

Slide scanning and analysis.

Cell counts and morphology on microarray chips were imaged and quantified using a Cellomics high-content scanner (Cellomics) with ArrayScan VTI software

(Thermo Fisher). Microarray chips were adhered to four-well rectangular culture plates

(Nunc). Cell counts were quantified from 50×-magnified images (two fields per spot) with

53 calcein AM staining. Morphological quantification was performed after phalloidin staining as described above. The object-identification parameters were adjusted to outline the perimeters of individual cells in each image that were subsequently counted and analyzed for morphology using the convex hull area ratio. 2D microarrays stained with

Alizarin Red and histological slides containing 3D spheroid microarrays were scanned with a ScanScope AT (Aperio) at 200× magnification and viewed with ImageScope software (Aperio).

Image processing.

Cell counts for adhesion studies were analyzed from 50×-magnified images using

ImageJ (US National Institutes of Health, Bethesda, MD). The cell number for each spot

(n = 6) was normalized to that of control spots (soluble collagen) on the same microarray. We analyzed color images of 2D and 3D Alizarin Red–stained microarrays with ImageJ to quantify the percent area of Alizarin Red staining. Regions of interest

(ROIs) were traced around the perimeter of each individual sample in the array for measurements of total area. Microarrays were threshold color-filtered in the RGB color space to allow selection of areas of positive red staining, and the filtered image was converted to binary. The percent area of positive staining was calculated using the

“measure” command in the ROI manager.

Mechanical testing.

The elastic modulus and hardness of dried tissue spots were measured using a

Nano Indentor XP (MTS Systems) and Nanosuite software, which supports measurements of complex, thin materials. Samples were approached with a velocity of

10 nm/s, and samples were indented with a Berkovich tip to a depth of 400 nm at a strain rate of 0.05 with a peak hold time of 10 s and 90% unload. To minimize the

54 influence of the underlying glass coverslip on array mechanics, we increased the sample thickness by double spotting. We spotted 10 μL of particle suspension and dried it as described previously, and then we spotted and dried an additional 10 μL of particle suspension on top of the original spot. Because each test took several hours, measurements were made on samples in a dried state to avoid confounding factors of evaporation. Spots of six different tissues (bladder, brain, cardiac, liver, lung and spleen) were indented in 8–10 different locations per sample. Outlier measurements were excluded from analysis as determined by Grubbs' test (one measurement each was excluded for bladder, brain, lung and spleen) as described under “Statistical Analysis.”

Proteomics network analysis.

The abundance of extracellular proteins (identified at the gene level) in each tissue was correlated with in vitro assay results. In some cases, assay results were not available for all tissues, resulting in fewer than 11 tissues for a given assay. We calculated correlation coefficients for each of the 3,879 identified proteins in the 15 in vitro assays by using the assay results directly and replacing each assay result with its rank-ordered value. The P value and t-statistic for each protein-assay correlation was calculated from the linear model for both the original and rank-transformed assay data.

We visualized the assay and protein results as a heat map showing the correlation coefficients for all genes and assays (Cluster 3.0) and by creating an assay-protein network (Cytoscape v3.2.1). In images, proteins and assays are displayed as vertices, and edges connect proteins to assays when they are strongly correlated (P < 0.05 for both the original assay and the rank-transformed assay results). The full results of these analyses are provided in Supplementary Table 2.

We conducted gene-set enrichment analysis to determine whether defined groups of proteins improved the predictive value for each assay. This was done for

55 bone, cartilage, lung and spleen tissues using the Biological Processes Gene Ontology database. ECM proteins from each of these tissues were evaluated for enrichment terms related to osteogenesis, and the corresponding P values for enrichment were compared.

Terms related to osteogenesis included skeletal system development (GO:0001501), ossification (GO:0001503), bone morphogenesis (GO:0060349), osteoblast differentiation (GO:0001649) and chondrocyte differentiation (GO:0002062). Identified

ECM proteins annotated to these processes are listed in Supplementary Table 3.

Statistical analysis.

Statistical significance for quantitative in vitro cell assays was determined by a one-way analysis of variance with post hoc Tukey's test (Graphpad Prism v6, GraphPad

Software). The effect of different processing methods on macrophage morphology was determined by Student's t-test with Bonferroni correction. Statistical significance was defined as P < 0.05. Differences between tissue spots are represented as tissue symbols above each bar in figures (representing the mean ± s.d.) or included in

Supplementary Table 1. Statistical outliers for nano-indentation measurements were observed from Tukey box plots as values more than 1.5 times the interquartile range of the quartile borders and confirmed via Grubbs' outlier test using α = 0.05 (Graphpad

Prism v6). These outlier measurements were excluded from analyses of tissue spot elastic modulus and hardness. Power analysis was not conducted to determine sample size, and investigators were not blinded.

Accession codes.

The mass spectrometry proteomics data have been deposited to the

ProteomeXchange Consortium via the PRIDE partner repository with the data set identifier PXD002571.

Results

Fabrication of 2D and 3D tissue ECM arrays

We established a physiologically broad tissue ECM data set by harvesting 11 different porcine tissues and organs (spleen, small intestine, bladder, bone, brain, cartilage, heart, kidney, liver, lung and adipose); chemically treating the tissues using a combination (unless noted otherwise) of acid, detergent and DNase (41); and then lyophilizing them to mechanically break them down into microparticles using a cryomill.

The tissue ECM particles contained numerous intact proteins reminiscent of the original tissue (Supplementary Fig. 1), a feature that is lost with alternative methods of enzymatic digestion.

Array spots (3-mm diameter) consisted of tissue matrix particles spotted on acrylamide-coated glass (Fig. 1a) with a layer of purified soluble type I collagen to promote homogeneous particle binding, as confirmed by scanning electron microscopy

(Supplementary Fig. 1b). Collagen controls included acid-soluble type I collagen protein without the addition of tissue ECM particles and insoluble collagen particles that match the structure of processed tissue ECM particles. We prepared up to 40 spots of tissue

ECM per array (Fig. 1b and Supplementary Fig. 2a) and then seeded cells uniformly on the array using a confined flow device (Supplementary Fig. 2b), which limited cell spot- seeding variability to 13% of the mean per row and column (n = 10 rows, n = 4 columns per 40-spot array) (Supplementary Fig. 2c). Spot locations and seeding direction were organized so that replicates of each tissue were distributed evenly throughout the seeding area to limit the effects of flow patterns in the seeding device (Supplementary

Fig. 2a).

3D systems provide an additional means to evaluate and predict cell-tissue interactions in an environment and may allow for better prediction of some in vivo

57 behaviors. With this in mind, we also developed hanging droplet arrays of 3D tissue

ECM spheroids in which each spheroid contained 10,000–20,000 cells and ECM particles at a concentration of 0–10 ng per cell in 40 μl of culture medium. Spontaneous cell-matrix assembly resulted in the formation of large agglomerations after 24 h in culture, with continued self-assembly over the course of 2–6 d (Fig. 1c). Tissue particle- to-cell ratios were optimized to maximize tissue ECM content without disruption of compact spheroid formation or cell viability (Supplementary Fig. 3a–c). Compact spheroid formation and microtissue size were consistent across all ECM types tested at concentrations of 2 ng per cell or less after 6 d of culture. Cells were viable at ECM concentrations of up to 2 ng per cell but decreased at higher particle concentrations for some tissues.

To form 3D spheroids, we seeded human adipose-derived stem cells (hASCs) with ~17,000 cells and 16 μg of tissue particles to create compact spheroids that had a uniform diameter of ~460 μm (±40 μm; n = 8) despite having different tissue particle compositions (purified type I collagen particles, bone, brain, cartilage, adipose, lung or spleen). To enable high-throughput morphological, histological and immunohistochemical analyses of the 3D microtissue arrays, we developed a method similar to the tissue microarray technology used in tumor pathology (82). We covered cell-tissue spheroids that were arranged in a microarray mold with agarose gel to fix the location of the microtissues and make possible sectioning of the spheroids. Microtissue cross-sections showed a relatively uniform distribution of cells and tissue particles throughout each spheroid (Fig. 1d).

Characterization of tissue ECM arrays

We then characterized the physical and biochemical properties of the tissue- specific ECMs to understand tissue-specific attributes and to confirm the reproducibility

58 of the array fabrication.

Physical properties:

After complete drying, spot surfaces for brain, bladder and small intestine were generally smooth, whereas those for other tissues, such as spleen, bone and liver, showed more variation in roughness and texture (Supplementary Fig. 1b). The elastic moduli of the dry tissue spots were two orders of magnitude higher and demonstrated less variation between tissues types than stiffness values reported for fresh hydrated tissue (83) (Supplementary Fig. 1c). For example, after being processed into particles and spotted in 2D, cardiac tissue spots were twofold stiffer than brain spots; by comparison, there is an estimated ~20-fold difference in stiffness between fresh heart and brain tissues (83).

Reproducibility.

Masson's trichrome stain was used to contrast the collagen composition between tissue spots. Spots derived from connective tissues, cartilage, bone and bladder contained qualitatively more collagen than did spots from solid organs such as liver and muscle. We observed differences in collagen composition between ECMs from different tissue sources, as well as between healthy and diseased tissues (cancer and muscular dystrophy), as ECM changes in composition and structure often occur during pathogenesis (Supplementary Fig. 4). The variability in the staining intensity of the most prominent color in tissue spot replicates remained below 25% (coefficient of variation; n

= 3 spots per tissue type) for 10 of the 11 tissue types analyzed with ImageJ

(Supplementary Fig. 5a,b). We confirmed the reproducibility of tissue particle composition using mass spectrometry and proteomic analysis, comparing replicates of small intestine (Supplementary Fig. 5c) with similar peptide abundances in triplicate

59 runs. Approximately 95% of proteins identified by at least two peptide-spectrum matches

(PSMs) were found in all small intestine samples, whereas only 42% of less abundant proteins (one PSM) were identified across all three runs (Supplementary Fig. 5d). Taken together, the results showed low tissue interspot variability and high tissue specificity in tissue array chips.

We demonstrated broad utility of our tissue array with a range of tissue- preparation methods by exposing cardiac and lung tissues to a range of chemical processing conditions that produced ECMs of various molecular weights (Supplementary

Fig. 6a). Spots from the same tissues processed with different methods were clearly differentiated by their collagen content (Supplementary Fig. 6b). Using an automated algorithm, we also looked for any changes in the morphology of primary mouse bone marrow–derived macrophages (BMDMs) seeded on the variably processed tissue ECMs

(Supplementary Fig. 7a). Lung and cardiac tissues (and combinations of the two) that underwent different types of preparation produced distinct cell morphologies

(Supplementary Fig. 7b,c), with changes in the processing of lung tissue yielding the greatest effect on cell morphology (Supplementary Fig. 7d). Thus, the arrays can be used with multiple tissue-processing methods, and, importantly, variable cell responses to the arrays can capture differences in tissue processing and thus in the tissues' resulting compositions.

Proteomic analysis.

The tissue particles used to fabricate the 2D and 3D arrays retained a complex protein composition that was representative of the native tissues from which they were derived. Nearly 4,000 unique proteins were identified across the 11 processed porcine tissue types, of which 111 were ECM and 98 were ECM associated. To understand the effect of processing on tissue composition in the primary arrays, we compared several

60 processed tissues from this study to native tissues from a recently published draft map of the human proteome6. Processed tissue particles showed strong similarities in ECM- specific proteins compared to the native tissues; however, processing reduced the protein diversity of non-ECM proteins (Fig. 2a). The total abundance of ECM and non-

ECM proteins varied by tissue type, although at similar ratios, with the exception of bone and cartilage, which had lower amounts of non-ECM proteins (Fig. 2b).

Subclassification of ECM and ECM-associated proteins in tissue particles revealed tissue-specific differences. Proteoglycans were enriched in adipose, cartilage and brain tissues, whereas amounts of secreted factors (including growth factors) were miniscule in most tissues except for brain, where they accounted for 8% of ECM total protein (Fig. 2c). Collagen subtypes accounted for the largest ECM fraction in all tissues except for brain and heart and represented more than 50% of the ECM in bladder, bone and cartilage (Fig. 2c). Detailed analysis of collagen composition showed the expected tissue specificity (Fig. 2d); for example, type X collagen is associated with endochondral ossification and was identified only in bone and cartilage.

Human stem cell response to 2D and 3D tissue ECM arrays

We used both 2D and 3D tissue arrays to study stem cell interactions with different ECMs. We seeded hASCs on 2D tissue ECM arrays or cultured them in 3D cell-

ECM spheroids, in both instances in either control or osteogenic medium. The percent area of calcified matrix on each spot or microtissue represented osteogenic differentiation. hASCs cultured on bone, bladder and lung 2D tissue ECMs demonstrated nearly 100% matrix deposition, whereas other tissue ECMs (adipose and collagen control) induced minimal osteogenic differentiation after only 6 d of culture (Fig. 3a–c). In

3D cell-ECM spheroids, osteogenic medium induced stem cell differentiation

61 predominantly in bone and lung, with minimal matrix production observed in cartilage, adipose tissue and spleen after 11 d of culture (Fig. 3d,e).

Overall osteogenic differentiation was similar in the 2D and 3D tissue arrays

(Spearman's rank coefficient, 0.79 (P < 0.05)), but the amount of calcified matrix present in the 3D cellular microtissues was less than that in the 2D arrays (Fig. 3f). As expected, bone tissue matrix supported and stimulated osteogenesis. However, newly formed calcified matrix was also abundant on lung ECM in both array systems, so differentiation was confirmed in bone and lung spheroids by increased expression of the gene osteocalcin (Bglap), a late-stage marker of osteogenesis (Supplementary Fig. 8a).

Cancer and immune cell behavior on tissue ECM arrays

We then demonstrated that the arrays are broadly applicable to cancer and immune cell types and their relevant biological outputs. For proof of concept, we screened human cancer cell lines in 2D and 3D to measure cell adhesion and proliferation, and we screened mouse primary macrophages on the 2D tissue ECM arrays to quantify cell morphology and phenotype changes.

Cancer cell adhesion and proliferation.

The tissue microenvironment has a major role in tumor progression and metastasis (84), and tissue-specific ECMs could provide insight into how the local environment affects tumor growth or metastatic preferences. To this end, we screened the adhesion of three different human cancer cell lines (lung, breast and skin) after 2 h on 2D ECM microarrays. Cell adhesion on different tissue matrices varied by up to tenfold for all cell types (Fig. 4a,b and Supplementary Table 1), but notably the cancer cells exhibited minimal adhesion to kidney- and liver-derived ECMs, whereas they

62 adhered to all of the mesenchymal tissues (adipose, bone, cardiac and cartilage) except spleen.

We next compared human breast cancer cell proliferation on 2D and 3D tissue

ECMs. In 2D, bone ECM suppressed cell proliferation, whereas cardiac tissue amplified proliferation (Fig. 4c and Supplementary Fig. 8b). However, proliferation in 3D cell-ECM spheroids over 7 d was similar in bone, kidney and cardiac ECMs (Supplementary Fig.

8c). We also compared the adhesion of three types of mouse B16 melanoma cells to evaluate tissue preferences of primary and metastatic tumor lines: the B16-F0 parent cell line; the B16-F10 line, associated with lung metastasis; and the B16-KY8 line, associated with liver metastasis. F0 cells demonstrated the highest overall attachment, and KY8 the lowest, as expected from their known in vivo behavior. Cartilage and spleen tissues promoted adhesion of all three melanoma cell types; conversely, liver, kidney and collagen control did not support the attachment of melanoma cells, which was similar to the response of human breast cancer cells. The most substantial difference between cell lines was observed for attachment to adipose tissue. Cluster analysis showed similarities in tissue adhesion between the two metastatic melanoma lines (F10 and KY8) in comparison to the primary cells, despite the fact that the metastatic lines have distinct metastatic profiles (Fig. 4d and Supplementary Table 1).

Immune-cell morphology and polarization.

Macrophages have a key role in normal tissue repair and response to disease.

For example, the tissue regenerative response has been linked to macrophage skewing toward a remodeling phenotype called M2; conversely, the M2 phenotype is associated with negative outcomes in a tumor environment (16). Recent studies have also demonstrated the importance of the local tissue environment on macrophage phenotype

(58). We used our 2D tissue array method to probe M2 and M1 (proinflammatory)

63 macrophage responses to specific tissue microenvironments (Fig. 4e). We skewed mouse BMDMs toward an M1 phenotype (interferon-γ + Escherichia coli lipopolysaccharide) or an M2 phenotype (interleukin 4 (IL-4)) and then quantified cell morphology, as a surrogate marker of immune polarization, on tissue ECMs.

Macrophage shape (convex hull area ratio) varied by up to 40% across the tissue matrix types in both cytokine environments, suggesting that tissue-specific matrix environments can modulate cell response (Fig. 4e,f and Supplementary Fig. 9). Macrophages changed morphology dramatically when cultured on bone tissue ECM, with M1-stimulated cells elongated and spread compared with the more compact M2-stimulated cells (Fig. 4e,f).

We further investigated macrophage polarization on the 2D tissue arrays by immunostaining M1 and M2 cytokine-stimulated BMDMs for the M1 marker inducible nitric oxide synthase (iNOS, also known as NOS2) and the M2 marker arginase 1 (Arg-

1) on both bone ECM and control collagen (insoluble particles) spots. Macrophages on the bone ECM spots showed increased Arg-1 staining in M2 conditions and increased iNOS staining in M1 conditions compared with collagen, suggesting enhanced polarization and responsiveness to either M1 or M2 conditions (Supplementary Fig.

10a). To provide a correlation between in vitro and in vivo results and demonstrate the predictive value of in vitro results for in vivo responses, we implanted bone ECM and control collagen matrices subcutaneously in C57BL/6 mice. Both scaffolds showed substantial mononuclear cell infiltration after implantation, and many of these cells were

F4/80+ macrophages expressing iNOS (Supplementary Fig. 10b). Similar to the increased macrophage polarization of the bone ECM observed in vitro, more macrophages expressed iNOS in the bone ECM implants than in collagen in vivo.

Lastly, we evaluated the effect of 2D tissue substrates on macrophage polarization by examining the expression of M1- and M2-associated genes (Supplementary Fig. 10c).

Macrophage gene expression was dependent on the source tissue and cytokine

64 stimulation, and we noted a number of statistically significant differences between tissue

ECM sources (Supplementary Table 1). Bone ECM induced extensive expression of M1- associated genes in M0 cytokine conditions (Il1b, P = 0.0193; Nos2, P = 0.0009), whereas liver did not alter phenotype relative to uncoated wells. In contrast, all tissues resulted in decreased expression of M2-associated genes in M2 cytokine conditions (P <

0.05) and did not substantially alter the expression of M1-associated genes in M1 conditions, although cardiac tissue increased the expression of Il10 in M2 conditions (P

< 0.0001).

Systems analysis of tissue composition and outcomes

An advantage of the tissue arrays is that they enable the user to perform multiplexed assays with a variety of cell types and matrix compositions. For example, to identify potential mediators of in vitro biologic function, we integrated the proteomic composition of tissue array spots with the outputs of the various in vitro assays. A heat map representation of correlation coefficients showed that different ECM and ECM- related proteins clustered with in vitro osteogenesis, cancer cell adhesion and macrophage polarization—the three outputs we investigated in this study (Fig. 5a). We selected the most statistically significant correlations (n = 28 proteins with P < 0.05 for both original and rank-transformed data; Supplementary Table 2) for network analysis to further probe common proteins that might influence cell behavior. Several ECM components, including S100A9, SERPINB10, CTSB, HAPLN3 and PRSS2, were strongly linked with multiple in vitro assay outcomes (Fig. 5b). The S100A family, for instance, correlated with both osteogenic and immunomodulatory assay outcomes.

Some of the more abundant ECM proteins, such as collagens, could not be used to distinguish different in vitro outcomes between tissues. Conversely, ECM glycoproteins showed correlation with several biological assay outputs in the network analysis—

65 specifically, positive correlations to cancer cell adhesion and osteogenesis and negative correlations to macrophage morphology.

To further investigate the unexpected stem cell osteogenesis that we observed on lung tissue in 2D and 3D cultures (Fig. 3), we probed the ECM and ECM-associated proteomic composition using gene ontology (GO) enrichment, specifically, terms related to skeletal development or bone formation. Lung ECM components showed significant enrichment for these terms in lung compared to spleen (Fig. 5c), including ossification

(GO:0001503) and bone morphogenesis (GO:0060349), suggesting that factors were retained in lung after processing that individually or in combination might have promoted osteogenesis (Supplementary Table 3). Cartilage was also enriched in these pro- osteogenic factors to an extent similar to that of bone, although it did not show any in vitro osteogenic potential, indicating that GO enrichment analysis alone is not sufficient to explain this behavior, or that there are inhibitors of osteogenesis present in processed cartilage. Correlation networks and GO analysis revealed individual molecules and groups of molecules, respectively, in the complex tissue ECMs.

Discussion

Microarrays designed to recreate the ECM use a reductionist approach that incorporates select components to modulate cell behavior (85, 86). Strategies for studying single or paired protein combinations have been unable to mimic the complex composition of healthy and diseased tissues. Tissue-derived scaffolds that maintain natural tissue complexity have been under investigation for more than 25 years (87) and, during that time, have been implanted in more than 1 million patients (88). However, the biochemical diversity of tissue- and organ-derived matrices renders it difficult to fully understand the material composition, specific interactions with cells and therapeutic action. Our mechanically processed ECMs retain many of the inherent physical and

66 biochemical cues. Although tissues are processed with a wide range of mechanical, chemical and/or enzymatic treatments (34), it may not be possible to completely remove all cell-associated components from a tissue matrix. These cell-associated proteins, both intracellular and secreted, may be relevant for the biological and even therapeutic outcomes of these materials, and should therefore be preserved in in vitro models.

Tissue ECMs from different sources exhibited relatively minor differences in mechanical and structural properties but large differences in composition, which is likely to be the primary influence in cell response.

The effects of specific tissues on stem cell osteogenesis were consistent in both

2D and 3D cultures, although the 2D assays showed an estimated 50-fold higher ratio of tissue matrix to cells. Bone tissue ECM enhanced osteogenesis, which was expected because demineralized bone materials are used clinically to stimulate bone healing.

However, lung tissue ECM unexpectedly enhanced osteogenesis compared with most other tissues. These unexpected results may lead to further study and understanding of lung tumor calcification (89). GO analysis showed that processed lung ECM was enriched for components associated with osteogenesis and chondrogenesis relative to other tissues, including ECM proteins such as type II and type XII collagen, matrilins, thrombospondin 1 and asporin (Supplementary Table 3).

Correlating in vitro assay outcomes with in vivo behavior will be an important link between our method and biology, with the hope that specific components of the tissue matrix can be linked to cell-signaling pathways. The extensive network analysis we performed to correlate proteomics and in vitro functional outcomes is ideal for identifying individual ECM candidates and for discovering potentially overlapping mechanisms across different cell processes. Conversely, GO enrichment is well suited for deriving complex combinations of ECM components for further investigation using previously defined annotations. The two analysis methods are complementary, and both are easily

67 applied to data sets generated by these in vitro arrays to discover drivers of the in vivo biological responses. Of particular interest was the connection between immune phenotype and osteogenic potential based on shared correlation of these outcomes to the S100A family of proteins. Consistent with our findings, S100A8 and S100A9 are damage-associated molecular pattern proteins that exert immunoregulatory functions via interaction with the receptor for advanced glycation end products, in addition to activating osteogenic gene programs in vivo (90). Whereas the dual immunoregulatory and osteogenic functions of S100A proteins have been partially characterized, other protein-function correlations are not as obvious. Application of ECM correlation networks to the cancer cells showed that both metastatic melanoma lines, but not the parent, were positively correlated with PRSS2, a serine protease that was enriched in bladder, spleen and brain tissue arrays. PRSS2 and other enzymes have long been associated with cancer and promote metastasis by cleaving cell-adhesion molecules and ECM barriers.

However, our data suggest that PRSS2 has a direct effect on metastatic cancer adhesion and might promote initial metastasis to these tissues through mechanisms such as shedding of other membrane-bound growth factors or enzymes (76). Although these data do not define mechanisms of tissue ECM modulation of processes such as osteogenic differentiation, cancer metastasis and immune modulation, the array screening and resulting analysis were able to establish correlations and provide promising candidate molecules for further investigation—either isolated components with novel, noncanonical functions or complex combinations of proteins that are not otherwise apparent.

The 2D and 3D arrays both present advantages and may complement each other; 2D microarrays are compatible with high-throughput microscope slide scanning and are simple to culture, whereas 3D microarrays better represent the natural microenvironment, do not have sample cross-talk, can be sectioned multiple times and

68 can be used with varying ratios of cells to tissue matrix to measure dose dependence.

Three-dimensional microtissues are also compatible with metabolic assays, single-cell analysis or sorting, and gene expression analysis. Each method, however, offers the unprecedented ability to probe mechanisms of cell interactions with complex tissue microenvironments, support the discovery of new therapeutic targets and applications and ultimately transform the therapeutic potential of ECM-based therapeutics.

Figure 2.1: 2D and 3D tissue and organ model ECM arrays.

(a) Microarrays were defined by a silicone gasket placed on acrylamide-coated coverslips. Each well was spotted with a particle suspension that was dried to form circular tissue substrates on the chip before the gasket was removed. (b) Masson's trichrome, which stains collagen blue, highlights the variation in biochemical composition of 11 different tissues and two different collagen controls spotted in three replicates on a microarray chip. Image is representative of n = 3 chips. (c) 3D microtissues formed in hanging drop culture from cell–tissue particle suspensions. Cells and tissue particles settled at the bottom of the droplet, where they subsequently self-assembled into a cell- tissue spheroid. (d) Cells and tissue particles distributed throughout histological sections of spheroid microtissues stained with Masson's trichrome (day 17; red denotes cells, and blue denotes collagen). Scale bars, 1 mm. Images are representative of n = 3 spheroids.

Figure 2.2: Tissue ECM protein composition and analysis.

(a) Results of mass spectrometry and proteomics analysis comparing processed tissues to native human tissues from a draft map of the human proteome. Hierarchical clustering analysis was conducted for proteins mapped to ECM and ECM-associated genes. The protein abundance for each gene was normalized (0–1) across all tissues. (b) The total abundance of ECM, ECM-associated and non-ECM proteins was determined from total normalized PSMs mapped to gene-level annotation; results from a single mass spectrometry run are shown. (c) The relative proportions of different categories of ECM and ECM-associated proteins in different tissues. SI, small intestine. (d) Tissue-specific distribution of fibrillar, fibril-associated, sheet-forming and other types of collagen as identified by proteomic analysis of processed tissues.

Figure 2.3: Stem cell–tissue ECM interactions in 2D and 3D arrays.

(a) hASCs stained with calcein AM (green, left and right) or Alizarin Red (middle) were cultured on 2D tissue microarray substrates for 6 d in osteogenic media (OM) or control or growth media (CM). Images are representative of n = 3 microarrays. (b) Osteogenic differentiation, as quantified by the percent area of each spot that stained positively with

Alizarin Red. Data are mean and s.d. (n = 9 spots from three microarrays). (c) Chips incubated in OM and CM. Images are representative of n = 3. (d,e) hASCs were cultured with the indicated types of tissue particles in 3D microtissue spheroids for 11 d in OM or

CM. (d) Histological sections of individual spheroids (images representative of n = 5 spheroids) stained with Alizarin Red. (e) Quantification of the percent area stained by

Alizarin Red. Data are mean and s.d. (n = 5 spheroids). (f) Normalized staining comparison between tissue types in 2D and 3D culture conditions. Spearman's rank correlation coefficient: 0.79 (P < 0.05). One-way analysis of variance with a post hoc

Tukey test was used for statistical analysis (b,e). *P < 0.05. Bl, bladder; Bo, bone; Br, brain; Ct, cartilage; Col, purified collagen I particles; Ad, adipose; Cdc, cardiac; Kd,

72 kidney; Lv, liver; Lg, lung; Sp, spleen; SI, small intestine; Ctl, control soluble collagen only; All, all other tissues.

Figure 2.4: Cancer cell and macrophage interactions with tissue microarrays.

(a) Adhesion of three different human cancer cell lines and a control mouse fibroblast line to different tissue ECM types. Cells (black) were stained with calcein AM and counted with ImageJ software (representative of n = 6 array spots). (b) Cell number normalized to the average number of cells on soluble collagen control spots located on the same microarray. Data are mean and s.d. (n = 6 array spots). (c) The breast cancer cell doubling rate calculated from the cell number after 1, 2 and 7 d of culture on each tissue microarray spot. Data are mean and s.d. (n = 9 array spots). (d) Heat map and hierarchical clustering of normalized (0–1) B16 mouse melanoma cell adhesion between

F0 (parent) and metastatic KY8 (liver) and F10 (lung) cell lines seeded on a tissue array

(n = 6 tissue array spots and n = 4 soluble collagen control spots). (e,f) The immunological properties of different tissue array spots as screened by mouse BMDMs chemically polarized toward M1 or M2 phenotypes. (e) Cell morphology visualized after

24 h with actin-phalloidin staining (n = 3 array spots; scale bar, 200 μm). (f) Cell

74 morphology quantified as the convex hull area ratio. LPS, lipopolysaccharide; IFN-γ, interferon-γ. Data are mean and s.d. (n = 3 array spots). *P < 0.05; abbreviations for tissue types defined as in Figure 3.

Figure 2.5: Systems biology analysis of tissue proteomic composition and in vitro function.

(a) Correlation between proteomics data and cell responses to 11 porcine tissues across

15 in vitro assays. Hierarchical clustering and heat map of the correlation coefficients between ECM proteins and each in vitro assay outcome (a portion of the full heat map is shown; the complete data set is available in Supplementary Table 2). Clusters of positive

(yellow) and negative (blue) correlations showed association of osteogenic assays with

M2 macrophage morphology and association of cancer cell adhesion with M1 macrophage morphology. (b) Network analysis showing the statistically significant correlations (edges) between in vitro outcomes and ECM proteins (vertices). Statistically significant correlations were defined as those with P < 0.05 for both the original assay and the rank-transformed assay results. Mac, macrophage; adh, adhesion; osteo, osteogenesis; mel, melanoma; fibro, fibroblast. (c) GO enrichment was performed on

ECM proteins in tissues that promoted in vitro osteogenesis (bone and lung) compared with those that did not (cartilage and spleen). All identified ECM and ECM-associated proteins were entered into enrichment analysis for biological-process GO, and terms

76 related to osteogenesis or skeletal development were selected with corresponding P values (a list of ECM proteins identified for each GO term is provided in Supplementary

Table 3).

Supplementary Figure 2.1: Characterization of tissue ECM spots.

(a) SDS-PAGE gel of intact tissue particles and pepsin digested particles (representative of n = 2 technical replicates). Qualitative protein composition of different tissue ECM types before and after pepsin processing is visualized as dark bands and regions with decreasing protein size from top to bottom. High molecular weight bands visible in intact tissue particles are not present after pepsin digestion (arrows). (b) Topographical features of array spots were examined using scanning electron microscopy and showed structural differences between tissue types (representative of 3 images from 1 array spot). Scale bar, 10 μm. (c) Nanoindentation testing was performed to measure the surface mechanical properties of the dried particle spots on tissue arrays. The elastic modulus and hardness are shown for 6 different tissue sources (8-10 technical replicate

78 measurements were conducted per tissue, and measurements defined as outliers by

Grubbs’ test were excluded; n = 9 bladder, n = 8 brain, n = 9 cardiac, n = 9 liver, n = 7 lung, n = 7 spleen). Data are mean ± s.d. Mechanical properties were analyzed by one- way ANOVA with a post-hoc Tukey test. Statistically significant increases (P < 0.05) are signified with (*) and the abbreviations of the tissues they are greater than: Bl=bladder,

Br-brain, Cdc=cardiac, Lv=liver, Lg=lung, Sp=spleen.

Supplementary Figure 2.2: Tissue array cell-seeding method optimization.

(a) Tissue microarrays were designed to randomize the location of each type of tissue spot to limit error associated with cell seeding. When the two spot patterns shown are used in duplicate (n = 6 array spots per tissue type), each spot type is distributed between the inner and outer columns and inlet and outlet rows. (b) A polydimethylsiloxane (PDMS) compartment was designed to confine medium flow and promote uniform cell seeding. (c) Variability associated with cell seeding in PDMS compartments on control microarray chips (all spots soluble collagen control) was less

80 than 15% between rows and columns (coefficient of variation, n = 10 rows, n = 4 columns, n = 2 array chips). Data are mean.

Supplementary Figure 2.3: Optimization of cell-tissue ECM for formation of breast cancer–tissue ECM 3D spheroids.

Cells were suspended at 500,000 cells/mL and added to tissue suspended at 0.25–4 mg/ml at a 1:1 v/v ratio in a 40 μL hanging drop. (a) Tissue particles at high concentrations could disrupt spheroid formation, but all tissues screened were compatible with formation of compact cell-tissue spheroids at tissue particle concentrations of 1 mg/ml or less after 6 days of culture (representative images of n = 3 array spots for 1 mg/ml & 0.5 mg/ml or n = 2 for 4 mg/ml, 2 mg/ml & 0.25 mg/ml). (b)

Spheroid formation began with the development of several small spheroids that coalesced to form a single compact spheroid after 6 days with a tissue concentration of

1 mg/ml (representative images of n = 3 array spots). (c) The viability/proliferation of cells after 6 days in culture increased with increasing tissue concentration up to 1 mg/ml

(n = 3 array spots for 1 mg/ml & 0.5 mg/ml or n = 2 for 4 mg/ml, 2 mg/ml & 0.25 mg/ml).

At concentrations of 2-4 mg/ml cell viability/proliferation decreased for several tissue types. Data are mean ± s.d. Scale bars = 1 mm.

Supplementary Figure 2.4: Array spots generated from diseased tissues.

Diseased tissue arrays were fabricated by processing tissue taken from murine mammary tumors and skeletal muscle of mutant dystrophic mice. Tissue was processed similar to normal tissue and spotted in a microarray with normal tissue controls.

Histological staining patterns with hematoxylin and eosin (H&E) and Masson’s trichrome are different for normal mammary, early stage tumor, and late stage tumor tissue.

Increased blue staining with Masson’s trichrome indicates higher levels of collagen content (n = 3 array spots). Scale bar = 20 μm.

Supplementary Figure 2.5: Fidelity of tissue processing, array spotting and proteomic analysis.

(a) Images of Masson’s Trichrome stained tissue spots (representative of n = 3 array spots) were compared to determine the uniformity of tissue array spots and the variability associated with array fabrication. (b) Staining characteristics were quantified by calculating different hue levels in ImageJ, and demonstrated that variability remained below 25% for the predominant color of all tissues except brain, kidney, and small intestine (coefficient of variation, n = 3 array spots). Data are mean ± s.d. (c) Proteomic comparison of n = 3 different spot suspensions of processed small intestine tissue confirmed consistency in the abundance (PSMs) and number of total unique ECM and non-ECM proteins identified. (d) Proteomics analysis confirmed most proteins were identified in at least two of three different small intestine spot suspensions with greater

85 consistency for higher abundance proteins (greater than 1 PSM) than rare proteins (1

PSM).

Supplementary Figure 2.6: Characterization and protein composition of tissue arrays prepared with multiple processing methods.

Porcine lung and cardiac tissues were treated using various degrees of chemical processing; mild (10 min peracetic acid & 18 hr DNAse), moderate (3 hr peracetic acid,

18 hr Triton X-100 & 18 hr DNAse), harsh (moderate treatment + 16 hr SDS), and digest

(moderate treatment + pepsin digestion). (a) The protein composition of tissue ECM prepared with different methods of processing show distinct banding patterns on an

SDS-PAGE gel (representative of n = 2 technical replicates). (b) Decellularized particles of each type were spotted onto arrays. Cardiac and lung tissue array spots were stained with Masson’s trichrome to show compositional variability following different degrees of processing. An example of combining particulate mixtures of moderately processed lung and cardiac ECM at different ratios (wt/wt) is also provided (representative images of n =

3 array spots for each condition).

Supplementary Figure 2.7: Comparison of macrophage morphology on arrays prepared with different processing methods.

(a) Macrophage morphology on each tissue ECM formulation was segmented using an automated algorithm (cell perimeter/segmentation outlined in blue). (b) Cell spreading on mixtures of moderately processed lung:cardiac tissue and (c) on single tissue spots with different degrees of processing (representative images of n = 3 array spots). (d)

Quantification of cell morphology by the convex hull perimeter to area ratio from segmented images of cells cultured on fresh, digested, and treated cardiac and lung tissue. Data are mean ± s.d. Scale bars = 200 μm. Morphology for each tissue type was

88 analyzed by a student’s t-test with the Bonferroni correction. Statistically significant increases (P < 0.05) between different processing methods within a tissue are signified with (*) and the label/abbreviations of the tissues they are greater than: fresh, mild, mod=moderate, harsh, or digest. Ratios of lung to cardiac tissue particles are signified as 1:3, 3:1, and 1:1.

Supplementary Figure 2.8: Cellular differentiation and viability in 3D spheroids.

(a) Osteogenic gene expression of hASCs cultured in 3D bone and lung tissue ECM microtissues. hASC microtissue spheroids were formed in growth media (GM) for 7 days, and either kept in GM or switched to osteogenic media (OM) for an additional 14 days. Expression of the late stage osteogenesis gene osteocalcin was determined via qRT-PCR relative to the day 7 time point (3 technical replicates of n = 3 pooled spheroids) in both GM (top) and OM (bottom) media conditions. Data are means ± s.d.

Differences in osteocalcin expression between 7 and 21 days for each condition were determined by a Student’s t-test (*P < 0.05). (b) Breast cancer cells labeled with Cell

Tracker or calcein AM were fluorescently imaged on 2D tissue arrays over 7 days to quantify cell number. Cardiac tissue supported the greatest number of cells after 7 days, while bone and kidney tissues had the fewest (n = 9 array spots). Data not available for

90 day 7 soluble collagen control (cells formed a monolayer and peeled from spot during culture) or day 3 liver (autofluorescence tissue too high for Cell Tracker imaging). (c) Cell viability and proliferation as assessed by Alamar Blue assay for 3D tissue spheroids decreased for all tissue types over 7 days but was greatest for cartilage tissue containing spheroids (n = 3 spheroids). Data are mean ± s.d.

Supplementary Figure 2.9: Attachment of bone marrow–derived macrophages to different tissue substrates on array.

Cells were polarized toward an M1 or M2 phenotype with cytokine stimulation, allowed to attach for 24 h, and fixed. Actin cytoskeleton is stained with phalloidin (representative image of n = 3 array spots). Scale bars, 200 μm.

Supplementary Figure 2.10: In vitro and in vivo macrophage polarization in response to tissue ECM.

(a) Macrophage phenotype on bone and collagen array spots after 24 hours in polarizing conditions was determined by immunofluorescence staining for Arginase-1 (green) and iNOS (red) (representative of n=3 array spots, scale bar=100 μm). (b) Bone tissue ECM and collagen were hydrated and injected subcutaneously into wild type C57BL/6 mice.

Implants were harvested after one week for immunofluorescence staining for iNOS

(green) and F4/80 (red) (representative of 4 images from n=2 animals). Scale bars=100

μm. Heat map representation of the effect of 2D tissue ECM substrates on macrophage gene expression. Primary mouse bone marrow derived macrophages were cultured on

2D bone, cardiac, liver, lung, or spleen ECM and compared to uncoated polystyrene in

M0 (non-polarized), M1 (LPS + IFNγ), and M2 (IL4) cytokine environments for 24 hours.

Relative gene expression of IL1β, iNOS TNFα, Arg1, Fizz1, and IL10 was determined via qRT-PCR, and is displayed as ΔΔCt over the tissue culture plastic (TCP) control in each media condition (triplicate wells of n = 3 macrophage isolations). Hierarchical

93 clustering analysis was conducted for each cytokine environment. Macrophage gene expression was analyzed by one-way ANOVA with a post-hoc Tukey test, and these statistical results are provided in Supplementary Table 1.

Supplementary Table 2.1: One-way ANOVA results for cancer cells and macrophages on 2D tissue arrays.

* indicates p < 0.05

Breast cancer adhesion Bladder Bone Brain Collagen Cartilage Control Adipose Cardiac Kidney Lung Liver Sm IntestineSpleen mean 1.18195 1.18195 1.18195 1.18195 1.18195 1.18195 1.18195 1.18195 1.18195 1.18195 1.18195 1.18195 1.18195 Bladder - Bone - Brain - Collagen - Cartilage * - Control * * * - Adipose - 0.69439 Cardiac * - Kidney * - Lung - Liver * - Sm Intestine * - Spleen * * * * * * -

L929 adhesion Bladder Bone Brain Collagen Cartilage Control Adipose Cardiac Kidney Lung Liver Sm IntestinSpe leen mean 0.41394 0.83166 0.30046 0.34049 0.3212 1 0.86579 0.76685 0.07262 0.52287 0.08569 0.50858 0.61998 Bladder - Bone - Brain - Collagen - Cartilage - Control * * * * - Adipose - Cardiac - Kidney * * * * - Lung - Liver * * * * - Sm Intestine - Spleen -

Lung cancer adhesion Bladder Bone Brain Collagen Cartilage Control Adipose Cardiac Kidney Lung Liver Sm IntestinSpe leen mean 0.5457 0.23045 0.33862 0.43446 1.03648 1 0.54486 0.78924 0.02601 0.23571 0.04172 0.31413 0.4469 Bladder - Bone - Brain - Collagen - Cartilage * * - Control * * - Adipose - Cardiac - Kidney * * * - Lung * * - Liver * * * - Sm Intestine * * - Spleen -

Skin cancer adhesion Bladder Bone Brain Collagen Cartilage Control Adipose Cardiac Kidney Lung Liver Sm IntestinSpe leen mean 0.31729 0.33859 0.1621 0.26245 1.04817 1 0.5583 0.56303 0.0136 0.0875 0.01908 0.20251 0.37097 Bladder - Bone - Brain - Collagen - Cartilage * * * * - Control * * * * - Adipose * - Cardiac * - Kidney * * * * - Lung * * * * - Liver * * * * - Sm Intestine * * - Spleen * * -

B16-F0 cancer adhesion Bladder Bone Brain Control Cartilage Collagen Adipose Cardiac Kidney Liver Lung Spleen Sm Intestine mean 0.6463 0.87037 1.45 1 0.86296 0.38148 3.48148 0.63519 0.24074 0.55185 1.17407 0.92778 0.79259 Bladder - Bone - Brain - Control - Cartilage - Collagen - Adipose * * * * * * - Cardiac * - Kidney * - Liver * - Lung * - Spleen * - Sm Intestine * -

B16-F10 cancer adhesion Bladder Bone Brain Control Cartilage Collagen Adipose Cardiac Kidney Liver Lung Spleen Sm Intestine mean 0.79413 0.0758 0.566 1 1.16973 0.16634 0.02196 0.36365 0.27785 0.08704 0.40651 0.93929 0.51612 Bladder - Bone * - Brain - Control * - Cartilage * - Collagen * * - Adipose * * * - Cardiac * - Kidney * - Liver * * * - Lung * - Spleen * * * * * - Sm Intestine -

B16-KY8 cancer adhesion Bladder Bone Brain Control Cartilage Collagen Adipose Cardiac Kidney Liver Lung Spleen Sm Intestine mean 0.56121 0.1023 0.49463 1 1.05524 0.03971 0.06488 0.28813 0.25537 0.02517 0.30129 0.50284 0.11155 Bladder - Bone - Brain - Control - Cartilage * - Collagen * - Adipose * - Cardiac - Kidney - Liver * - Lung - Spleen - Sm Intestine * -

Supplemental Table 2.1 continued

M0 (unstimulated) Tnfα IL1β iNOS Arg1 Fizz1 IL10 Control vs. Bone * * Control vs. Cardiac * * Control vs. Liver Control vs. Lung * * Control vs. Spleen * * * Bone vs. Cardiac * * * Bone vs. Liver * * Bone vs. Lung * * * Bone vs. Spleen * * * * Cardiac vs. Liver * * Cardiac vs. Lung Cardiac vs. Spleen * Liver vs. Lung * Liver vs. Spleen * * Lung vs. Spleen *

M1 (LPS + IFNγ) Tnfα IL1β iNOS Arg1 Fizz1 IL10 Control vs. Bone * Control vs. Cardiac * * * * Control vs. Liver * Control vs. Lung * * * Control vs. Spleen * * Bone vs. Cardiac * * * * Bone vs. Liver Bone vs. Lung * * * Bone vs. Spleen * * Cardiac vs. Liver * * * * * Cardiac vs. Lung * * * Cardiac vs. Spleen * * * Liver vs. Lung * * Liver vs. Spleen * Lung vs. Spleen *

M2 (IL-4) Tnfα IL1β iNOS Arg1 Fizz1 IL10 Control vs. Bone * * * * Control vs. Cardiac * * * * * Control vs. Liver * * * Control vs. Lung * * Control vs. Spleen * * Bone vs. Cardiac * * * * * Bone vs. Liver * * * * Bone vs. Lung * * * * Bone vs. Spleen * * * * * Cardiac vs. Liver * * * * * Cardiac vs. Lung * * * Cardiac vs. Spleen * * * * Liver vs. Lung * * * Liver vs. Spleen * * * * Lung vs. Spleen *

Supplemental Table 2.2: Results of correlation analysis between ECM protein quantity and in vitro assay outcomes.

This table can be found online at: http://www.nature.com/nmeth/journal/v12/n12/extref/nmeth.3619-S3.xlsx

Supplemental Table 2.3: Osteogenesis gene ontology enrichment analysis results for tissue ECMs.

1 n 2

I A s e 2 G B e e L F P l n S O G p e S T H C g

n o 1 2 1 e i g 1 t A A A 1 3 ) 2 a a 1 1 I 7 A i i 2 N N l N 1 1 2 2 G t B 6 T T t L L L L A F P n r 0 A A S C e a O O O O G 2 r C A C M C T H C M C 0 e 0 f f 0 i : 1 2 d

O 1 A A 3 e 2 1 1 A G t N ( N e 1 1 2 G y T L L L c A n P A S C o o O O O r B A C C H M C d n o h 1 1 3 C 2 A I N N 2 G B g T T L F P n A A S O G u L M T H M C

1 1 n A A e 6 1 s e L L C l e O O N p n S C T C e g

n e o g 1 1 i ) t a A A i 9 a l 6 1 i 4 t L L t C r 6 n a O O N 1 e C C T C 0 r 0 e f 0 f : i O 1 1 d

A A G t ( s e 6 1 1 P a L L C n P l S o O O N P b B B C S T C I o e t s O 1 1 A A N 6 1 g L L C A n C O O N u L C V T C ) 9

1 1 4 n 2 1 A A 3 e S 1 2 G 0 e L L B P l 6 S O O H p 0 S H T C C 0 : O 1 1 e G ( g 1 1 1 A A 1

2 a s 3 7 A 1 A A 0 P i N l e S S 2 9 1 2 1 G T t M L L L L L n B B P r A e S a O O O O O O H H g C M C T H C C T C C C

s i s 1 e 1 1 1 n A 2 e A A A 0 1 P g S e 9 1 2 1 G M L L L L o B n P S h o O O O O O H p B C H C T C C C r o m

1 1 1 1 e 2 A A A 1 P N n S 9 1 2 G g T o M L L L B P n A B S O O O O H u L M C H C T C C ) 3

1 2 1 1 0 n 2 A A A A 5 e 6 5 1 2 1 G e L L L L C 0 P l S 0 O O O O N p 0 S C H C T C C : O G 1 2 1 e (

1 g 1 2 1 1 A A A

2 a C 1 A 1 A A A 0 3 C

l S R 1 6 1 5 1 2 1 G R P

t L L L L L L L

C B P r H A

S O T O O O O O O a

H N P

C C C T C C H M S T C C C C

1 2 1 s 1 1 2 1 1 A A A e 2 C 1 A 1 A A A 0 C n R e 1 6 1 5 1 2 1 G R 1 P e P L L L L L L L C n P H A P g G S S o O T O O O O O O N P P B n B C C C C H M S S T C C C I C o i t a c 1 1 2 1 i f 2 A A A A i N s 1 2 5 6 G N g s L L L L C A P P n S S C O O O O O N u L A C C C C H T V 1

1 2 2 1 1 A

n 2 A I 2 A 1 A A A s P e 1 S 3 1 1 5 1 2 G e B L e L L L L L L F B P M E n l N S e O O O O O O R G H p B U g S C L F T H C C T C P C C

t n e 1 2 1 1 1 e 3 1 g 1 2 1 2 2 1 1 A A A A A 1 3 m ) 2 a N N 1 A A 1 2 7 A A A 0 A A 3 I 1 C P i P 1 p N N D l L L N S S 1 3 9 1 1 2 9 1 5 1 1 2 G B R L P T T 0 o t L L P P M L L L L L L L L L L A l F A B B P M E r A 5 A G A e A A S C a O O O O O O O O O H O O O O R H G H P 1 U v C C A L C M H H T C C C H M T S C C C M T C P C C C C C 0 e 0 d 0

: 1 2 1 1 m 1 O 1 1 2 2 1 1 A A A A 3 e 2 N 1 A 1 2 A A A A A 0 G 1 C P t P N ( D L N S e s 1 3 1 1 9 1 5 1 2 1 G R L P T L L P M L L L L L L L L y A A B n P M E A G A s A S C o O O O O O O O H O O O O R H P U l B L C A C H C C H M S C C M T C P C C C C C a t e l 1 e 1 1 2 1 1 2 1 A 1 3 k 2 N 2 A A A A A A N I 1 P S P N N L N 1 S T 1 1 1 2 3 5 9 G B L g T T P L L L L L L L M S F B A P M E N n A A A S C O O O O O O O O O R G H B u U L C C C C C C C C F H H L M M P P T T V

The Scaffold Immune Microenvironment: Biomaterial-Mediated Immune

Polarization in Traumatic and Non-Traumatic Applications

This work has not been previously published and is currently in preparation.

Authors: Kaitlyn Sadtler, Brian Allen, Kenneth Estrellas, Franck Housseau, Drew Pardoll and Jennifer Elisseeff

Abstract

Immune responses to biomaterials were historically associated with a negative outcome and a foreign body response. The goal of early materials selection was geared towards immune ignorance of the biomaterial. As the field evolved, biomaterials are now being designed to interact with surrounding environment including stem cells and vascular-related cells. It has also been appreciated that immune cells may play a positive role in materials response due to their critical role in wound healing and tissue regeneration and should thus be considered. Current approaches to immune analysis of scaffolds should focus on the immune microenvironment created by the scaffold. Here, we present a methodology to quantitatively evaluate immune cell populations in tissue- derived extracellular matrix (ECM) scaffolds in non-traumatic subcutaneous and traumatic muscle-injury models. We detected innate and adaptive immune cells, more specifically macrophages, dendritic cells, polymorphonuclear cells (PMNs), T cells and B cells. We further define a specific scaffold-associated macrophage (SAM), which relies on signals from the adaptive immune system. Building on this work, it would be possible to characterize in detail the scaffold immune microenvironment (SIM) of a given biomaterial to determine the effect of scaffold changes on immune response and subsequent regenerative potential.

Introduction

The immune system is divided into two large classes, the innate and adaptive immune systems. The innate system, composed of macrophages, dendritic cells, and polymorphonuclear cells (PMNs, neutrophils, basophils and eosinophils) is the first responder to infection, tissue damage or biomaterial implantation. These cells scavenge debris and provide the first barrier to any disruption in homeostasis. They interact with local tissue and foreign antigens to educate the adaptive arm of the immune system.

Adpative immune cells, T lymphocytes and B lymphocytes, are presented pathogen or damage-specific antigens by innate immune cells (mainly dendritic cells and to a lesser extent macrophages) which induces a highly specific response catered to the antigen and tissue location (2, 43). B cells secrete immunoglobulin antibodies, which target pathogens for destruction and activate other immune pathways. T cells are divided into two main groups, CD4+ helper T cells and CD8+ cytotoxic T lymphocytes. CD4+ T cells receive signals from extracellular antigen sources and secrete cytokines that act upon surrounding cells, creating a specific immune environment based on their secretome.

CD8+ T cells receive signals from intracellular antigen sources (mainly viral) and target the antigen presenting cell for destruction through perforin and granzyme secretion. The innate and adaptive immune systems work together to quickly and efficiently form a specific response to the disruption in homeostasis ultimately clearing the pathogen or healing the wounded tissue. As guardians of body homeostasis, these cells are the first to interact with implanted biomaterials and greatly determine their regenerative potential

(19, 20, 91, 92).

Interactions of biomaterials with a host were historically viewed in the context of biocompatibility, with a focus on the foreign body response (27-29). This immune

100 response was viewed as a destructive phenomenon that ultimately ended in either fibrous capsule formation or inflammation and scaffold degradation. Application of a biomaterial, in the presence or absence of trauma, will initiate recruitment of immune cells. The immune system surveys the body for disruptions in homeostasis, such as introduction of synthetic polymers or biologic scaffolds. Therefore, many studies have involved the protection of biomaterials from the immune response and decreasing recognition of the materials as foreign to create a more regulatory response that didn’t lead to inflammation or fibrosis (93). The concept of avoiding immune-mediated pathologies is relevant and important to materials design; however, ignorance of the immune system to the material is not necessarily ideal.

The immune system can impact many processes in the body ranging from embryonic development to wound healing. There are cells of the immune system spread throughout the body that can relay a signal of tissue damage or material implantation. In the case of injury, immune cells are critical in wound healing and tissue regeneration.

Previously, researchers have noted the importance of macrophages in complex tissue regeneration in amphibians (13), and muscle regeneration in higher organisms.

Eosinophils, macrophages and T cells have been implicated in regenerative processes

(21, 22, 94). Additionally, macrophages are important in the outcome of biologic scaffold remodeling, specifically tissue-derived extracellular matrix (ECM) scaffolds (16, 18).

These scaffolds recruit M2, or anti-inflammatory macrophages, as opposed to M1 inflammatory macrophages. Biochemically, macrophages, among other cells, secrete enzymes responsible for ECM degradation and remodeling. Without these cells present, old matrix cannot be degraded and therefore new matrix cannot be deposited in its place. This is especially important in tissue engineering, as the host must be able to successfully incorporate and remodel the scaffold into a functional replacement tissue.

101

Analysis of the complex dynamics of immune cell recruitment and polarization in response to biomaterial scaffolds has been limited partially due to a lack of experience in intense immune analytic techniques. Here, we present an analysis of cellular recruitment and polarization to ECM scaffolds in non-traumatic (subcutaneous) and traumatic

(muscle injury) models through the use of flow cytometry. Using flow cytometry to analyze the scaffold immune microenvironment (SIM) revealed the presence of a specific set of scaffold-associated macrophages (SAMs). Additionally, we observed an adaptive immune-dependent MHCII and CD11c upregulation by scaffolds in the volumetric muscle injury model. Through further understanding of flow cytometric analyses of the SIM it would be possible to thoroughly characterize the immune infiltrate of a biomaterial scaffold and promote a more rational scaffold design.

Methods

Tissue Decellularization & Plate Preparation

Tissues were derived from porcine (Wagner Meats) processed following a standard protocol. Firstly, tissue was formed into a paste through use of a knife-mill processor

(Retsch, Germany) to form pieces no larger than 2 mm2 to ensure thorough diffusion of reagents used in processing. Resultant material rinsed thoroughly with running distilled water until blood was cleared from samples and the water ran clear. Bone samples were decalcified in 10% formic acid (Sigma) for 3 days, and tested for any residual calcium with a colorimetric calcium test (STANBIO). Tissues were then incubated in 3.0% peracetic acid (PAA; Sigma) on a shaker at 37oC for 4 hours, after the first hour, the solution was switched to new PAA and continued for the next 3 hours. Tissues were rinsed with phosphate buffered saline (PBS, Life Technologies) until reaching pH 7

102 according to colorimetric pH paper. After pH was equilibrated, samples were incubated for 3 days in 1% Triton-X100 (Sigma) + 2 mM Sodium EDTA solution on a shaker or stir plate at 400 rpm, changing the solution daily. Material was rinsed thoroughly with distilled water to remove any residual detergent. Finally, they were transferred to 600

U/ml DNase I (Roche Diagnostics) + 10 mM MgCl2 + 10% Antifungal-Antimycotic

(Gibco®) for 24 hours at room temperature. Tissues were rinsed 3 times with distilled water, then frozen at -80oC and lyophilized for 2-3 days. The dry sample was turned into a particulate form using SPEX SamplePrep Freezer/Mill. For in vitro studies, these particles were retrieved, rehydrated, and sonicated (GE 130PB, Cole-Parmer, Vernon

Hills, IL) before being passed through a 100 μm cell strainer (BD Falcon) to ensure similar particle size across all tissue types. Concentrations were adjusted to 4-5 mg/ml in distilled water for use in plate coating. 6-well plates were coated with 1ml of 5 mg/ml

ECM solution and allowed to air dry for 24 hours. After drying plates were sterilized under UV light for 1 hour on each side, then rinsed with PBS directly before cell culture to remove any non-adhered particles. ECM powder was stored between -20oC and -

80oC.

Cell Culture

Murine immortalized bone marrow macrophages (iBMM, Michele de Palma, Ecole

Polytechnique Federale Lausanne) were cultured as per developer’s protocol in IMDM

(Gibco®) media containing 20% FBS (Hyclone, GE Healthcare Life Sciences), L-

Glutamine (Gibco®), PenStrep (Life Technologies), and 50ng/ml M-CSF (Recombinant

Mouse, BioLegend). iBMM’s were cultured on plates coated with decellularized ECM powder for 24 hours in growth media.

103

Fluorescence Microscopy

Coverslips were functionalized by acid etching in piranha solution (3:1 30% Hydrogen

Peroxide: Concentrated H2SO4) then amidated using 3-(aminopropyl) trimethoxy-silane

(Sigma). ECM was coated as per previously described and cells were allowed to adhere for 24 hours prior to fixation and permeabilization with 3% Paraformaldehyde (Electron

Microscopy Sciences) and Trition-X100 (Sigma). NFkB signaling was done by staining with DAPI, Phalloidin-AlexaFluor647 and p65 rabbit anti-mouse (Cell Signaling

Technologies), followed by an anti-rabbit AlexaFluor594 secondary (Life Technologies).

Samples were imaged on a Zeiss Axiocam a2 upright microscope. Nuclear translocation was measured through JACoP analysis in ImageJ image analysis software (NIH), reported as a correlation coefficient between DAPI and AlexaFluor594 pixels after converting to 8-bit images. Positive control sample was stimulated with LPS (200ng/ml) and IFNγ (20 ng/ml).

RT-PCR

RNA was isolated by a combined TRIzol (Life Technologies)-RNeasy column (Qiagen) approach and converted into cDNA with SuperScriptRTIII (Life Technologies) using manufacturers instructions. cDNA was loaded into a 96 well plate and ran on a Applied

Biosystems Step One Plus Real-Time PCR machine using a SYBR Green Reporter dye and the following primers: : B2m forward CTC GGT GAC CCT GGT CTT TC, B2m reverse GGA TTT CAA TGT GAG GCG GG; Tnfα forward GTC CAT TCC TGA GTT

CTG, Tnfα reverse GAA AGG TCT GAA GGT AGG; Il1β forward GTA TGG GCT GGA

CTG TTT C, Il1β reverse GCT GTC TGC TCA TTC ACG; Arg1 forward CAG AAG AAT

GGA AGA GTC AG, Arg1 reverse CAG ATA TGC AGG GAG TCA CC; Retnla forward

CTT TCC TGA GAT TCT GCC CCA G, Retnla reverse CAC AAG CAC ACC CAG TAG

CA; Il10 forward TCT CAC CCA GGG AAT TCA AA, Il10 reverse AAG TGA TGC CCC

104

AGG CA; Inos forward GAC GAG ACG GAT AGG CAG AG, Inos reverse GTG GGG

TTG TTG CTG AAC TT.

Subcutaneous ECM Implantation

ECM was hydrated with 1X PBS to form implants of 0.3 mg/ml. 0.2 ml of the ECM paste was injected twice subcutaneously into female 6.5 week old wild type C57BL/6 mice

Volumetric Muscle Loss Surgery

6-8 week old female C57BL/6 or B6.129S7.Rag1tm1Mom/J (Rag-/-) were induced under

4.0% isoflurane and maintained at 2 – 3.0% isoflurane and 2 % O2 during the surgery.

Both hind limbs were cleared of hair using an electric razor. Skin was sterilized with 70% ethanol prior to surgery. A 1cm incision was created above the quadriceps muscle using surgical scissors. The underlying fascia was also cut and the fat pad was pushed toward the hip joint to reveal the quadriceps muscle group. Using surgical scissors, a 3mm x

4mm portion of muscle was removed and back-filled with either saline or 50 ul of a 200 –

300 mg/ml particulate biomaterial. The skin incision was closed with surgical staples and the procedure was repeated on the contralateral leg. Prior to removal from anesthesia, animals were given 5 mg/kg Rimadyl for pain management. For the duration of the study animals were maintained under antibiotic feed (275 ppm Trimethoprim and 1365 ppm

Sulfadiazine, Harlan Laboratories). At day 7 post-injury the scaffold and surrounding

105 muscle was harvested for FACS analysis by running a scalpel blade from the knee along the femur to the hip joint and dicing the muscle and scaffold into small pieces before proceeding with enzyme digestion. All animal procedures were done in accordance with the Johns Hopkins University ACUC guidelines.

Histology

Implants were fixed overnight in 10% formalin prior to dehydration and paraffin embedding. 5 μm sections were rehydrated then prepared for immunohistochemistry

(IHC) or direct staining with hematoxylin and eosin (Sigma-Aldrich). IHC samples were treated with a citrate antigen retrieval buffer, 10 mM sodium citrate (J.T. Baker) + 0.05%

Tween20 (Sigma-Aldrich) at pH 6, for 30 minutes in a vegetable steamer. Sections were stained with primary antibodies against CD11b, CD3, Neutrophil Elastase, or CD11c

(AbCam) overnight at 4oC then visualized using SuperPictureTMPolymer Detection Kit,

HRP-DAB (Life Technologies). These samples were then counterstained with hematoxylin (Sigma-Aldrich).

Flow Cytometry

Subcutaneous ECM implants were harvested then finely diced using a scalpel in

1XPBS. Resultant material was digested for 45 minutes at 37oC in 5 mg/ml Liberase TL

(Roche Diagnostics) + 0.2 mg/ml DNase I (Roche Diagnostics) in serum-free RPMI

(Gibco). Digest was filtered through a 100 μm cell strainer then washed twice with

1XPBS + 0.5 mM EDTA and once with 1XPBS. Cells were resuspended in 5 ml 1XPBS and carefully layered atop 5 ml Lympholyte-M (Cedarlane), then spun for 20 minutes at

1200 xg. Cellular interphase was washed twice with 1XPBS then transferred to a 96-well plate for antibody staining. Isolated cells were stained with the following antibody panel:

Viability Aqua (Life Technologies), CD19 BrilliantViolet 421 (BioLegend), CD3

106

AlexaFluor 488 (BioLegend), CD11c APC-Cy7 (BD Biosciences), F4/80 PE-Cy7

(Biolegend), CD86 AlexaFluor700 (BioLegend), CD206 APC (BioLegend). After staining cells were fixed and analyzed as per previously described. Viability Aqua negative (live) cells were evaluated based off of percent population of T Cells (CD3+), B Cells (CD19+),

Dendritic Cells (CD11c+), Progenitor Cells (CD34+) and Macrophages (F4/80+). As with in vitro studies, macrophages were further analyzed for polarization by mean fluorescence intensity of F4/80+ cells in CD86 AlexaFluor700 and CD206 APC channels.

All analysis was performed in FlowJo Flow Cytometry Analysis Software (Treestar). T cell analysis was performed using the following panel: CD3 AlexaFluor488 (Biolegend),

CD4 PE/Cy7 (Biolegend), CD8 AlexaFluor700 (Biolegend), FoxP3 Pacific Blue

(Biolegend) and Fixable Viability Dye eFluor780 (eBioscience). Myeloid compartment analysis in the volumetric muscle wound at 1 week post-injury was done with the following antibody panel: Fixable Viability Dye eFluor®780 (eBioscience), F4/80 PE-Cy7

(BioLegend), CD11b AlexaFluor700 (BioLegend), CD11c APC (BioLegend), Ly6C

Per/CP-Cy5.5 (BioLegend), Ly6G PacificBlue (BioLegend), CD86 BrilliantViolet510

(BioLegend), CD206 PE (BioLegend), MHCII I-A/I-E AlexaFluor488 (BioLegend).

Statistical Analysis

ANOVA and Student’s T-tests were performed using Prism GraphPad software at P ≤

0.05.

Results

Subcutaneous implantation of ECM derived from various tissue sources

To model the immune response to ECM scaffolds in a non-traumatic setting, we injected 0.2 cc of a 300 mg/ml ECM scaffold subcutaneously in wild type C57BL/6 mice.

107

After 1 and 3 weeks the implants were harvested for analysis via histology. In all tissue sources tested (Bone, Cardiac, Liver, Lung and Spleen) a 100 to 200 micron fibrous capsule formed around the implant by 1 week post injection (Fig. 1a) which thickened and increased in cellularity by 3 weeks post-injection (Fig. 1b). Implants decreased in size over time from 1 to 3 weeks as the scaffold was being degraded and remodeled

(Fig. 1a-b). Dense cellular tissue was detected both at the skin (dorsal) and capsular

(ventral) interfaces with cellular infiltration in most implants through the center by 1 week post-injection (Fig. 1c). There was not a significant difference in capsule thickness (Fig.

1d) or intra-implant cellularity (Fig. 1e) between the various tissues ECM sources.

Further characterization of the immune infiltrate was achieved by staining for cells of the innate (macrophages, dendritic cells) and adaptive (T cells) immune system.

The immune response to subcutaneous scaffolds was dominated by CD11b+ macrophages with scattered CD11c+ dendritic cells and a large number of CD3+ T cells

(Fig. 2). Most macrophages occupied the capsule including dense clusters of cells at the skin interface (Fig. 2a-b) with minimal CD11b staining in the more central regions.

Dendritic cells were found in the dermis, characteristic of resting Langerhans cells, as well as at the skin and capsular interfaces. These cells were excluded from the clusters containing CD11b+ macrophages. CD3+ T cells were found surrounding the implant much like macrophages and dendritic cells, and were densely present around vascular structures (Fig. 2a, Bone CD3). Additionally, T cells co-localized with CD11b+ macrophages in cellular aggregates at the skin interface (Fig. 2b). The clustering of macrophages and T cells but not dendritic cells suggests communication between T cells and macrophages as opposed to the canonical communication with dendritic cells

(Fig. 2c).

108

Flow cytometric analysis of the scaffold immune microenvironment

Thorough analysis of the scaffold immune microenvironment (SIM) was achieved through flow cytometry (FACS). After enzymatic digestion, isolated cells were stained with an 8-color panel to evaluate cell recruitment and polarization (Figs. 3-4). Presence of T Cells, dendritic cells, and macrophages first observed in histological staining was confirmed with FACS (Fig. 3). B Cells and CD34+ stem cells were also detected (Fig. 3).

There was a statistically significant variation of immune response dependent upon scaffold source. Recruitment of T cells, B cells and progenitor cells was more prominent at 3 weeks (Fig. 3b-c) suggesting vascularization and induction of acquired immune responses was occurring. The overall SIM profile was similar between scaffolds with a high proportion of myeloid cells (> 25% of total infiltrate) followed by CD3+ T cells (3-5

%) and minimal levels of CD19+ B cells (<0.1 %). By 3 weeks, Cardiac matrix recruited the lowest fraction of F4/80+CD11c+/- myeloid cells, and the overall profile was significantly different between several tissue sources (Fig. 3c).

As the M1/M2 axis of macrophage polarization has been associated with scaffold remodeling, wound healing and tissue regeneration, we further characterized myeloid cells present in the SIM. Myeloid cells were by their expression of CD86 (costimulatory molecule in antigen presentation; M1 marker) and CD206 (mannose receptor; M2 marker). Three distinct myeloid cell populations were present and varied in their

CD86/CD206 expression. Scaffold associated macrophages (SAMs) were

F4/80+CD86+CD206hi (Fig. 4) where the more mature macrophages (F4/80hi) expressed higher levels of CD206 and CD86 than F4/80lo macrophages. Dendritic cells (CD11c+) did not express CD206, but had high levels of CD86 expression. CD11c+F4/80+ macrophages had high levels of both CD86 and CD206 (CD86hiCD206+). As with in vitro studies, ECM scaffolds induced a mixed M1/M2 phenotype and the SAMs expressed

109 both the M1 and M2 markers (CD86 and CD206). Further dissection of the myeloid compartment could reveal more specific subtypes present in the SIM.

Additionally, implants were stained with a 5-color panel to determine presence of more specific T cell subtypes based on their expression of CD4 (helper T cells), CD8

(cytotoxic T lymphocytes) and FoxP3 (regulatory T cells) (Fig. 5). Helper T cells were the most abundant CD3+ T cell subtype in both scaffolds tested (> 60% of CD3+) compared to CD8+ cytotoxic T lymphocytes (<15% of CD3+) (Fig. 5b). Cardiac ECM recruited slightly skewed the ratio of CD4 to CD8 T cells. Both scaffolds recruited FoxP3+ T cells, but Bone ECM recruited far more than Cardiac ECM (24.13 ± 4.33 vs 1.24 ± 0.22, P =

0.0019, Fig. 5c).

A scaffold-associated immune profile is detected in ECM-treated volumetric muscle wounds

ECM scaffolds are used clinically to treat traumatic wounds such as skin lesions and muscle loss (33, 38, 87). To model a traumatic muscle wound, we used a murine volumetric muscle injury (VML). After injury, mice received 0.05 ml of saline, collagen, decalcified bone (B-ECM) or ventricular cardiac muscle (C-ECM) (Fig. 6a). At 7 days post-injury, the scaffold and surrounding musculature was harvested and prepared for analysis with flow cytometry. As this is an early injury response window, we stained for a comprehensive panel of innate immune cells, or myeloid cells. From this panel we could detect F4/80+CD11b+ macrophages (Fig. 6b), CD11b+/-F4/80-CD11c+ dendritic cells (Fig.

6c), CD11b+Ly6c+ immature monocytes and CD11b+Ly6g+ polymorphonuclear cells (Fig.

6d, PMNs; neutrophils, eosinophils, and basophils). Additionally, we included markers for M1/M2 polarization (CD86 and CD206) and antigen presentation to CD4+ T cells

(MHCII I-A/I-E). Previously, we have demonstrated the critical role for adaptive immune

110 cells, specifically CD4+ Th2 T cells, in polarizing the scaffold immune microenvironment

(SIM) and promoting an M2-macrophage phenotype. Therefore, we analyzed the myeloid profile of B6.129S7.Rag1tm1Mom/J (Rag-/-) mice to further dissect the interactions of adaptive immune cells with the myeloid compartment of the SIM. As with previous studies, the SIM comprised a large proportion of F4/80+CD11b+ macrophages (40-50 % at 1 week post-injury; Fig. 6e). Rag-/- mice recruited a higher proportion of macrophages compared to WT counterparts and fewer PMNs (Fig. 6f).

There was a characteristic adaptive-immune dependent M2-phenotype in F4/80+CD11b+ macrophages with elevated CD206 and decreased CD86 at 1 week post-injury (Fig. 7a).

As noted previously, Rag-/- mice had an increased number of CD11b+F4/80+ macrophages, which correlated with a decrease in CD11b+F4/80- myeloid cells suggesting a possible differentiation or maturation of F4/80- cells to F4/80+ macrophages

(Fig. 7b). Interestingly, in scaffold-mediated CD11c upregulation in CD11b+F4/80+ macrophages is dependent upon adaptive immune cells (Fig. 7c). MHCII expression was highest on CD11chiCD11b+F4/80+ macrophages followed by CD11b+/-F4/80-CD11c+ dendritic cells and CD11cloCD11b+F4/80+ macrophages (Fig. 7d). As with CD86 and

CD206 expression, MHCII expression was dependent upon adaptive immune cells and decreased in Rag-/- mice. PMNs, monocytes and dendritic cells expressed low levels of

CD86 and CD206 when compared to macrophages.

Discussion

Every biomaterial will create a distinct scaffold immune microenvironment (SIM) based on its composition (synthetic versus biologic) and the location in which the material is implanted. The SIM is determined by the cells that are recruited to the scaffold and the activation/polarization of those cells. We have previously defined a pro-

111 regenerative skeletal muscle SIM that is characterized by a Th2-dependent M2 macrophage population. Through histological examination, it is possible to determine the extent of cellularity as well as the locational distribution of cells of certain lineages. For example, in subcutaneous ECM implants we describe an abundance of CD11b+ macrophages that colocalize with CD3+ T cells in cellular aggregates at the skin interface. This suggests a communication of T cells with macrophages and possibly an activation of the adaptive immune system through this highly phagocytic cell type. To further define the scaffold immune microenvironment (SIM) created by extracellular matrix scaffolds, we employed the use of multi-color flow cytometry. Flow cytometry

(FACS) provides a method of high-throughput quantitative immune phenotyping. FACS is critical in fields such as tumor immunology where they study immune cell recruitment and polarization in tumors to determine the most efficacious immunotherapy techniques for cancer treatment. Just as tumor immunologists use FACS to define a tumor microenvironment, we used FACS to define the SIM. As with histological staining, we were able to detect macrophages, dendritic cells and T cells as well as B cells. By staining for CD86 (M1 marker) and CD206 (M2 marker) we were able to characterize the polarization of myeloid cells along the M1-M2 axis. Three myeloid cell types with different M1/M2 profiles were described. These include M1-like dendritic cells, M2-like macrophages and CD11c+F4/80+ myeloid cells which represent an intermediate phenotype expressing high levels of both CD86 and CD206.

Macrophage polarization is a complex spectrum of different phenotypes that are specific for the challenge that they are faced with in the body. These cells can adopt a continuum of different surface markers, morphological characteristics, and gene expression dependent upon their polarization phenotype. It is best to consider macrophage polarization in context of function as opposed to categorization.

112

Functionally, surface markers can be associated with processes, such as signal transduction from cytokines and damage-associated pattern recognition, chemotaxis, and adhesion, as well as many others (7, 8, 10). If we take this into consideration, and each protein or gene is described not as a categorical marker, but a functional protein

(as it is biologically), then we will gain a greater understanding of the specific immune functions elicited by biomaterials and necessary for scaffold integration and tissue regeneration.

Macrophage colocalization with T cells in skin interface cellular aggregates suggests a possible communication with the adaptive immune system. Previously, researchers have shown that subcutaneous implants can induce a Th2-like phenotype characterized by systemic IL-4 upregulation in the peripheral blood. Connecting the MHCII+ M2- macrophage phenotype with the knowledge of a Th2-associated ECM response, we can assume that there is communication between M2 macrophages and Th2 T cells in the

SIM. We have defined a scaffold-associated macrophage (SAM) that is F4/80+CD11c+/-

CD206hiCD86+MHCII+. CD206+ SAMs could potentially be recognizing the fragmented extracellular matrix (ECM) composed of proteins, proteoglycans and polysaccharides that have been milled to a fine particulate. More specifically CD206 can recognize N- acetylglucosamine, fucose and mannose. These ECM fragments recapitulate the damaged tissue that is produced in a wounding response and therefore can magnify the immune response to DAMPs that are augmented in with the ECM scaffolds. Additionally,

CD206 and CD86 expression on SAMs is dependent upon the presence of adaptive immune cells. Here, we show that adaptive immune cells regulate CD11c expression on

SAMs. CD11c is an integrin that is involved in adherence of activated cells to the endothelium of blood vessels and binding of complement-coated particles. CD11c and

113

CD206 are important sensors of the environment, and can dictate the myeloid response to a scaffold or surrounding tissue.

As with alterations in tissue-source for extracellular matrix scaffold, one would assume a different SAM phenotype in synthetic or hybrid (biologic-synthetic) materials.

Macrophages have been implicated in multiple aspects of immune response to biomaterials as well as the outcome of scaffold remodeling, where M2-type macrophages are critical in ECM scaffold remodeling. Additionally, macrophages have been associated with several developmental processes including salamander limb regeneration. Other immune cells, such as eosinophils, have been implicated in wound healing of higher organisms. These connections between biomaterials, immune cells and development are important considerations for scaffold design to ensure that the biomaterial will promote regeneration through proper polarization of the immune response.

When analyzing cell recruitment and polarization in biomaterial scaffolds via flow cytometry, several factors must be addressed and acknowledged to ensure reliable data acquisition. For example, when digesting a tissue sample to isolate single cells, there are various enzymatic reagents that can be used. These enzymes vary in target residues and strength, which change how thoroughly they digest the tissue. A stronger enzyme cocktail would be more useful in preparation of a dense tissue such as dermal tissue, however these enzymes have the ability to cleave residues that reside on the surface of cells. Cleavage of surface proteins can confound results and decrease signal for certain proteins that are sensitive to proteolytic cleavage. Therefore it is necessary to balance cell isolation with integrity of surface proteins. When analyzing many parameters in a given cell population, it is advisable to divide the cell isolate into multiple

114 panels to avoid over-crowding of fluorophores in a single flow cytometry panel. In some scenarios, what can be achieved in a 10 color panel can be easily recapitulated in two-5 or 6-color panels.

Detailed immune analysis of scaffold responses will elucidate in further the details of the foreign body response and mechanisms of immune-mediated regeneration.

As the goal of tissue engineering is to create a functional replacement tissue, it is important to integrate the material used to create the tissue with the host. The first cells to respond to trauma or biomaterial implantation are immune cells, which can greatly dictate downstream events such as stem cell differentiation. The analyses conducted in this study present a future direction for tissue engineering and regenerative medicine in understanding the immune dynamics of scaffolds.

Over the past 10 years there has been a gradual shift in perception of immune response to biomaterial scaffolds. As the immune system can dictate many processes and surveys the body continuously for disruptions in homeostasis, it is an important consideration in regenerative medicine. Immune cells are critical mediators of wound healing and regeneration. When a biomaterial scaffold is applied to a wound or as a tissue filler, several factors will dictate the immune microenvironment that is created by that scaffold. This includes the material that the scaffold is made of, the location in which the material is implanted and the state of that implantation site (wounded or non- traumatic). In the future, it will be necessary to analyze the immune cells that are present in different scaffolds and different tissue locations as they will dictate the outcome of regeneration. In this manuscript we have described the different immune microenvironments created by extracellular matrix scaffolds when applied in a non-

115 traumatic (subcutaneous) space. As each tissue scaffold differs in its constitution, we see unique differences from tissue-to-tissue, but overall a similar immune microenvironment. If one were to apply these analyses to synthetic scaffolds, it would be possible to detect minute differences in cell populations upon differing modifications such as additions of functional groups or growth factors. When the immune profile of a scaffold is defined, we can use that information to determine the best application or tissue location for its use. Understanding the immunomodulatory nature of these materials will allow for a more rational scaffold design to promote immune-mediated regeneration.

116

Figure 3.1: Subcutaneous injection of particulate ECM scaffolds induces encapsulation cellular infiltration.

(a) H&E composite image of 1 week post-injection, liver ECM implant. (b) Composite image of 3 weeks post-injection, lung ECM implant. (c) High magnification of skin

(dorsal) and capsule (ventral) interfaces and center of implant. (d) Quantification of the capsule width and subcutaneous fat pad width in microns. (e) Cellular infiltration displayed as cell count per mm2.

117

Figure 3.2: Immune response to scaffolds is dominated by CD11b+ Macrophages communicating with CD3+ T cells.

(a) Immunohistochemical staining of the dorsal (skin) interface of the implant for CD11b+ macrophages, CD11c+ dendritic cells and CD3+ T cells. (b) Cellular aggregates present at the dorsal interface stain positive for CD11b and CD3 but exclude CD11c. (c)

Colocalization of CD11b+ macrophages and CD3+ T cells suggests an increased communication between macrophages and T cells as opposed to canonical CD11c+ dendritic cell.

118

Figure 3.3: Immune infiltrate profile depends on tissue source and time post-injection

(a) 8-color flow cytometry was used to define the lymphoid compartment: CD3+ T and

CD19+ B cells and the myeloid compartment: F4/80+ macrophages and CD11c+ dendritic cells. (b) Immune profile at 1 week post-injection (c) Immune profile at 3 weeks post- injection. Data are mean ± SEM, n = 4.

119

Figure 3.4: Myeloid subtypes defined by F4/80, CD11c, CD206 and CD86 expression.

F480+ macrophages are CD86+CD206hi, CD11c+ dendritic cells are CD86hiCD206-, and

CD11c+F4/80+ macrophages are CD86hiCD206+. (a) Representative FACS plots of the mean fluorescence intensity quantification in panel (b) for the 3 myeloid subtypes.

120

Figure 3.5: T cell profile is dominated by CD4+ T cells

(a) Representative plots of CD4 vs CD8 profile and FoxP3 gating. (b) CD4/CD8 ratio in

Bone and Cardiac derived scaffolds. (c) FoxP3+ Tregs are detected at higher levels in

Bone ECM than Cardiac ECM.

121

Figure 3.6: Detailed profile of myeloid cells in a scaffold-treated volumetric muscle wound

(a) A 3mm x 4mm portion of the quadriceps muscle is excised and replaced with a biomaterial scaffold: Saline vehicle control, Collagen, B-ECM (decalcified bone ECM), and C-ECM (ventricular cardiac muscle derived ECM). (b) F4/80+CD11b+ Macrophage infiltrate. (c) CD11c+F4/80- Dendritic cell infiltrate (d) CD11b+Ly6c+ (Immature monocytes) or Ly6g+ (Polymorphonuclear cells; neutrophils, eosinophils, basophils). (e-f)

Myeloid infiltrate in WT (e) or Rag-/- (f) mice, which lack mature T and B cells.

122

Figure 3.7 Myeloid profile is dependent upon scaffold and presence of adaptive immune cells

(a) Characteristic CD86 increase and CD206 decrease in the absence of adaptive immune cells (Rag-/- ; red) (b) Decrease in CD11b+F4/80- cells in Rag-/- mice. (c) CD11c expression on F4/80+CD11b+ macrophages is dependent on adaptive immune cells. (d)

Multi-parametric analysis of surface markers on different myeloid subtypes. Light blue line = CD11b+F4/80-Ly6c+ Monocytes; Green line = CD11b+F4/80-Ly6g+

Polymorphonuclear cells (PMNs, neutrophils, basophils eosinophils); Orange line =

CD11b+/-F4/80-CD11c+ Dendritic Cells; Red line = CD11b+F4/80+CD11clo Macrophages;

Blue line = CD11b+F4/80+CD11chi Macrophages.

123

Composition and immune environment of urinary bladder matrix scaffolds

This work has been not published and is currently in preparation.

Authors: Kaitlyn Sadtler, Sven Sommerfeld, Matthew Wolf, Xiaokun Wang, Shoumyo

Majumdar, and Jennifer Elisseeff

Abstract

Urinary bladder matrix (UBM) is used clinically for treatment of diabetic ulcers, abdominal wall reconstruction, soft tissue filling among other applications. Composed mainly of collagen, the material is a mix of fibrous collagen structures and sheet-forming structures characteristic of the urinary bladder matrix basal lamina. Upon application of a biomaterial in a traumatic or non-traumatic setting, there is a cascade of immune cells that react to the damaged tissue and biomaterial. Here, through the use of multicolor flow cytometry, we describe the various cells that infiltrate the UBM scaffold in a subcutaneous and volumetric muscle injury model. A wide variety of immune cells are found in the UBM scaffold immune microenvironment (SIM) including F4/80+ macrophages, CD11c+ dendritic cells, CD3+ T cells and CD19+ B cells. UBM induced a systemic IL-4 up-regulation and a local M2-macrophage response. The recruitment and activation of these cells is dependent upon signals from the scaffold and communication between the different cell types present.

124

Introduction

Urinary bladder matrix (UBM), a decellularized extracellular matrix, is used clinically in a variety of applications (32, 33, 37, 39, 95). These include abdominal wall repair, diabetic ulcer treatment, skin wounding, and treatment of other soft tissue defects. Extracellular matrix (ECM) scaffolds are created by treating native tissue, often porcine or human, with a variety of acids and detergents to remove the majority of cellular components and leave behind a complex structural and signaling scaffold (32,

34). A large array of tissues have been decellularized for tissue engineering applications including but not limited to, urinary bladder, small intestinal submucosa, cardiac muscle, demineralized bone, and aminon. They are synthesized either as a sheet (used for large surfaces) or particulate (used in ulcer treatments) and currently investigated in other forms such as solubilized gels. As they are derived from native tissue, ECM scaffolds carry a structural and biochemical complexity that cannot be mimicked synthetically.

ECM is mainly composed of collagens and proteoglycans interspersed with sequestered growth factors and cytokines (54, 96). The constituents of these scaffolds can greatly affect their biologic activity and performance in the clinical setting. Currently the best mechanisms of determining components of ECM scaffolds are through histological staining and ELISA of target proteins such as growth factors. Further understanding of the components of these scaffolds can be achieved through proteomic analysis (77). Contrary to typical substrates for proteomic analysis, ECM scaffolds are highly hydrophobic, difficult to solubilize, and require a more thorough digestion to properly create fragments that can be detected by mass-spectrometry. However, with a distinct profile of the proteins found in ECM scaffolds, we can better characterize the results seen in the clinic with the scaffold composition.

125

The proteomic composition of a material can be correlated with alterations in cell function, including immune responses (55, 97, 98). Immune cells are the first responders to injury and biomaterial implantation. The location and environment in which the material is placed will also affect the cells that are recruited. The immune microenvironment created by a scaffold will alter the presence of various cytokines and growth factors that can contribute to stem cell differentiation and tissue regeneration. A wide variety of immune cells have been implicated in regeneration of murine muscle, liver, and salamander limbs (13, 21, 22). Characterizing the scaffold immune microenvironment (SIM) of clinical grade materials would allow insight into the cell dynamics that could be expected in a clinical setting. ECM scaffolds are associated with a pro-regenerative M2-macrophage phenotype that is dependent upon signaling from

Th2 T cells (30, 44).

Materials and Methods

Mechanical testing

Rheology of the samples was performed using an Ares G2 rheometer (TA

Instruments New Castle, DE) with 25 mm serrated parallel plate geometry. Experiments were conducted at physiological temperatures (37oC) and performed in sequence. For each test, 400 mg (wet) of each sample was loaded on the rheometer stage. The gap distance between the upper and lower plates was set to 1 mm. Following a temperature and axial force equilibration step, the linear viscoelastic properties of the samples were determined. The experimental design followed a frequency sweep from 0.1 rad/s to 15 rad/s, oscillation frequency (15 rad/s constant), creep test (.5% strain) and finally, compression testing with constant deformation until failure (-1.0 mm delta lenth).

Rheological measurements for each sample were run in triplicate and were conducted

126 on samples with ECM concentrations ranging from 100 to 300 mg/mL. Samples were kept hydrated during these rheology experiments using a solvent trap.

Differential Scanning Calorimetry

ECM samples were were weighed 20 mg each, loaded into 50 uL aluminum sample pans, and crimp sealed. A sealed empty pan was used as reference. The denaturation temperature of each sample was determined using a differential scanning calorimeter

(DSC 8000; Perkin Elmer, Waltham, MA) over 10–95 degrees C at a rate of 10 degrees

C/min under a 20 mL/min nitrogen flow. The peak temperature, specific heat, heat of fusion (ΔH*) for each sample was determined,and the thermograms were analyzed using the Pyris Series thermal analysis software (Perkin Elmer), version 10.1.

Proteomics

ECM proteomic composition was determined by mass spectometry of tryptic peptides derived from soluble and insoluble matrix components. ECM was suspended in

9 M urea (pH 8, Hepes buffer) at a concentration of 2.5 mg ECM/ml, vortexed for 60 s, and centrifuged for 10 min at 16,000x g. The supernatant was collected and henceforth referred to as the soluble ECM fraction, and was further reduced with DTT, alkylated with IAA, and digested with 20 ug Trypsin (promega Gold) overnight at 37C. The resulting pellet was washed with 1M Urea (pH 8), and resuspended in 1 M urea with 20 ug of trypsin for mechanically assisted trypsinization. Sample with trypsin was transferred to a barocycler (NEP2320, Pressure BioSciences, Inc), which cyclically alternated between high and low pressures. The resulting peptides are referred to as derived from the insoluble fraction.

Proteomic analysis was conducted as per Beachley et. al. Nature Methods 2015.

In short the soluble and insoluble fraction were dried then reconstituted in 40 μL of 2%

127 acetonitrile, 0.1% formic acid, injected 6 μL. Liquid chromatography–tandem mass spectrometry (LC-MS/MS) identification of proteins and analysis of peptides was performed using a Q-Exactive interfaced with a Thermo Easy-nLC 1000 system (Thermo

Scientific) or a Velos Orbitrap (Thermo Scientific) interfaced with a NanoACQUITY

UPLC system (Waters). Peptides were fractionated by reversed-phase HPLC on a 75

μm × 12 cm column with a 15-μm emitter tip (New Objective, Woburn, MA) packed in- house with Magic C18AQ (5 μm, 120 Å, Michrom Bioresources) using a 0–90% acetonitrile, 0.1% formic acid gradient over 90 min at 300 nL/min. Eluting peptides were sprayed directly into the Q-Exactive or Velos at 2.0 kV. Q-Exactive survey scans were acquired from 350–1800 m/z with up to 15 peptide masses (precursor ions) individually isolated with a 2.0-Da window and fragmented (MS/MS) using a collision energy of 27 and 30-s dynamic exclusion. Precursor and fragment ions were analyzed at 70,000 and

17,500 resolution, respectively. Velos survey scans were acquired at 350–1,800 m/z with up to eight peptide masses (precursor ions) individually isolated with a 1.9-Da window and fragmented (MS/MS) using a collision energy of 35 and 30-s dynamic exclusion. Precursor and fragment ions were analyzed at 30,000 and 15,000 resolution, respectively.

The mass spectrometry–derived data were searched against a combined human and porcine using the SEQUEST HT search algorithm through Proteome Discoverer

(version 1.4.1.14, Thermo Scientific). Precursor mass tolerance was set to 20 ppm, the fragment mass tolerance was set to 0.05 Da, and the maximum peptide length was seven amino acids. Peptides that passed the 1% false discovery rate threshold were used for protein identification.

128

Cell culture iBMM, immortalized bone marrow macrophages (Squadrito et. al. Cell Reports 2014) were cultured under manufacturers recommended conditions in IMDM + 50 ng/ml M-

CSF. Tissue culture plates were coated with 1 ml of a 4 mg/ml UBM particulate (per well in a 6 well plate) and dried overnight in a fume hood. Plates were UV sterilized prior to use. Macrophages were cultured on UBM-coated plates for 24 hours prior to analysis via

RT-PCR or flow cytometry.

Subcutaneous implantation

Particulate UBM was hydrated with PBS to 300 mg/ml. 0.2cc of the hydrated scaffold was injected subcutaneously into the dorsal region of C57BL/6 female wild type mice at 6-7 weeks of age. Mice were anesthestized under 3% isoflurane and maintained at 2% isoflurane and oxygen for the duration of the surgery. Prior to injection, the dorsal side of the mouse was sterilized with 70% ethanol. After 1 and 3 weeks post-injection, animals were sacrificed and implants and surrounding tissue were dissected with or without the skin for histology and flow cytometry, respectively.

Volumetric muscle loss surgery

6 week old wild type female C57BL/6 mice (Charles River Laboratories) were used in all studies. Mice were anesthetized under 3.0% isoflurane and maintained at

2.0% isoflurane for the duration of the surgery. Hair was removed from both hindlimbs through the use of an electric razor. Skin was sanitized with 70% ethanol before making a 1 cm incision in the skin above the quadriceps muscle group. Using surgical scissors, a 3mm x 4mm wound was created and back-filled with 50 ul of a 300 mg/ml UBM particulate paste. The wound was stapled closed and procedure was repeated on the contralateral leg. After surgery, mice were injected subcutaneously with 5 mg/kg Rimadyl

129 for pain relief. And monitored until amubulatory. All procedures were done in accordance with guidelines provided by the Johns Hopkins University Animal Care and Use

Committee.

Histology

Subcutaneous implants were incubated in 10% Formalin overnight prior to dehydration and mounting in paraffin. 5 micron sections were taken on a microtome

(Leica) and stained with hematoxylin and eosin (Sigma).

Flow Cytometry

Subcutaneous implants and muscle samples were diced finely with a scalpel before incubation in 0.5 mg/ml Liberase + 0.2 mg/ml DNase I in RPMI media for 45 minutes, shaking at 400 rpm and 37oC. Resultant material was filtered through a 100 um cell strainer and washed twice with 1XPBS. To remove debris the resultant material was subjected to density separation with Lympholyte-M as per manufacturers instructions.

Interphase was washed twice with PBS before staining with the following panels. Cell population analysis: CD3 AlexaFluor488, CD19 BrilliantViolet421, CD11c APC-Cy7,

F4/80 PE/Cy7, CD86 AlexaFluor700, CD206 APC and Viability Aqua. Cells were then fixed with BD Cytofix and stored overnight prior to analysis on a BD LSRII flow cytometer.

Gene Expression Analysis

Cell culture samples were harvested in 1 ml TRIzol, and mixed with 200 ul

Chloroform before spinning at 12000 xg for 15 minutes. The aqueous phase was added to 500 ul 70% ethanol and run through an RNeasy Spin column. The RNA was cleaned and eluted as per manufacturers instructions. cDNA was synthesized using SuperScript

130

RTIII reagents. Gene expression was determined using the following primers with a

SYBR Green reporter dye. : B2m forward CTC GGT GAC CCT GGT CTT TC, B2m reverse GGA TTT CAA TGT GAG GCG GG; Tnfα forward GTC CAT TCC TGA GTT

CTG, Tnfα reverse GAA AGG TCT GAA GGT AGG; Il1β forward GTA TGG GCT GGA

CTG TTT C, Il1β reverse GCT GTC TGC TCA TTC ACG; Arg1 forward CAG AAG AAT

GGA AGA GTC AG, Arg1 reverse CAG ATA TGC AGG GAG TCA CC; Retnla forward

CTT TCC TGA GAT TCT GCC CCA G, Retnla reverse CAC AAG CAC ACC CAG TAG

CA; IL-10 forward TCT CAC CCA GGG AAT TCA AA, IL-10 reverse AAG TGA TGC

CCC AGG CA; iNOS forward GAC GAG ACG GAT AGG CAG AG, iNOS reverse GTG

GGG TTG TTG CTG AAC TT.

Statistics

ANOVA and Students T-test were performed in GraphPad Prism operating with a P value < 0.05.

Results

Mechanical and proteomic properties of UBM scaffolds

UBM was hydrated at 100, 200 and 300 mg/ml to form pastes that were used for subsequent in vivo studies. UBM particulates were composed of both fibrous and sheet- forming connective tissue characteristic of basal membrane and collagen strands present in the urinary bladder connective matrix (Fig. 1a). Proteomic analysis was conducted to determine the protein components within the UBM scaffold. We detected high levels of collagen and other extracellular matrix proteins, as well as residual cellular components such as desmin, actin and myosin (Fig. 1b). The UBM scaffold, as it is highly hydrophobic, was not amenable to typical proteomic preparation methods. We

131 observed a large difference in the urea-soluble fraction and insoluble fraction (solubilized by high-pressure enzymatic digestion) of extracellular matrix proteins (Fig. 1c). Though some were detected in the soluble fraction a large quantity remained. Thus, proteomic characterization of this scaffold may not correctly demonstrate the proportion of highly hydrophobic extracellular matrix proteins such as collagen. Rheologic characterization of storage and loss moduli, viscosity, creep and axial force showed concentration- dependent mechanical properties of the UBM pastes, achieving varying stiffness and elasticity (Fig. 1d-e). Differential scanning calorimetry revealed a lower melting point for collagen compared to UBM in the dry particulate form, however when hydrated, the peaks underwent a characteristic broadening, but the pattern was reversed (Fig. 1f-g).

Immunomodulation of macrophage behavior by UBM scaffolds

Macrophages are phagocytic cells of the innate immune system that respond to stimuli in their environment such as pathogen presence or wounding. As these cells form part of the body’s first line of defense, we cultured murine bone marrow derived macrophages (BMDM) on particulate UBM to determine if there is a direct effect of the scaffold on macrophage behavior. Macrophages were cultured on UBM-coated tissue culture plastic for 24 hours prior to gene expression analysis via RT-PCR. We analyzed genes associated with M1 (inflammatory, classical activation) and M2 (anti-inflammatory, alternative activation) polarization phenotypes, Tnfa, Il1b and Inos or Arg1, Retnla and

Il10, respectively (Fig. 2). In M0, unstimulated conditions, UBM upregulated M1 inflammatory genes, but did not alter M2 gene expression. However, in M2 media conditions (+ IL-4) UBM induced a higher expression of Arg1, a canonical M2 marker, suggesting a mixed phenotype that does not fit into a specific M1- or M2-promoting material (Fig. 2a). To further test macrophage response to UBM, we analyzed the expression of CD86 (M1) and CD206 via flow cytometry. As with RT-PCR, in vitro, UBM

132 promotes expression of CD86, an M1 marker, and decreases expression of CD206, an

M2 marker. However, UBM did up-regulate expression of IL-4Rα in all 3 media conditions, again confirming the mixed M1/M2 phenotype (Fig. 2b).

Scaffold immune microenvironment of subcutaneous UBM

The immune response is a complex coordination of local and systemic effects from multiple cell types involving many signaling molecules. Therefore, it is hard to recapitulate in an in vitro setting. To define the scaffold immune microenvironment (SIM) of UBM in non-injury setting, we injected 0.2 cc of a 300 mg/ml particulate UBM paste in the subcutaneous space of C57BL/6 WT mice (Fig. 3-4). Histologically, by 1 week post- injection a capsule had formed around the implant adhering it tightly to the underside of the skin (Fig. 3a). Cellular infiltration was present through the center of the scaffold by 3 weeks post-injection (Fig. 3b). Using flow cytometry, we were able to identify these cells that formed the UBM-SIM. After mechanical and enzymatic separation, a single cell suspension was stained for the following markers: CD3 (T cells), CD19 (B cells), CD34

(progenitor cells, mainly vascular), CD11c (dendritic cells), F4/80 (macrophages), CD86

(M1 macrophages) and CD206 (M2 macrophages). We detected a high proportion of

F4/80+CD11c+/- macrophages in the UBM-SIM at 1 and 3 weeks post-injection (Fig. 3c- d). Low numbers of CD11c+F4/80- dendritic cells were also detected. In addition to macrophages and dendritic cells (cells of the innate immune system) we detected both

CD3+ T cells and CD19+ B cells suggesting activation of the adaptive immune response

(Fig. 3c-d).

To further evaluate the UBM-SIM, we analyzed the expression of CD86 and

CD206 on myeloid cells (Fig. 4). F4/80+ macrophages expressed intermediate levels of

CD206, with more mature macrophages (F4/80hi) expressing higher levels of CD206

(CD206hi). These macrophages also expressed CD86, an M1 marker, which decreased

133 over time (Fig. 4b). Expression of the M2 and M1 macrophage markers was not mutually exclusive suggesting a complex phenotype that does not neatly categorize into a distinct

“type-1” or “type-2” macrophage. Additionally, CD11c+ dendritic cells were detected with low CD206 expression and high CD86 expression. CD11c+F4/80+ macrophages also expressed both CD86 and CD206.

UBM macrophage and T cell recruitment in traumatic muscle injury

UBM is clinically applied in cases of tumor resection and soft tissue loss, which often follow traumatic injury, either from the surgical removal of cancerous tissue or from wounding. To model the traumatic environment of these clinical applications, we applied the UBM scaffold to a murine model of volumetric muscle loss (VML). A 3mm x 4 mm portion of the quadriceps muscle was removed using surgical scissors, and the resulting defect was filled with 0.05 cc of 300 mg/ml UBM. After 1 and 3 weeks post-injury, the scaffold and surrounding area were harvested for analysis of the SIM via flow cytometry.

As with the subcutaneous implantation, response to the UBM scaffold was dominated by

F4/80+CD11c+/- macrophages, followed by CD11c+ dendritic cells, CD3+ T cells and

CD19+ B cells (Fig. 5a). There was a peak level of macrophage infiltration at 1 week post-injury which decreased by 3 weeks post-injury. CD3+ T cells were present at 1 week post-injury, and persisted after 3 weeks. Of these CD3+ T cells, the majority were

CD4+ helper T cells (Fig. 5b, 66.3%) compared to CD8+ cytotoxic T lymphocytes (7.23%) and double negative cells (25.9%, mainly natural killer T cells). Additionally, CD4+FoxP3+

+ Tregs were present in low numbers (2.39% of total CD3 cells).

As the SIM recruited high numbers of CD3+ T cells, we measured gene expression in the local (inguinal) and distal (axillary/brachial) lymph nodes to determine if there was a systemic activation of the immune system by the UBM scaffold. Compared to the saline-treated control, mice treated with UBM induced a systemic up-regulation of

134

IL-4 in the axillary and brachial lymph nodes. Genes associated with Th1 T cells, such as Tnfa and Il1b were not up-regulated, suggesting an activation and polarization of T cells towards a Th2 lineage. To probe the activation of the adaptive immune system, we evaluated the expression levels of MHCII (I-A/I-E), a protein complex that is responsible for antigen presentation to CD4+ T cells. In the UBM-SIM, macrophages expressed higher levels of MHCII than the canonical CD11c+ dendritic cells, suggesting communication between the local macrophages and lymph node Th2 T cells.

Discussion

Clinically used biomaterials were previously designed in the context of a foreign body response by the immune system. However, as research has continued into immunology and tissue development, the idea of an immunomodulatory biomaterial for regenerative purposes has come into focus. Synthetic nanoparticle materials are used currently as delivery vehicles for drugs and peptides that can alter the immune response, especially in the case of cancer immunotherapy. Solid scaffolds meant to educate the immune system, acting as peripheral hubs of activation and polarization, are also being investigated with applications in oncology. Immune ignorance of a material was once considered the gold standard for a regenerative scaffold. The foreign body response, first characterized by James Anderson in the 1980’s, described immune infiltration by neutrophils and macrophages that ended with oxidation and degradation of the biomaterial, foreign body giant cell formation, tissue-damaging inflammation, and fibrous encapsulation (27-29). Certain applications would favor immune ignorance and tolerance, such as drug delivery materials that require a certain rate of degradation to deliver an effective dose of the drug, or permanent materials such as pacemakers or synthetic intraocular lenses and steps are being taken to screen materials that promote dampening of the immune response (93). However, for degradable scaffolds meant to

135 integrate with tissue and promote regeneration, immune ignorance is not ideal. Over the past 10 years researchers have begun work on regenerative immunology, understanding the precise activation of the immune system that promotes scaffold remodeling and tissue development.

Both subcutaneous and muscle-injury applications of UBM recruited large numbers of F4/80+ macrophages, that were CD86+CD206hi. Additionally, UBM recruited cells of the adaptive immune system including T cells and B cells, predominately CD4+ T cells. In addition to the local M2 macrophage phenotype, there was a systemic up- regulation of IL-4, a canonical Th2 cytokine, in the distal lymph nodes. The local M2 macrophages expressed high levels of MHCII, and were the dominant antigen- presenting cell in the UBM-SIM. Normally, M1 macrophages are more associated with high levels of MHCII, however in the UBM-SIM, there is an up-regulation of M2- macrophage communication with the adaptive immune system, which correlates with the systemic IL-4 detection. This suggests a macrophage-dependent T cell activation that in turn feeds back onto the macrophage population to further promote a pro-regenerative niche.

Formulated into different concentrations, UBM paste adopts dose-dependent rheological properties. These differing mechanical properties could imbue the material with different immunological properties. Further research would involve determining the immunologic effects of scaffold modification, from differing concentrations of material, to physical modifications such as gelation (36) or combination with synthetics (40).

Understanding the structure-function relationship of scaffold modification could allow researchers to leverage material delivery and structural properties with immune response to determine the optimal configuration for the desired application.

136

Figure 4.1: UBM materials characterization.

(a) SEM of particulate UBM. (b) Proteomic evaluation of the proteins with highest abundance in UBM. (c) Non-urea soluble ECM components that required further solubilization processing to detect on mass spectrometry. (d-e) Storage and Loss moduli of 100, 200 and 300 mg/ml ECM pastes. (f) Differential scanning calorimetry of dry versus hydrated 300 mg/ml paste UBM compared to a milled collagen I control.

137

Figure 4.2: Particulate UBM induces alterations in macrophage phenotype in vitro.

(a) RT-PCR analysis of M1 (Tnfa, Il1b, Inos) and M2 (Arg1, Retnla, Il10) genes in bone- marrow derived macrophages cultured on UBM for 24 hours in M0 (growth, unstimulated), M1 (inflammatory, LPS + IFNg) or M2 (anti-inflammatory, IL-4) media conditions. (b) Flow cytometric analysis of CD86 (M1), CD206 (M2) and IL4ra (M2) on macrophages at 24 hours post-seeding on UBM coated tissue culture plastic.

138

Figure 4.3: The scaffold immune microenvironment of UBM.

C57BL/6 mice received 200 ul of a 200 mg/ml subcutaneous UBM implant. (a) Cross- section of subcutaneous UBM implant at 1 week post-injection. (b) Dorsal (skin), center, and ventral (capsule) sections of UBM implant at 1 week post-injection. (c) FACS analysis of resident immune cells at 1 and 3 weeks post-injection showing high presence of F4/80+ macrophages and an increase in CD3+ T cells over time. (d) Representative

FACS plots from data quantified in (c).

139

Figure 4.4: UBM promotes an M2-macrophage phenotype that matures over time.

(a) CD206 and CD86 expression on 3 myeloid subtypes detected in the implant at 3 weeks post-injection. (b) CD86 expression quantified as mean fluorescence intensity

(MFI) at 1 and 3 weeks post-injection. (c) CD206 expression quantified as MFI.

140

Figure 4.5: UBM-treated muscle wounds recruit a diverse immune cell repertoire.

C57BL/6 WT mice received a bilateral 3mm x 4mm muscle removal from their quadriceps muscle group which was back filled with 50 ul of a 200 mg/ml UBM paste and analyzed via flow cytometry. (a) Immune cell populations at 1 and 3 weeks post- injury. (b) CD4:CD8 ratio at 3 weeks post-injury.

141

Figure 4.6: UBM induces a systemic IL-4 upregulation correlated with local antigen- presenting M2-macrophages.

(a) RT-PCR of distal lymph nodes (axillary/brachial) at 3 weeks post-injury displayed as

RQ to saline treated VML control. (b) Percent (%) of MHCII+ cells that are present in the scaffold immune microenvironment at 1 week post-injury. (c) MHCII expression is selectively detected on CD11b+ myeloid cells. (d) MHCII+ cells are mainly

F4/80+CD206+ M2 macrophages, representative FACS plot of data quantified in (b).

142

Chapter 4: The Role of the Adaptive Immune System In Formation of A Pro-

Regenerative Scaffold Immune Microenvironment

143

Summary

Immune responses to biomaterials create a distinct signaling microenvironment that we have shown. Additionally, immune cells can greatly alter stem cell differentiation and tissue development. The integration of signals from surrounding immune cells determines the regenerative success of a scaffold. Here, we investigate the immune microenvironment created by extracellular matrix scaffolds in a murine model of volumetric muscle loss. Through these studies, we identify several cells and signals required for formation of a pro-regenerative scaffold immune microenvironment.

Integration of signals from the innate and adaptive immune systems govern immune cell recruitment, scaffold remodeling, and tissue regeneration. We identify Th2 T cells, dependent upon mTORC2/Rictor signaling and an IL-4 effector cytokine, as critical players in formation of this niche. Without Th2 T cells present, scaffold associated macrophages (SAMs) lose their pro-regenerative phenotype, scaffold is not properly remodeled, and there is an imbalance in fibro/adipogenic lineage commitment in the healing muscle.

144

Developing a Pro-Regenerative Biomaterial Scaffold Microenvironment Requires T

helper 2 cells

This work has been published and is reprinted here with permission from AAAS.

Sadtler K, Estrellas K, Allen BW, Wolf MT, Fan H, Tam A, Patel C, Luber B, Wang H,

Wagner K, Powell J, Housseau F, Pardoll D, Elisseeff JH. Science 2016. Accepted

This is the author’s version of the work. The definitive version was published in Science.

For final published version, please visit www.sciencemag.org

Abstract

Immune mediated tissue regeneration driven by a biomaterial scaffold is emerging as an innovative regenerative strategy to repair damaged tissues. We investigate herein how biomaterial scaffolds shape the immune microenvironment in traumatic muscle wounds to improve tissue regeneration. The scaffolds induced a pro-regenerative response, characterized by an mTOR/Rictor-dependent T helper 2 pathway that guides interleukin-

4-dependent macrophage polarization, which is critical for functional muscle recovery.

Manipulating the adaptive immune system using biomaterials engineering may support the development of therapies that promote a systemic and local pro-regenerative immune responses, ultimately stimulating tissue repair.

Methods

Tissue ECM Selection and Preparation

Porcine derived tissues (Wagner Meats, Mt. Airy MD) were processed following a standard protocol. Samples were formulated into a paste through the use of a knife-mill

145 processor (Retsch, Germany) with particle sizes no larger than 5 mm2 and rinsed thoroughly with running distilled water until blood was cleared from samples. Bone samples were pre-treated for decalcification by incubation in 10% formic acid (Sigma) for

Freezer/Mill (SPEX CertiPrep). ECM powder was stored between -20oC and -80oC and

UV sterilized prior to use. Collagen from bovine tendon (Sigma) was cryomilled using the

SPEX SamplePrep Freezer/Mill to form a particulate similar to the whole tissue ECM samples. This would serve as the “single-component” control for the studies compared to tissue-derived complex ECM scaffolds.

Broad tissue ECM source screening for macrophage response was previously performed using a tissue array and cell morphological analysis (98). A focused screen of

5 tissues (lung, cardiac, bone, spleen, and liver) was then performed in vitro to evaluate macrophage M1 (CD86) and M2 (CD 206) marker expression (as described in detail in the cell culture and in vitro flow cytometry methods). These tissues were selected for

146 the focused screen due to the range of morphological results in the previously referenced publication (98) and also the desire to include a sampling of tissue properties

(highly cellular tissues versus dense connective tissues). For translation to the in vivo traumatic wound model, cardiac and bone tissue was selected due to their diverse response in M1 and M2 markers FACS and their varying proteomic composition (cellular non ECM components versus a connective tissue with primarily ECM components, respectively). Furthermore, a clinically used material (Matristem, urinary bladder matrix) was tested to compare in vivo CD86 levels.

Cell Culture

Murine immortalized bone marrow macrophages (iBMM, (99)) were cultured as per developer’s protocol in IMDM (Gibco®) media containing 20% FBS (Hyclone, GE

Healthcare Life Sciences), 2.5 mM L-glutamine (Gibco®), 1% PenStrep (Life

Technologies), and 50 ng/ml M-CSF (Recombinant Mouse, BioLegend). iBMM cells were compared to primary (BMDMs) cells using RT-PCR after 24 hours of polarization in conditioned media (M0, M1, or M2 media) and confirmed to be a reliable in vitro comparison to primary macrophages (Fig. S1). Specifically, iBMM macrophages were cultured on plates coated with ECM powder for 24 hours in growth medium, or medium supplemented with 200 ng/ml E. coli lipopolysaccharide (LPS 055:B5, Sigma) and 20 ng/ml interferon gamma (IFNγ, Peprotech) or 20 ng/ml interleukin-4 (IL-4, Peprotech), for

M1 and M2 polarizations, respectively. Prior to cell seeding, ECM powder was resuspended to 4-5 mg/ml in distilled water, and coated (1 ml/well) on 6-well plates by allowing the solution to air dry. After dry, plates were sterilized under UV and rinsed with

1XPBS directly before cell seeding to remove non-adhered particles. iBMM macrophages performed similarly to BMDM in gene expression studies (upregulating

Tnfa Il1b and Inos in M1 conditions, and Arg1 and Retnla in M2 conditions). Additionally

147 they adopted characteristic morphological changes and expressed CD86 during M1 stimulation and CD206 during M2 stimulation (see Flow Cytometry: In vitro screening section for more detail).

Flow Cytometry: In vitro screening

In vitro samples were harvested using Accutase (Life Technologies) and washed with cold 1X PBS. Then, cells were stained with the following antibody panel: F4/80 PE-Cy7

(BioLegend), CD11b Pacific Blue (BioLegend), CD11c APC-Cy7 (BD Biosciences),

CD86 AlexaFluor700 (BioLegend), MHCII (I-A/I-E) AlexaFluor488 (BioLegend), CD206

APC (BioLegend) and LIVE/DEAD® Fixable Aqua Dead Cell Stain Kit (Life

Technologies). Samples were fixed using the BD Cytofix/Cytoperm TM kit (BD

Biosciences), and run on BD LSRII Cell Analyzer, data was analyzed using FlowJo Flow

Cytometry Analysis Software (Treestar). M1/M2 polarization levels were determined by mean fluorescence intensity of CD86 and CD206 in LIVE/DEAD® Fixable Aqua Dead

Cell Stain -F4/80+CD11c- cells.

Volumetric Muscle Loss (VML) Surgery

Six- to eight-week-old female wild type C57BL/6 (Charles River), B6.129S7-

Rag1tm1Mom/J, BALB/c-Il4ratm1Sz/J, or B6.129S2-Cd4tm1Mak/J (Jackson Laboratories) mice were anesthetized with 4.0% isoflurane and maintained under 2.5% isoflurane. Hair was removed from the lower extremities with an electric razor (Oster). After ethanol sterilization of the surrounding skin, a 1.5-cm incision was created between the knee and hip joint to access the quadriceps femoris muscle. Through the use of surgical scissors, a 3 mm x 3 mm deep defect was created in the quadriceps femoris muscle group. The resulting bilateral defects were filled with 0.05 cc of a 200 - 300 mg/ml biomaterial scaffold (UV-sterilized ECM (manufactured in house) or Collagen (Sigma)) or 0.05 cc of

148 a vehicle (saline) control. Mice were under anesthesia for 10 – 15 minutes during surgical preparation and procedure before return to cage and monitored until ambulatory. Directly after surgery, mice were given subcutaneous carprofen (Rimadyl®,

IACUC protocol.

T cell Adoptive Transfer

CD4+ T cells were isolated from lymph nodes and spleens of wild type C57BL/6 and

B6.Rictor-/- mice (a gift from Jonathan Powell, created by crossing B6.RictorF/F with

B6.Cd4-cre) using MACS CD4+ T Cell Isolation Kit (Miltenyi Biotec) as per manufacturers instructions. Purity was confirmed by staining with the following FACS antibody panel: CD3 AlexaFluor488, CD4 PE-Cy7, CD8 APC (Biolegend). Three million

CD4+ T cells were injected into B6.129S7-Rag1tm1Mom/J. After 12 days post-injection, mice were tested for T cell presence and CD4/CD8 purity in peripheral blood to confirm repopulation. After 2 weeks, muscle surgery was performed as per previously described.

RT-PCR

In vivo inguinal and axillary/brachial lymph node samples from volumetric muscle loss

(VML) studies were homogenized in TRIzol and RNA was extracted using a combination

149 of TRIzol and RNeasy Mini (Qiagen) column-based isolations. cDNA was synthesized through the use of SuperScript Reverse Transcriptase III (Life Technologies) as per manufacturer’s instructions. RT-PCR was conducted on an Applied Biosystems Real

Time PCR Machine using SYBR Green (Life Technologies) as a reporter and the following primers: B2m forward CTC GGT GAC CCT GGT CTT TC, B2m reverse GGA

TTT CAA TGT GAG GCG GG; Tnfα forward GTC CAT TCC TGA GTT CTG, Tnfα reverse GAA AGG TCT GAA GGT AGG; Il1β forward GTA TGG GCT GGA CTG TTT C,

Il1β reverse GCT GTC TGC TCA TTC ACG; Retnla forward CTT TCC TGA GAT TCT

GCC CCA G, Retnla reverse CAC AAG CAC ACC CAG TAG CA; Ifnγ forward TCA AGT

GGC ATA GAT GTG GAA, Ifnγ reverse TGA GGT AGA AAG AGA TAA TCT GG; Il4 forward ACA GGA GAA GGG ACG CCA T, Il4 reverse ACC TTG GAA GCC CTA CAG

A. Whole muscle samples were processed similarly to lymph nodes to isolate RNA and produce cDNA. Primers used included those previously described and: Arg1 forward

CAG AAG AAT GGA AGA GTC AG, Arg1 reverse CAG ATA TGC AGG GAG TCA CC;

Col1a1 forward CTG GCG GTT CAG GTC CAA T, Col1a1 reverse TTC CAG GCA ATC

CAC GAG C; Fabp4 forward TCA CCT GGA AGA CAG CTC CT, Fabp4 reverse AAT

CCC CAT TTA CGC TGA TG; AdipoQ forward TCC TGG AGA GAA GGG AGA GAA

AG, AdipoQ reverse TCA GCT CCT GTC ATT CCA ACA T; Lep forward TTC ACA CAC

GCA GTC GGT AT, Lep reverse ACA TTT TGG GAA GGC AGG CT; Actb forward ATG

TGG ATC AGC AAG CAG GA, Actb reverse AAG GGT GTA AAA CGC AGC TCA

(Integrated DNA Technologies). F4/80+ and CD3+ cells from volumetric muscle wounds were sorted directly into RNA lysis buffer; RLT buffer (Qiagen) + β-mercaptoethanol

(Sigma). RNA was isolated using an RNeasy Micro Kit (Qiagen) with carrier RNA and on-column DNase treatment. cDNA synthesis was performed with a High Capacity cDNA Reverse Transcription Kit (Applied Biosystems). Isolated RNA underwent preamplification prior to plating in custom 96-well TaqMan® Array Fast Plates (Life

150

Technologies) and gene expression was detected on an Applied Biosystems StepOne

Real-Time PCR System.

Histology

Inguinal lymph nodes were harvested and fixed in formalin overnight before dehydration and paraffin embedding, microtome sectioning, then histological examination via hematoxylin and eosin staining. Muscle samples were prepared as fresh-frozen samples for cryosectioning by flash freezing in isopentane after mounting in Tragacanth gum

(Sigma Life Science). A Microm HM 550 cryostat (Fisher Scientific) was used to obtain

10 μm cryosections from 5-7 different regions of each muscle roughly 300 μm apart.

Sections were stained with a Hematoxylin and Eosin protocol (Sigma Aldrich) or with a

Modified Masson’s Trichrome protocol.

Flow Cytometry

Muscle wounds and surrounding area were harvested at 1 (7 days), 3 (24 days) and 6

(42 days) weeks post-surgery by cutting the quadriceps muscle from the hip to the knee and finely diced in 1X PBS. Resultant material was digested for 45 minutes at 37oC in

1.67 Wünsch U/ml Liberase TL (Roche Diagnostics) + 0.2 mg/ml DNase I (Roche

Diagnostics) in serum-free RPMI-1640 medium (Gibco) on a shaker at 400 rpm. Digest was filtered through a 100 μm cell strainer (Fisher) then washed twice with 1X PBS.

Cells were resuspended in 5 ml 1X PBS and layered atop 5 ml Lympholyte-M

(Cedarlane), then spun for 20 minutes at 1200 x g. Cellular interphase was washed twice with 1X PBS. Isolated cells were stained with the following antibody panel:

LIVE/DEAD® Fixable Aqua Dead Cell Stain Kit (Life Technologies), CD19 BrilliantViolet

421 (BioLegend), CD3 AlexaFluor 488 (BioLegend), CD11c APC-Cy7 (BD Biosciences),

F4/80 PE-Cy7 (BioLegend), CD86 AlexaFluor700 (BioLegend), CD206 APC

151

(BioLegend). After staining cells were fixed and analyzed on a BD LSR Analyzer (BD

Biosciences). LIVE/DEAD® Fixable Aqua Dead Cell Stain negative (live) cells were evaluated based upon percent population of T cells (CD3+), B cells (CD19+), dendritic cells (CD11c+), and macrophages (F4/80+). Macrophages were further analyzed for polarization by mean fluorescence intensity of F4/80+, CD11c+ and F4/80+CD11c+ cells in CD86 AlexaFluor700 and CD206 APC channels. All analyses were performed in

FlowJo Flow Cytometry Analysis Software (Treestar). The T cell panel included:

LIVE/DEAD® Fixable Aqua Dead Cell Stain Kit (Life Technologies), CD3 AlexaFluor488

(BioLegend), CD4 PE-Cy7 (BioLegend), CD8 AlexaFluor 700 (BioLegend), FoxP3

Pacific Blue (BioLegend), IL4ra PE (BioLegend) and CCR5 APC (BioLegend). FoxP3 staining followed fixation and permeabilization with BD CytoFix/CytoPerm Kit (BD

Biosciences). Samples prepared for sorting of F4/80+ and CD3+ cells followed the same isolation, then were stained with Fixable Viability Dye eFluor®780 (eBioscience), F4/80

PE-Cy7 (BioLegend), CD11c APC-Cy7 (BD Biosciences) and CD3 AlexaFluor488

(BioLegend). Samples were run on a BD FACS Aria and collected directly into RLT lysis buffer (Qiagen) containing β-mercaptoethanol (Sigma), and stored at -80OC until RNA isolation.

Statistical Analysis

All samples are representative of n = 4 mice and are representative of at least 2 independent experiments unless otherwise stated. Data are displayed as mean ± standard error of the mean. Statistical outliers were removed using Grubbs’ outlier test at alpha = 0.05 using GraphPad Prism v6 Software (GraphPad Software Inc., La Jolla, CA).

Two-way ANOVAs were performed (GraphPad Prismv6), with statistical significance designated at p ≤ 0.05. For multiple comparisons, Tukey or Dunnet post-test corrections were applied. For gene expression analyses of sorted CD3+, F4/80+ WT, and F4/80+

152

Rag1-/- cells, scatter plots, heat maps, and correlation matrices of gene expression levels were used to compare across different materials: Saline, Bone, Cardiac, Collagen. To distinguish which groups of genes were differentially expressed based on material vs. saline, we used a re-sampling based permutation test based on the maximum Wilcoxon

Rank Sum statistic within the gene group. Individual gene expression was also compared across material and saline using the Wilcoxon Rank Sum test. We compare expression between F4/80 WT and Rag1-/- for each material using the Wilcoxon Rank

Sum test. We compare the difference of each material and saline between F4/80 WT and Rag1-/- using linear regression models (material by Rag1-/--status interaction). Due to the exploratory nature and the small sample size, adjustment for multiple comparisons was not considered. Statistical analyses were performed using the R statistical package

(version 2.15.1). Power analysis was not conducted to determine sample size.

Micro-CT Imaging

Imaging was conducted using the Sedecal SuperArgus 4R PET/CT system. We acquired 720 projection images over 360 degrees in 0.5 degree increments; the maximum resolution mode was used, which means each acquired projection image is magnified 5.5 times when compared with the object. The x-ray tube was set with 50 kVp and 100 µA. Exposure time for each projection was 350 ms. Projection images were stored in a matrix with dimensions 1536x972, with 0.15 mm pixel size. CT images were reconstructed using Cobra reconstruction software. Each CT image was stored in a matrix with size 1344x1344x864 with voxel size 0.031 mm. Each image in figure was contrast-enhanced to show defects, the same enhancements were applied for each image.

Treadmill Testing for Muscle Function

153

Forty-eight hours prior to testing, mice were trained on treadmill apparatus running at 5 m/min and increased by 1 m/min every minute for a total of 5 minutes. Mice were run to exhaustion starting at 5 m/min and increased by 1 m/min every minute. Exhaustion was defined as when the mouse stayed on the pulsed shock grid for a continuous 30 seconds (Treat NMD). Animals were tested at least 48 hours prior to harvesting for analysis via FACS, PCR, or histology.

Main Text

Immune homeostasis is indispensable to tissue development, regeneration and repair

(19). Trauma initiates a cascade of local and systemic immune events that trigger the mobilization of cells into the damaged site to initiate host defense and tissue repair. The limited success achieved to date in rebuilding human tissues may be due in part to the tendency for therapeutic strategies to target later processes in wound healing and regeneration, such as stem cell differentiation. Conversely, the immune system is a highly flexible network that serves as a guardian of tissue integrity, and is adapted to the nature of the local microenvironment (2). The immune system participates in tissue repair by scavenging debris and dead cells (100), recruiting and supporting proliferation of tissue progenitor cells (101), and inducing vascularization (102). Previously, immune responses to biomaterials were related to rejection (27-29), however subsets of innate immune cells have been identified as important mediators of scaffold remodeling (17, 37,

38) and can be targeted for immune-mediated tissue regeneration. Here we explore the role of adaptive immunity in tissue regeneration, identifying Th2 responses as critical in driving repair of traumatic tissue injury.

To model a traumatic wound, we surgically excised a portion of the quadriceps muscle group in C57BL/6 mice, provoking an irreversible volumetric muscle loss (VML) injury

154

(103). Based on the regenerative potential and clinical use of tissue-derived extracellular matrix (ECM) scaffolds (17, 38), we screened and selected bone- and cardiac muscle- derived tissue ECM scaffolds (B-ECM and C-ECM) for their immunomodulatory properties (Fig. S1, (98)). The presence of scaffolds in damaged muscle significantly increased the number of myeloid cells (F4/80+ macrophages and CD11c+ dendritic cells;

P < 0.0001) and lymphocytes (CD3+ T cells and CD19+ B cells; P < 0.05) present at the injury site compared to a saline treated control after 1 and 3 weeks (Fig. 1a, Fig. S2). At

1 week, collagen treated wounds recruited the highest number of immune cells into the defect region (36.0% of total live cells, 13.6 million cells) followed by B-ECM and C-ECM treated wounds (39.3%, 5.32 million and 45.4%, 5.44 million), with saline treated wounds containing significantly fewer cells (36.4%, 0.97 million). The proportion of myeloid cells in the damaged muscle peaked at 1 week post injury and the T cell fraction, consisting of both CD4+ and CD8+ cells, peaked in all treatment groups at 3 weeks post injury. In the muscle wound, biomaterial scaffolds skewed the ratio of CD4:CD8 T cells towards a higher fraction of CD4+ helper T cells (~70% in scaffold treated, versus ~50% in saline treated wounds) at 1 week post injury (Fig. 1b). CD4+FoxP3+ regulatory T cells were also present in low levels and increased over time (Fig. S3).

Expression of Interleukin 4 (Il4), a canonical type 2 helper T cells (Th2) cytokine that is also important in muscle healing (12, 21, 104, 105), increased in the presence of the scaffold (Fig. 1c). Therefore we sought to understand the role of cells of the adaptive immune system on formation of the regenerative immune microenvironment. When scaffolds were implanted into B6.129S7-Rag1tm1Mom/J (Rag1-/-) mice, which lack mature T and B cells. In Rag1-/- mice, scaffold-mediated Il4 up-regulation was lost, suggesting a

Th2-driven scaffold immune microenvironment. CD3+ cells were sorted out of muscle injuries at 1 week post-injury for detailed gene expression analysis (Fig. 1d; Fig. S4

155

Table S1). Scaffolds induced a Th2-type gene expression profile as characterized by increased Il4 expression and decreased expression of Ifng and Tbx21 (Th1 canonical genes). In addition, Jag2, which encodes the Notch ligand Jagged 2, was elevated.

Jagged2 helps direct Th differentiation away from Th1 and toward Th2 (106). Il10, which encodes a general anti-inflammatory cytokine that is not Th-specific, was also up- regulated. Other genes that are more selectively expressed by Th1 cells, such as Fasl and Cd28 (the co-stimulatory receptor for CD86), were likewise down-regulated.

The regenerative outcome of tissue-derived ECM scaffolds in animals and humans is correlated with an immunoregulatory M2 macrophage phenotype during remodeling (17,

37, 38). Biomaterials increased expression of genes associated with a pro-regenerative type 2 immune response including hallmark genes of M2 myeloid cells, more specifically macrophages that are stimulated by IL-4, known as M(IL-4) macrophages (Fig. S5)(3).

As with Il4 expression, induction of these M(IL-4) markers was almost completely lost in

Rag1-/- mice (Fig. S5). In the presence of adaptive immune cells, biomaterial scaffolds inhibited macrophage CD86 up-regulation (a co-stimulatory molecule expressed at high levels by classical “M1” macrophages) at 3 weeks post surgery (Fig. 2a, SFig 6h). In

Rag1-/- mice, however, down-regulation of CD86 expression was mitigated, and ECM scaffold treated wounds returned to a macrophage polarization profile resembling that of saline treated control animals (Fig. 2a), On the other hand, CD206 expression (a mannose receptor and classical “M2” marker) was similar between ECM scaffold-treated and saline-treated mice post-implantation at one week, with increased expression at 3 weeks (Fig. 2b). However, this increase in CD206 was ablated in Rag1-/- mice in both scaffold- and saline-treated wounds, suggesting that the adaptive immune system also has a scaffold-independent role in shaping the wound healing response. Moreover, this

CD206 up-regulation was also impaired in B6.129S2-Cd4tm1Mak/J mice (Cd4-/-), which

156 maintain B cells and CD8+ T cells but lack CD4+ helper T cells (Fig. S6a). Additionally, in

Cd4-/- mice, the recruitment of B cells (a commonly Th2-driven adaptive effector cell) was diminished (Fig. S6g).

To further elucidate the role of CD4+ T cells, and more specifically Th2 T cells on polarization of myeloid cells, we evaluated myeloid CD206 expression in Rag1-/- mice that were repopulated with either WT CD4+ T cells or Rictor-/- CD4+ T cells (Fig. 2c-d;

Fig. S7). Rictor is a critical component of the mTORC2 complex that integrates signals from the environment and drives the polarization of Th2 cells (107). Myeloid cells in

Rag1-/- mice expressed lower levels of CD206 compared to WT mice; however, when repopulated with WT CD4+ T cells (T-WT), this phenotype was rescued. When mice received Th2-deficient T cells (T-Rictor-/-) CD206 expression was not rescued, proving that Th2 T cells, dependent upon mTORC2 signaling, are necessary for pro-regenerative myeloid polarization. To confirm the role of IL-4 in Th2-dependent myeloid polarization, we characterized the phenotype of macrophages in BALB/c-Il4ratm1Sz/J (Il4ra-/-) mice that cannot receive signals from IL-4 (Fig. 2c-d). Compared to WT controls, myeloid cells in

Il4ra-/- wounds expressed far lower levels of CD206, suggesting that the macrophage activation was controlled by IL-4, and verifying that the pro-regenerative profile is associated with M(IL-4) cells.

The pleiotropic nature of immune responses typically results in complex expression profiles beyond stereotypical M1 versus M2 “poles” (3). The expression of CD86 and

CD206 on macrophages (classically M1 versus M2) were not mutually exclusive, however these scaffold-associated macrophages also up-regulated the expression of genes encoding Arg1 and Retnla (encoding Fizz1), similar to the results from qRT-PCR analyses of the whole wound (Fig. 2e; Fig. S5). Additionally, Cebpb, and Timp1 were up-

157 regulated, while Mmp16 and Mmp9 were down-regulated further suggesting a pro- regenerative function of the scaffold-associated macrophages (108-110). In Rag1-/- mice, which cannot mount a Th2 immune response, scaffold-associated macrophages lost their pro-regenerative transcriptome (Fig. 2f; Fig. S8; Table S2). Several genes directly implicated in muscle regeneration such as Igf1 (insulin-like growth factor-1) (111-113) and Vegfa (vascular endothelial growth factor) (114) decreased significantly in Rag1-/- mice. Gene ontology enrichment analysis of genes differentially expressed in Rag1-/- versus WT macrophages enriches programs associated with morphogenesis and differentiation, suggesting a reliance on the adaptive immune system for up-regulation of developmentally active immune genes (Fig. S9).

The detection of a local scaffold-associated Th2 polarization led us to investigate the potential systemic T cell response (115). Subcutaneous scaffold implants produce a systemic Th2-like response in the bloodstream, however the connection to wound healing and regeneration is unknown (115, 116). (Fig. 3; Fig. S10). Scaffold treatment induced hypertrophy of local draining lymph nodes (Fig. 3a), which accompanied a robust increase in Il4 expression (Fig. 3b; Fig. S10). This Il4 induction was absent at 1 week post-injury in Rag1-/- mice but present after 3 weeks, suggesting an early adaptive immune-dependent Il4 up-regulation followed by an innate immune driven Il4 up- regulation later in the wound healing and regeneration processes. Additionally, Cd4-/- mice displayed a significant decrease in scaffold-mediated Il4 up-regulation in inguinal lymph nodes at 3 weeks post-injury in C-ECM treated animals (Fig. 3b). This Il4 expression level was higher than that of Rag1-/- mice, demonstrating an important role of

CD4+ T cells in scaffold-induced systemic type 2 immunity, but with potential further contributions by B cells or CD8+ T cells.

158

Functionally, WT animals recovered to run distances similar to healthy uninjured counterparts after 6 weeks (Fig. 4a). However, this restoration in running capacity was ablated in the absence of T and B cells (Rag1-/-) in ECM scaffold treated wounds. At 3 weeks post-injury, repopulation of Rag1-/- mice with WT T cells rescued their functional capacity and the animals could run greater distances compared to mice lacking the CD4 subset (Fig. 4b; 91.11±3.83 vs 60.06±9.69, P=0.0032). Furthermore, Rag1-/- mice repopulated with wild type CD4 T cells performed better than those repopulated with

Rictor-/- CD4+ T cells (72.31±7.40, P=0.0368), confirming the role of Th2 CD4+ T cells in functional muscle regeneration.

Muscle structure correlated with the differences in functional capacity. Histologically at 6 weeks post-injury, the quadriceps muscle treated with C-ECM scaffold appeared similar to that of healthy controls, with minimal scaffold visible and repair tissue fully integrated within the surrounding musculature. A large region of fibrous tissue with active inflammation was present in muscles treated with the collagen scaffold (Fig. 4c, Fig.

S11). Rag1-/- mice displayed increased adipose deposition, fibrosis, scaffold persistence, and smaller diameter muscle fibers compared to their wild type counterparts. At 3 weeks post-injury, centrally nucleated muscle fibers, which are indicative of active regeneration or recovery from injury, were present within the biomaterial scaffold and around the defect site (Fig. 4d, Fig. S12). Wild type mice produced muscle with larger, more rounded fibers, whereas Rag1-/- mice muscle contained smaller, irregularly shaped fibers, indicating a defect in muscle regeneration. In addition, the pathologic Rag1-/- histo-morphology was recapitulated in Cd4-/- mice, confirming the role of CD4+ T cells in fibro-adipogenic lineage commitment (Fig. 4d). Increased gene expression of Adipoq

(adiponectin) confirmed ectopic adipogenesis in Rag1-/- whole muscle. Similarly, the expression of Col1a1 (Type I Collagen) increased in Rag1-/- mice muscle highlighting

159 increased fibrosis (Fig. 4e; Fig. S13). While scaffold treatment reduced fibro- and adipogenesis markers in WT animals, this benefit was lost in Rag1-/- mice.

Here, we demonstrated that biomaterial scaffolds enhance the development of a pro- regenerative immune environment and implicate adaptive immune cells, specifically mTORC2-dependent CD4+ Th2 T cells, in the process of functional tissue restoration

(Fig. S14). Just as cancer research has made great strides in T cell therapies, these concepts can be translated to biomaterials design to improve tissue repair and regeneration (117-120).

160

Figure 5.1: Biomaterial scaffolds induce a Th2 response in volumetric muscle wounds

C57BL/6 (WT) and Rag1-/- mice received a critical size quadriceps muscle injury and were treated immediately with 0.05 ml of saline, particulate collagen, bone ECM (B-

ECM), or cardiac ECM (C-ECM). (A) Proportions of myeloid (F4/80+ Macrophages and

CD11c+ Dendritic Cells) and lymphoid (CD3+ T cells and CD19+ B cells) cell populations in WT wound environment, determined by flow cytometry (% = mean fraction of live cells across all treatments, peak level in bold text). Greatest cell numbers in scaffold treated wounds. (B) Proportion of CD3+ T cells that are CD4+ Helper T cells or CD8+ Cytotoxic T lymphocytes at 1 week post-injury treated with saline, collagen, B-ECM or C-ECM by flow cytometry. (C) qRT-PCR analysis of Interleukin 4 gene expression in WT and Rag1-

/- mice at 1 week post-injury. (D) One week post-injury transcriptome of CD3 cells sorted from wounded muscles treated with saline, collagen, B-ECM or C-ECM determined by

161 qRT-PCR. Data are displayed as RQ to Saline-treated wounds. Data are means ± SEM, n = 4 mice (2 legs pooled per mouse, representative of at least 2 independent experiments), ANOVA ****= P<0.0001, ***= P<0.001, **= P<0.01, * P<0.05

162

Figure 5.2: M(IL-4) pro-regenerative myeloid polarization induced by scaffolds is Th2- dependent.

(A-B) Macrophages in wounded muscle are characterized for CD86 (A) and CD206 (B) expression by flow cytometry at 1 and 3 weeks post-injury in presence of saline or ECM scaffold in WT (Blue bars) and Rag1-/- (red bars) mice. Mean of fluorescence (MFI)(C)

163

CD206 expression at 3 weeks post-injury in C-ECM treated WT, Il4ra-/-, Rag1-/- and

Rag1-/- mice reconstituted with either WT CD4+ T cells (T-WT, n = 2) or Rictor-/- CD4+ T cells (T-Rictr-/-; Th2-deficient). (D) Representative comparison of CD206 expression between WT, Il4ra-/-, Rag1-/- and Rag1-/- reconstituted with WT and Rictor-/- CD4+ T cells.

(E) qRT-PCR Gene expression analysis in cell sorted macrophages from wounded muscles 1 week post-injury and treated with collagen (light grey striped bars), B-ECM

(black solid bars) and C-ECM (grey solid bars) compared to saline control. RQ to saline

= 2-ΔΔCt. (F) RQ to saline in WT and Rag1-/- mice when wounds were treated with C-

ECM. The figure shows a loss of scaffold-mediated macrophage polarization in Rag1-/- mice. WT = blue bars, Rag1-/- = red bars. Data are means ± SEM, n = 4 mice unless otherwise stated (representative of 1 to 2 independent experiments); ANOVA (A-B) and

Students T-test (D) **** = P <0.0001, *** = P <0.001, ** = P <0.01, * = P <0.05

164

Figure 5.3: Systemic immune homeostasis is modified by application of biomaterial scaffolds.

(A) Inguinal lymph node morphology at 1 week post-injury in saline (left) and C-ECM

(right) treated WT type animals. Hematoxylin and eosin (H&E) staining. (B) qRT-PCR analysis of Il4 gene expression in local draining lymph nodes (inguinal, top bar graphs) and distal lymph nodes (axillary/brachial, bottom bar graphs) in WT, Rag1-/-, and Cd4-/- mice at 1 and 3 weeks after wound treatment with collagen, B-ECM and C-ECM. RQ to saline is 2-ΔΔCt. Data are means ± SEM, n = 4 mice (representative of at least 2 independent experiments); ANOVA **** = P <0.0001, ** = P <0.01, * = P <0.05.

165

Figure 5.4: Th2/M(IL-4) responses to biomaterial-treated muscle wound promote functional tissue regeneration

(A) Treadmill exhaustion assay of mice at 6 weeks post-injury to test muscle function in

WT (blue bars) and Rag1-/- (red bars) mice. Normalized to uninjured control (= 100 m). n

= 5 mice per condition and genotype. (B) Treadmill exhaustion at 3 weeks in Cd4-/-, and

Rag1-/- mice repopulated with WT (T-WT) or Rictor-/- (T-Rictr-/-; Th2 deficient) CD4+ T cells. n = 4 mice (Cd4-/-) or n = 10 mice (T-WT and T-Rictr-/-) (C) Transverse section of quadriceps muscle at 6 weeks post-injury in collagen and C-ECM treated WT and Rag1-

/- mice. Black arrow = injury/treatment area. A = anterior, P = posterior, H&E. (D) C-ECM treated VML at 3 weeks post-injury in WT, Rag1-/-, and Cd4-/- mice stained with H&E.

Small muscle fibers and ectopic adipogenesis are present in Rag1-/- and Cd4-/- wounds.

Scale bars = 50 μm. (E) Gene expression (qRT-PCR) of Adipoq (adipose marker) and

Col1a1 (collagen I) showing increased adipose gene expression in Rag1-/- as well as increased collagen gene expression suggesting alterations in connective tissue deposition and possible scarring. n = 4 mice unless otherwise stated (representative of

166 at least 2 independent experiments). Data are means ± SEM; ANOVA (A, D) and

Students T-test (E) **** = P <0.0001, *** = P <0.001, ** = P <0.01, * = P <0.05 *.

167

Supplementary Figure 5.1: Materials characterization and selection.

(A) Extracellular matrix (ECM) scaffold preparation. (B) Histological staining

(hematoxylin & eosin) of tissues pre- and post- ECM processing, top row = native cellular tissue (Native), bottom row = isolated extracellular matrix (Decell). (C-D) In vitro flow cytometric analysis of iBMM (immortalized bone marrow macrophage; Squadrito et. al. 2014) cell line cultured on varying ECM substrates identifies Bone (B-ECM) and

Cardiac (C-ECM) as strong immunomodulatory scaffolds. CD86 = type-1 inflammatory macrophage, CD206 = type-2 alternative macrophage. (E-H) Verification of iBMM cell line. (E-F) qRT-PCR comparing gene expression between iBMM and primary bone- marrow derived macrophages (BMDM) in control M1 (E, LPS + IFNγ) and M2 (F, + IL-4) media conditions. Tnfa, Il1b, Inos = M1 markers. Arg1, Retnla = M2 markers. (G)

Morphological characterization of iBMM in M0, M1 and M2 media conditions. (H) Flow cytometric analysis of iBMM in control polarizing conditions. Data in (D & H) are

168 expressed as fold change over TCP control in the corresponding media condition (M1,

M2, or M0). (E-F) are expressed as fold change over gene expression (2-ΔΔCt) in M0 unstimulated growth media. Data are means ± SEM n = 3.

169

Supplementary Figure 5.2: Cell recruitment to muscle injury.

(A) Gross images of mouse quadriceps muscle at 3 weeks post-operation. (B) Total number of cells infiltrating Saline- and scaffold-treated wounds. (C) Percent of overall cell population identified as F4/80+ macrophages, CD11c+ dendritic cells, CD3+ T cells or

CD19+ B cells. Data are means ± SEM n = 4.

170

+ Supplementary Figure 5.3: FoxP3 Treg populations at 1 and 3 weeks post-operation.

(A) Proportion of FoxP3+CD4+ T cells in the defect at 1, 3, and 6 weeks post-operation.

(B) ANOVA of FoxP3+ cell infiltration over time. Data are means ± SEM n = 4

171

Supplementary Figure 5.4: Data spread of gene expression profiling of CD3+ cells sorted from 1 week post surgery muscle defects.

dCt of WT CD3+ cells. Saline = Black dots. B-ECM = blue dots, C-ECM = red dots,

Collagen = green dots.

172

Supplementary Figure 5.5: M2/M(IL4) Gene expression in scaffold-treated muscle wounds

Biomaterial scaffolds induced the expression of two M2/M(IL4) myeloid genes, Retnla, encoding Fizz1 and Arg1 encoding Arginase 1. ANOVA *** = P <0.001, * = P <0.05.

Data are means ± SEM n = 4.

173

Supplementary Figure 5.6: Myeloid polarization in WT, Rag1-/- and Cd4-/- mice.

(A) Confirmation of participation of CD4+ T cells in M2-myeloid polarization as determined in Rag1-/- studies. CD206 mean fluorescence intensity in F4/80+ macrophages from Cd4-/- mice compared to WT and Rag1-/- mice at 3wks post-injury. (B-

E) Further analysis of WT versus Rag1-/- myeloid polarization. (B) Statistical analysis of overall effect of genotype and scaffold on expression of CD86 and CD206 at 1 and 3 weeks post surgery (C) Two-Way ANOVA comparing CD86 and CD206 expression in scaffold treatment to Saline control wounds at 1 and 3 weeks post surgery (D) Mean

CD86 fluorescence intensity at 1 and 3 weeks post surgery in CD11c+F4/80- and

CD11c+F4/80+ dendritic cells. (E) Mean CD206 fluorescence intensity at 1 and 3 weeks post surgery in CD11c+F4/80- and CD11c+F4/80+ dendritic cells (F) Two-way ANOVA of

CD86 and CD206 expression at 1 and 3 weeks post surgery for CD11c+F4/80- and

CD11c+F4/80+ dendritic cells. (G) CD19+ B cell recruitment, characteristic of Th2 phentoype, dependent on CD4+ T cells. (H) B-ECM and C-ECM behave similarly to

174 clinically used urinary-bladder matrix (UBM) material (Matristem). Decreased CD86 expression on F4/80+ macrophages at 3 weeks post-injury in WT mouse displayed as fold change over Saline control. Data are means ± SEM n = 4.

175

Supplementary Figure 5.7: Adoptive Transfer of CD4+ T cells into Rag1-/- mice.

(A) Timeline of adoptive transfer studies. (B) Purity confirmation of CD4+ T cells after isolation from WT and RictorF/FCd4-Cre mice. (C) Confirmation of adoptive transfer at 11 days post-injection.

176

Supplementary Figure 5.8: Data spread of gene expression profiling of cells sorted from 1 week post surgery muscle defects.

(A) dCt of WT F4/80+ cells. (B) dCt of Rag1-/- F4/80+ cells. Saline = black dots, Bone = blue, Cardiac = red, Collagen = green.

177

Supplementary Figure 5.9: Gene ontology analysis of adaptive immune dependent gene expression changes in SIM F4/80+ macrophages associated with wound healing and tissue regeneration.

Data displayed for genes significant in Fig. S7b (F4/80+ Macrophages), input into

STRING interaction database (121). (A) Gene interaction network. (B) GO processes that are significantly enriched (FDR P-value < 0.05) from genes that alter expression in

Rag1-/- mice related to development and tissue regeneration. (C) Word map showing common terms in GO processes related to development and tissue regeneration.

178

Supplementary Figure 5.10: Gene expression in draining lymph nodes at 1 and 3 weeks post-operation.

Gene expression was measured in local (A, inguinal) and distal (B, axillary/brachial) lymph nodes at 1 and 3 weeks post-operation to measure type-1 (Tnfa, Il1b, Ifng) and type-2 (Il4, Retnla) gene changes dependent upon scaffold application. (C) ANOVA of

WT versus Rag1-/- effect on gene expression. Data are means ± SEM. (n = 4, Saline, B-

ECM, C-ECM; n = 3, collagen).

179

Supplementary Figure 5.11: Computed Tomography imaging reveals irregular muscle density in Rag1-/- mice.

CT imaging of mice at 6 week post-injury shows non-uniform muscle density in Saline treated and C-ECM treated Rag1-/- mice, but uniform muscle in C-ECM treated WT mice.

Top row of images shows location (white box) of zoomed in image below. Dense tissue

= white arrowheads, Less dense tissue = black arrowheads.

180

Supplementary Figure 5.12: Quadriceps muscle at 3 weeks post-operation in WT and Rag1-/- mice.

(A) Hematoxylin and eosin-stained histological sections of unaffected and affected quadriceps muscle immediately after volumetric muscle loss surgery. (B) Increased fibrosis and decreased cellularity in Collagen treated scaffolds in absence of adaptive immune cells (Rag1-/-) at 3 weeks post-injury. (C) Mosaic of quadriceps muscle at 3 weeks post-operation stained with Masson’s trichrome (top) and Hematoxylin and Eosin

(bottom). Scale bars = 50 microns in (b) and 500 microns in (c).

181

Supplementary Figure 5.13: Collagen and adipose-related gene expression increases in Rag1-/- mice.

(A) Col1a1 gene expression at 1 and 3 weeks post injury (B) Adipogenesis gene expression shown as a fold change over WT in corresponding scaffold treatment at 3 weeks post-injury (C-E) Adipogenesis gene expression displayed as a fold change over saline control in (C) collagen, (D) B-ECM and (E) C-ECM treated injuries at 3 weeks post-injury. Data are means ± SEM n = 4. ANOVA *** = P <0.001, ** = P <0.01, * = P

<0.05 WT versus Rag1-/-. Panel A, Ψ = ANOVA vs Saline control.

182

Supplementary Figure 5.14: T cell participation in muscle regeneration and fibro/adipogenic lineage commitment.

T cell activation and polarization induce local Th2/M(IL-4) polarization of the SIM, promoting regenerative phenotypes such as wound healing and myotube fusion, and inhibit intramuscular adipose formation and collagen deposition. We hypothesize that the process begins with an innate response, in which ECM components induce a partial M2- like macrophage differentiation and simultaneously present ECM protein-derived peptides to T cells together with IL-4 production that drives Th2 differentiation. Th2 cells then significantly recruit and enhance M2 responses at the site of wound healing, forming a feed-forward amplification

183

Supplementary Table 5.1. Wilcoxon Rank Sum Test Results on sorted CD3+ T cells

(A) CD3+ Genes analyzed in scaffold treated vs. saline control. (B) CD3+ gene group analysis.

184

Supplementary Table 5.2. Wilcoxon Rank Sum and Linear Regression Test Results on sorted F4/80+ macrophages in WT and Rag1-/- mice.

(A) Wilcoxon Rank Sum test on F4/80+ Genes analyzed in scaffold treated vs. saline control. (B) Linear Regression model on material by genotype interaction. WT vs. Rag1-/- comparison in effect of scaffold on gene expression changes versus saline treated control, p-values displayed for significant comparisons (P < 0.05 in at least one scaffold treatments).

185

Chapter 6: Conclusions and Future Directions

186

Conclusions and Future Directions

Implications of these data presented span from the laboratory to the clinic. In the laboratory, many materials are tested in immunodeficient mice, be it fully immunodeficient or Rag1/2-/- (adaptive immune deficient) mice. As shown in the aforementioned studies, we have proven a critical role of the adaptive immune system in recognition and response to biomaterials, as well as the regenerative outcome of those scaffolds. Without adaptive immune cells present, regeneration in a murine muscle loss model is severely hindered. From the literature, we know that cells from the innate immune system are also critical in complex tissue development and wound healing. If materials are tested in immunodeficient mice, their immune microenvironment and resulting effects on stem cell differentiation and tissue development cannot be determined. Therefore, more pre-clinical work must focus on immune-competent animals to accurately describe the role of the material in regeneration. The main usage that warrants immunodeficient animals is in the case of implantation of xenogeneic stem cells, often human, into a model organism such as the mouse. In this case, immune cells that we have recognized as important in biomaterial responses and tissue regeneration can detect the foreign stem cells and target them for destruction. Without the immune system present, cells are able to develop under physiologic conditions that are more translatable than an in vitro system. Given this necessity, immunodeficient strains will prove to be important in stem cell differentiation studies involving a xenogeneic donor, but it is important to consider any side-effects of an absent immune system, and non-cell laden material, if possible, should be used in addition to take into account any responses by the adaptive immune system.

Additionally, further studies will involve a further dissection and analysis of the scaffold immune microenvironment (SIM) created by biomaterial scaffolds. We have a broad overview of the different cells infiltrating a scaffold-treated wound at different time

187 points, but focus directly on the mTORC2-dependent Th2 lineage CD4+ T cells.

Determination of a more precise M1/M2 profile, with an increased number of polarization and functional markers will improve our understanding of the macrophage dynamics in the SIM. As macrophages receive signals from the environment through pattern recognition receptors such as sugar/carbohydrate receptors (such as mannose receptor) and TLRs (Toll-like receptors, TLR2 can recognize fragmented hyaluronic acid found in damaged extracellular matrix) it will be of interest to analyze the effect of macrophages on the Th2 T cell polarization discussed in this work. We hypothesize that scaffold- associated macrophages (SAMs) integrate signals from their environment and polarize towards a pro-M2 phenotype with low levels of M2-associated gene and protein expression. These pro-M2 SAMs then interact with T cells by presenting processed antigen, educating the T cells to polarize in a Th2 fashion. After the Th2 T cells have been activated and begin secreting large amounts of IL-4, there is feedback to the SAMs present in the environment that further polarize them towards a full pro-regenerative

SAM phenotype. Thus far, we have defined the ECM-associated pro-regenerative SAM as F4/80+CD11b+CD11c=/-CD86+CD206hiMHCII+ expressing high levels of Arg1, Retnla,

Cebpb, and Ccl5. In addition to innate immune cells, further investigation is necessary for other cells of the adaptive immune system, such as B cells and CD8+ T cells.

When speaking of “biomaterials” and “scaffolds” we are referring to a broad spectrum of different materials used in bioengineering. Our focus was on ECM scaffolds as they have shown promise in the clinic and promote constructive immune responses.

However, there are a large variety of materials used in bioengineering both synthetic and naturally-derived as discussed in the introduction. The immune microenvironment of synthetic scaffolds could differ greatly compare to naturally-derived ECM scaffolds. In the case of synthetics, one could imagine altering the base composition (using PLGA,

PCL, PE, PEG, etc.) or chemical modifications and altering the immune response to that

188 scaffold. The same principle applies to ECM-derived scaffolds. The strength of the

Decellularization process will alter the molecular components of the ECM. With a stronger detergent, such as SDS, you will remove much more cellular material, but at the same time you will be removing extracellular matrix molecules such as glycosaminogylcans and proteoglycans, thereby modifying the ECM structure. This also applies to acid treatments. With increasing acidification of the ECM, you will remove more of the acid-soluble collagens while leaving behind the acid insoluble collagens. As with synthetics, you can chemically modify the ECM scaffolds, adding functional groups and biologically active small molecules or proteins. All of these modifications will in turn modify the immune response thereby altering the SIM associated with that specific material formulation. All immune cell responses to this material may be altered in different manners. A modification that alters the response of a macrophage may not have a direct effect on T cells, but through interactions of the different cells, it could cause a downstream alteration through cell-cell signaling and cytokine secretion.

Clinically, we have defined a new avenue of manipulation to promote regeneration in biomaterial scaffolds. Future work will include identification of immunomodulatory reagents focused on the adaptive arm of the immune system and how to leverage immune activation and polarization with regeneration.

189

References

1. M. E. Bianchi, DAMPs, PAMPs and alarmins: all we need to know about danger. Journal of leukocyte biology 81, 1-5 (2007). 2. P. Matzinger, T. Kamala, Tissue-based class control: the other side of tolerance. Nature reviews. Immunology 11, 221-230 (2011). 3. P. J. Murray et al., Macrophage activation and polarization: nomenclature and experimental guidelines. Immunity 41, 14-20 (2014). 4. H. L. Weiner, Oral tolerance: immune mechanisms and the generation of Th3- type TGF-beta-secreting regulatory cells. Microbes and infection / Institut Pasteur 3, 947-954 (2001). 5. A. T. Clark et al., Successful oral tolerance induction in severe peanut allergy. Allergy 64, 1218-1220 (2009). 6. C. Berlin et al., alpha 4 integrins mediate lymphocyte attachment and rolling under physiologic flow. Cell 80, 413-422 (1995). 7. M. R. Bono et al., The essential role of chemokines in the selective regulation of lymphocyte homing. Cytokine & growth factor reviews 18, 33-43 (2007). 8. S. J. Ono et al., Chemokines: roles in leukocyte development, trafficking, and effector function. The Journal of allergy and clinical immunology 111, 1185-1199; quiz 1200 (2003). 9. J. Y. Niederkorn, See no evil, hear no evil, do no evil: the lessons of immune privilege. Nature immunology 7, 354-359 (2006). 10. J. Ma et al., Regulation of macrophage activation. Cellular and molecular life sciences : CMLS 60, 2334-2346 (2003). 11. A. B. Roberts et al., Transforming growth factor-beta: multifunctional regulator of differentiation and development. Philosophical transactions of the Royal Society of London. Series B, Biological sciences 327, 145-154 (1990). 12. V. Horsley, K. M. Jansen, S. T. Mills, G. K. Pavlath, IL-4 acts as a myoblast recruitment factor during mammalian muscle growth. Cell 113, 483-494 (2003). 13. J. W. Godwin, A. R. Pinto, N. A. Rosenthal, Macrophages are required for adult salamander limb regeneration. Proceedings of the National Academy of Sciences of the United States of America 110, 9415-9420 (2013). 14. J. Leleux, K. Roy, Micro and nanoparticle-based delivery systems for vaccine immunotherapy: an immunological and materials perspective. Advanced healthcare materials 2, 72-94 (2013). 15. J. A. Hubbell, S. N. Thomas, M. A. Swartz, Materials engineering for immunomodulation. Nature 462, 449-460 (2009). 16. S. F. Badylak, J. E. Valentin, A. K. Ravindra, G. P. McCabe, A. M. Stewart-Akers, Macrophage phenotype as a determinant of biologic scaffold remodeling. Tissue engineering. Part A 14, 1835-1842 (2008). 17. B. N. Brown et al., Macrophage phenotype as a predictor of constructive remodeling following the implantation of biologically derived surgical mesh materials. Acta biomaterialia 8, 978-987 (2012). 18. B. N. Brown, B. D. Ratner, S. B. Goodman, S. Amar, S. F. Badylak, Macrophage polarization: an opportunity for improved outcomes in biomaterials and regenerative medicine. Biomaterials 33, 3792-3802 (2012). 19. T. A. Wynn, A. Chawla, J. W. Pollard, Macrophage biology in development, homeostasis and disease. Nature 496, 445-455 (2013).

190

20. J. A. Stefater, 3rd, S. Ren, R. A. Lang, J. S. Duffield, Metchnikoff's policemen: macrophages in development, homeostasis and regeneration. Trends in molecular medicine 17, 743-752 (2011). 21. J. E. Heredia et al., Type 2 innate signals stimulate fibro/adipogenic progenitors to facilitate muscle regeneration. Cell 153, 376-388 (2013). 22. Y. P. Goh et al., Eosinophils secrete IL-4 to facilitate liver regeneration. Proceedings of the National Academy of Sciences of the United States of America 110, 9914-9919 (2013). 23. P. M. Kou, J. E. Babensee, Validation of a high-throughput methodology to assess the effects of biomaterials on dendritic cell phenotype. Acta biomaterialia 6, 2621-2630 (2010). 24. N. Artzi et al., In vivo and in vitro tracking of erosion in biodegradable materials using non-invasive fluorescence imaging. Nature materials 10, 704-709 (2011). 25. A. T. Hillel et al., Photoactivated composite biomaterial for soft tissue restoration in rodents and in humans. Science translational medicine 3, 93ra67 (2011). 26. B. Reid et al., PEG hydrogel degradation and the role of the surrounding tissue environment. Journal of tissue engineering and regenerative medicine 9, 315-318 (2015). 27. J. M. Anderson, Biological responses to materials. Annual Review of Materials Research 31, 81-110 (2001). 28. J. M. Anderson, K. M. Miller, Biomaterial biocompatibility and the macrophage. Biomaterials 5, 5-10 (1984). 29. J. M. Anderson, Inflammatory response to implants. ASAIO Journal 34, 101-107 (1988). 30. D. M. Faulk et al., ECM hydrogel coating mitigates the chronic inflammatory response to polypropylene mesh. Biomaterials 35, 8585-8595 (2014). 31. M. T. Wolf et al., Polypropylene surgical mesh coated with extracellular matrix mitigates the host foreign body response. Journal of biomedical materials research. Part A 102, 234-246 (2014). 32. T. W. Gilbert, T. L. Sellaro, S. F. Badylak, Decellularization of tissues and organs. Biomaterials 27, 3675-3683 (2006). 33. S. F. Badylak, D. O. Freytes, T. W. Gilbert, Extracellular matrix as a biological scaffold material: Structure and function. Acta biomaterialia 5, 1-13 (2009). 34. P. M. Crapo, T. W. Gilbert, S. F. Badylak, An overview of tissue and whole organ decellularization processes. Biomaterials 32, 3233-3243 (2011). 35. M. H. Zheng et al., Porcine small intestine submucosa (SIS) is not an acellular collagenous matrix and contains porcine DNA: possible implications in human implantation. Journal of biomedical materials research. Part B, Applied biomaterials 73, 61-67 (2005). 36. D. O. Freytes, J. Martin, S. S. Velankar, A. S. Lee, S. F. Badylak, Preparation and rheological characterization of a gel form of the porcine urinary bladder matrix. Biomaterials 29, 1630-1637 (2008). 37. B. M. Sicari et al., An acellular biologic scaffold promotes skeletal muscle formation in mice and humans with volumetric muscle loss. Science translational medicine 6, 234ra258 (2014). 38. V. J. Mase, Jr. et al., Clinical application of an acellular biologic scaffold for surgical repair of a large, traumatic quadriceps femoris muscle defect. Orthopedics 33, 511 (2010). 39. L. Zhang et al., Effect of an inductive hydrogel composed of urinary bladder matrix upon functional recovery following traumatic brain injury. Tissue engineering. Part A 19, 1909-1918 (2013).

191

40. J. J. Stankus, D. O. Freytes, S. F. Badylak, W. R. Wagner, Hybrid nanofibrous scaffolds from electrospinning of a synthetic biodegradable elastomer and urinary bladder matrix. Journal of biomaterials science. Polymer edition 19, 635-652 (2008). 41. I. Wu, Z. Nahas, K. A. Kimmerling, G. D. Rosson, J. H. Elisseeff, An injectable adipose matrix for soft-tissue reconstruction. Plastic and reconstructive surgery 129, 1247-1257 (2012). 42. L. W. Dunne et al., Human decellularized adipose tissue scaffold as a model for breast cancer cell growth and drug treatments. Biomaterials 35, 4940-4949 (2014). 43. P. Matzinger, Friendly and dangerous signals: is the tissue in control? Nature immunology 8, 11-13 (2007). 44. B. M. Sicari et al., The promotion of a constructive macrophage phenotype by solubilized extracellular matrix. Biomaterials 35, 8605-8612 (2014). 45. B. N. Brown, B. M. Sicari, S. F. Badylak, Rethinking regenerative medicine: a macrophage-centered approach. Frontiers in immunology 5, 510 (2014). 46. D. V. Pechkovsky et al., Alternatively activated alveolar macrophages in pulmonary fibrosis-mediator production and intracellular signal transduction. Clinical immunology (Orlando, Fla.) 137, 89-101 (2010). 47. A. T. Hillel et al., Dysregulated Macrophages Are Present in Bleomycin-Induced Murine Laryngotracheal Stenosis. Otolaryngology--head and neck surgery : official journal of American Academy of Otolaryngology-Head and Neck Surgery 153, 244-250 (2015). 48. J. D. Bryers, C. M. Giachelli, B. D. Ratner, Engineering biomaterials to integrate and heal: the biocompatibility paradigm shifts. Biotechnology and bioengineering 109, 1898-1911 (2012). 49. J. L. Zakrzewski, M. R. van den Brink, J. A. Hubbell, Overcoming immunological barriers in regenerative medicine. Nature biotechnology 32, 786-794 (2014). 50. V. Ballotta, A. Driessen-Mol, C. V. Bouten, F. P. Baaijens, Strain-dependent modulation of macrophage polarization within scaffolds. Biomaterials 35, 4919- 4928 (2014). 51. J. N. Beck, A. Singh, A. R. Rothenberg, J. H. Elisseeff, A. J. Ewald, The independent roles of mechanical, structural and adhesion characteristics of 3D hydrogels on the regulation of cancer invasion and dissemination. Biomaterials 34, 9486-9495 (2013). 52. M. M. Martino et al., Engineering the growth factor microenvironment with fibronectin domains to promote wound and bone tissue healing. Science translational medicine 3, 100ra189 (2011). 53. A. Singh et al., Enhanced lubrication on tissue and biomaterial surfaces through peptide-mediated binding of hyaluronic acid. Nature materials 13, 988-995 (2014). 54. E. D. Hay, Extracellular matrix. The Journal of Cell Biology 91, 205-223 (1981). 55. F. M. Watt, W. T. Huck, Role of the extracellular matrix in regulating stem cell fate. Nature reviews. Molecular cell biology 14, 467-473 (2013). 56. C. Bonnans, J. Chou, Z. Werb, Remodelling the extracellular matrix in development and disease. Nature reviews. Molecular cell biology 15, 786-801 (2014). 57. G. Y. Lee, P. A. Kenny, E. H. Lee, M. J. Bissell, Three-dimensional culture models of normal and malignant breast epithelial cells. Nat Methods 4, 359-365 (2007).

192

58. Y. Lavin et al., Tissue-resident macrophage enhancer landscapes are shaped by the local microenvironment. Cell 159, 1312-1326 (2014). 59. M. S. Kim et al., A draft map of the human proteome. Nature 509, 575-581 (2014). 60. M. Uhlen et al., Proteomics. Tissue-based map of the human proteome. Science (New York, N.Y.) 347, 1260419 (2015). 61. D. G. Anderson, S. Levenberg, R. Langer, Nanoliter-scale synthesis of arrayed biomaterials and application to human embryonic stem cells. Nature biotechnology 22, 863-866 (2004). 62. Y. Mei et al., Combinatorial development of biomaterials for clonal growth of human pluripotent stem cells. Nature materials 9, 768-778 (2010). 63. C. J. Flaim, S. Chien, S. N. Bhatia, An extracellular matrix microarray for probing cellular differentiation. Nat Methods 2, 119-125 (2005). 64. N. E. Reticker-Flynn et al., A combinatorial extracellular matrix platform identifies cell-extracellular matrix interactions that correlate with metastasis. Nature communications 3, 1122 (2012). 65. S. Gobaa et al., Artificial niche microarrays for probing single stem cell fate in high throughput. Nat Methods 8, 949-955 (2011). 66. A. Dolatshahi-Pirouz et al., A combinatorial cell-laden gel microarray for inducing osteogenic differentiation of human mesenchymal stem cells. Scientific reports 4, 3896 (2014). 67. K. Woodruff, L. M. Fidalgo, S. Gobaa, M. P. Lutolf, S. J. Maerkl, Live mammalian cell arrays. Nat Methods 10, 550-552 (2013). 68. A. Ranga et al., 3D niche microarrays for systems-level analyses of cell fate. Nature communications 5, 4324 (2014). 69. J. K. Mouw, G. Ou, V. M. Weaver, Extracellular matrix assembly: a multiscale deconstruction. Nature reviews. Molecular cell biology 15, 771-785 (2014). 70. A. Atala, S. B. Bauer, S. Soker, J. J. Yoo, A. B. Retik, Tissue-engineered autologous bladders for patients needing cystoplasty. Lancet 367, 1241-1246 (2006). 71. P. Macchiarini et al., Clinical transplantation of a tissue-engineered airway. Lancet 372, 2023-2030 (2008). 72. H. C. Ott et al., Perfusion-decellularized matrix: using nature's platform to engineer a bioartificial heart. Nature medicine 14, 213-221 (2008). 73. B. E. Uygun et al., Organ reengineering through development of a transplantable recellularized liver graft using decellularized liver matrix. Nature medicine 16, 814-820 (2010). 74. T. H. Petersen et al., Tissue-engineered lungs for in vivo implantation. Science (New York, N.Y.) 329, 538-541 (2010). 75. J. J. Song et al., Regeneration and experimental orthotopic transplantation of a bioengineered kidney. Nature medicine 19, 646-651 (2013). 76. J. A. Vizcaino et al., ProteomeXchange provides globally coordinated proteomics data submission and dissemination. Nature biotechnology 32, 223-226 (2014). 77. A. Naba et al., The matrisome: in silico definition and in vivo characterization by proteomics of normal and tumor extracellular matrices. Molecular & cellular proteomics : MCP 11, M111.014647 (2012). 78. P. Yan, W. Seelentag, A. Bachmann, F. T. Bosman, An agarose matrix facilitates sectioning of tissue microarray blocks. The journal of histochemistry and cytochemistry : official journal of the Histochemistry Society 55, 21-24 (2007). 79. K. Yoshimura et al., Integrin alpha2 mediates selective metastasis to the liver. Cancer research 69, 7320-7328 (2009).

193

80. J. B. Mitchell et al., Immunophenotype of human adipose-derived cells: temporal changes in stromal-associated and stem cell-associated markers. Stem cells (Dayton, Ohio) 24, 376-385 (2006). 81. A. M. Bratt-Leal, R. L. Carpenedo, M. D. Ungrin, P. W. Zandstra, T. C. McDevitt, Incorporation of biomaterials in multicellular aggregates modulates pluripotent stem cell differentiation. Biomaterials 32, 48-56 (2011). 82. J. Kononen et al., Tissue microarrays for high-throughput molecular profiling of tumor specimens. Nature medicine 4, 844-847 (1998). 83. A. J. Engler, S. Sen, H. L. Sweeney, D. E. Discher, Matrix elasticity directs stem cell lineage specification. Cell 126, 677-689 (2006). 84. M. J. Bissell, W. C. Hines, Why don't we get more cancer? A proposed role of the microenvironment in restraining cancer progression. Nature medicine 17, 320- 329 (2011). 85. C. A. DeForest, B. D. Polizzotti, K. S. Anseth, Sequential click reactions for synthesizing and patterning three-dimensional cell microenvironments. Nature materials 8, 659-664 (2009). 86. M. W. Tibbitt, K. S. Anseth, Dynamic microenvironments: the fourth dimension. Science translational medicine 4, 160ps124 (2012). 87. S. F. Badylak, G. C. Lantz, A. Coffey, L. A. Geddes, Small intestinal submucosa as a large diameter vascular graft in the dog. The Journal of surgical research 47, 74-80 (1989). 88. R. M. Nerem, Regenerative medicine: the emergence of an industry. Journal of the Royal Society, Interface / the Royal Society 7 Suppl 6, S771-775 (2010). 89. H. Kuribayashi et al., Clinicopathological analysis of primary lung carcinoma with heterotopic ossification. Lung cancer (Amsterdam, Netherlands) 64, 160-165 (2009). 90. M. M. Averill, C. Kerkhoff, K. E. Bornfeldt, S100A8 and S100A9 in cardiovascular biology and disease. Arteriosclerosis, thrombosis, and vascular biology 32, 223- 229 (2012). 91. B. Chazaud, Macrophages: supportive cells for tissue repair and regeneration. Immunobiology 219, 172-178 (2014). 92. A. B. Aurora, E. N. Olson, Immune modulation of stem cells and regeneration. Cell stem cell 15, 14-25 (2014). 93. A. J. Vegas et al., Combinatorial hydrogel library enables identification of materials that mitigate the foreign body response in primates. Nature biotechnology, (2016). 94. S. Reinke et al., Terminally differentiated CD8(+) T cells negatively affect bone regeneration in humans. Science translational medicine 5, 177ra136 (2013). 95. T. W. Gilbert, D. B. Stolz, F. Biancaniello, A. Simmons-Byrd, S. F. Badylak, Production and characterization of ECM powder: implications for tissue engineering applications. Biomaterials 26, 1431-1435 (2005). 96. R. O. Hynes, The extracellular matrix: not just pretty fibrils. Science (New York, N.Y.) 326, 1216-1219 (2009). 97. H. K. Kleinman, D. Philp, M. P. Hoffman, Role of the extracellular matrix in morphogenesis. Current opinion in biotechnology 14, 526-532 (2003). 98. V. Z. Beachley et al., Tissue matrix arrays for high-throughput screening and systems analysis of cell function. Nat Meth 12, 1197-1204 (2015). 99. M. L. Squadrito et al., Endogenous RNAs modulate microRNA sorting to exosomes and transfer to acceptor cells. Cell reports 8, 1432-1446 (2014). 100. Y. F. Peng et al., Innate and Adaptive Immune Response to Apoptotic Cells. Journal of autoimmunity 29, 303-309 (2007).

194

101. J. S. Otis et al., Pro-inflammatory mediation of myoblast proliferation. PloS one 9, e92363 (2014). 102. S. Frantz, K. A. Vincent, O. Feron, R. A. Kelly, Innate immunity and angiogenesis. Circulation research 96, 15-26 (2005). 103. B. M. Sicari et al., A Murine Model of Volumetric Muscle Loss and a Regenerative Medicine Approach for Tissue Replacement. Tissue engineering. Part A 18, 1941-1948 (2012). 104. V. Salmon-Ehr et al., Implication of interleukin-4 in wound healing. Laboratory investigation; a journal of technical methods and pathology 80, 1337-1343 (2000). 105. W. C. Gause, T. A. Wynn, J. E. Allen, Type 2 immunity and wound healing: evolutionary refinement of adaptive immunity by helminths. Nature reviews. Immunology 13, 607-614 (2013). 106. T. C. Fang et al., Notch directly regulates Gata3 expression during T helper 2 cell differentiation. Immunity 27, 100-110 (2007). 107. G. M. Delgoffe et al., The kinase mTOR regulates the differentiation of helper T cells through the selective activation of signaling by mTORC1 and mTORC2. Nature immunology 12, 295-303 (2011). 108. D. Ruffell et al., A CREB-C/EBPbeta cascade induces M2 macrophage-specific gene expression and promotes muscle injury repair. Proceedings of the National Academy of Sciences of the United States of America 106, 17475-17480 (2009). 109. J. Blackwell et al., Changes in CEBPB expression in circulating leukocytes following eccentric elbow-flexion exercise. The journal of physiological sciences : JPS 65, 145-150 (2015). 110. M. E. Davis, J. P. Gumucio, K. B. Sugg, A. Bedi, C. L. Mendias, MMP inhibition as a potential method to augment the healing of skeletal muscle and tendon extracellular matrix. Journal of Applied Physiology 115, 884-891 (2013). 111. F. Mourkioti, N. Rosenthal, IGF-1, inflammation and stem cells: interactions during muscle regeneration. Trends in immunology 26, 535-542 (2005). 112. A. Musaro et al., Localized Igf-1 transgene expression sustains hypertrophy and regeneration in senescent skeletal muscle. Nature genetics 27, 195-200 (2001). 113. J. P. Liu, J. Baker, A. S. Perkins, E. J. Robertson, A. Efstratiadis, Mice carrying null mutations of the genes encoding insulin-like growth factor I (Igf-1) and type 1 IGF receptor (Igf1r). Cell 75, 59-72 (1993). 114. N. Arsic et al., Vascular endothelial growth factor stimulates skeletal muscle regeneration in vivo. Molecular therapy : the journal of the American Society of Gene Therapy 10, 844-854 (2004). 115. A. J. Allman et al., Xenogeneic extracellular matrix grafts elicit a TH2-restricted immune response. Transplantation 71, 1631-1640 (2001). 116. A. J. Allman, T. B. McPherson, L. C. Merrill, S. F. Badylak, D. W. Metzger, The Th2-restricted immune response to xenogeneic small intestinal submucosa does not influence systemic protective immunity to viral and bacterial pathogens. Tissue engineering 8, 53-62 (2002). 117. S. A. Rosenberg, J. C. Yang, N. P. Restifo, Cancer immunotherapy: moving beyond current vaccines. Nature medicine 10, 909-915 (2004). 118. O. A. Ali et al., Identification of immune factors regulating antitumor immunity using polymeric vaccines with multiple adjuvants. Cancer research 74, 1670- 1681 (2014). 119. J. Kim, D. J. Mooney, In Vivo Modulation of Dendritic Cells by Engineered Materials: Towards New Cancer Vaccines. Nano today 6, 466-477 (2011).

195

120. O. A. Ali, P. Tayalia, D. Shvartsman, S. Lewin, D. J. Mooney, Inflammatory cytokines presented from polymer matrices differentially generate and activate DCs. Advanced functional materials 23, 4621-4628 (2013). 121. D. Szklarczyk et al., STRING v10: protein-protein interaction networks, integrated over the tree of life. Nucleic acids research 43, D447-452 (2015).

196

Curriculum Vitae Johns Hopkins University School of Medicine Kaitlyn N Sadtler 17. March. 2016

Educational History Ph.D. expected 2016 Program in Cellular Johns Hopkins and Molecular Medicine School of Medicine Mentor: Jennifer Elisseeff

B.S. 2011 Biological Sciences University of Maryland Baltimore County

Professional Experience Johns Hopkins University School of Medicine, Baltimore, MD Aug. 2012 – Mar. 2016 Laboratory of Dr. Jennifer Elisseeff Graduate Student  Conducted research in Biomedical Engineering, focusing on immune-mediated tissue regeneration  Aided and assisted graduate students & postdocs in immunological analysis, including introduction to flow cytometry techniques.  Created platform of flow cytometric analyses to investigate immune environment of wounds & scaffolds  Experience in product development in creation of an injectable gel treatment for alopecia.  Research mentor to Brian Allen (Undergrad 2014-2015), Liam Chung (Ph.D. Student 2015 - 2016), Chris Moad (Ph.D. Student 2016), Natalie Campbell (Undergrad 2015 - 2016) and Fernando Vicente (Undergrad 2015)

National Institutes of Health, Bethesda, MD June 2011 – Aug. 2012 National Institute of Allergy and Infectious Disease Postbaccalaureate IRTA  Laboratory of Cellular and Molecular Immunology (LCMI), Dr. Ron Schwartz  Research in JAK/STAT cytokine signaling in naïve and in vitro activated murine T cells  Investigating the property of cross-inhibition in common gamma chain family cytokines Interleukin-4 (IL-4), IL-7, and IL-21.

Johns Hopkins University Applied Physics Lab, Laurel, MD June 2009 – Aug 2010 National Security and Technology Department Intern  Research in biodefense-based applications, development of a modular sample screening device for biohazards in foods and soils, development of a protein-ligation based assay to detect staphylococcal enterotoxin, identification of indigenous soil bacteria capable of degrading trace explosives to improve improvised explosive device (IED) detection and decrease false positives on current detectors

Green Valley Animal Hospital, Ijamsville, MD June 2004 – Aug 2008 Veterinary Clinic Kennel/Veterinary Technician  Aided in routine and emergency care for household animals  Shadowed veterinarian during appointments and surgeries

197

Honors and Awards Hyaluronan International Conference, Florence Italy, Travel Award 2015 Tissue Engineering and Regenerative Medicine (TERMIS) AM Travel Award 2014 Graduate Student Association Travel Award 2013 Pollard Scholar in Molecular Biology and Genomics 2013-2014 Graduation Honors: Summa Cum Laude (GPA 3.95 or above) 2011 Outstanding Graduating Senior in the Biological Sciences 2011 Inducted into Phi Kappa Phi Honor Society 2011

Publications  Hillel A, Ding D, Samad I, Ma G, Sadtler K, Powell J, Lane A, Horton M. “Dysregulated macrophages are present in murine and human laryngotracheal stenosis". Otolaryngology - Head and Neck Surgery. 2015 Aug;153(2):244-50  Beachley VZ*, Wolf MT*, Sadtler K*, Manda SS, Jacobs H, Blatchley M, Bader J, Pandey A, Pardoll D, Elisseeff JH. “Tissue and organ matrix arrays for high throughput screening and systems analysis of cell function". Nature Methods. 2015 Dec;12(12):1197-1204 *Authors contributed equally.  Sadtler K, Estrellas K, Allen BW, Wolf MT, Fan H, Tam A, Patel C, Luber B, Wang H, Wagner K, Powell J, Housseau F, Pardoll D, Elisseeff JH. “Developing a Pro- Regenerative Biomaterial Scaffold Microenvironment Requires T helper 2 cells " Science. 2016. Accepted.  Singh A*, Sadtler K*, Wolf MT, Wang X, Pardoll DM, Elisseeff JH. “Tissue-specific biomaterials and immune considerations in regenerative medicine”. Nature Reviews Materials. 2016 *Authors contributed equally. Accepted.

Publications in Submission or Preparation  Sadtler K, Allen BW, Garza L, Elisseeff JH. “Low molecular weight hyaluronan induces anagen through TLR2." PLOS One. 2016. Submitted  Sadtler K, Sommerfeld S, Wang X, Wolf MW, Majumdar S, Elisseeff JH. “Composition and immune microenvironment of urinary bladder matrix scaffolds.” 2016 In Preparation.  Sadtler K, Allen BW, Estrellas K, Housseau F, Pardoll DP and Elisseeff JH. “The Scaffold Immune Microenvironment: Biomaterial-Mediated Immune Polarization in Traumatic and Non-Traumatic Applications." 2016. In Preparation  Gonnord P, Angermann BR, Sadtler K , Gombos E, Chappert P, Meier-Schellersheim M, Varma R. “IL-7 receptor ligation limits the availability of the common gamma chain to other cytokines receptors”. 2016. In Preparation.

Oral Presentations  Sadtler K, Estrellas K, Allen BW, Wolf MT, Fan H, Tam A, Patel C, Luber B, Wang H, Wagner K, Powell J, Housseau F, Pardoll D, Elisseeff JH. (2016) Th2 T cells are required for extracellular matrix-mediated functional muscle regeneration. Oral Presentation, 9th Symposium on Biologic Scaffolds for Regenerative Medicine. Napa Valley, CA.  Sadtler K, Allen B, Estrellas K, Wolf MT, Housseau F, Pardoll D, Elisseeff JH. (2015) Johns Hopkins University, Baltimore, MD 21231. T cells are Required for M2- Macrophage Polarization in ECM Scaffold-Treated Volumetric Muscle Injury. Oral Presentation, Biomedical Engineering Society (BMES) Annual Meeting. Tampa, FL.  Sadtler K, Allen B, Garza L, Elisseeff JH. (2015) Johns Hopkins University, Baltimore, MD 21231. Induction of hair growth through subcutaneous extracellular

198

matrix scaffold injection. Oral and Poster Presentation, Hyaluronan (ISHAS) International Conference. Florence, Italy  Sadtler K, Elisseeff JH. (2015) Johns Hopkins University, Baltimore, MD 21231. Defining the regenerative immunology niche: cells, signals and scaffolds. Oral Presentation, Institute of Cell Engineering Research in Progress Seminar Series, Johns Hopkins Univ. School of Med. Baltimore, MD.  Sadtler K. (2015) Johns Hopkins University, Baltimore, MD 21231. Tissue Engineering and Regenerative Medicine. Oral Presentation, McDaniel College. Westminster, MD.  Sadtler K, Beachley VZ, Elisseeff JH. (2014) Johns Hopkins University, Baltimore, MD 21231. Characterization of Immune Cell Recruitment and Polarization in Response to Extracellular Matrix Derived Scaffolds. Oral Presentation, Tissue Engineering and Regenerative Medicine International Society Americas (TERMIS- AM) Annual Conference. Washington, DC.  Sadtler K, Utria A, Elisseeff JH. (2014) Regenerative Medicine in Transplant: Soft Tissue Reconstruction. Oral Presentation, Partnering Towards Discovery Lecture Series, Johns Hopkins University School of Medicine. Baltimore, MD.

Published Abstracts  Sadtler K, Allen B, Estrellas K, Wolf MT, Housseau F, Pardoll D, Elisseeff JH. T cells are Required for M2 Macrophage Polarization in ECM Scaffold-Treated Volumetric Muscle Injury. Annals of Biomedical Engineering (2015) Biomedical Engineering Society (BMES) Annual Meeting.  Sadtler K, Allen B, Estrellas K, Wolf MT, Housseau F, Pardoll D, Elisseeff JH. Immune Profiles of Particulate ECM Scaffolds in Volumetric Muscle Wounds. Tissue Engineering Part A (2015) Tissue Engineering and Regenerative Medicine International Society (TERMIS) World Congress.  Wolf MT, Krill, JD, Wang TL, Sadtler K, Kim C, Elisseeff JH. Immunomodulatory Extracellular Matrix Nanoparticles. Tissue Engineering Part A (2015) Tissue Engineering and Regenerative Medicine International Society (TERMIS) World Congress.  Sadtler K, Beachley VZ, Elisseeff JH. Characterization of Immune Cell Recruitment and Polarization in Response to Extracellular Matrix Derived Scaffolds. Tissue Engineering Part A (2014) Tissue Engineering and Regenerative Medicine International Society Americas (TERMIS-AM) Annual Conference.  Beachley VZ, Sadtler K, Jacobs H, Blatchley M, Elisseeff JH. Tissue and Organ Microarrays for Probing Extracellular Matrix-derived Materials. Tissue Engineering Part A (2014) Tissue Engineering and Regenerative Medicine International Society Americas (TERMIS-AM) Annual Conference.

Patents  Sadtler K, Allen BW, Garza L, Elisseeff JH. 2015. Subcutaneous complex and single- component ECM compositing for induction of hair growth and follicular regeneration. JHTV Disclosure C13298, U.S. Patent Application 62111815, Filed February 2015. Patent Pending.  Sadtler K, Housseau F, Pardoll D, Elisseeff JH. 2015. Compositions and Methods for Modulating Wound Healing and Regeneration. JHTV Disclosure C13655, U.S. Patent Application 62202537, Filed August 2015. Patent Pending.

199

Memberships and Volunteering Global Engineering Innovation Team Member, Laos 2015-16 American Association for the Advancement of Sciences Aviculture Volunteer, National Aquarium in Baltimore Graduate Coating Ceremony Committee Member

200

Permission Letters to Reprint or Use Copyrighted Material

Book Chapter Title: "Integrating Tissue Microenvironment with Scaffold Design to Promote Immune-Mediated Regeneration.” Biomaterials in Regenerative Medicine and the Immune System. Springer International Publishing, 2015. 35-51. Publisher: Springer International Publishing

The authors of this chapter have obtained a license to re-print the information contained in the above chapter through Springer International Publishing, Copyright Clearance Center License #3823780882688

Manuscript Title: Tissue matrix arrays for high-throughput screening and systems analysis of cell function doi:10.1038/nmeth.3619 Journal: Nature Methods Publisher: Nature Publishing Group

The authors of manuscripts prepared by NPG retain the right to re-print their article in whole or in part within a dissertation or thesis providing the article and journal are cited. Further information regarding NPG author permissions can be found online at: http://www.nature.com/reprints/permission-requests.html

Manuscript Title: Developing a Pro-Regenerative Biomaterial Scaffold Microenvironment Requires T helper 2 cells (Accepted Version) Journal: Science Publisher: American Association for the Advancement of Science (AAAS)

This is the author’s version of the work. The definitive version was published in Science. The authors of manuscripts prepared by AAAS retain the right to re-print their article in whole or in part within a dissertation or thesis providing the article and journal are cited. Further information regarding AAAS permissions can be found online at: http://www.sciencemag.org/help/reprints-and-permissions.

All other text and figures are pre-submission and not bound by license regulations at this time.

201