The Therapeutic Potential of Indian Hedgehog (Ihh) for Tendon-to-Bone Repair
A dissertation submitted to the
Graduate School of the University of Cincinnati
in partial fulfillment of the requirements for the degree of
DOCTOR OF PHILOSOPHY (Ph.D.)
in the Biomedical Engineering Program of the College of Engineering and Applied Science
2015
by
Steven David Gilday
B.S., University of Virginia, Charlottesville, VA, 2007
Committee Chair: Jason T. Shearn, Ph.D.
Abstract
Tendon injuries are common, debilitating, and often difficult to treat. Reattaching ruptured tendons to their bony insertions has been a fundamental challenge in orthopaedics for decades, yet effective solutions that restore normal fibrocartilaginous enthesis architecture and mechanical function are still lacking. In our tissue engineering laboratory, we believe that the developmental signals governing tendon differentiation and patterning can be strategically reintroduced and/or manipulated during adult tendon repair in order to achieve better functional outcomes. In recent years, Indian hedgehog (Ihh) signaling has emerged as a key regulator of enthesis differentiation, growth, and mineralization. Given Ihh’s importance during development, the overall objective of this dissertation was to examine the role of hedgehog signaling in mature tendons and evaluate the potential therapeutic effects of recombinant Ihh during enthesis healing.
In aim 1, we developed and biomechanically characterized a new murine model of patellar tendon (PT) enthesis injury. Unlike the larger animal models that have been traditionally used for studies of tendon-to-bone healing, the murine model provides us the opportunity to conduct both basic and translational tissue engineering studies in transgenic strains relatively quickly and at low cost. In aim 2, we defined the natural patterns of endogenous hedgehog signaling in the mature murine PT. We found that hedgehog signaling remained active in the unmineralized entheseal fibrocartilage even in 46 week old mice, thereby suggesting a role for
Ihh in enthesis homeostasis throughout life. Prominent hedgehog signaling activity was also seen in regions of tendon undergoing fibrocartilaginous metaplasia. This observation, coupled with our finding that direct stimulation of cultured tenocytes with Ihh caused the cells to adopt a more chondrocytic phenotype, suggests that hedgehog signaling may regulate fibrocartilage formation
ii in tendons. In aim 3, we attempted to translate these discoveries into a clinically relevant biologic therapy that would promote regeneration of a zonal fibrocartilaginous enthesis during tendon-to-bone healing. In collaboration with materials scientists, we designed and fabricated
Ihh-infused polymeric drug delivery scaffolds and tested their ability to improve repair outcomes in our murine injury model. To our knowledge, this represents the first ever study to evaluate the effects of Ihh on tendon-to-bone healing in vivo. Although Ihh-infused scaffolds appeared to increase fibrocartilaginous cellular morphology at the healing enthesis, this unfortunately did not translate into improved biomechanical properties at 5 weeks post-surgery. Further work is needed in order to fully characterize the effects of Ihh on enthesis healing, but efficacious therapies for tendon-to-bone repair will likely remain elusive unless more precise methods for controlling the spatiotemporal delivery of biologic factors to the site of injury are developed.
As our knowledge of tendon biology continues to expand, the ongoing challenge for clinicians and bioengineers will be to translate this growing knowledge into effective treatments for tendon disorders. This dissertation paves the way for future tissue engineering studies in which the Ihh signaling pathway is targeted during tendon repair. Ultimately, we hope the continuation of our work by others will eventually lead to new therapeutic strategies for tendon disorders via modulation of hedgehog signaling.
iii
iv
Acknowledgements
I would like to express my deep appreciation and gratitude to the many individuals who have helped and encouraged me on my path to becoming a physician-scientist. To all my current and former teachers, mentors, collaborators, and fellow lab members, I give you my sincerest thanks. This dissertation would not have been possible without you. The following list of people deserve special acknowledgement for their contributions to my research, their mentorship during my physician-scientist training, and their constant support as I navigated the trials and tribulations of graduate school.
First and foremost, I would like to thank my two primary research advisors, Dr. David
Butler and Dr. Jason Shearn. Your willingness to take me into your tissue engineering and biomechanics laboratory as an MD/PhD student, introduce me to the field of musculoskeletal research, and jointly mentor me throughout the course of my dissertation has been critical to my development as a scientist. Although different in your mentoring style, together you have guided me through the scientific process and taught me many valuable lessons, ranging from how to best design an experiment to how to communicate more effectively. Most importantly, you have constantly challenged me to reach my full potential and encouraged me to persevere in the face of hardship. I enjoyed working with both of you and look forward to continued collaboration in the future.
I would also like to thank Dr. Rulang Jiang and Dr. Keith Kenter for serving on my dissertation committee. As a developmental biologist and an orthopaedic surgeon, respectively, your diverse perspectives on the biological and clinical aspects of my research are greatly valued. Your expertise, input, and suggestions have improved the quality of my work and forced me to think about research questions from more than one angle.
v
During my time as a graduate student, I have been fortunate enough to work side-by-side in the lab with some of the kindest, smartest, and most helpful people I have ever met. Dr.
Nathaniel Dyment, Dr. Andrew Breidenbach, Dr. Andrea Lalley, Dr. Rebecca Spatholt, and
Cindi Gooch not only provided me with invaluable technical training and experimental assistance, but also made the lab a fun place to work. Our daily discussions and general camaraderie helped break up the drudgery and dissipate the stresses of graduate school. I not only consider you scientific colleagues, but also close friends. Thank you so much for all you have done for me.
Many of the studies presented in this dissertation could not have been completed without the help of a multidisciplinary team of researchers. In addition to my fellow lab members mentioned above, I am indebted to the many collaborators that I have been privileged to work with and learn from over the past four years, including: Dr. Chris Wylie, Dr. Rulang Jiang, Dr.
Chia-Feng Liu, Dr. Han Liu, and Lindsey Aschbacher-Smith from the Division of
Developmental Biology at Cincinnati Children’s Hospital Medical Center; Dr. Keith Kenter and
Dr. Chris Casstevens from the Department of Orthopaedic Surgery at the University of
Cincinnati; Dr. Samer Hasan from the Cincinnati Sports Medicine and Orthopaedic Center; Dr.
Heather Powell from the Department of Materials Science and Engineering at The Ohio State
University; Dr. Marepalli Rao from the Division of Biostatistics and Bioinformatics at the
University of Cincinnati; and Dr. Daria Narmoneva from the Biomedical Engineering Program at the University of Cincinnati.
My research successes would not have been possible without the committed support of the University of Cincinnati Medical Scientist Training Program (MSTP). Thank you to Dr.
Patrick Tso for recruiting me into the program and thereby allowing me to pursue my goal of
vi becoming a physician-scientist. I would also like to thank Dr. Gurjit (Neeru) Khurana Hershey,
Dr. Kathryn Wikenheiser-Brokamp, Dr. George Deepe, Dr. Andrew Herr, Dr. Timothy LeCras,
Laurie Mayleben, Andrea DeSantis, and Amy Flanary for your dedication to the UC MSTP and for your tireless efforts to create a nurturing environment that fosters both scientific and professional development. I am also constantly motivated and inspired by my MSTP classmates
Rahul D’Mello, Mike Horwath, Jed Kendall, Julie Lander, and Dr. Martine Lamy, all of whom have supported and encouraged me throughout my training.
I want to extend a special thank you to my former teachers Lois Schultz, Linda Noble,
Jon Corum, and Mary Rhein for instilling in me a lifelong love of learning and science. I also want to thank Dr. Jay Hove for hiring me as an undergraduate lab assistant and thus introducing me to the wonders of modern biomedical research many summers ago. The positive experience I had working in the Hove lab with Dr. Michael Craig, Kate Olukalns, Mitul Desai, Dr. Robert
Littleton, and Dr. Mikah Coffindaffer-Wilson cemented my desire to pursue a graduate degree.
Most importantly, thank you to my wonderful family and friends who have stood by my side throughout the ups and downs of my research and graduate education. My parents, Dave and
Cheryl, my brothers, Dan and Scott, and my wife, Sarah, have provided limitless love and encouragement. Your unwavering support has given me the strength and motivation to work hard, pursue my passions, and “finish strong.” I love you all very much.
vii
Organization of Dissertation
This dissertation is composed of seven chapters and includes an introduction to the clinical problem of tendon-to-bone repair and a review of the relevant background literature
(Chapter 1), a focused discussion of the Indian hedgehog signaling pathway and our laboratory’s recent efforts to understand its role in enthesis development (Chapter 2), a description of the overall research approach and specific objectives of the dissertation (Chapter 3), a summary of the experimental results in the form of three scientific manuscripts (Chapters 4-6), and a discussion of the major conclusions, limitations, and future directions of this work (Chapter 7).
Chapter 1 begins by providing background about the burden of tendon injuries and our laboratory’s functional tissue engineering approach to tendon repair, including our strategy to develop both mechanical and biological success criteria. The structure and function of the tendon enthesis is also reviewed, as is the recent literature regarding tendon-to-bone healing. Chapter 2 is focused specifically on the Indian hedgehog (Ihh) signaling pathway and its recently discovered role in enthesis development, a critical finding which provided the rationale for much of this dissertation. With the background firmly established, Chapter 3 goes on to describe the research approach as well as the specific aims, hypotheses, and significance of this dissertation.
In Chapters 4, 5, and 6, the experimental research findings are reported in the form of traditional scientific manuscripts with introduction, methods, results, and discussion sections.
Chapter 4 introduces a new murine model of patellar tendon (PT) enthesis injury and characterizes healing biomechanics and regional strains in this model. Chapter 5 examines patterns of hedgehog signaling activity in the tendons of adult mice. Additionally, in an attempt to understand the effects of recombinant Ihh on mature tendon cells, preliminary studies are conducted which explore the dose-dependent effects of Ihh stimulation on tendon fibroblasts in
viii vitro. Chapter 6 describes the design and fabrication of bioresorbable, polymeric, Ihh-infused scaffolds and reports the results of our attempts to use these implantable scaffolds to augment tendon-to-bone healing in the mouse PT.
In Chapter 7, a summary of the principal research findings is presented along with a discussion of their significance to the field of tendon-to-bone healing and repair. The major limitations of this work and suggested alternative approaches are examined. At the end of the dissertation, a number of key unanswered questions regarding Ihh signaling at the healing enthesis are discussed, and specific recommendations for future studies to address these important new research questions are proposed.
ix
Table of Contents
ABSTRACT………………………………………………………………………………...….....ii ACKNOWLEDGEMENTS………………………………………………………………...…...... v ORGANIZATION OF DISSERTATION……………………………………………...……….viii TABLE OF CONTENTS…………………………………………………………….………...... x LIST OF TABLES AND FIGURES…………………………………………………………....xiii
CHAPTER 1: Background and Review of the Literature…………………………………………1
1.1. Introduction………………………………………………………………………………..1 1.2. Burden of tendon and ligament injuries……………………………………...……………2 1.3. Functional tissue engineering: A new paradigm for tendon repair……………….……….3 1.3.1. Mechanical success criteria……………………………………………………...….4 1.3.2. Biological success criteria…………………………………………….…………….8 1.3.2.1. Cellular phenotype: Scleraxis-expressing cells…………………………..10 1.3.2.2. Extracellular matrix: Collagen organization and alignment……………...12 1.3.2.3. Tissue ultra-structure: Zonal fibrocartilaginous enthesis…………………14 1.3.3. Challenges and alternative approaches……………………………………………15 1.3.4. Summary……………………………………………………………………….….17 1.4. The tendon enthesis: A brief literature review………………………………………...…18 1.4.1. Structure and function of the enthesis……………………………………………..19 1.4.2. Enthesis injury, healing, and repair………………………………………………21 1.5. Using tendon development to guide tendon tissue engineering………………………….25
CHAPTER 2: Indian Hedgehog Signaling in Bone and Tendon Development…………………27
2.1. The hedgehog signaling pathway…………………………………………………..…….27 2.2. Ihh signaling regulates endochondral bone growth…………………………………..….28 2.3. Ihh signaling regulates enthesis development………………………………………..…..32 2.4. Summary……………………………………………………………………………...….39
CHAPTER 3: Research Approach and Objectives………………………..……………………..42
3.1. Research approach and scope……………………………………………………………42 3.2. Specific aims and hypotheses……………………………………………………………45 3.3. Significance and impact……………………………………………………………….…49
CHAPTER 4: Murine Patellar Tendon Biomechanical Properties and Regional Strain Patterns during Natural Tendon-to-Bone Healing after Acute Injury…………………………….51
4.1. Abstract…………………………………………………………………………………..52 4.2. Introduction………………………………………………………………………………53 4.3. Materials and methods…………………………………………………………………...55 4.3.1. Experimental design……………………………………………………………….55 4.3.2. Murine patellar tendon injury model……………………………………………...55 4.3.3. Biomechanical testing and analysis……………………………………………….57
x
4.3.4. Statistical analysis…………………………………………………………………59 4.4. Results…………………………………………………………………………………....60 4.4.1. Gross morphology and tendon dimensions………………………………………..60 4.4.2. Structural and material properties…………………………………………………60 4.4.3. Regional tissue strains……………………………………………………………..62 4.4.4. Failure location……………………………………………………………………66 4.5. Discussion………………………………………………………………………………..67 4.6. Acknowledgements………………………………………………………………………70
CHAPTER 5: Hedgehog Signaling is Active in Unmineralized Enthesis Fibrocartilage but also in Regions of Fibrocartilaginous Metaplasia in Adult Murine Tendons……………71
5.1. Abstract…………………………………………………………………………………..72 5.2. Introduction………………………………………………………………………………73 5.3. Materials and methods…………………………………………………………………...75 5.3.1. Animal model……………………………………………………………………...75 5.3.2. Whole mount X-Gal staining and histology………………………………………76 5.3.3. Tendon cell isolation and culture………………………………………………….77 5.3.4. In vitro experiments……………………………………………………………….77 5.3.5. Imaging……………………………………………………………………………78 5.4. Results……………………………………………………………………………………78 5.4.1. Hedgehog signaling is active in the mature PT enthesis…………………………..78 5.4.2. Hedgehog signaling activity at the PT enthesis does not decrease with age……...81 5.4.3. Hedgehog-responsive fibrocartilage cells at the PT enthesis are Scx-negative…...83 5.4.4. Hedgehog signaling is active in regions of fibrocartilaginous metaplasia………..84 5.4.5. Changes in hedgehog signaling activity affect cultured tenocyte phenotype……..86 5.5. Discussion……………………………………………………………………………..…89 5.6. Acknowledgements……………………………………………………………………....96
CHAPTER 6: Delivery of Recombinant Indian Hedgehog Protein to the Healing Patellar Tendon Enthesis Does Not Improve Functional Repair Outcomes in a Murine Model…97
6.1. Abstract…………………………………………………………………………………..98 6.2. Introduction………………………………………………………………………………99 6.3. Materials and methods………………………………………………………………...101 6.3.1. Scaffold design and fabrication………………………………………………..101 6.3.2. In vitro studies……………………………………………………………………105 6.3.2.1. Ihh release kinetics……………………………………………………....105 6.3.2.2. Verification of bioactivity of released Ihh………………………………106 6.3.3. In vivo studies……………………………………………………………………106 6.3.3.1. Experimental design……………………………………………………..106 6.3.3.2. Murine patellar tendon injury model……………………………………108 6.3.3.3. Treatment groups………………………………………………………..109 6.3.3.4. Dissection and gross examination……………………………………….111 6.3.3.5. Histological sectioning and staining…………………………………….111 6.3.3.6. Biomechanical testing and analysis……………………………………..112 6.3.4. Statistical analysis………………………………………………………………..112 6.4. Results…………………………………………………………………………………..113
xi
6.4.1. Scaffold characterization and in vitro testing ………………………………...…113 6.4.2. Surgical outcomes and gross examination…………………………………….…115 6.4.3. Structural and material properties………………………………………………..118 6.4.4. Regional strain patterns…………………………………………………………..120 6.4.5. Failure mode and location………………………………………………………..120 6.4.6. Histologic outcomes……………………………………………………………...121 6.5. Discussion………………………………………………………………………………123 6.6. Acknowledgements……………………………………………………………………..129
CHAPTER 7: Conclusions and Recommendations……………………………………..……...131
7.1. Summary and discussion of principal findings…………………………………………131 7.2. Unanswered questions and recommendations for future studies……………………….136 7.3. Overall impact on the field of tendon tissue engineering…………………………..…..145
BIBLIOGRAPHY………………………………………………………………………………147
xii
List of Tables and Figures
Tables
Table 3.1: Experimental approaches for targeting the Ihh signaling pathway…………………..44 Table 4.1: Biomechanical properties of injured and sham tendons at 2, 5, and 8 weeks………..61
Figures
Fig 1.1: Functional tissue engineering improves tendon repair in a rabbit model………………..6 Fig 1.2: Proposed strategy to establish biological and mechanical success criteria……………....9 Fig 1.3: Our three chosen biological success criteria for tendon repair…………………………10 Fig 1.4: Structure of the fibrocartilaginous enthesis…………………………………………….20 Fig 1.5: Tendon insertion site injuries heal via scar formation in animal models……………….23 Fig 1.6: Enthesis architecture and tendon biomechanics are not restored during healing………24
Fig 2.1: The hedgehog signaling pathway……………………………………………………….28 Fig 2.2: The role of Ihh signaling during endochondral bone development………………….…31 Fig 2.3: Active hedgehog signaling at the developing enthesis………………………………....33 Fig 2.4: Hedgehog-responsive cells populate the mature enthesis………………………………33 Fig 2.5: Histological differences between control and Smo tKO entheses at 12 weeks……...…36 Fig 2.6: Biomechanical differences between control and Smo tKO tendons at 12 weeks………37 Fig 2.7: Ablating Ihh-responsive cells impairs fibrocartilage formation and mineralization...…39
Fig 4.1: Acute surgical injury and contralateral sham procedure………………………………..57 Fig 4.2: Biomechanical testing setup and regional strain measurement technique……………...59 Fig 4.3: Average stress-strain curves for injured and sham tendons at 2, 5, and 8 weeks………61 Fig 4.4: Biomechanical properties of injured and sham tendons at 2, 5, and 8 weeks…………..62 Fig 4.5: Average regional strains in normal murine patellar tendon…………………………….63 Fig 4.6: Average regional strains in injured and sham tendons at 2, 5, and 8 weeks……………65 Fig 4.7: Average regional strains in injured and sham tendons at physiologic load levels……...66 Fig 4.8: Failure location and failure mechanism of injured and sham tendons………………….67
Fig 5.1: Patterns of active hedgehog signaling in the 12 week old murine PT………………….79 Fig 5.2: Hedgehog signaling is localized to the unmineralized entheseal fibrocartilage………..80 Fig 5.3: Hedgehog signaling activity at the PT enthesis does not change with age……………..82 Fig 5.4: Hedgehog-responsive fibrocartilage cells at the PT enthesis are Scx-negative………...83 Fig 5.5: Hedgehog signaling is active in regions of tendon subjected to compressive loads…....84 Fig 5.6: Age-related changes in hedgehog signaling in the murine Achilles tendon……………86 Fig 5.7: Changes in hedgehog signaling activity affect cultured tenocyte phenotype…………..88
Fig 6.1: Flowchart for selecting final scaffold material and design…………………………....104 Fig 6.2: Final Ihh-infused scaffold design……………………………………………………...105 Fig 6.3: Mouse breeding scheme and baseline biomechanical properties……………………...107 Fig 6.4: Experimental design, treatment groups, and response measures……………………...108
xiii
Fig 6.5: PT surgical injury and scaffold implantation………………………………………….110 Fig 6.6: Ihh release kinetics in vitro……………………………………………………………114 Fig 6.7: Bioactivity of Ihh does not change as a result of infusion into scaffolds……………..115 Fig 6.8: Frequency of spontaneous PT rupture at 5 weeks post-surgery……………………….116 Fig 6.9: Gross appearance of intact and ruptured tendons at 5 weeks post-surgery…………...117 Fig 6.10: Average stress-strain curves for all treatment groups at 5 weeks post-surgery……...118 Fig 6.11: Structural and material properties at 5 weeks post-surgery………………………….119 Fig 6.12: Treatment with Ihh did not change PT failure mechanisms………………………....121 Fig 6.13: Repair tissue morphology differs between natural healing and Ihh-treated tendons...122 Fig 6.14: Morphology of the tendon-bone interface in control and Ihh-treated PTs…………..123 Fig 6.15: Ihh-infused scaffolds do not induce ectopic mineralization in the PT……………….124
Fig 7.1: Proposed mechanism of fibrocartilage metaplasia in tendons………………..……….139
xiv
CHAPTER 1
Background and Review of the Literature
1.1. Introduction
Tendon injuries are common and are often difficult to treat. Reattaching ruptured soft tissue to bone has been a fundamental challenge in orthopaedics for decades, yet effective solutions that restore normal biology and mechanical function are still lacking. A major reason for this lies in the multifaceted nature of the problem itself. Tendon-to-bone repair is not only a clinical challenge for surgeons, but it also represents a significant biological and engineering problem. From a biological perspective, the tendon-bone junction (termed the “enthesis”) has a complex natural architecture, with multiple different cell types, extracellular matrix proteins, and signaling pathways interacting in a precisely controlled manner in order to maintain a healthy and functional attachment site. Furthermore, our incomplete understanding of the basic molecular mechanisms governing enthesis development, homeostasis, and healing has limited our attempts to develop therapies for enthesis repair. From an engineering perspective, attaching two materials with drastically different mechanical properties is inherently challenging due to the accumulation of potentially damaging stress concentrations at the interface1-3. Additionally, given their predominantly mechanical function, repaired tendons are often subjected to dynamic in vivo loading patterns immediately following surgery, but the complex interplay between mechanics and biology (mechanobiology) during the healing process is still poorly understood.
Unfortunately, adult tendon-to-bone healing does not restore a normal, zonal insertion site following injury. The distinct fibrocartilage transition region which is normally interposed
1 between tendon and bone does not regenerate. Instead, deposition of scar tissue at the tendon- bone interface results in a morphologically abnormal and biomechanically inferior enthesis that is susceptible to further damage. Despite advances in surgical technique and rehabilitation protocols, re-rupture of repaired tendons remains a serious complication for many patients. Thus, there exists a critical need for innovative strategies to improve tendon-to-bone repair.
1.2. Burden of tendon and ligament injuries
Tendon and ligament injuries represent nearly half of all musculoskeletal injuries in the
US and account for billions of dollars annually in direct and indirect medical costs4. The aging of the US population has increased the prevalence of many musculoskeletal conditions, including both soft tissue and bone disorders. By the year 2040, the number of individuals in the US older than the age of 65 years is projected to grow from the current 15% of the population to 21%, and persons age 85 years and older will double from the current <2% to 4%4. Musculoskeletal impairments will continue to increase because they are most prevalent in these elderly segments of the population. Although cardiovascular disease, cancer, and stroke often dominate the headlines due to their higher mortality rates, the socioeconomic burden of musculoskeletal disorders must not be ignored. Bone and joint disorders actually account for more than one-half of all chronic conditions in people older than 50 years of age in developed countries, and are the most common cause of severe, long-term pain and disability4.
In this dissertation, our research focus is specifically on tendon injuries. Nonetheless, much of our work may also be applicable to ligament repair since both of these soft tissues share a similar structure, including an enthesis where they attach to bone. Basic research on tendon biology has lagged behind that of other musculoskeletal tissues such as bone and muscle, and as
2 a result, effective biological strategies for improving tendon healing are not yet available for clinical use. There is an urgent clinical need for new therapies for tendon repair. The incidence of rotator cuff tears, flexor tendon ruptures, and traumatic injuries to the Achilles and patellar tendons continues to increase, especially in aging individuals. Large studies have shown that rotator cuff tears are present in approximately 20% of adults in the general population5 and in
30% of cadaveric shoulders6. Over 250,000 rotator cuff repairs are performed each year in the
US alone, with rates of re-rupture ranging from 11.4% to 94% depending on the patient population, size of the initial tear, and other factors7. In elderly patients over age 65, re-rupture rates commonly exceed 40%8. Although outcomes after flexor tendon repair in the hand have improved with modern treatment, a recent meta-analysis revealed rates of re-operation of 6%, rupture of 4%, and adhesions of 4%9. Acute patellar and Achilles tendon ruptures are less common but can be extremely debilitating, often rendering the patient unable to ambulate.
Furthermore, managing chronic tendinopathy or neglected tendon ruptures presents a significant clinical challenge10, 11. A unifying feature of all these injuries is the poor natural healing of damaged tendon tissue. In particular, the tendon enthesis, once disrupted, does not regenerate its complex zonal arrangement, leaving the tendon susceptible to further injury.
1.3. Functional tissue engineering: A new paradigm for tendon repair
Note: This section is adapted with permission from the following published manuscript: [Breidenbach AP*, Gilday SD*, Lalley AL*, Dyment NA, Gooch C, Shearn JT, Butler DL. 2014. Functional tissue engineering of tendon: Establishing biological success criteria for improving tendon repair. J Biomech 47: 1941-1948. *co-first authors]12
As previously described, tendon and ligament injuries continue to burden the U.S. population and economy. Repairing these injuries remains a challenge, often resulting in long-
3 term impairments such as chronic tendinopathy and osteoarthritis13. Tissue engineering represents a novel approach to potentially improve tendon and ligament repair outcomes by combining cells, biomaterials, and chemical and mechanical stimulation to restore or even regenerate damaged tissue. Taking this concept a step further, functional tissue engineering
(FTE) establishes the importance of biomechanical aspects of the design process by focusing on how soft and/or hard tissues are normally loaded during activities of daily living (ADLs)14, 15. By incorporating FTE principles into the design and evaluation process, musculoskeletal tissue engineers can produce more effective tendon repairs that meet the mechanical demands of the in vivo setting. However, if we are to design truly successful repairs, we must also understand the biological processes that influence tendon healing. This will allow us to expand the FTE paradigm to also include a set of biological design criteria for tissue-engineered tendon repair.
1.3.1. Mechanical success criteria
Since FTE was first described in 200014, our laboratory has been designing tissue- engineered constructs (TECs) to improve tendon healing following injury with the ultimate goal of clinical translation. Tendons contain compositionally and structurally distinct, but mechanically interconnected, regions that are regulated by the normal loading environment and the location within the body16, 17. When tendons are not repaired after injury, the natural tendon healing process often fails to restore normal mechanical properties, leading to increased rates of re-injury18 as well as tendinopathy and osteoarthritis in the long-term13. The standard of care for many acute tendon injuries is surgical repair, but post-operative clinical outcomes have been variable19-21. As such, the concept of developing a tissue-engineered repair to augment the healing process remains an attractive alternative.
4
Guided by FTE principles14, our laboratory established two primary mechanical success criteria to evaluate the effectiveness of our tissue-engineered constructs (composed of autologous, mesenchymal progenitor cells seeded in collagen scaffolds) during repair of a rabbit central patellar tendon (PT) defect22:
1. Exceeding Peak In Vivo Forces. One aspect of FTE is the importance of characterizing
the normal mechanical properties of native tissue for benchmarking engineered constructs
and in vivo repairs. We measured peak in vivo forces in the rabbit and goat, finding that
tendons experience quite different levels of peak in vivo force (IVF). In the rabbit
Achilles’, flexor digitorum profundus, and patellar tendons, in vivo loads range from 11-
28% of a tissue’s failure force during moderate ADLs23-25. However, in the goat model,
the percentage approaches 40% of tensile failure load26. These results suggested that
TECs must be designed to accommodate the differences between species and tissue sites.
2. Matching Normal Tangent Stiffness up to Peak IVFs with a Safety Factor. Using the
rabbit central PT defect model, we concluded that any TEC repair must match the tangent
stiffness of normal PT up to the peak IVFs but also incorporate a safety factor to account
for potential overloading during more strenuous activities. Working towards this goal, we
strategically improved our repairs by optimizing the cell density, scaffold material, and
mechanical preconditioning of the TEC in culture before implantation22. Not only did we
generate 12-week repairs of the central PT that exceeded peak IVFs recorded in this
model, but we also matched the tangent stiffness of the normal tendon up to 50% beyond
peak IVFs (Fig. 1.1).
5
Figure 1.1: Our laboratory’s iterative efforts to improve tendon repair in a rabbit central patellar tendon defect model using a functional tissue engineering approach. Natural healing (NH) results in significantly inferior biomechanical outcomes when compared to normal tendon (N). In an attempt to improve repair outcomes, our research group has developed different tissue- engineered constructs (TECs) for augmenting tendon healing by incorporating cells, collagen scaffolds, and in vitro mechanical stimulation. Our best TEC repair (denoted as SCS) consisted of a collagen sponge, an autologous mesenchymal stem cell population, and cyclic in vitro tensile stimulation. This TEC repair matched the tangent stiffness of normal patellar tendon up to 32% of failure force, a load level which is 50% greater than the largest measured in vivo force (IVF) in the rabbit model. Adapted with permission from (Butler et al, 2008)22.
While our previous studies have shown that our TECs produce an improved mechanical repair outcome compared to natural healing, we still have not fully restored normal tendon structure and function in the following two ways:
6
1. Repair tissue does not match tangent stiffness up to at least 40% of failure force, which
corresponds to the highest measured in vivo loads we have recorded in the goat26.
Although matching tangent stiffness up to 32% of failure force is 50% above the peak
IVFs recorded in the rabbit PT (Fig. 1.1), there is still significant room for improvement.
Focusing on mechanics alone may be insufficient to reach our ultimate goal.
2. We do not adequately understand the biological mechanisms which led to the successful
repairs seen previously. There are several questions that still remain which may help
explain our results. Specifically, what was the TEC cell phenotype at the time of harvest,
after preconditioning in culture, and following repair? How was the extracellular matrix
assembled during the repair process and how does its composition change over time?
How do interactions between cells, matrix, and the local mechanical environment lead to
regional differences in tissue properties? These are all biological questions that would not
only help explain our mechanical results, but also provide potential predictive success
criteria for future experiments.
Ultimately, tendon repairs must be able to withstand in vivo mechanical forces to be truly considered a clinical success. Mechanical outcomes are, in turn, a function of the sequence of biological processes that occur during TEC creation, maturation, implantation, and healing.
Therefore, we believe it is critical for the musculoskeletal tissue engineering field to begin establishing biological parameters which consistently lead to mechanical and clinical benchmarks of successful repair. In our laboratory, we have already begun adapting the FTE paradigm to establish these important biological success criteria.
7
1.3.2. Biological success criteria
We have previously discussed how our lab has used FTE principles to create mechanical success criteria by (1) measuring normal in vivo tendon forces, (2) selecting mechanical parameters based on sub-failure tendon mechanics, and (3) prioritizing a subset of these parameters to establish mechanical success criteria for functional assessment of our tissue- engineered repairs. We can now use this same approach to generate biological success criteria for tendon repair. First, we must define what constitutes normal tendon biology. Second, we need to identify which of these biological parameters are critical to tendon function. Finally, we should assess how deviations in these biological parameters affect mechanical outcomes in order to prioritize which biological properties to focus on in future tendon repair studies.
Mature tendon cells (tendon fibroblasts or tenocytes) express necessary transcription factors and signaling ligands to maintain their phenotype and synthesize extracellular matrix proteins which form the scaffolding for the living tendon tissue. The cellular phenotype and extracellular matrix composition varies along the tendon length, generally dividing the tissue into a myotendinous junction, tendon midsubstance, and enthesis27, 28 and producing regional variations in mechanical properties29-31. Such compositional variation necessitates that we define normal tendon biological parameters based upon the (1) cellular phenotype, (2) extracellular matrix, and (3) regionalization of these two parameters within the tissue ultra-structure.
Recently, our lab has been using a strategy of comparing and contrasting normal tendon development to natural tendon healing to identify potential biological targets to assess and modulate in future repair studies (Fig. 1.2). Understanding tendon development is critical to defining what makes a “tendon a tendon” and allows for the identification of biological parameters which are required for proper tendon formation and potentially necessary for
8 successful repair. Prioritizing these biological parameters based on their respective contributions to tendon function is essential for improving the efficiency of the tissue engineering process. Based on our current understanding of tendon biology, here we propose three biological success criteria which we believe are essential for normal tendon function (Fig. 1.3).
Figure 1.2: Our strategy to establish biological and mechanical success criteria for functional assessment of tissue-engineered tendon repairs. Building on the FTE paradigm we used to develop mechanical design criteria, we now seek to establish biological design criteria to more fully characterize these repairs. Our strategy is to investigate the normal biological processes responsible for tendon development and maturation, identify and categorize the biological parameters that may be important, and finally, assess how these chosen biological parameters affect the mechanical outcomes of tissue repair. This approach can be applied to other tissue systems and injury models to begin developing stronger connections between biology and mechanics. Adapted with permission from (Breidenbach, Gilday, and Lalley et al, 2014)12.
9
Figure 1.3: Our proposed biological success criteria include (1) Scleraxis-expressing cells situated between (2) densely-packed collagen fibers aligned along the axis of tension and (3) a zonal enthesis with unmineralized and mineralized fibrocartilage regions. As can be seen histologically, all three of these proposed biological success criteria are prominent features of normal tendons. Toluidine blue staining (A,C,G) and ScxGFP fluorescence (B,D,H) in a 4-week old murine patellar tendon depict highly aligned tenocytes within the tendon midsubstance that are predominately ScxGFP+ and stacked, rounded fibrochondrocytes within the zonal insertion site. Two photon images of tendon midsubstance (E,F) and enthesis (I,J) depict highly-aligned, densely-packed collagen fibers (second harmonic signal in blue and grey) within the tendon midsubstance that extend through the enthesis into the underlying bone. The tidemark (red dotted line) indicates the junction between the unmineralized and mineralized fibrocartilage. Scale bars = 100m. Adapted with permission from (Breidenbach, Gilday, and Lalley et al, 2014)12.
1.3.2.1. Cellular phenotype: Scleraxis-expressing cells
The developmental biology field has begun to identify mechanisms leading to normal tendon development, namely transcription factors which aid in characterizing cell phenotype and function to regulate the expression of other important tendon genes, such as extracellular matrix proteins. Scleraxis (Scx)32-35, mohawk homeobox (Mkx)36, 37, and early growth response 1
10
(Egr1)38, 39 are three of the more extensively studied tenogenic transcription factors. All have been shown to be important regulators of normal tendon development and maturation, as loss of expression of these markers results in phenotypes ranging from severely impaired tendon function in load-bearing tendons33 to abnormalities in collagen fibrillogenesis and impaired biomechanical properties36-39. Moreover, spatial variations in the expression of these transcription factors can alter the development of the tendon enthesis. For example, a chondrogenic-tenogenic progenitor pool of cells resides at the enthesis during embryonic development. These cells dually express Scx and the chondrogenic transcription factor SRY-box containing gene 9 (Sox9) and give rise to the fibrocartilage in the enthesis40, 41. Scx-expressing progenitors are required for enthesis formation, as Scx knockout mice exhibit impaired bone tuberosity formation42. Furthermore, conditionally knocking out Sox9 or disrupting chondrogenic signaling pathways in Scx-expressing cells also results in defective enthesis formation in limb tendons41, 43.
Of the aforementioned markers, Scx appears to be the most critical regulator of tendon formation known to date. Scx expression defines tendon progenitors during limb condensation32,
35, and it continues to be expressed in mature tenocytes (Fig. 1.3). It regulates expression of other tenogenic markers including type I collagen44, tenascin-C35, 45, and tenomodulin46. Its expression during development is so critical that knocking it out results in disorganized and reduced collagen content and a loss of function in many axial and limb tendons33. However, its role in tendon healing and repair are only beginning to be clarified.
Scleraxis expression during tendon healing exhibits distinct spatiotemporal patterns. Although Scx remains slightly down-regulated during early stages of tendon natural healing (compared to normal tendon)47-49, its expression increases in the tendon callus during
11 later stages of healing49. Furthermore, mechanical loading of the healing tendon during late-stage remodeling can further increase Scx expression48. Interestingly, Scx expression during healing appears to originate from a paratenon source of progenitors47. These cells migrate to the wound site then express Scx as the tenascin-rich provisional matrix transitions to a more mature collagenous matrix. While direct correlations with mechanics have yet to be made, increased Scx expression during remodeling appears to be commensurate with increased biomechanical properties49, 50.
Scx expression is necessary, but certainly not sufficient, for the formation of mechanically functional tendon. Although Scx has a role in tendon development and healing, it is also up-regulated in the fibrotic response of many other tissues51-53. As fibrotic scar is generally characterized by disorganized collagen assembly, another critical biological benchmark of successful tendon repair is the formation of an organized and competent extracellular matrix to resist tensile loads.
1.3.2.2. Extracellular matrix: Collagen organization and alignment
Tendons are collagenous tissues composed predominantly of collagen type I54, along with lesser amounts of collagen types III55, V56, 57, VI58, 59, and X60, and additional matrix proteins including proteoglycans, such as decorin61 and lumican62. Hierarchical in nature, the tendon matrix is composed of collagen microfibrils and fibrils, which aggregate to form fibers. Given their highly organized structure, tendons exhibit high tensile strength, resisting loading in the axial direction, vital to normal tendon function in the body63, 64.
The complex process through which collagen fibrils are processed, assembled, and organized is known as collagen fibrillogenesis65, 66. Collagen fibrillogenesis is highly regulated
12 during tendon development, involving the interaction of integrins, collagens, and collagen binding proteins67. Early in development, small diameter fibrils predominate. As the tendon grows and begins to experience loading, collagen fibrils become longer and larger in diameter, resulting in a matrix that assumes a bimodal distribution of both large (~100-150 nm) and small
(~40-75 nm) collagen fibrils to form the mature tendon68. The larger diameter fibrils are thought to resist tensile loading and the smaller diameter fibrils to negate creep and improve fibril binding strength69. Proteoglycans, such as decorin61, biglycan70, fibromodulin71, and lumican62, also play a key role in collagen fibrillogenesis by regulating collagen fibril interactions and ultimately tendon mechanics.
Researchers disagree about why bimodal fibril distributions are not restored after injury.
Some attribute the inferior mechanical properties to an altered collagen fibrillogenesis process, with a predominance of small diameter fibrils that never form large diameter fibrils72-74. Others contend it may not be the size of the fibrils, but their total number and/or packing density that ultimately affects the mechanical outcome75, 76. Regardless, natural tendon healing typically results in scar tissue formation, consisting of small collagen fibrils that are not oriented along the direction of loading. Further complicating the healing process and rehabilitation is the formation of adhesions between tendons and other structures. Tendon adhesions generally affect intrasynovial flexor tendons, such as in the finger, and can reduce joint range of motion, often leading to pain and discomfort77, 78.
Given that tendons are highly loaded structures vital to skeletal and joint movement, understanding their matrix composition is vital to developing effective tissue-engineered repairs.
While an aligned collagen matrix is necessary for normal mechanical function, the composition of the matrix varies along the length of the tendon, producing regional variations in mechanical
13 properties. Therefore, as we develop biological success criteria for tendon repair, we must also consider regional differences in the tissue ultra-structure.
1.3.2.3. Tissue ultra-structure: Zonal fibrocartilaginous enthesis
Tendons provide the physical linkages between muscle and bone and serve to transmit muscular forces to the skeleton. Thus, a functional tendon actually consists of three specialized tissue regions: the myotendinous junction, the tendon midsubstance, and the tendon-to-bone insertion site (enthesis). The enthesis is particularly critical for proper mechanical function because it facilitates force transmission between the compliant tendon and the much stiffer bone while also ameliorating potentially damaging stress concentrations that would otherwise accumulate at this interface3, 79. Uninjured entheses exhibit gradations in cell phenotype, biochemical composition, matrix organization, and mineral distribution along their length28, 80.
The enthesis contains rounded fibrochondrocytes81 and large amounts of type II and X collagen as well as proteoglycans and glycoproteins such as aggrecan, biglycan, and tenascin C82-84. Due to differences in collagen fiber organization and crimp pattern85-87, local strains near the insertion are often 2-3 times greater than in the tendon midsubstance under sub-failure tensile loads30, 88.
As discussed previously, the transcription factors Scx and Sox9 appear to be critical for enthesis formation40, 41. Recent evidence also indicates that Indian hedgehog (Ihh) signaling is involved in enthesis differentiation (discussed in detail in Chapter 2), and knocking out this pathway in Scx-expressing cells during development results in morphologic and biomechanical deficits that persist into adulthood43, 89. However, even if the correct biological cues are in place, experiments using botox to inhibit muscular contraction have shown that muscle loading is also required for normal enthesis formation90. In fact, the presence of mature fibrocartilage at the
14 enthesis is correlated with increased compressive loads in this region27, indicating that both biologic and mechanical factors are important players in enthesis development and maturation.
Unfortunately, once damaged, the complex structure of the enthesis is not regenerated during natural healing. Deposition of disorganized scar tissue at the healing tendon-bone interface and the absence of a morphologically normal fibrocartilage transition region results in altered biomechanical properties and premature failures at the enthesis91-94. Thus, restoring the fibrocartilaginous interface between tendon and bone represents our third biological success criterion for tissue-engineered tendon repair and is the primary focus of this dissertation.
1.3.3. Challenges and alternative approaches
Establishing a set of biological success criteria to accompany existing mechanical design goals for tendon repair would greatly benefit both the clinical and tissue engineering communities. The idea of creating biological design standards for engineered tissues is not new95,
96, but linking structure to function in biological systems remains a significant challenge. This is certainly true in the case of tendon healing and repair as the complexity of these processes makes it difficult to parse out which biological criteria make the largest contributions to tendon mechanics and long-term repair outcomes. To address this problem, we have attempted to define aspects of normal tendon biology relative to cellular phenotype, extracellular matrix, and tissue ultra-structure that we would like to reproduce in our tissue-engineered repairs. We are prioritizing these biological criteria by examining their role and relative importance during natural tendon formation and tendon healing. In future tissue engineering studies, we will assess whether achieving these chosen biological criteria improves repair biomechanics.
15
Although straightforward in principle, this strategy does have challenges. (1) Many aspects of normal tendon biology are not well understood and other biological markers of normal tendon function are likely still waiting to be discovered. Even very fundamental questions such as “what defines a tenocyte?” are still being debated. (2) Tendons are dynamic tissues whose biologic properties depend on anatomical and mechanical cues, which adds a layer of complexity to our strategy97, 98. (3) From a clinical perspective, patient factors such as age99, 100, gender101,
102, disease103, 104, and activity level105, 106 affect normal tendon properties, making it even more difficult to establish a set of biological success criteria that would be broadly applicable. (4)
Prioritizing biological criteria is currently a rather subjective process since few published studies have linked a particular biological criterion with a functional mechanical outcome. In our laboratory, we have chosen to compare and contrast normal tendon development (a model of successful tendon formation) with inadequate natural healing. We reason that biological events occurring during normal development but absent in natural healing might be effective initial targets to modulate during repair. However, this approach to prioritizing biological success criteria is predicated on the assumption that successful tendon healing should spatially and temporally resemble normal tendon development, which may not be the case.
While our approach represents one possible strategy for establishing a set of biological success criteria for tendon repair, other approaches should also be investigated. Recent evidence indicates that some animal models, such as the Murphy Roth’s Large (MRL) mouse, display
“super healing” capabilities107. Our laboratory has begun investigating tendon healing in adult
MRL mice and has found that this murine strain exhibits significantly improved biomechanical healing outcomes compared to C57Bl6 wild type controls108. However, the biological mechanisms underlying the improved healing response are still unclear. Uncovering the specific
16 biological differences between super healers and normal healers may reveal new biological criteria that significantly affect tendon repair. Similarly, studies investigating the underlying genetic and/or structural differences between human patients with good and poor surgical outcomes could identify which biological criteria are most highly correlated with clinical success.
Establishing clear linkages between biology and mechanics in tissue-engineered tendon repairs will require well-designed and appropriately controlled studies that isolate specific biological criteria and quantify their effects on mechanical repair outcomes. However, such studies are inherently difficult because biological criteria are often qualitative in nature and may be difficult to measure. Biological criteria also vary widely across both spatial and temporal scales. Small, early changes in one biological criterion (for example, expression of a certain transcription factor) may result in drastic changes in mechanics later on, but these correlations are hard to detect. Furthermore, biological processes are often interrelated and many compensatory mechanisms are activated if normal biology is altered in any way, making it difficult to isolate the effects of a single factor. Finally, establishing biological and mechanical homology across species is challenging but will be required if novel tendon repair strategies are to be translated towards the clinic. This is a particularly difficult task for functional tissue engineers, since investigators use different injury models and assess different mechanical repair outcomes at different time points.
1.3.4. Summary
To achieve success, the tissue engineering field needs to develop better strategies and adopt more unified approaches for the identification, prioritization, and evaluation of biological
17 success criteria for tendon repair. Our laboratory has developed a general strategy in which we:
(1) identify and categorize biological parameters of normal tendon based on cellular phenotype, extracellular matrix, and tissue ultra-structure; (2) select a subset of biological parameters by examining their relative importance during both normal development and natural healing; and (3) prioritize these parameters by experimentally assessing whether they affect mechanical outcomes in a tissue engineering scenario. This paradigm, although presented here only in the context of tendon repair, could be applied to any load-bearing tissue in the body. We have begun using this strategy to select specific biological success criteria critical to normal tendon function. In our view, a successful tendon repair must not only meet the stated mechanical design limits but also exhibit (1) scleraxis-expressing cells embedded within (2) a well-organized and axially-aligned collagen matrix that is (3) securely attached to bone via a fibrocartilaginous enthesis as a protection against long term failure.
In the remainder of this dissertation, we focus almost all our attention on the third of these three biological success criteria, namely the restoration of a functional fibrocartilaginous enthesis following injury. However, in order to create effective repair strategies, we first must understand the unique structure and function of the native tendon-to-bone insertion site.
1.4. The tendon enthesis: A brief literature review
Tendons connect muscles to bones and transmit muscular forces to the skeleton in order to provide structural support and allow for movement. The specialized area where tendon meets bone is often referred to as the osteotendinous junction, insertion site, or enthesis. Anatomically, tendon entheses can be grouped into two broad categories according to their structure109. Fibrous entheses are characterized by tendon fibers that insert either directly into the mineralized bone or
18 indirectly into the periosteum across a wide footprint. When viewed histologically, a prominent feature of fibrous entheses are the perforating collagenous fibers (Sharpey’s fibers) that interdigitate with the underlying bone in order to securely anchor the tendon in place. Examples of fibrous entheses include the insertions of the deltoid tendon and the medial collateral ligament. In contrast, fibrocartilaginous entheses, such as those found at the insertions of the anterior cruciate ligament (ACL) and rotator cuff, Achilles, and patellar tendons, contain a distinct region of fibrocartilage tissue interposed between the tendon and its bony attachment site. This dissertation will focus exclusively on fibrocartilaginous entheses and our laboratory’s attempts to restore normal function to these specialized tissue regions following injury.
1.4.1. Structure and function of the enthesis
Fibrocartilaginous entheses exhibit a gradual transition between soft, compliant tendon and hard, mineralized bone via a fibrocartilage transition region109. As such, the insertion site can be described in terms of four histologic zones: 1) tendon proper, 2) unmineralized fibrocartilage,
3) mineralized fibrocartilage, and 4) bone (Fig. 1.4). Tissue organization, matrix composition, cell phenotype, and biomechanical properties vary widely between these zones110. The tendon proper (also called the tendon midsubstance) is composed mainly of well-aligned type I collagen fibers designed to resist tensile loads. The midsubstance contains tenocytes, which are fibroblast- like cells that produce the extracellular matrix. In contrast, the fibrocartilage at the insertion site contains large amounts of type II and X collagen as well as proteoglycans (e.g. aggrecan, biglycan) and glycoproteins (e.g. tenascin C)82, 84. The fibrochondrocyte cells in this region are larger, rounder, and reside within lacunae, often appearing in pairs or stacks. Collagen fibers originating in the tendon midsubstance pass through the fibrocartilage region and anchor into the
19 mineralized bone. Due to the aforementioned differences in tissue structure and regional variations in collagen crimp patterns, local strains at the insertion site are often 3-4x greater than in the midsubstance under sub-failure tensile loading30. The fibrocartilage interface between tendon and bone is particularly critical for proper function because it facilitates force transmission between mechanically dissimilar tissues and ameliorates potentially damaging interfacial stress concentrations3, 79, 80.
Figure 1.4: Structure of the fibrocartilaginous enthesis. Tendons connect muscles to bones via a fibrocartilaginous insertion site (left panel) that can be described in terms of four distinct histologic zones: tendon proper, unmineralized fibrocartilage, mineralized fibrocartilage, and bone (right panels). Tissue organization, composition, and mechanical properties vary along the length of the insertion in order to provide a gradated transition between the compliant tendon and the much stiffer bone. Adapted with permission from (Thomopoulos et al, 2010)111 and (Thomopoulos, 2011)112.
20
1.4.2. Enthesis injury, healing, and repair
Injuries involving the enthesis can occur via several different mechanisms. These include acute tendon ruptures, often seen in healthy tendons that are acutely overloaded or lacerated.
Tendons and their entheses can also become inflamed and painful due to mechanical overuse or rheumatologic disease, a condition termed tendinitis or enthesitis. The hallmark of these conditions is the accumulation of inflammatory cells within the tendon tissue. Chronic overuse, repetitive microtrauma, and age-related degeneration are also thought to contribute to non- inflammatory tendinosis or tendinopathy. The exact etiology and pathogenesis of degenerative tendinopathy is not fully understood, but it is characterized histopathologically by mucoid degeneration of the matrix, disordered arrangement of collagen fibers, increased cellularity, increased vascularity, and decreased tendon biomechanical properties, all in the absence of tendon inflammation113, 114. Both tendinitis and tendinopathy increase the risk of tendon rupture, and orthopaedic surgeons are often faced with acute-on-chronic tendon injuries in which a chronically degenerative tendon suddenly experiences an acute tear. These types of injuries are particularly difficult to surgically repair given the poor quality of the native tendon tissue.
The healing potential of an injured tendon varies based on anatomic location and the local environment, but the general process of tendon healing is largely conserved and has classically been described as a series of three overlapping phases: an early inflammatory phase, a proliferative or reparative phase, and finally a lengthy remodeling or maturation phase115, 116.
During the inflammatory phase, which lasts approximately one week, vascular permeability increases and inflammatory cells (mainly neutrophils) extravasate and enter the wound site.
These cells release cytokines and growth factors that lead to recruitment of circulating monocytes, which infiltrate the wound and differentiate into tissue macrophages. Macrophages
21 serve many important functions during early wound healing including removal of damaged cells and matrix as well as promotion of angiogenesis, fibroblast proliferation, and matrix synthesis117.
In the proliferative phase of tendon healing, which lasts several weeks, fibroblasts in and around the wound site proliferate and begin producing a provisional matrix rich in collagen type 3. The production of collagenous matrix during healing is driven in part by transforming growth factor beta (TGFβ) signaling118. During the remodeling or maturation phase, which can last for months or even years, the provisional matrix is slowly replaced by a more mature matrix rich in collagen type 1. The resident tendon fibroblasts are critical in this process because they not only produce this matrix, but also assemble, organize, and crosslink the mature collagen fibers.
In addition to the temporal phases of healing discussed above, clinicians and researchers often describe two major theories of tendon healing based on the spatial origin of the cells involved in the reparative process. Intrinsic healing occurs within the body of the tendon itself and is mediated by the proliferation of resident tendon fibroblasts and/or tendon stem cells. For example, the ends of a lacerated flexor tendon in the hand can be sewn back together and allowed to heal via intrinsic mechanisms. In contrast, extrinsic healing relies on the contribution of cells and signals that are external to the tendon. During extrinsic healing, cells from the paratenon, tendon sheath, synovium, or other surrounding tissues migrate into the defect site where they proliferate and differentiate. Extrinsic healing often results in scar tissue and/or adhesion formation. This is often beneficial in the short term since it provides mechanical support and allows for vascular ingrowth to the damaged area, but can be detrimental in the long term due to disruptions in normal tendon architecture which affect the mechanical properties of the tendon and limit range of motion. It is commonly accepted that both intrinsic and extrinsic
22 healing occur simultaneously to varying degrees in any given tendon injury, yet the theoretical distinction still provides a useful conceptual framework.
Unfortunately, tendon injuries involving disruptions to the insertion site generally do not heal well naturally. Even after surgical intervention, the complex zonal arrangement of the insertion site is not regenerated, resulting in an inferior connection between tendon and bone.
Studies of natural tendon-to-bone healing in sheep94, goats119, canines93, 120, 121, rabbits92, 122, and rats91, 123-126 have all shown disappointing outcomes, including: (1) deposition of disorganized scar tissue at the tendon-bone interface (Fig. 1.5), (2) the absence of a morphologically normal fibrocartilage transition region, including the aberrant expression of normal fibrocartilage markers (Fig. 1.6A), and (3) reduced biomechanical properties of the tendon-bone unit (Fig.
1.6B). Although the healing insertion site does continue to adapt and remodel over time, normal tissue architecture is not regained even years after the injury94.
Figure 1.5: Tendon insertion site injuries heal via scar formation in animal models. In this example, normal insertion site morphology (a) is not regenerated following surgical injury and repair of the rat supraspinatus tendon. Instead, disorganized scar tissue forms at the interface between tendon and bone by 4 weeks after the initial injury (b). Although the healing insertion site does continue to remodel over time, normal tissue architecture is never regained. T, tendon; B, bone; U-Fc, unmineralized fibrocartilage; M-Fc, mineralized fibrocartilage; IF, fibrovascular scar tissue at the interface. Adapted with permission from (Rodeo, 2007)127.
23
Figure 1.6: The fibrocartilage interface between tendon and bone exhibits abnormal tissue architecture following natural healing or autograft repair, leading to decreased biomechanical properties of the tendon-bone unit. (A) Compared to normal strut (NS), natural healing (NH) and patellar tendon autograft repair (PTA) both exhibit a disorganized enthesis and the presence of fibrous scar tissue following healing of an insertion site injury in a rabbit central-third patellar tendon defect model. The poor reintegration of the tendon with the bone due to the lack of an organized fibrocartilage region is highlighted by collagen type 2 (Col2) immunostaining, which specifically marks fibrocartilage. (B) As a result, the biomechanical properties of repaired tendons were significantly decreased compared to normal. T, tendon; B, bone; FC, unmineralized fibrocartilage; MFC, mineralized fibrocartilage; FS, fibrous scar; IVD, in vivo displacement level; IVF, in vivo force level. Adapted with permission from (Kinneberg et al, 2011)92.
Tendon ruptures are often treated by direct surgical repair or by implanting a tendon graft. Autografts, allografts, and synthetic grafts128 are all available for clinical use.
Unfortunately, these surgical repairs are often weak and prone to re-rupture due to their low intrinsic healing capacity and subsequent failure to reintegrate with the native bone. In recent years, clinicians have become increasingly interested in innovative new therapeutic approaches
24 to stimulate tendon-to-bone healing. A vast number of different cell, scaffold, gene, and growth factor-based tissue engineering strategies aimed at restoring a functional insertion site have been proposed [reviewed in 129]. From a translational perspective, the delivery of exogenous growth factors to injured tendons as an adjunct to surgical repair is a particularly attractive strategy because this approach avoids many of the regulatory issues associated with cell- and gene-based therapies. Numerous pre-clinical animal studies employing a wide variety of growth factors
(BMP, TGFβ, bFGF, PDGF, VEGF, IGF, GDF, HGF, and others) have been performed
[reviewed in 130-133], but to date no single treatment has consistently produced positive outcomes.
As a result, none of these individual growth factors are FDA-approved for clinical use in the context of tendon repair.
1.5. Using tendon development to guide tendon tissue engineering
Clinicians and researchers alike have devoted significant amounts of time and resources in order to study the problem of tendon-to-bone repair, so why are effective treatments still lacking? In our view, one of the major limiting factors has been an incomplete understanding of basic tendon biology. As a result, tissue engineers resorted to a trial and error approach, often delving into experiments without a solid biological rationale to support their hypotheses.
Recognizing this, our biomedical engineering research group at the University of Cincinnati established a formal partnership with developmental biologists at Cincinnati Children’s Hospital
Medical Center via an NIH Bioengineering Research Partnership grant titled “A developmentally-based tissue engineering approach to improve tendon repair” (AR056943 to
D.L. Butler, PI). We believed that in order to successfully engineer a replacement tissue or develop effective treatments for tendon-to-bone healing, it would be necessary to first understand
25 the biological signals that govern how the natural interface between tendon and bone develops111,
134. With this goal in mind, we began investigating tendon development in the mouse in an attempt to identify novel genes or signaling pathways that regulate enthesis differentiation and maturation. As is discussed in Chapter 2, we (and now also others) have shown in a series of recent publications that Indian hedgehog signaling is a critical regulator of enthesis development.
Naturally, this exciting finding also led us to consider the therapeutic potential of Ihh in the context of adult tendon-to-bone repair, an ongoing endeavor which forms the basis for this dissertation.
.
26
CHAPTER 2
Indian Hedgehog Signaling in Bone and Tendon Development
2.1. The hedgehog signaling pathway
The hedgehog signaling pathway is one of the key regulators of vertebrate development135 and has also been implicated in the pathology of many diseases136. Mammals have three known hedgehog homologues: Sonic hedgehog (Shh), Desert hedgehog (Dhh), and
Indian hedgehog (Ihh). These secreted proteins can signal in an autocrine or paracrine manner by binding to the cell surface receptor patched (Ptch1). In the absence of hedgehog ligands, Ptch1 constitutively inhibits the G protein-coupled receptor smoothened (Smo), but the binding of a hedgehog ligand to Ptch1 relieves Smo inhibition and leads to the intracellular activation of zinc finger transcription factors in the Gli family, namely Gli1, Gli2, and Gli3 (Fig. 2.1). These Gli proteins perform context-dependent positive and negative functions during both development and disease137, 138, the details of which are beyond the scope of this dissertation. Importantly, in addition to their other effector functions, Gli proteins participate in a feedback loop by upregulating the expression of genes in the hedgehog pathway, including Ptch1 and Gli1. Thus, since Gli1 gene transcription provides a direct readout of hedgehog pathway activation, mice containing a knock-in of the common bacterial reporter gene β-galactosidase (LacZ) within the
Gli1 locus (termed Gli1-LacZ reporter mice)139 are particularly useful for studying hedgehog signaling and are used extensively in this dissertation.
27
Figure 2.1: The hedgehog signaling pathway. (a) In the absence of hedgehog ligands, patched constitutively inhibits smoothened and no intracellular signal is generated. (b) When hedgehog ligands are present, they bind to patched, which relieves the inhibition on smoothened and allows it to initiate an intracellular signaling cascade that results in activation of zinc finger transcription factors in the Gli family. Adapted with permission from (Weitzman, 2002)140.
2.2. Ihh signaling regulates endochondral bone growth
Ihh is best known for its important role in vertebrate skeletogenesis, where it acts as a master regulator of endochondral bone formation141-143. The skeletal disorder brachydactyly type
A1, first described in 1903 and characterized by shortening of the middle phalanges of all digits and short stature, has been linked to mutations in the human Ihh gene144, 145. Homozygous mutations in Ihh have also been found to cause acrocapitofemoral dysplasia, an autosomal recessive skeletal disorder characterized clinically by short stature with short limbs and radiographically by cone-shaped epiphyses, mainly in the hands and hips146. Not surprisingly,
Ihh-null mice display a severe skeletal phenotype characterized by markedly reduced
28 chondrocyte proliferation, premature chondrocyte hypertrophy, and a complete failure of osteoblast development in long bones147. These observations clearly suggest a role for Ihh in skeletal development, and over the last 20 years, the collective results of many experimental studies have begun to elucidate the specific mechanisms by which Ihh acts to regulate growth plate dynamics and bone growth.
In the growth plate, Ihh is expressed by pre-hypertrophic and early hypertrophic chondrocytes148 and acts via multiple different mechanisms (Fig. 2.2):
1) Ihh induces the expression of parathyroid hormone-related protein (PTHrP) in the
perichondrium, and PTHrP acts to inhibit the transition of proliferating chondrocytes
into hypertrophic chondrocytes. PTHrP synthesis in the growth plate is controlled by
Ihh148 and PTHrP expression is absent from the growth plate in Ihh-deficient mice147.
Although increased levels of exogenous Ihh inhibit chondrocyte hypertrophy, the
deletion of PTHrP or its receptor abolishes this effect148, 149. Notably, the expression
of constitutively active PTHrP receptor in the growth plate of Ihh-null mice reverses
premature chondrocyte hypertrophy but fails to rescue decreased chondrocyte
proliferation150, indicating that not all of Ihh’s effects are mediated through PTHrP-
dependent mechanisms.
2) Ihh acts directly as a positive regulator of chondrocyte proliferation. In support of a
direct role for Ihh, the conditional deletion of Smo in a cartilage-specific manner
decreases chondrocyte proliferation150. In the same study, Ihh was shown to promote
chondrocyte proliferation by increasing the expression of cyclin D1, a positive
regulator of cell cycle progression150. Ihh has also been shown to promote the
29
transition of small round chondrocytes into proliferating chondrocytes by inhibiting
the repressor activity of Gli3, an effect that is independent of PTHrP151, 152.
3) Ihh signaling is directly required for the osteoblast lineage in the endochondral
skeleton. In addition to its regulation of chondrocyte proliferation and hypertrophy,
Ihh is absolutely required for osteoblast formation in endochondral bones147, 153. Ihh
has been shown to up-regulate the expression and enhance the transcriptional activity
of Runt-related transcription factor 2 (Runx2), also known as core-binding factor
subunit alpha-1 (CBF-alpha-1), an essential transcription factor for
osteoblastogenesis154. This pro-osteogenic effect is thought to be mediated via the
activity of Gli2154, 155.
In summary, Ihh is a master regulator of both chondrocyte and osteoblast differentiation during endochondral bone formation. Ihh stimulates chondrocyte proliferation directly in order to maintain a sufficient pool of immature chondrocytes in the proliferative zone of the growth plate.
Through stimulation of PTHrP synthesis, Ihh also determines the distance from the end of the bone at which chondrocytes stop proliferating and undergo hypertrophic differentiation. Finally,
Ihh is required for the differentiation of osteoblasts, which first appear in the bone collar and eventually replace the primary cartilaginous template with mature, mineralized bone. Not surprisingly, Ihh acts synergistically with other signaling pathways in the regulation of osteoblast development, most notably BMP153, 154, 156-158 and Wnt/β -catenin159-161 signaling. Through integration of these multiple different functions, Ihh signaling couples chondrogenesis to osteogenesis during endochondral bone development and controls both the rate and spatial
30 orientation of bone growth162-164. The main functions of Ihh at the growth plate are summarized in Fig. 2.2.
Figure 2.2: The role of Ihh signaling during endochondral bone growth. In the growth plate, Ihh is produced by hypertrophic chondrocytes (purple) and acts via multiple different mechanisms: (1) Ihh induces PTHrP synthesis in the perichondrium at the ends of long bones (orange), and (2) PTHrP acts to inhibit chondrocyte maturation and hypertrophy. (3) Ihh also acts directly to promote proliferation of immature chondrocytes. Together, these mechanisms regulate the spatial and temporal pattern of chondrocyte proliferation and hypertrophic differentiation at the growth plate, thereby maintaining a sufficient pool of proliferating chondrocytes to sustain bone growth. In addition, (4) Ihh acts on nearby perichondrial cells to convert these cells into osteoblasts of the bone collar. Adapted with permission from (Kronenberg, 2003)141.
31
In addition to its roles during bone development, multiple studies have shown that Ihh expression is highly upregulated during the early stages of fracture repair165-168, a process which recapitulates many key elements of embryonic skeletal formation169, 170. This supports one of our research group’s most general hypotheses, namely that restoration of structure and function to damaged adult musculoskeletal tissues likely will require many of the same biological cues and signals that are important during normal development. In fact, it was this hypothesis that led to our initial studies on tendon development and our subsequent investigations into the role of Ihh signaling in the development and maturation of the murine patellar tendon enthesis.
2.3. Ihh signaling regulates enthesis development
Ihh and PTHrP regulate chondrocyte proliferation and differentiation at the growth plate, but recent evidence indicates that they are also expressed in cells at developing tendon-to-bone insertion sites111. Using Gli1-LacZ reporter mice to screen for hedgehog signaling activity during normal tendon development, our research group found that active hedgehog signaling occurs in cells at the immature enthesis (but not in the midsubstance) of the patellar tendon during late embryonic and early postnatal development89 (Fig. 2.3). This finding was our first indication that
Ihh signaling might play an important role in enthesis development. Other research groups have also become interested in hedgehog signaling in the context of tendon development. For example, a recent study by Schwartz, Long, and Thomopoulos171 used a Gli1-CreERT2;mTmG mouse model to identify hedgehog-responsive cells at the developing supraspinatus tendon enthesis. Lineage tracing experiments in this model demonstrated that this hedgehog-responsive cell population goes on to populate the unmineralized fibrocartilage region of the mature enthesis
(Fig. 2.4).
32
Figure 2.3: Active hedgehog signaling at the developing murine PT enthesis. At late embryonic and early postnatal time points (E17.5, P1, P7, and P14), active hedgehog signaling is occurring in cells at the tendon insertion but not in the tendon midsubstance. In these sagittal tissue sections from Gli1-LacZ reporter mice, cells responding to hedgehog are stained blue while non- responsive cells are counterstained red. Adapted with permission from (Liu et al, 2012)89.
Figure 2.4: Hedgehog-responsive cells populate the unmineralized fibrocartilage zone in murine entheses. Lineage tracing experiments using Gli1-CreERT2;mTmG mice show that the unmineralized fibrocartilage in mature (16 week old) tendon and ligament entheses contains cells derived from the hedgehog-responsive cell population identified during early postnatal development. In this mouse model, hedgehog-responsive cells and their progeny fluoresce green, while all other cells fluoresce red. SS, supraspinatus tendon; HH, humeral head; ACH, Achilles tendon; CAL, calcaneus; ACL, anterior cruciate ligament; TIB, tibia. Dotted white lines show the mineralization tidemark. Adapted with permission from (Schwartz et al, 2015)171.
33
Interestingly, Ihh signaling at the developing enthesis appears to be regulated by the mechanical environment. In fact, the entire process of enthesis formation is highly dependent on mechanical factors111, 172-174. Experiments using botulinum toxin to paralyze the supraspinatus muscle (thus decreasing the amount of load on the supraspinatus tendon) in developing mice have shown impaired fibrocartilage formation, reduced mineralization, disorganized collagen fiber distribution, and inferior mechanical properties at the insertion site90, 172, 175. These impairments may be directly related to altered hedgehog signaling, because botox-induced muscle unloading has also been shown to increase the hedgehog-responsive cell population and hedgehog signaling activity at the enthesis171. Additionally, mechanical strain has been shown to directly contribute to fibrochondrogenic differentiation in the Achilles tendon enthesis of miniature pigs by modulating Ihh/PTHrP signaling176. These results complement previous studies that have shown that Ihh expression is modulated by the loading environment177-180.
In order to better understand the functional role of hedgehog signaling in enthesis development, our research group performed a series of gain- and loss-of-function experiments both in vivo and also ex vivo using an organ culture system43, 181. Using both genetic and immunohistochemical strategies, we first verified that Ihh (and not Shh) was the ligand responsible for activating the hedgehog pathway during murine patellar tendon enthesis development43. Next, we generated mice in which all cells expressing Scx also expressed a constitutively active mutant of Smo (CA-Smo). We found that forced activation of the hedgehog signaling pathway in Scx-positive cells drastically altered the expression patterns of insertion site markers, including tenascin-C, biglycan, and collagen type 2. These markers, which are usually restricted to the enthesis, were ectopically expressed throughout the tendon midsubstance in CA-
Smo animals, indicating that their spatial expression patterns are controlled, at least in part, by
34 hedgehog signaling43. Since the CA-Smo animals die at birth due to respiratory failure, we also developed an ex vivo organ culture system to facilitate the study of Ihh’s effects on tendon differentiation at postnatal time points. By culturing tendon explants in this system and exposing them to recombinant Ihh protein (500 ng/ml) for three days, we confirmed that ectopic activation of Ihh signaling resulted in upregulation of insertion site genes and increased expression of insertion site proteins throughout the tendon43. Taken together, these data indicate that Ihh regulates differentiation of the enthesis in the growing murine patellar tendon.
We next carried out loss-of-function experiments in vivo by deleting Smo in Scx- expressing cells and examining the resulting effects on enthesis differentiation and maturation.
Cre-mediated deletion of Smo in tenocytes caused markedly reduced hedgehog signaling at the patellar tendon insertion site but did not disrupt hedgehog signaling in the nearby articular cartilage of the tibia43. Loss of Ihh signaling at the developing enthesis in the tissue-specific Smo knockout animals (Smo tKO) resulted in morphologically abnormal insertion sites as well as phenotypic alterations in the fibrochondrocytes at the tendon-bone interface. These differences were not noticeable at early time points (P1-P7), but subtle changes were apparent by P14 and became more dramatic with increasing age. Compared to age-matched controls, Smo tKO animals exhibited less differentiated cells within cartilaginous lacunae, suggesting that loss of
Ihh signaling caused reduced chondrogenesis at the enthesis43. In further support of this hypothesis, Smo tKO animals also displayed a reduction of cartilage glycosaminoglycans
(GAGs) at the insertion site, as evidenced by decreased alcian blue staining43. At 12 weeks, mineralization deficits were also observed in the Smo tKO animals, most noticeably the absence of a discrete mineralization tidemark, which normally denotes the boundary between the unmineralized and mineralized fibrocartilage zones of the enthesis43. Although collagen fiber
35 orientation appeared overtly normal in the patellar tendons of these animals, ablating Ihh signaling at the enthesis caused a 32% reduction in mineralized fibrocartilage area, leading to less collagen embedded within mineralized tissue181. These histological differences (summarized in Fig. 2.5) indicate that Ihh signaling plays a critical role in chondrogenic differentiation and post-natal mineralization of the fibrocartilaginous enthesis.
Figure 2.5: Histological differences between control and Smo tKO entheses at 12 weeks. Features of the patellar tendon enthesis in control mice include a defined mineralization tidemark (arrow in A), abundant GAGs (arrow in B), and a prominent mineralized fibrocartilage zone (arrows in C and D). In contrast, the enthesis in Smo tKO mice lacks a tidemark (E), has reduced GAG content (F), and exhibits a 32% reduction in mineralized fibrocartilage area (G and H). Adapted with permission from (Liu et al, 2013)43 and (Breidenbach et al, 2015)181.
36
Importantly, we found that a lack of Ihh signaling at the enthesis during development also has functional consequences in the adult. In addition to the morphologic differences discussed above, 12 week old Smo tKO mice displayed significantly reduced biomechanical properties of the patellar tendon-bone unit. Tensile failure testing revealed significantly decreased linear stiffness, increased displacement, and increased failure strain in Smo tKO tendons compared to controls43 (Fig. 2.6A). Smo tKO tendons also showed sub-failure mechanical abnormalities at the enthesis, most notably ~2.3 fold greater strain in the insertion region at load levels greater than
20% of normal failure load181 (Fig. 2.6B). However, midsubstance strains did not significantly differ between Smo tKO and control tendons at any load level181 (Fig. 2.6C). These results further support our conclusion that Ihh signaling is a critical regulator of enthesis development, and that any disturbances in this pathway will lead to chronic functional deficits in the adult.
Figure 2.6: Biomechanical differences between control and Smo tKO tendons at 12 weeks. Smo tKO tendons showed increased strain at failure compared to controls (A). Regional strain measurements revealed increased insertion strains in Smo tKO tendons at all load levels greater than 20% of normal failure load (B, gray area represents p<0.05) but no significant differences in midsubstance strains (C). Adapted with permission from (Breidenbach et al, 2015)181.
37
A recent study by Schwartz, Long, and Thomopoulos171 has confirmed many of our findings regarding the role of Ihh signaling during enthesis development. Not only did they show that mature enthesis fibrocartilage is derived from a population of hedgehog-responsive cells (see
Fig. 2.4), but they also examined the long term effects of ablating this hedgehog-responsive cell population. By crossing Gli1-CreERT2 mice with Rosa-DTA mice, they generated a tamoxifen- inducible mouse model in which Cre-expressing cells produce diphtheria toxin A (DTA), an extraordinarily potent toxin that kills the cells producing it. Using this model, they were able to show that early ablation of the hedgehog-responsive cells at the enthesis resulted in decreased proteoglycan content, a reduction in the number of fibrochondrocytes, and a decrease in mineralized fibrocartilage at 6 weeks171 (Fig. 2.7). In this same study, they also examined the phenotype of Smo tKO mice at 8 weeks of age and found severely disrupted mineralization in the entheseal fibrocartilage as well as significantly reduced biomechanical properties of the tendon-bone unit relative to controls171, thus confirming our research group’s similar findings.
Notably, their work focused on the supraspinatus tendon while our group primarily examined the patellar tendon. The consistency of the results from these two independent studies indicates that the role of Ihh signaling in entheseal fibrocartilage development is likely conserved across different tendons regardless of their anatomical location.
38
Figure 2.7: Loss of Ihh-responsive cells impairs fibrocartilage formation and mineralization at the murine supraspinatus insertion site. Ablating the Ihh-responsive cell population at P6 using a Gli1-CreERT2;Rosa-DTA mouse model resulted in decreased proteoglycan content (A and D, safranin O staining), a reduction in the number of fibrochondrocytes (B and E, DAPI staining), and a decrease in mineralized fibrocartilage (C and F, von Kossa staining) at P42 compared to control animals. Adapted with permission from (Schwartz et al, 2015)171.
2.4. Summary
In addition to its established role as a regulator of endochondral bone formation at the growth plate, the recent studies discussed above43, 89, 171, 181 have convincingly shown that Ihh signaling is a critical regulator of enthesis development. A unique population of hedgehog- responsive (Gli1-positive) cells exists at the immature enthesis in the perinatal period. These cells first appear at late embryonic stages (~E16-18) and go on to populate the fibrocartilage
39 region of the enthesis during postnatal development. Although these cells are derived from the
Scx-positive lineage, they are spatially and morphologically distinct from tenocytes and epiphyseal chondrocytes. Importantly, if these cells are destroyed or if hedgehog signaling is lost during development, then enthesis fibrocartilage differentiation and growth is impaired.
After birth, secondary ossification begins in long bone epiphyses, a process which seems to coincide with the initiation of mineralization in the enthesis. The molecular mechanisms controlling enthesis mineralization are not completely understood, but both mechanical and biological cues are thought to be critical for establishing the mineralization gradient at the insertion site. Mineral apposition begins in the subchondral bone and progresses into the entheseal fibrocartilage, eventually establishing a discrete zone of mineralized fibrocartilage which serves to anchor the collagenous tendon fibers into a stiff, mineralized matrix. Ihh appears to regulate the mineralization process, since decreased hedgehog signaling or loss of the Gli1- positive cell population results in significant reductions in the amount of mineralized fibrocartilage at the enthesis.
Although the mechanistic details can be complex, in a general sense, the roles of Ihh at the growth plate and at the developing enthesis are actually very similar. In the simplest terms,
Ihh signaling performs three major functions:
(1) Stimulate chondrocyte proliferation
(2) Regulate chondrocyte differentiation/maturation
(3) Promote chondrocyte mineralization
If Ihh signaling is removed from the growth plate, the result is a reduction in the number of proliferative chondrocytes, spatiotemporal dysregulation of chondrocyte differentiation and hypertrophy, and a complete failure of bone mineralization. Similarly, if Ihh signaling is
40 removed from the developing enthesis, the result is a smaller and less well differentiated fibrocartilage region that fails to mineralize properly.
It is now clear that Ihh signaling is an important developmental pathway in tendons and is critical for the formation of a functional fibrocartilaginous enthesis. However, despite the mounting evidence that implicates Ihh as one of the major regulators of enthesis development, almost nothing is known about the role of Ihh signaling during adult tendon homeostasis or during enthesis healing following injury. Furthermore, given its ability to stimulate chondrocyte proliferation, differentiation, and mineralization, Ihh is intriguing as a potential therapeutic molecule for augmenting fibrocartilage healing during adult tendon-to-bone repair. In the remainder of this dissertation, we describe our initial attempts to evaluate the therapeutic effects of Ihh delivery during tendon-to-bone healing.
41
CHAPTER 3
Research Approach and Objectives
3.1. Research approach and scope
For over a decade, the main goal of our research laboratory has been to design and test tissue-engineered therapies for tendon repair. Our group has successfully utilized a functional tissue engineering (FTE) approach in which we combine mesenchymal progenitor cells (MPCs), biomaterials, growth factors, and mechanical stimulation to generate tissue engineered constructs
(TECs) in vitro22. These TECs can then be surgically implanted in order to help restore function to damaged tendon tissues (for example, see Fig. 1.1).
Despite our successes using TECs for tendon repair in animal models, this strategy has several important limitations when considering translation to human patients. TECs are created using autologous MPCs, which must be harvested, cultured, and expanded in vitro prior to in vivo use. Once created, TECs would have to be kept viable in a sterile environment and therefore would have a very limited shelf life. As a cell-based therapy, TECs would face very stringent regulation by the FDA, another significant deterrent to potential commercialization. Lastly,
TECs were designed as a synthetic replacement for damaged tendon tissue and mimic the tendon midsubstance in terms of their composition and organization. However, TECs lack the specialized adaptations in structure that are found at the fibrocartilaginous enthesis in many tendons. Thus, TECs might be useful for repairing a midsubstance defect in the body of a tendon, but they have limited utility for injuries involving the tendon-to-bone insertion site. In fact, one
42 recurring observation from our prior tendon repair studies in the rabbit was that the TEC repairs almost always failed at the tendon-bone junction due to poor integration with the native bone.
In recent years, our approach to tendon tissue engineering has evolved. Our initial FTE approach emphasized the biomechanical aspects of tendon healing and repair, but in our current work, we have begun to focus heavily on understanding normal tendon biology and establishing biological design criteria to complement our mechanical design goals. As discussed in chapter 1, one of our primary biological design goals is the restoration of a fibrocartilaginous enthesis in repaired tendons. We believe that in order to make progress towards this goal, it is first necessary to understand the biological cues that govern how the natural interface between tendon and bone develops. Furthermore, we believe that restoration of normal structure and function to the damaged insertion site will require many of the same biological signals that were critical during enthesis development. Through our collaborations with developmental biologists at Cincinnati
Children’s Hospital, we have been investigating the molecular mechanisms that regulate tendon differentiation and growth. As discussed in chapter 2, our work has yielded important mechanistic insights into the role of Ihh signaling during enthesis development.
We now seek to use what we have learned from studying normal tendon development and apply it to an injury-and-repair scenario. In principle, targeting of the Ihh signaling pathway is an attractive strategy for augmenting enthesis repair. However, in practice, translating basic research findings into applied therapies that are both efficacious and clinically feasible is challenging. There are many possible experimental methods for therapeutically targeting the Ihh pathway, such as delivering recombinant Ihh protein to the healing enthesis or modulating hedgehog signaling using a cell- or gene-therapy approach (Table 3.1). Each of these strategies has advantages and disadvantages, and some strategies are much more amenable to clinical
43 translation than others. As will be described later, we chose to design and fabricate Ihh-infused scaffolds and then implant these scaffolds in a murine PT injury model to determine if they could improve tendon-to-bone healing. To our knowledge, this work is the first ever investigation into the therapeutic effects of Ihh protein on enthesis healing in an animal model.
Table 3.1: Experimental approaches for targeting the Ihh signaling pathway in vivo.
Example Pros Cons
Direct drug -Bolus injection of Ihh -Simple and low cost -Poor spatiotemporal control leading to delivery protein or a hedgehog low efficacy or potential off-target agonist effects
Scaffold-based -Infusion of Ihh protein -Better spatiotemporal -Scaffold design, fabrication, and testing drug delivery or a hedgehog agonist control of drug delivery increases time and cost into an implantable, since scaffolds can be controlled-release drug tailored to meet specific -Presence of scaffold may negatively delivery scaffold needs affect healing process; cytotoxicity or potential immune responses can occur
Cell-based -Implantation of cells -Implanted cells can -Requires in vitro cell culture, leading to therapies conditioned to produce integrate into the healing increased time and cost or respond to Ihh tissue and positively contribute to the healing -Cell death, unexpected changes in cell process by producing phenotype, and immune rejection can cytokines, ECM proteins, occur etc. -Potentially serious adverse effects, such -Stem/progenitor cells as cancer have the capability to differentiate into multiple -Clinical translation is more difficult due cell types and regenerate to more stringent FDA regulation entire tissues
Genetic -Targeted genetic -Commonly used in mice -Complex, expensive, and time approaches editing of Ihh signaling to study signaling consuming pathway components pathways -Not directly clinically translatable to -Precise spatiotemporal humans at this time control can be achieved via tissue-specific, inducible genome editing technologies
44
3.2. Specific aims and hypotheses
The work presented in this dissertation seeks to further our long term goal of developing novel therapies for tendon repair by focusing specifically on augmenting tendon-to-bone healing.
We believe that the normal developmental signals governing tenogenesis, such as Ihh signaling at the enthesis, can be strategically reintroduced and/or manipulated during adult tendon healing in order to achieve better functional outcomes. This type of biologic augmentation could one day be combined with traditional surgical repair, thus allowing orthopaedic surgeons to treat tendon injuries with a more effective, multi-faceted approach.
Given Ihh’s importance at the insertion site during development, the overall objective of this dissertation is to evaluate the therapeutic potential of Ihh for adult tendon-to-bone repair. In this work, we will examine the natural patterns of endogenous Ihh signaling in adult tendon entheses, assess the effects of Ihh stimulation on tendon fibroblasts in vitro, and then attempt to therapeutically upregulate the Ihh signaling pathway by delivering exogenous Ihh to the damaged enthesis. Our central hypothesis is that increasing Ihh signaling activity at the healing insertion site will promote the restoration of a zonal fibrocartilaginous enthesis, thereby improving tendon biomechanical repair outcomes. Our specific aims are as follows:
AIM 1: Develop a standardized and repeatable surgical model of acute murine PT enthesis injury and biomechanically characterize natural tendon-to-bone healing in this model.
Before we can begin examining the therapeutic effects of Ihh, we first need to develop and characterize an appropriate animal model of adult tendon-to-bone healing. Although much of our lab’s prior work was done in the rabbit PT, murine models are advantageous due to their lower cost and genetic tractability. However, the small size of the murine PT makes surgical
45 procedures and biomechanical testing much more challenging. In aim 1, we seek to characterize the global mechanical properties of native murine PT, the regional differences in tissue strains during loading, and the biomechanical outcomes of natural murine PT-to-bone healing following a surgical enthesis injury. Based on data from larger animals and humans, as well as our own prior experience working with an empty central-third PT defect model in the mouse50, we hypothesize the following:
Hypothesis 1: Native murine PT will exhibit significantly increased tissue strains in
the insertion region compared to the midsubstance during loading.
Hypothesis 2: Following enthesis injury, naturally healing tendons will exhibit
inferior material properties and increased insertion strain compared to controls.
Hypothesis 3: In injured tendons allowed to heal naturally, biomechanical failures
will initiate at the enthesis.
These hypotheses are rooted in the fact that the native tendon enthesis contains an organized fibrocartilage transition zone which endows the insertion site with increased elasticity and toughness to prevent damage during loading. This fibrocartilage tissue is not regenerated following injury and is replaced by large amounts of disorganized scar, thus permanently altering the biomechanics of the tendon-bone unit and making the insertion site the weakest link in the force transfer chain. By first defining the mechanical properties of the native murine PT and how these properties change following a surgical enthesis injury, we will be well positioned to move forward and assess whether or not Ihh delivery affects tendon-to-bone healing and improves functional success in this model system.
46
AIM 2: Identify the distribution of Ihh-responsive cells in adult murine tendons in vivo and assess the effects of Ihh stimulation on tendon-derived cells in vitro.
Ihh signaling has been shown to regulate enthesis development, but its role during adult tendon homeostasis or following an enthesis injury is unknown. In aim 2, we seek to examine
Ihh signaling in the adult murine patellar and Achilles tendons by defining which cells are actively responding to hedgehog in vivo. Based on our own studies of hedgehog signaling during enthesis development as well as experiments conducted by others to trace the hedgehog- responsive (Gli1-positive) cell lineage throughout post-natal tendon growth171, we hypothesize:
Hypothesis 4: Adult murine tendons will contain Ihh-responsive cells in the enthesis
fibrocartilage, but endogenous hedgehog signaling activity at the enthesis will
decrease over time concomitant with the slowing of longitudinal bone growth and
decreasing mineral apposition rate at the tendon-bone junction.
Hypothesis 5: Hedgehog signaling activity will not be detected in the midsubstance of
adult murine tendons.
Although we hypothesize that hedgehog signaling will not be active in the tendon midsubstance, this does not necessarily mean that the tendon fibroblasts which reside in this region are incapable of responding to Ihh. In fact, we would like to know the specific effects of
Ihh stimulation on tendon fibroblasts, since they are the main cell type in healthy tendon and also act as primary mediators of tendon repair following injury. Are these differentiated cells capable of responding to Ihh? If so, what are the phenotypic changes that result from Ihh stimulation? To try and answer these questions, we will isolate tendon fibroblasts from adult mice, culture them in monolayer, expose them to different concentrations of recombinant Ihh protein, and assess the resulting effects. This work is novel because there are currently no published studies describing
47 the effects of Ihh on tendon-derived cells in vitro. Based on our in vivo studies of Ihh function during enthesis development, we propose the following hypothesis:
Hypothesis 6: Ihh stimulation will induce proliferation, chondrogenic
transdifferentiation, and mineralization of tendon fibroblasts in vitro.
Collectively, the studies in aim 2 will give us a better understanding of Ihh signaling activity in mature tendons. Understanding the patterns of Ihh signaling in adult tendons and the effects of Ihh on tendon-derived cells will aid us in the design of our Ihh-based therapies and provide clearer indications as to how delivery of exogenous Ihh might need to be precisely localized to a specific cell population in order to be effective. Additionally, the in vitro component of aim 2 will allow us to directly assess how cultured tendon fibroblasts respond to different concentrations of Ihh under controlled conditions. Taken together, the knowledge gained from this aim could greatly aid our translational efforts and allow us to design more efficacious and targeted therapies for tendon-to-bone repair.
AIM 3: Design and fabricate biodegradable, polymeric, Ihh-infused scaffolds, then test their ability to improve tendon-to-bone repair outcomes in our murine PT injury model.
In our final aim, we attempt to deliver recombinant Ihh protein to the healing murine PT enthesis with the expectation that this therapeutic intervention will improve functional healing outcomes in our murine injury model. By collaborating with an expert materials scientist (Dr.
Heather Powell, The Ohio State University), we will design and fabricate biodegradable, polymeric, Ihh-infused scaffolds which will serve as drug delivery devices. The composition and properties of these novel scaffolds can be tailored in order to control the spatiotemporal drug release profile. Once prototype scaffolds have been fabricated, we will test their performance in
48 vitro. Based on our desired design parameters and the performance of similar scaffolds created for other drug delivery applications, we hypothesize:
Hypothesis 7: Our Ihh-infused scaffolds will release bioactive Ihh protein for a
sustained time period (>2 weeks), thereby upregulating hedgehog signaling locally.
Once we have settled on a final design, the Ihh-infused scaffolds will be surgically implanted adjacent to the healing murine PT enthesis. Based on our current knowledge of Ihh’s function at the enthesis during tendon development, we hypothesize:
Hypothesis 8: Delivery of Ihh to the healing enthesis will stimulate the proliferation,
differentiation, and mineralization of neo-fibrocartilage tissue.
Hypothesis 9: Delivery of Ihh to the healing enthesis will result in improved
biomechanical repair outcomes compared to controls.
Ultimately, aim 3 will allow us to determine if delivery of recombinant Ihh protein to the healing PT enthesis improves functional healing outcomes in our murine injury model. To our knowledge, this represents the first ever attempt to discern whether recombinant Ihh can be used therapeutically in the context of tendon healing. Thus, our work paves the way for future tissue engineering studies in which the Ihh signaling pathway is modulated during tendon repair.
3.3. Significance and impact
This work is significant because it addresses the common problem of failed tendon-to- bone repair from a biological, clinical, and an engineering perspective. Using our knowledge of tendon development coupled with our expertise in musculoskeletal tissue engineering, we have developed a therapeutic strategy in which we deliver signaling molecules (e.g. recombinant Ihh
49 protein) to the damaged tendon in order to upregulate specific biological pathways (e.g. hedgehog signaling) that are known to be important during normal development. The work described in this dissertation is very innovative because Ihh has never before been investigated as a potential mediator of tendon-to-bone healing. Although hedgehog signaling is clearly important during enthesis development, our work represents the first investigation into the ability of exogenous Ihh to improve tendon-to-bone healing in the adult. Furthermore, using a custom- designed and fabricated scaffold infused with recombinant Ihh to augment the natural healing process represents a novel strategy to improve tendon-to-bone repair.
The work described in this dissertation will be impactful in three major ways: 1) our experiments will help clarify Ihh’s functional effects, both on cultured murine cells and during in vivo tendon-to-bone healing; 2) our approach will test the feasibility of using surgically- implanted, biodegradable polymeric scaffolds as a means of delivering Ihh (or potentially other signaling molecules) to specific regions of the tendon for sustained time periods; and 3) our results will determine whether or not delivery of exogenous Ihh improves tendon-to-bone repair, thus validating its therapeutic potential and warranting further studies in larger animal models.
50
CHAPTER 4
Murine Patellar Tendon Biomechanical Properties and Regional Strain Patterns during Natural Tendon- to-Bone Healing after Acute Injury
Steven D. Gilday a,b, E. Chris Casstevens c, Keith Kenter c, Jason T. Shearn a, David L. Butler a
a Biomedical Engineering Program, College of Engineering and Applied Science, University of Cincinnati, Cincinnati, OH
b Medical Scientist Training Program, College of Medicine, University of Cincinnati, Cincinnati, OH
c Department of Orthopaedic Surgery, College of Medicine, University of Cincinnati, Cincinnati, OH
Note: This manuscript is published in the Journal of Biomechanics88.
51
4.1. Abstract
Tendon-to-bone healing following acute injury is generally poor and often fails to restore normal tendon biomechanical properties. In recent years, the murine patellar tendon (PT) has become an important model system for studying tendon healing and repair due to its genetic tractability and accessible location within the knee. However, the mechanical properties of native murine PT, specifically the regional differences in tissue strains during loading, and the biomechanical outcomes of natural PT-to-bone healing have not been well characterized. Thus, in this study, we analyzed the global biomechanical properties and regional strain patterns of both normal and naturally healing murine PT at three time points (2, 5, and 8 weeks) following acute surgical rupture of the tibial enthesis. Normal murine PT exhibited distinct regional variations in tissue strain, with the insertion region experiencing approximately 2.5 times greater strain than the midsubstance at failure (10.80 ± 2.52% vs. 4.11 ± 1.40%; mean ± SEM). Injured tendons showed reduced structural (ultimate load and linear stiffness) and material (ultimate stress and linear modulus) properties compared to both normal and contralateral sham-operated tendons at all healing time points. Injured tendons also displayed increased local strain in the insertion region compared to contralateral shams at both physiologic and failure load levels.
93.3% of injured tendons failed at the tibial insertion, compared to only 60% and 66.7% of normal and sham tendons, respectively. These results indicate that 8 weeks of natural tendon-to- bone healing does not restore normal biomechanical function to the murine PT following injury.
52
4.2. Introduction
Tendon injuries occur in a diverse patient population and commonly result in pain, disability, and significant healthcare costs4. Frequently injured tendons include the Achilles, patellar, and quadriceps tendons of the lower extremity, the biceps and rotator cuff tendons of the upper extremity, and the flexor and extensor tendons of the hands182. Treatment of such injuries routinely requires reattachment of a ruptured tendon to its bony insertion, but this presents a challenge due to the extreme difference in material properties between soft and hard tissue. To cope with this material mismatch, the uninjured tendon-to-bone insertion site, also known as the enthesis, exhibits a gradual transition between the compliant tendon and the much stiffer bone via a fibrocartilaginous transition region109. Gradations in matrix composition, collagen alignment, cell phenotype, and mineralization28, 79, 80 help facilitate optimal force transmission while also dissipating potentially damaging interfacial stress concentrations between these mechanically dissimilar materials3, 183. Unfortunately, once disrupted, the insertion site does not regenerate its complex natural architecture and is instead replaced by scar tissue, resulting in a mechanically inferior interface that is susceptible to further injury91-94.
Functional tissue engineering (FTE), an evolving discipline which emphasizes the restoration of normal mechanical function in damaged load-bearing tissues, has been proposed as a promising alternative to traditional tendon repair strategies14, 15. Fundamental to the FTE paradigm is the need to measure the biomechanical properties of normal and naturally healing tissues under physiologic as well as failure loads in order to establish quantitative benchmarks against which tissue-engineered repairs can be compared14. Working towards this goal, our research group determined that in vivo patellar tendon (PT) forces in the rabbit23 and goat26 reached 21% and 40% of normal PT failure force, respectively, during simulated activities of
53 daily living (ADLs). Then, using these physiologic force thresholds as mechanical benchmarks, we evaluated the relative success of various tissue-engineered tendon repairs in a full-length rabbit PT defect model22. Although this FTE approach did yield improved mechanical outcomes compared to natural healing alone, the vast majority of our tissue-engineered PT repairs still fail prematurely at the distal insertion, indicating a need for better strategies to stimulate tendon-to- bone healing.
More recently, our attempts to regenerate functional tendon-bone interfaces have necessitated moving from large animal models such as the rabbit to the more genetically tractable mouse. The availability of transgenic and knockout mice has permitted detailed studies of PT enthesis development41, 43, 89 and PT natural healing47, 49, 50. Lower costs and higher throughputs also make murine models an attractive option for screening the efficacy of novel therapeutic treatments for tendon-to-bone healing before scaling up to more clinically relevant model systems. However, applying the FTE paradigm to PT repair in the mouse has proven difficult because the peak in vivo forces in the murine PT are unknown and the mechanical properties of native murine PT, specifically the regional (insertion site versus midsubstance) differences in tissue strains during loading, have not been adequately described. Furthermore, the biomechanical outcomes of natural tendon-to-bone healing after murine PT enthesis injury have not been well characterized.
Thus, the objective of this study was to analyze the global biomechanical properties and regional strain patterns of 1) normal murine PT and 2) naturally healing murine PT at three time points (2, 5, and 8 weeks) following acute surgical rupture of the tibial enthesis. We hypothesized that normal murine PT would exhibit regional variations in tissue strain, with the more compliant insertion region experiencing larger strain than the stiffer midsubstance. We also
54 hypothesized that at all time points following enthesis injury, healing tendons would exhibit reduced global biomechanical properties and increased strain in the insertion region compared to contralateral shams, resulting in failure initiation at the insertion site.
4.3. Materials and methods
4.3.1. Experimental design
Patellar tendon dimensions, structural and material properties, regional strain patterns, and failure locations were assessed at three different post-injury time points (2, 5, and 8 weeks) in a cohort of 30 twenty-week-old (20.3 ± 0.5 weeks; mean ± SD) male CD-1 wild-type mice.
Twenty-week-old mice were chosen for this study because they are skeletally mature adults whose patellar tendons are large enough to allow for the creation of standardized, repeatable surgical injuries and the biomechanical testing of normal and healing tissues in vitro. The study time points were carefully selected in order to capture both the proliferative and remodeling phases of tendon healing and to keep consistent with our group’s previous work on natural healing of murine PT50. Following surgical injury, naturally healing tendons (n=10 per time point) were directly compared with contralateral shams (n=10 per time point). Inter-animal comparisons were also made using a separate group of normal, unoperated patellar tendons
(n=10) from healthy twenty-week-old male CD-1 mice.
4.3.2. Murine patellar tendon injury model
All murine surgeries were performed by one coauthor (ECC) and were approved by the
University of Cincinnati Institutional Animal Care and Use Committee. Mice were anesthetized with 4% isoflurane, subcutaneously injected with 1 mg/kg buprenorphine, and both hind limbs
55 were shaved and aseptically prepped. Using surgical loupes (2.5x), small (0.5-1 cm) longitudinal skin incisions were made to expose the PT in each limb. An acute surgical injury was then created in the left PT while the contralateral PT was subjected to a sham procedure.
Surgical injury (Fig. 4.1A): Using a previously described surgical technique50, two full- length longitudinal incisions were created in the left PT in order to isolate the central-third portion of the tendon from adjacent medial and lateral struts. The central-third of the PT was then transected at its distal insertion into the tibia. Any remaining soft tissue at the insertion site was removed with microsurgical scissors and the enthesis was further disrupted by using a small jigsaw blade to create a shallow bony defect. Care was taken not to damage the intact struts. The tendon’s proximal patellar insertion was also left intact. The transected central-third was laid back in its normal anatomic position between the struts with the distal end in close proximity to its original insertion site on the tibia, but no attempt was made to physically reattach the tendon tissue to the bone.
Contralateral sham (Fig. 4.1B): The central-third of the right PT was isolated from the struts as described above but was not transected at the distal insertion. Thus, sham-operated tendons retained a structurally intact tendon-bone interface at both the tibial and patellar ends.
In both injured and sham limbs, skin incisions were closed with 5-0 prolene suture. Mice were allowed full range of motion and unlimited cage activity immediately following surgery.
No gait alterations or behavioral changes were noted as a result of the surgical procedure. At the designated post-surgical time point (2, 5, or 8 weeks), mice were euthanized by carbon dioxide asphyxiation and frozen at -20 °C to await biomechanical testing.
56
Figure 4.1: Murine patellar tendons were subjected to either an acute surgical injury or a contralateral sham procedure. To create the surgical injury (A), two full-length longitudinal incisions were made to isolate the central third of the tendon, which was then transected at the tibial insertion. A shallow bony defect was created in the tibia and the transected central-third was laid back in its normal anatomic position between the medial and lateral struts to facilitate tendon-to-bone healing. For the contralateral sham procedure (B), longitudinal incisions were made to isolate the central third of the tendon, but the tibial insertion was left intact.
4.3.3. Biomechanical testing and analysis
On the day of testing, murine hind limbs were thawed and dissected to expose the PT.
After noting gross morphological appearance, the central-third of each PT was isolated by dissecting away the medial and lateral struts. Using 6-0 silk suture soaked in Verhoeff’s stain, tendons were marked with three horizontal stain lines located just distal to the tibial insertion, approximately 1 mm proximal to the tibial insertion, and approximately 2 mm proximal to the tibial insertion, respectively. The stain lines clearly delineated the insertion region (defined in this study as the distal 1/3 of the PT) from the midsubstance region (the central 1/3 of the PT).
57
Tibia-central-third PT-patella units were mounted in a materials testing system (100R, Test
Resources) by embedding the tibia in the upper grip with polymethylmethacrylate and fixing it in place with a metal staple, then securing the patella in a pre-existing, conical-shaped lower grip
(Fig. 4.2A). Once mounted, tendons were submerged in PBS at 37 °C and preloaded to 0.02 N50.
High resolution (~6 um/pixel) digital images were taken in both the frontal and sagittal planes in order to calculate initial tendon dimensions. Following a preconditioning phase (25 cycles, 0-1% strain, 0.003 mm/s), specimens were failed in uniaxial tension at a rate of 0.003 mm/s while recording grip-to-grip displacement and load50. Images of the tendon’s anterior surface were captured at 15 second intervals throughout the failure test in order to optically measure regional tissue strains and assess failure location.
Ultimate load, displacement at failure, ultimate stress, and strain at failure were recorded for each PT specimen. Using an automated linear regression algorithm, stiffness and modulus were calculated from the linear region of the load-displacement and stress-strain curves, respectively. Regional tissue strains were calculated by optically tracking the applied stain lines.
For each specimen, the digital images captured during the failure test were stacked, cropped, and thresholded so that frame-by-frame centroid displacements of the stain lines could be automatically tracked using the MTrack2 plugin for ImageJ (Fig. 4.2B-D). This raw displacement data was used to calculate and plot the tensile strain in both the distal insertion and midsubstance regions as a function of load. Failure location (tibial insertion, midsubstance, or patellar grip) was also noted for each specimen.
58
Figure 4.2: Central-third patellar tendon specimens were marked with three stain lines and loaded into a tensile testing system by embedding the tibia in polymethylmethacrylate, fixing the bone in place with a metal staple, then securing the patella in a conical-shaped grip (A). To calculate local tissue strain in the insertion and midsubstance regions, high resolution images were captured at 15 second intervals during the failure test (B) and thresholded (C) so that the frame-by-frame displacement of each optical strain marker could be tracked using the MTrack2 plugin for ImageJ (D).
4.3.4. Statistical analysis
All data were verified to be normal and homoscedastic prior to statistical testing. Two- way factorial ANOVA was used to determine the main effects of surgical treatment and time post-surgery on PT cross-sectional area, structural properties, material properties, and regional tissue strains. At each time point (2, 5, and 8 weeks), significance between treatment groups
(normal, sham, injury) was assessed using one-way ANOVA followed by Fisher’s least significant difference (LSD) post-hoc comparisons. The effect of treatment on failure location was evaluated using multinomial logistic regression. Significance was set at p < 0.05. All statistical testing was performed using IBM SPSS Statistics 21.0.
59
4.4. Results
4.4.1. Gross morphology and tendon dimensions
At 2 weeks post-injury, gross observation revealed dark pink granulation tissue at the healing tendon-bone interface and a thickened, fibrotic paratenon on both the anterior and posterior surfaces of the PT. At 5 and 8 weeks post-injury, the scar-like repair tissue at the insertion site appeared more integrated with the underlying bone and adjacent struts but remained discolored. Fibrous adhesions to subcutaneous connective tissue or the infrapatellar fat pad were present in some specimens. At all post-surgical time points, the injured tendons exhibited significantly increased cross-sectional area compared to both normal and contralateral sham tendons (p < 0.05; Table 4.1). The sham procedure resulted in mild fibrosis on the surface of the PT but did not significantly affect tendon cross-sectional area.
4.4.2. Structural and material properties
Surgical treatment and time post-surgery each significantly affected PT structural and material properties. Injured tendons displayed significantly reduced ultimate load, linear stiffness, ultimate stress, and linear modulus at both 2 and 5 weeks post-surgery compared to normal and sham tendons (p < 0.05; Table 4.1, Fig. 4.3A and Fig. 4.4). Although ultimate load had returned to 87% of normal values by 8 weeks post-surgery (p = 0.139), linear stiffness, ultimate stress, and linear modulus only reached 79%, 49%, and 42% of normal values, respectively (p < 0.05; Fig. 4.4). The ultimate load and ultimate stress of the injured tendons increased linearly over time, whereas linear stiffness and linear modulus increased significantly only between 2 and 5 weeks post-surgery (p < 0.001 in both cases), plateauing between the 5 and
8 week time points (p = 0.817 and p = 0.784, respectively; Fig. 4.4). None of the structural or
60 material properties of the sham-operated tendons were significantly different from normal at any of the post-surgical time points (Table 4.1, Fig. 4.3B and Fig. 4.4).
Table 4.1: Structural and material properties of normal, sham, and injured central-third murine patellar tendons at 2, 5, and 8 weeks post-surgery (n = 10 per group; mean ± standard deviation).
Cross-Sectional Ultimate Ultimate 2 Stiffness (N/mm) Modulus (MPa) Area (mm ) Load (N) Stress (MPa) Normal PT 0.26 ± 0.05 4.73 ± 1.03 11.55 ± 2.32 17.96 ± 3.09 140.04 ± 19.60 Sham (2 week) 0.32 ± 0.07 5.19 ± 1.30 10.84 ± 2.68 16.84 ± 4.85 115.04 ± 30.29 Injury (2 week) 0.59 ± 0.09 a,b 2.11 ± 0.74 a,b 4.28 ± 1.63 a,b 3.02 ± 0.93 a,b 20.49 ± 6.53 a,b Sham (5 week) 0.30 ± 0.05 5.25 ± 1.13 11.86 ± 1.58 17.54 ± 3.81 129.37 ± 26.88 Injury (5 week) 0.49 ± 0.10 a,b 2.90 ± 0.64 a,b 7.43 ± 2.28 a,b 6.15 ± 1.71 a,b 53.05 ± 19.81 a,b Sham (8 week) 0.31 ± 0.04 5.36 ± 1.09 11.53 ± 1.32 17.45 ± 3.13 121.95 ± 14.97 Injury (8 week) 0.47 ± 0.07 a,b 4.10 ± 0.92 b 8.29 ± 1.34 a,b 8.82 ± 1.51 a,b 58.60 ± 11.43 a,b a Significantly different compared to normal PT. b Significantly different compared to contralateral sham at the same time point.
Figure 4.3: Average stress-strain curves for injured and sham tendons at 2, 5, and 8 weeks post- surgery. (A) Injured tendons showed significantly decreased material properties at all time points compared to normal PT (p < 0.05). (B) The sham procedure had no effect on material properties at any time point. Error bars indicate SEM; n = 10 per group.
61
Figure 4.4: Structural and material properties of injured and sham tendons plotted as a percent of normal. With the exception of ultimate load at 8 weeks, injured tendons showed significantly reduced ultimate load (A), linear stiffness (B), ultimate stress (C), and linear modulus (D) compared to both normal and sham at all time points (p < 0.05). The ultimate load and ultimate stress of injured tendons increased linearly over time (A and C), whereas linear stiffness and linear modulus increased significantly between 2 and 5 weeks but then plateaued between 5 and 8 weeks (B and D). The sham procedure had no effect on structural or material properties at any time point. Error bars indicate SD; n = 10 per group.
4.4.3. Regional tissue strains
Normal, unoperated murine PTs exhibited distinct regional variations in tissue strain
(Fig. 4.5). At all load levels greater than 0.5 N, average local strains in the insertion region were
2-3 times greater than corresponding strains in the tendon midsubstance. In both the insertion
62 and midsubstance regions, a non-linear relationship appeared to exist between load and strain. As load increased from 0 to 1 N, strain in the insertion and midsubstance regions increased rapidly, but as load continued to increase past 1 N, the resulting increases in local strain became less pronounced. At failure, insertion strain had reached a maximum value of 10.80 ± 2.52% (mean ±
SEM) whereas midsubstance strain had plateaued at a maximum value of only 4.11 ± 1.40%.
Figure 4.5: Normal murine patellar tendons exhibited distinct regional variations in tissue strain. At all load levels greater than 0.5 N, local strains in the insertion region (top curve) were 2-3 times greater than corresponding strains in the tendon midsubstance (bottom curve). At failure, insertion strain reached a maximum value of 10.80 ± 2.52% (mean ± SEM) compared to only 4.11 ± 1.40% in the midsubstance. The curves represent the average of 10 specimens; error bars indicate SEM.
63
Tendons subjected to an enthesis injury showed increased local strain in the insertion region compared to normal tendons at all time points (Fig. 4.6A). As expected, the healing tissue at the insertion site was more compliant than the normal enthesis, with strain at failure reaching a maximum of 18.85 ± 4.37% after 8 weeks of healing. The sham procedure did not affect local strain in the insertion region (Fig. 4.6C). Surprisingly, both the injury and sham treatments produced increases in midsubstance strain compared to normal tendons (Fig. 4.6B and D).
Factors such as inflammation, disruption of normal blood supply, altered mechanical loading, or tissue fibrosis in response to the surgical procedure may have contributed to the increased local strain in the midsubstance.
Since the small size of the mouse prevents direct in vivo measurement of PT forces, we chose to assess local strains in the insertion and midsubstance at 21% and 40% of normal PT failure force, which correspond to the peak in vivo forces measured in the rabbit23 and goat26 patellar tendon, respectively. At these physiologic load levels, local strains in the insertion region were significantly increased in injured tendons compared to contralateral shams at the 5 and 8 week time points (p < 0.05; Fig. 4.7), indicating that the natural healing process was unable to regenerate a mechanically normal enthesis over this time period. Local strains in the tendon midsubstance did not differ between the injured and sham groups (Fig. 4.7).
64
Figure 4.6: Regional strains in injured and sham tendons compared to normal. At all post- surgical time points, insertion strains were increased in injured tendons (A) but not in contralateral shams (C). In contrast, midsubstance strains were increased in both the injured (B) and sham (D) tendons following surgery. Error bars indicate SEM; n = 10 per group.
65
Figure 4.7: Regional strains differed significantly between injured and sham tendons at physiologic load levels. (A) At 21% of normal PT failure force, injured tendons showed increased local strain in the insertion region (solid black bars) compared to contralateral shams (solid grey bars) at both the 5 and 8 week time points (*p < 0.05). (B) At 40% of normal PT failure force, injured tendons showed increased local strain in the insertion region (solid black bars) compared to contralateral shams (solid grey bars) at all time points (*p < 0.05). No significant differences in midsubstance strain were detected between the injured (hashed black bars) and sham (hashed grey bars) groups at any time point. a Peak in vivo forces measured in the rabbit patellar tendon23. b Peak in vivo forces measured in the goat patellar tendon26. Error bars indicate SEM; n=10 per group.
4.4.4. Failure location
Surgical treatment was found to be a significant predictor of failure location (p = 0.043).
Twenty-eight out of thirty injured tendons (93.3%) failed at the tibial insertion compared to only
60% and 66.7% of normal and sham tendons, respectively (Fig. 4.8A). Although the insertion was the most common site of failure in all treatment groups, tendons subjected to a surgical injury were 1.4 times as likely to fail at the insertion site compared to sham tendons (p = 0.041).
Normal and sham tendons often failed via a delamination mechanism in which the anterior and posterior portions of the tendon separated and slid past one another (Fig. 4.8B). In contrast,
66 surgically injured tendons most often failed via a transverse rupture of the repair tissue at the tendon-bone junction (Fig. 4.8C).
Figure 4.8: Failure location and failure mechanism differed between injured and sham tendons. (A) A higher percentage of injured tendons failed at the tibial insertion compared to contralateral shams (93.3% vs. 66.7%; *p = 0.041). (B) Normal and sham tendons often failed via a delamination mechanism in which the anterior and posterior portions of the tendon separated and slid past one another. (C) Surgically injured tendons most often failed via a transverse rupture of the repair tissue at the tendon-bone junction.
4.5. Discussion
The objective of this study was to analyze the global biomechanical properties and regional strain patterns of both normal and naturally healing murine PT at 2, 5, and 8 weeks following acute surgical rupture of the tibial enthesis. Our first hypothesis was that normal murine PT would exhibit regional variations in tissue strain, with the more compliant insertion region experiencing larger strain than the stiffer midsubstance. By optically tracking surface strain markers during biomechanical testing, we found that local strains in the insertion region were 2-3 times greater than corresponding strains in the midsubstance at all load levels greater
67 than 0.5 N. At failure, the insertion region had experienced 10.80 ± 2.52% strain (mean ± SEM) compared to only 4.11 ± 1.40% in the midsubstance. In addition to our findings in the murine
PT, regional variations in local tissue strain have been recorded in human patellar30, 184 and supraspinatus185 tendon specimens, as well as in rat tibialis anterior29, 186, murine Achilles31, and frog semitendonosis187 tendons. These results imply that normal tendon biomechanical properties vary along the tendon length, likely due to differences in tissue composition and structure between the midsubstance and insertion regions28 or as a result of adaptation to subtle differences in the in vivo loading environment27, 188, 189. In fact, differences in collagen crimp pattern at rest and fiber realignment during mechanical loading exist between the tendon insertion and midsubstance85-87, which may explain the observed differences in strain pattern.
Our second hypothesis was that at all time points following enthesis injury, healing tendons would exhibit reduced global biomechanical properties and increased strain in the insertion region compared to contralateral shams, resulting in failure initiation at the insertion site. Although ultimate loads returned to 87% of normal levels in the injured tendons at 8 weeks post-surgery, linear stiffness, ultimate stress, and linear modulus remained significantly inferior to normal and shams at all time points. Injured tendons also displayed increased local strain in the insertion region at both physiologic and failure load levels. Additionally, 93.3% of injured tendons failed at the tibial insertion, compared to only 60% and 66.7% of normal and sham tendons, respectively. Taken as a whole, these results indicate that 8 weeks of natural tendon-to- bone healing does not restore normal biomechanical function to the murine PT following an acute injury to the tibial enthesis. Studies of tendon-to-bone healing in other injury models have reported similar results. For example, the rat supraspinatus tendon also exhibits impaired biomechanical properties after 8 weeks of tendon-to-bone healing following acute injury91.
68
Interestingly, biomechanical outcomes in our injury model (in which the central-third of the PT was transected at the tibial insertion and the damaged flap of tendon tissue left in place to heal back to the bone) were no better than in a full-length, full-thickness central-third murine PT defect model (in which the central-third of the tendon was completely removed, leaving an empty defect)50. This finding, coupled with the fact that cross-sectional area increased significantly in the injured tendons, indicates that failed healing is not the result of a lack of new tissue formation, but instead is due to the inability of the healing soft tissue to reintegrate with bone.
There were limitations to our study. (1) After creating the acute injury, we did not attempt to repair or reattach the injured tendon to the underlying bone. Thus, this particular model is not representative of a surgical tendon repair, but instead mimics unaided tendon-to- bone healing in a load-protected environment in which the healing central-third PT is under no initial tension. (2) Only the central-third portion of the PT was biomechanically tested, and so any changes in the struts in response to the injury or sham procedure would not have been detected. This is an important consideration, since the intrinsic tendon healing response may actually initiate in the paratenon surrounding the defect47, and any adaptations due to altered loading in the struts following injury could contribute significantly to overall tendon biomechanics. (3) In this study, we calculated global tissue strain based on grip-to-grip displacement measurements, but we calculated local tissue strain by optically tracking stain lines on the anterior surface of the PT during the failure test. It is well known that these two techniques for estimating strain do not always produce equivalent results186, 190, and in our study, normal murine PT experienced grip-to-grip strains in excess of 20% at failure but local strain measurements in the insertion and midsubstance regions never exceeded 12%. However, we did
69 not measure local strains in the most proximal region of the PT, and large strains in this tissue region could have accounted for the apparent discrepancy.
The results of this study indicate that 8 weeks of natural tendon-to-bone healing does not restore normal biomechanical function to the acutely injured murine PT enthesis. Although we focused only on biomechanical outcomes, biological repair outcomes also need to be assessed in future studies in order to better understand the linkages between tendon biology and mechanics.
Evaluating cell phenotype, gene expression, matrix composition, and collagen alignment in the healing tissue would help reveal the underlying biological differences compared to the normal enthesis, which in turn could help explain the observed alterations in mechanics. Moving forward, this murine PT injury model will be used to test the efficacy of novel biologic treatments for tendon-to-bone healing, including new functional tissue engineering strategies aimed at regenerating a normal tendon-bone interface.
4.6. Acknowledgements
We gratefully acknowledge the National Institutes of Health for providing research support (R01 AR056943) as well as student funding via the University of Cincinnati MSTP training grant (T32 GM063483). We also thank Dr. Andrew Breidenbach, Cindi Gooch, and Dr.
Andrea Lalley for their assistance during animal surgeries and Dr. Lou Soslowsky for his valuable contributions to our current biomechanical testing protocols.
70
CHAPTER 5
Hedgehog Signaling is Active in Unmineralized Enthesis Fibrocartilage but also in Regions of Fibrocartilaginous Metaplasia in Adult Murine Tendons
Steven D. Gilday a,b, David L. Butler a, Jason T. Shearn a
a Biomedical Engineering Program, College of Engineering and Applied Science, University of Cincinnati, Cincinnati, OH
b Medical Scientist Training Program, College of Medicine, University of Cincinnati, Cincinnati, OH
71
5.1. Abstract
Indian hedgehog (Ihh) signaling has recently been shown to play a critical role in the differentiation, growth, and mineralization of fibrocartilaginous tendon entheses. Despite its importance during development, very little is known about the role of hedgehog signaling in mature tendons. Thus, the objective of this study was to examine the patterns of active hedgehog signaling in healthy adult murine hind limb tendons using Scx-GFP;Gli1-LacZ double reporter mice. We found that hedgehog signaling was active in unmineralized entheseal fibrocartilage and also in growth plate cartilage and articular cartilage. More generally, hedgehog activity was consistently localized to the unmineralized side of interfaces between soft and hard tissue, in close association with mineralization fronts. Hedgehog signaling did not decrease with age and remained active in the aforementioned tissues even in 46 week old mice. Interestingly, prominent hedgehog signaling activity was also seen in regions of tendon undergoing fibrocartilaginous metaplasia, a phenomenon which often occurs as an adaptation to compressive load. This observation motivated us to conduct preliminary experiments on cultured tenocytes, which showed that these cells were responsive to Ihh and adopted a more chondrocytic phenotype as a result of Ihh stimulation in vitro. Our results suggest that Scx-positive tenocytes in the body of the tendon have the capability to transdifferentiate into Gli1-expressing fibrochondrocytes via a mechanism that is likely regulated by both mechanical forces and by hedgehog signaling. Taken as a whole, the data presented here indicates that hedgehog signaling is not only important during enthesis development, but also plays a fundamental role in the formation and maintenance of fibrocartilage in tendons throughout life. Ultimately, a better understanding of the mechanobiological factors which regulate hedgehog signaling in tendons could lead to new therapeutic strategies for treating tendon disorders or augmenting tendon-to-bone repair.
72
5.2. Introduction
Tendon to bone attachment sites, or entheses, are specialized tissues which facilitate the transfer of muscular forces to the skeleton. Many entheses, including those of the patellar,
Achilles, and supraspinatus tendons, contain a region of fibrocartilage interposed between the tendon and bone, an important structural adaptation which is critical for proper function27, 109.
This fibrocartilage first appears during early postnatal development and becomes partially mineralized as the tendon matures191. Gradations in structure, composition, and mechanical properties across the enthesis help dissipate forces and limit potentially damaging stress concentrations from forming at the tendon-bone interface3, 79, 80, 110. Unfortunately, injuries involving the tendon-bone junction are particularly difficult to treat because current repair strategies do not regenerate the zonal fibrocartilaginous insertion site. Instead, tendon-to-bone healing occurs via the formation of fibrovascular scar tissue that never attains the gross, histological, or mechanical characteristics of the normal enthesis88, 91, 192.
Recently, Indian hedgehog (Ihh) has been implicated as a key regulator of enthesis fibrocartilage differentiation, growth, and mineralization during tendon development43, 89, 171, 181.
Using Gli1 expression as a known downstream marker of active hedgehog signaling139, 193, studies of both the murine patellar and supraspinatus tendons have revealed a unique population of hedgehog-responsive (Gli1-positive) cells which exist at the immature enthesis in the perinatal period89, 171. These cells first appear at late embryonic stages and go on to populate the fibrocartilage region of the enthesis during postnatal development. Hedgehog signaling activity in this cell population appears to be regulated by the mechanical environment, since botox- induced muscle unloading increases the number of hedgehog-responsive cells and upregulates hedgehog pathway gene transcription at the enthesis171.
73
If the hedgehog-responsive cell population at the enthesis is destroyed or if hedgehog signaling is lost during development, then fibrocartilage differentiation, growth, and mineralization is impaired. Ablation of Gli1-expressing cells or Cre-mediated deletion of
Smoothened (Smo, a critical receptor-associated protein in the hedgehog signaling pathway) in the Scleraxis (Scx, a tendon marker) domain both result in less differentiated fibrochondrocytes, reduced glycosaminoglycan content, and significant reductions in mineralized fibrocartilage area at the enthesis43, 171, 181. Tensile failure testing of patellar tendons from Smo conditional knockout mice at 12 weeks of age also revealed significantly reduced biomechanical properties of the tendon-bone unit and sub-failure mechanical abnormalities at the enthesis, most notably ~2.3 fold greater strain in the insertion region at load levels greater than 20% of normal failure load43,
181. Together, these findings indicate that Ihh signaling plays a critical role in chondrogenic differentiation and postnatal mineralization of the fibrocartilaginous enthesis during tendon development and is necessary for the maturation of a functional tendon-bone attachment site.
Despite the growing evidence that Ihh signaling is critical during fibrocartilaginous enthesis development, very little is known about the role of hedgehog signaling in mature adult tendons. Thus, the primary objective of this study was to examine the spatiotemporal patterns of active hedgehog signaling in healthy adult murine hind limb tendons using Scx-GFP;Gli1-LacZ double reporter mice. Similar to our findings from murine tendon development, we hypothesized that adult murine tendons would contain hedgehog-responsive cells in the enthesis fibrocartilage but no active hedgehog signaling would be observed in the midsubstance. We also hypothesized that hedgehog signaling activity at the enthesis would decrease with age, concomitant with the cessation of tendon growth and the decline in enthesis mineral apposition rate.
74
In addition to investigating the natural patterns of hedgehog signaling in adult tendons, our secondary objective was to determine how activation or inhibition of hedgehog signaling activity affects the phenotype of mature tendon cells. Scx-expressing tendon fibroblasts, or tenocytes, constitute the vast majority of cells in healthy tendon and act as primary mediators of tendon repair following injury47, but their responsiveness to Ihh and the phenotypic changes that occur in these cells as a result of hedgehog pathway activation are unknown. We have optimized protocols for isolating tenocytes from murine tendons and culturing them in vitro, thus providing a well-controlled environment for studying the effects of Ihh stimulation. Based on our in vivo studies of Ihh function during enthesis development, we hypothesized that Ihh stimulation would induce proliferation, chondrogenic transdifferentiation, and mineralization of cultured tenocytes in a dose-dependent manner. Defining the patterns of endogenous Ihh signaling in adult murine tendons and assessing the effects of Ihh stimulation on tenocytes will be critical first steps towards the development of novel therapeutic strategies in which modulation of the Ihh signaling pathway is used as a biologic treatment to augment tendon-to-bone repair.
5.3. Materials and methods
5.3.1. Animal model
Double transgenic Scx-GFP;Gli1+/lacZ mice were used in this study. Male and female animals were used in equal proportion. Scleraxis-GFP (Scx-GFP) reporter mice, which have been described previously194, were generously provided by Dr. Ronen Schweitzer (Oregon
Health & Science University, Portland, OR) and maintained by crossing to wildtype CD1 mice.
Gli1tm2Alj/J (Gli1-LacZ) reporter mice were obtained from Jackson Laboratory (stock number:
008211). Scx-GFP female mice were crossed with Gli1-LacZ male mice to generate double
75 transgenic offspring. Pups were genotyped to distinguish between Scx-GFP;Gli1+/+ and Scx-
GFP;Gli1+/lacZ animals by staining ear punches with X-Gal (described below) to detect the presence of β-galactosidase (the protein product of the LacZ gene) in hair follicles, which are known to contain a large population of Gli1-expressing cells195. All animal procedures were approved by the University of Cincinnati Institutional Animal Care and Use Committee prior to study initiation.
5.3.2. Whole mount X-Gal staining and histology
Following euthanasia via CO2 asphyxiation, murine hind limbs were dissected and disarticulated at the hip. Limbs were fixed in fresh 1% paraformaldehyde for one hour at 4°C, washed three times for 10 minutes each with a detergent rinse solution (0.02% Igepal CA-630,
0.01% sodium deoxycholate, and 2 mM MgCl2 in 100 mM phosphate buffer, pH 7.5), then stained for 24 hours in the dark at 37°C with X-gal staining solution (detergent rinse solution plus 5 mM potassium ferricyanide, 5 mM potassium ferrocyanide, and 1 mg/ml 5-bromo-4- chloro-indolyl-β-D-galactopyranoside [X-Gal, Sigma-Aldrich], pH 7.5). Staining was stopped by washing limbs twice with PBS then post-fixing with 4% paraformaldehyde overnight at 4°C prior to imaging.
After whole-mount imaging, limbs were transferred to 30% sucrose solution overnight and embedded in Tissue-Tek OCT medium (Sakura Finetek). Thin (6-12 um) frozen sections were cut in the sagittal plane on a cryostat using Kawamoto’s cryofilm technique (Section-Lab,
Hiroshima, Japan)196, 197, which has been used previously to cut non-decalcified frozen sections of tendons and their bony insertion sites47, 171. Sections were washed with PBS and mounted with aqueous mounting medium prior to imaging.
76
5.3.3. Tendon cell isolation and culture
Tenocytes were isolated and cultured from the patellar tendon (PT) of 12 week old Scx-
+/lacZ GFP;Gli1 mice. Following euthanasia via CO2 asphyxiation, murine hind limbs were dissected and the PTs were cut away from surrounding structures, cleaned of any excess fat or connective tissue, and rinsed in PBS. PTs from 4-6 animals were pooled together and the tissue was minced into small pieces using a scalpel blade, centrifuged, washed with PBS, and resuspended in 0.25% trypsin for 30 minutes in a shaker bath at 37°C to dissociate cells from the matrix. Samples were then vortexed, centrifuged, washed again with PBS, and resuspended in digestion media consisting of 3 mg/ml collagenase D and 2 mg/ml dispase in cell culture medium (MEM alpha basal media plus 10% fetal bovine serum and 1% antibiotic/antimycotic) for 30-60 minutes in a shaker bath at 37°C. Once the tissue appeared thoroughly disrupted, the cell suspension was centrifuged, washed with PBS, and resuspended in fresh cell culture media.
The tendon-derived cells were then plated in monolayer (passage 0, P0) and cultured under standard conditions (37°C, 95% relative humidity, 5% CO2). Cells were allowed to grow to 80% confluence and then were successively passaged to P3 or P4 in order to obtain sufficient cell numbers for in vitro experiments.
5.3.4. In vitro experiments
To examine the effects of Ihh signaling on tendon-derived cells in vitro, cultured murine tenocytes at P3 or P4 were passaged into 12-well plates at a density of 0.1 x 106 cells/well. Four different treatments, each performed in triplicate, were assessed: (1) Control group: cells were incubated in standard cell culture media (MEM alpha basal media plus 10% fetal bovine serum and 1% antibiotic/antimycotic). (2) Cyclopamine group: media was supplemented with 5 uM
77 cyclopamine (Stemcell Technologies), an inhibitor of hedgehog signaling198. (3) Low Ihh group: media was supplemented with 20 ng/ml recombinant Ihh (R&D Systems, catalog number: 1705-
HH-025/CF). (4) High Ihh group: media was supplemented with 2 ug/ml recombinant Ihh.
Treated cells were incubated for 3 days, rinsed once with PBS, and fixed with a solution of 4% formaldehyde and 0.5% glutaraldehyde for 5 minutes on ice. The cells were then washed three times with PBS and stained overnight at 37°C with X-Gal staining solution (2 mM MgCl2, 5 mM potassium ferricyanide, 5 mM potassium ferrocyanide, and 1 mg/ml X-Gal in PBS, pH 7.5).
After aspirating the X-Gal staining solution, cells were washed once with PBS and counterstained with DAPI.
5.3.5. Imaging
Brightfield, phase contrast, and fluorescence imaging of frozen tissue sections and fixed cells was performed on an Olympus IX81 inverted microscope equipped with appropriate fluorescent filters and a color CCD camera. Images were acquired using the same settings for all samples within each experimental group and subsequently refined using Fiji image analysis software199.
5.4. Results
5.4.1. Hedgehog signaling is active in the mature PT enthesis
Although recent studies have examined hedgehog signaling during enthesis development, the patterns of active hedgehog signaling in mature tendons have not been characterized. Here we used adult Scx-GFP;Gli1-LacZ reporter mice to examine which cells and tissues in and around the PT were actively responding to hedgehog. In 12 week old mice, hedgehog signaling
78 was readily detected at the distal PT enthesis but was not present in the tendon midsubstance
(Fig. 5.1A,C,E). Hedgehog signaling was also active at the proximal insertion of the PT into the patella (Fig. 5.1A,B,D). Other locations where hedgehog signaling was detected include: (1) the proximal tibial growth plate (Fig. 5.1A,C, white arrowheads); (2) throughout the articular cartilage in the knee joint (Fig. 5.1E, yellow arrowheads); and (3) in blood vessels (for example, in small vessels in the paratenon, Fig. 5.1C, red arrowhead).
Figure 5.1: Patterns of active hedgehog signaling in the PT of mature (12 week old) Gli1-LacZ reporter mice. Blue stain indicates cells expressing Gli1, a known marker of hedgehog pathway activation. (A-C) Whole mount staining revealed active hedgehog signaling at both the proximal and distal insertions of the PT (black arrowheads) but not in the tendon midsubstance. Positive staining was also seen in the proximal tibial growth plate (white arrowheads) and in blood vessels (red arrowhead). (D-E) Histology confirmed the presence of active hedgehog signaling in cells at the proximal insertion of the PT into the patella (D) and the distal insertion of the PT into the tibia (E). Positive staining was also detected in articular cartilage (yellow arrowheads).
Looking more closely at the distal PT enthesis in 12 week old mice, we found that hedgehog signaling was active in the unmineralized entheseal fibrocartilage but was completely absent in the mineralized fibrocartilage and bone (Fig. 5.2). The cells responding to hedgehog
79 had a rounded morphology and often appeared in stacked columns at the enthesis, which is typical of clonal populations of chondrocytes. Similarly, hedgehog signaling was active in both growth plate chondrocytes and articular chondrocytes but was absent in the adjacent mineralized bone (Fig. 5.2). In general, hedgehog activity was consistently localized to the unmineralized side of interfaces between soft and hard tissue and closely followed the contours of the mineralization front.
Figure 5.2: Hedgehog signaling is localized to the unmineralized fibrocartilage of the adult murine PT enthesis. (A) Active hedgehog signaling was detected in three distinct populations of chondrocytes: unmineralized entheseal fibrochondrocytes, growth plate chondrocytes, and articular chondrocytes. (B) In the enthesis, hedgehog signaling was only active in unmineralized tissue, not in adjacent mineralized fibrocartilage or bone. Chondrocytic cells responding to hedgehog had a rounded morphology and often appeared in stacked columns. UFC, unmineralized fibrocartilage; MFC, mineralized fibrocartilage.
80
5.4.2. Hedgehog signaling activity at the PT enthesis does not decrease with age
In order to investigate the effects of aging on hedgehog signaling activity at the murine
PT enthesis, we examined different aged mice (12, 22, and 46 weeks old). Since many developmental signaling pathways become downregulated or disappear altogether in mature tissues, we hypothesized that hedgehog signaling activity at the enthesis would decrease with age. Contrary to our hypothesis, we found that hedgehog signaling remained active at all time points, with no appreciable differences in staining intensity or the spatial distribution of hedgehog-responsive cells over time (Fig. 5.3). Even in 46 week old mice, hedgehog signaling was still active at the proximal tibial growth plate, in articular chondrocytes, at the proximal insertion of the PT into the patella, and within the unmineralized entheseal fibrocartilage at the distal insertion of the PT into the tibia (Fig. 5.3).
81
Figure 5.3: Hedgehog signaling activity at the PT enthesis does not change with age. Whole mount staining (A-C) revealed active hedgehog signaling in the proximal tibial growth plate (white arrowhead) and at the proximal and distal insertions of the PT (black arrowheads) at all three time points examined (12, 22, and 46 weeks). Histology of the proximal PT insertion into the patella (D-F) and the distal PT insertion into the tibia (G-I) revealed no appreciable differences in the spatial distribution of hedgehog-responsive cells over time. Active hedgehog signaling was also detected in the articular cartilage (yellow arrowhead in G) at all time points.
82
5.4.3. Hedgehog-responsive fibrocartilage cells at the PT enthesis are Scx-negative
In order to better characterize the hedgehog-responsive cells in the adult murine PT enthesis, we examined whether or not these Gli1-expressing cells co-expressed the tenocyte marker Scx. By staining for the presence of the Gli1-LacZ reporter and then visualizing Scx-GFP expression in the same tissue sections using fluorescence microscopy, we found that the unmineralized fibrocartilage region of the enthesis was composed almost entirely of Gli1- positive, Scx-negative cells (Fig. 5.4). These cells had a rounded morphology typical of chondrocytes. In contrast, the tendon midsubstance was composed of Scx-positive cells, which normally have a spindle-shaped morphology (Fig. 5.4, white arrowheads). However, in response to compressive loads, such as those produced when the posterior aspect of the PT presses against the tibial articular cartilage, tenocytes can begin to undergo fibrocartilaginous metaplasia27, 200,
201, thereby becoming more chondrocyte-like in appearance while still retaining Scx expression
(Fig. 5.4, black arrowheads).
Figure 5.4: Hedgehog-responsive fibrocartilage cells at the PT enthesis are Scx-negative. X-Gal staining to detect the Gli1-LacZ reporter (A) and fluorescence imaging of Scx-GFP expression (B,C) in the same tissue section revealed a Gli1-positive, Scx-negative cell population in the unmineralized fibrocartilage (outlined in yellow) of the 12 week old murine PT enthesis. In contrast, the tendon midsubstance contained spindle-shaped, Scx-positive tenocytes (C, white arrowheads), which are capable of undergoing fibrocartilaginous metaplasia in response to compressive loading (C, black arrowheads).
83
5.4.4. Hedgehog signaling is active in regions of fibrocartilaginous metaplasia
Anatomical studies have shown that fibrocartilage is often found in regions of tendon where compressive stress is present, including the tendon enthesis27, 183, 200, 201. In this study, we found that hedgehog signaling is active in tendon regions undergoing fibrocartilaginous metaplasia in response to natural compressive loading. For example, active hedgehog signaling was observed grossly in tendons that wrapped around bony pulleys, as in the case of the peroneus brevis tendon, which wraps around the lateral malleolus in the mouse hind foot (Fig.
5.5A). At the cellular level, active hedgehog signaling was seen in regions of tendon that came into contact with cartilage or bone, such as the posterior aspect of the distal PT, which presses against the tibial articular cartilage when the knee is flexed (Fig. 5.5B).
Figure 5.5: Hedgehog signaling is active in regions of tendon subjected to compressive loads. (A) Whole mount X-Gal stain showing active hedgehog signaling in the dissected murine peroneus brevis tendon, which wraps around the lateral malleolus in vivo. The region of contact between the tendon and the bone is denoted by the black arrowhead. (B) Hedgehog signaling is active in cells in the posterior aspect of the distal PT (black arrowheads), which are compressed against the tibial articular cartilage (yellow arrowheads) during knee flexion. UFC, unmineralized fibrocartilage; MFC, mineralized fibrocartilage.
84
Interestingly, foci of active hedgehog signaling were also observed in the Achilles tendon midsubstance of older mice (Fig. 5.6). These well-circumscribed regions of positive X-Gal staining were not present in 12 week old mice, but were consistently observed in 22 week old animals and had expanded in both size and number by 46 weeks, ultimately involving much of the distal half of the tendon. In 12 week old murine Achilles tendons, active hedgehog signaling was confined to the calcaneal insertion (Fig. 5.6A and B, white arrowheads) and was not detected in the tendon midsubstance, similar to the pattern seen in the PT. However, in every 22 week old animal examined in this study, at least one foci of active hedgehog signaling was identified in the tendon midsubstance, usually on the anterior aspect of the distal half of the tendon, separate and distinct from the hedgehog-responsive cell population at the enthesis (Fig.
5.6D, black arrowhead). In 46 week old animals, positive X-Gal staining was observed in multiple large foci throughout the distal half of the tendon (Fig. 5.6E and F, black arrowheads).
Direct palpation revealed that these positively-staining regions felt harder and less compliant than adjacent unstained tendon tissue, suggesting that active hedgehog signaling was correlated with age-related changes in the tendon matrix, most likely due to fibrocartilaginous metaplasia
(and possibly ectopic ossification, although the presence of mineral was not assayed in this study). Of note, X-Gal staining was most intense at the leading edges of these foci of hedgehog activity (Fig. 5.6G, yellow arrowheads).
85
Figure 5.6: Age-related changes in hedgehog signaling in the murine Achilles tendon. Posterior (A,C,E), lateral (B,D,F), and anterior (G) views of whole mount X-Gal stained Achilles tendons revealed the appearance and expansion of foci of active hedgehog signaling over time. In 12 week old mice (A,B), hedgehog activity was confined to the calcaneal insertion (white arrowheads). By 22 weeks (C,D), discrete foci of positive X-Gal staining had appeared in the midsubstance (black arrowheads). At 46 weeks (E-G), extensive hedgehog signaling was present in multiple foci in the distal half of the tendon, with the most intense staining occurring at the leading edges of these regions (yellow arrowheads).
5.4.5. Changes in hedgehog signaling activity affect cultured tenocyte phenotype
Given our results showing that hedgehog signaling is active in regions of tendon undergoing fibrocartilaginous metaplasia, we next wanted to determine how activation or inhibition of the hedgehog pathway affects the phenotype of mature tendon cells. To do this, we harvested cells from the PT of 12 week old Scx-GFP;Gli1-LacZ mice and cultured them in vitro
86 in order to provide a well-controlled environment for studying the effects of hedgehog pathway modulation. We found that in normal monolayer culture, murine tendon-derived cells adopted a morphology that was intermediate between tenocytic (elongated, spindle-like) and chondrocytic
(rounded, cobblestone-like), stained weakly positive with X-Gal (indicating some low level of baseline hedgehog signaling activity), and retained Scx-GFP expression, albeit at a low level
(Fig. 5.7A-E). Interestingly, after incubation for 3 days in media containing the hedgehog inhibitor cyclopamine, these cells underwent a radical change in morphology, becoming much more elongated with long cellular processes, concomitant with increased Scx-GFP expression
(Fig. 5.7F-J). These changes are suggestive of a more tenocytic phenotype. Conversely, when incubated in media supplemented with either low or high concentrations of recombinant Ihh, the cells stained strongly positive with X-Gal and grew in a more compacted, cobblestone-like manner with minimal Scx-GFP expression (Fig. 5.7K-T), suggesting a more chondrocytic phenotype. No differences were observed between the low Ihh and high Ihh treatment groups.
Figure 5.7: Changes in hedgehog signaling activity affect cultured tenocyte phenotype. Compared to normal monolayer cultures of murine patellar tendon-derived cells (A-E), cyclopamine-treated cells (F-J) exhibited a more tenocytic morphology and increased Scx-GFP expression. In contrast, cells treated with low (K-O) or high (P-T) concentrations of recombinant Ihh displayed stronger X-Gal staining, a more chondrocytic morphology, and minimal Scx-GFP expression. Phase-contrast, brightfield, and fluorescence images were acquired using the same settings across all treatment groups to allow for fair comparisons.
87
88
5.5. Discussion
Musculoskeletal development is a complex process that involves the differentiation and coordinated patterning of multiple different tissue types, including muscle, tendon, cartilage, and bone. Numerous chemical and mechanical signals must be precisely integrated and regulated in space and time in order to produce mature structures with normal form and function. Indian hedgehog is one of the primary developmental morphogens involved in musculoskeletal tissue patterning due to its central role in the regulation of endochondral bone formation in the developing skeleton141, 202. Recent studies have also shown that hedgehog signaling is a key mechanism in the development of fibrocartilaginous tendon entheses43, 89, 171, 181. Ihh positively regulates the differentiation, growth, and mineralization of entheseal fibrochondrocytes, and loss of this signal results in a morphologically and functionally abnormal tendon-to-bone insertion.
Despite its importance during tendon development, the potential roles of hedgehog signaling in tendon homeostasis, in normal tendon aging, in healing following acute tendon injury, and in the etiology and pathogenesis of tendinopathies are all currently unknown.
In order to begin addressing these questions, our primary objective in this study was to examine the spatiotemporal patterns of active hedgehog signaling in healthy adult murine hind limb tendons. Lineage tracing of hedgehog-responsive (Gli1-positive) cells in the developing murine patellar and supraspinatus tendons has shown that these cells populate the fibrocartilage region of the enthesis and become restricted to the unmineralized fibrocartilage zone following the postnatal onset of mineralization171, 203. Thus, we hypothesized that adult murine tendons would contain hedgehog-responsive cells in the unmineralized entheseal fibrocartilage but no active hedgehog signaling would be observed in the midsubstance. We also hypothesized that hedgehog signaling activity at the enthesis would decrease with age, concomitant with the
89 cessation of tendon growth and the decline in enthesis mineral apposition rate. This latter hypothesis was based on the notion that hedgehog signaling is a critical regulator of embryonic cell fate during embryogenesis/morphogenesis but, in general, seems to be a relatively silent pathway during adult life204. However, a growing body of research has shown that hedgehog signaling is actually critical for maintaining homeostasis in many different mature tissues205 and disruptions to the hedgehog pathway contribute to the etiopathogenesis of many age related- diseases206.
We found that active hedgehog signaling was indeed present in unmineralized enthesis fibrocartilage, but also in growth plate chondrocytes and articular chondrocytes. In general, active hedgehog signaling was distinctly localized to chondrocytic cells on the unmineralized side of interfaces between soft and hard tissue. Of note, hedgehog signaling activity was not detected in the meniscal fibrocartilage, consistent with the findings of Pazin et al207, who showed that Gli1 expression is downregulated in the developing meniscus compared to articular and growth plate cartilage. Thus, while Gli1 may be a useful marker for some cartilages including the unmineralized entheseal fibrocartilage, it is not a ubiquitous chondrocyte marker. The observation that hedgehog signaling is active in mature chondrocytes and appears in close association with the mineralization tidemark in tendon entheses (but is generally absent from the tendon midsubstance), supports the paradigm that hedgehog signaling is intimately involved in both cartilage formation and the positive regulation of mineralization. Indeed, genetic studies of the function of Indian hedgehog signaling in the developing endochondral skeleton have revealed a direct role in the regulation of chondrocyte proliferation150 and in the differentiation of osteoblasts153. Similarly, gain- and loss-of-function studies during patellar and supraspinatus tendon development have revealed that active hedgehog signaling promotes chondrogenesis,
90 increases expression of cartilage markers (e.g. collagen type 2, biglycan, tenascin C), and positively regulates mineralization at the enthesis43, 171, 181, 203.
Many of Ihh’s functions are mediated via interactions with other genes and signaling pathways. For example, Ihh and Runx2 (Runt-related transcription factor 2) interact to coordinate early chondrogenesis208, chondrocyte maturation209, 210, and osteoblast differentiation154, 211. Furthermore, Ihh and BMPs (bone morphogenetic proteins) are co- expressed at many sites in the mouse212, 213 and BMP has been shown to induce the expression of
Ihh214-216. The crosstalk between these two signaling pathways is critical for integrating chondrocyte proliferation and differentiation during development and morphogenesis217-219.
Importantly, Ihh has been shown to be a chondro-osteogenic inducer of mesenchymal stem cells, an effect that is synergized by BMP157, 220, 221. This has implications for both cartilage and bone tissue engineering. In fact, Ihh has already shown promise as a potential therapeutic for the repair of articular cartilage222 and calvarial defects158. Understanding the complex interplay between hedgehog signaling and other pathways will likely be critical for developing effective Ihh-based therapeutics for tendon enthesis repair as well.
In this study, we showed that hedgehog signaling was active at the enthesis in 12, 22, and
46 week old mice, suggesting that this pathway is not downregulated in entheseal fibrocartilage during natural aging. This is interesting, since longitudinal bone/tendon growth and enthesis mineral apposition in the mouse decline to very low or undetectable levels well before 46 weeks.
Although murine growth plates do not fuse after sexual maturity as they do in humans, longitudinal growth slows dramatically at puberty, which occurs at 6-8 weeks. In outbred CD1 mice, femoral length does not significantly change between 8 and 18 months of age in males or females223. Similarly, in C57BL/6 mice, there was no change between 6 and 12 months of age in
91 either sex224, despite the fact that the growth plates had not closed. In addition, mineral apposition in the murine Achilles and PT entheses has been shown to decrease dramatically following puberty203. Thus, the fact that hedgehog signaling remains active even in 46 week old mice suggests that this pathway is not only active during development and growth, but plays a role in cartilage homeostasis throughout life. It remains to be seen whether these findings are unique to the murine model or also hold true in larger animals, including humans.
Given Ihh’s roles in chondrogenesis and mineralization, the fact that hedgehog signaling remains active at the enthesis even in mature tendons raises an interesting question: Why doesn’t the enthesis fibrocartilage continue to expand and mineralize throughout life? During development, enthesis fibrocartilage grows and appositional mineralization occurs directionally towards the tendon body191, 203. Ihh signaling is a positive regulator of this process, so if Ihh signaling remains active in mature tendons, why does fibrocartilage growth and mineralization dramatically decline with age? What prevents the mineralization front from steadily moving into the body of the tendon during normal aging? One possible explanation is that there are other developmental signals in addition to Ihh that are necessary for enthesis formation (such as BMP signaling40, 42), and age-related changes in these complementary signaling pathways may dictate the temporal decline in fibrocartilage growth and mineralization. Another possibility is that inhibitory signals, produced either in the maturing tendon itself or from neighboring tissues such as muscle, limit Ihh-driven fibrocartilage expansion in adult animals. Of note, these signals could be chemical or mechanical, since tendons are subjected to dynamic loading during activities of daily living.
Another intriguing result from this study was the observation that Gli1-expressing cells are present in regions of tendon where compressive loading is occurring, thus implicating
92 hedgehog signaling in the phenomenon of compression-induced fibrocartilaginous metaplasia. It is well established that the formation of fibrocartilage in tendons can occur as a natural adaptation to compressive loads27, 201, 225-227. This often occurs in sites where the tendon surface comes into repeated contact with nearby bone, such as when a tendon wraps around a bony pulley201, 228 or in the sesamoid fibrocartilages of enthesis organs82 (the prototypical enthesis organ being the insertion of the Achilles tendon into the calcaneus). Despite its prevalence, the molecular mechanisms governing fibrocartilaginous metaplasia in tendons have not been fully elucidated. Although only a correlative relationship was established in this study, our data suggest that hedgehog signaling is activated by compressive forces in tendons. In support of this theory, Ihh expression has been shown to be highly mechanosensitive, becoming upregulated in response to compression both in vivo and in vitro166, 177-180, 229-232. Thus, Ihh signaling may be a key biological intermediary that links compressive loading to the development of fibrocartilage in tendons. Further mechanistic studies which test this hypothesis directly should reveal whether compressive forces and hedgehog signaling are necessary and/or sufficient to induce fibrocartilaginous metaplasia in vivo.
In addition to being a normal adaptation to compressive loads, it has also been suggested that fibrocartilaginous metaplasia in tendons can be pathologic233. Rat supraspinatus tendon expresses cartilage markers with overuse234, and the appearance of fibrocartilage in the rotator cuff has been proposed as a pathogenic mechanism of tendon tears235. In this study, we observed the age-related formation and expansion of new foci of active hedgehog signaling in the midsubstance of the murine Achilles tendon (but not the patellar tendon). These foci developed in regions of the tendon in which no obvious source of compressive stress could be identified.
Furthermore, X-Gal staining was most intense at the leading edges of these foci of hedgehog
93 activity, which could imply a direct role for hedgehog signaling in mediating the transdifferentiation of tenocytes into fibrochondrocytes (or also in regulating the process of ectopic ossification in tendons if these regions become mineralized, a question which was not examined in this study but warrants further investigation). It is unclear whether this phenomenon represents a normal age-related change or is due to some type of pathology, such as degenerative tendinopathy. Additional studies making use of human tissues from patients with tendinosis and animal models of tendinopathy are warranted in order to discern the role of hedgehog signaling, if any, in the etiology and pathogenesis of age-related tendon degeneration.
The origin of the hedgehog-responsive fibrochondrocyte-like cells that appear in the tendon midsubstance as a result of compressive forces or age-related tendon pathology is a subject of debate. The healthy tendon midsubstance is composed almost exclusively of spindle shaped, Scx-positive tenocytes, but a seminal study by Bi et al236 showed that both human and mouse tendons also harbored a unique stem/progenitor cell population that had universal stem cell characteristics such as clonogenicity, multipotency and self-renewal capacity. Thus, the appearance of rounded, Gli1-positive fibrochondrocyte-like cells in the midsubstance could either be due to the expansion and chondrogenic differentiation of resident stem/progenitor cells or to the direct transdifferentiation of tenocytes into fibrochondrocytes. Our data seems to support the latter, because we visualized numerous Scx-positive, weakly Gli1-LacZ-positive cells with a rounded, chondrocytic morphology in regions of tendon known to experience compressive loading (for example, on the posterior surface of the distal PT, which presses against the tibial articular cartilage during knee flexion, see Fig. 5.4C, black arrowheads). Our preliminary in vitro experiments on Scx-positive tenocytes cultured in monolayer also revealed that upregulation of hedgehog signaling via stimulation with recombinant Ihh induced
94 chondrocytic changes in cell morphology, while inhibition of the hedgehog pathway with cyclopamine appeared to have a pro-tenogenic effect. Together, these findings imply that tenocytes can transdifferentiate into fibrochondrocytes in reaction to certain environmental stimuli, and hedgehog signaling likely plays a role in mediating this transition. However, additional in vitro studies are needed in order to more fully characterize the effects of hedgehog pathway activation on proliferation, chondrogenic gene expression, and mineralization of tenocytes in both 2D and 3D culture.
In this primarily descriptive study, we did not experimentally manipulate mechanical forces or hedgehog pathway activity in vivo, nor did we examine the effects of compressive mechanical stimulation of tenocytes in culture. Furthermore, we did not attempt to explicitly quantify the relative intensity of hedgehog signaling between different tissues or time points.
Nonetheless, our use of the Gli1-LacZ reporter was an effective approach for investigating the general spatiotemporal patterns of active hedgehog signaling within murine tendons. Mechanistic experiments that attempt to answer some of the questions raised here are certainly warranted in future studies and will be critical for developing a more complete understanding of the molecular mechanisms by which Ihh mediates its effects during tendon development, homeostasis, and disease. In addition, other signaling pathways known to be involved in enthesis formation, such as BMP signaling, should be investigated along with hedgehog, since interactions between these pathways are likely. From a translational perspective, more work is needed to understand the effects of hedgehog pathway activation and inhibition on tendon cell phenotype, both in vitro and in vivo. In the future, modulation of hedgehog signaling could have impactful consequences in the areas of tendon tissue engineering, tendon-to-bone repair, and the design of biologic therapies for combatting tendinopathy and age-related tendon degeneration.
95
5.6. Acknowledgements
We gratefully acknowledge the National Institutes of Health for providing research support (R01 AR056943) as well as student funding via the University of Cincinnati MSTP training grant (T32 GM063483). Student tuition and stipend support was also provided by the
University of Cincinnati Graduate School Dean’s Fellowship (to S.D. Gilday). We thank Dr.
Andrew Breidenbach, Dr. Andrea Lalley, Dr. Nathaniel Dyment, and Dr. Han Liu for technical assistance with histology and staining procedures. We also acknowledge Dr. Daria Narmoneva for providing microscopy and imaging support and Cindi Gooch for helping with animal husbandry, breeding, and genotyping.
96
CHAPTER 6
Delivery of Recombinant Indian Hedgehog Protein to the Healing Patellar Tendon Enthesis Does Not Improve Functional Repair Outcomes in a Murine Model
Steven D. Gilday a,b, E. Chris Casstevens c, Heather M. Powell d, Keith Kenter c, David L. Butler a, Jason T. Shearn a
a Biomedical Engineering Program, College of Engineering and Applied Science, University of Cincinnati, Cincinnati, OH
b Medical Scientist Training Program, College of Medicine, University of Cincinnati, Cincinnati, OH
c Department of Orthopaedic Surgery, College of Medicine, University of Cincinnati, Cincinnati, OH
c Department of Materials Science and Engineering, College of Engineering, The Ohio State University, Columbus, OH
97
6.1. Abstract
Reattaching ruptured tendons to their bony insertions is a fundamental challenge in orthopaedics. Once disrupted, the fibrocartilaginous enthesis does not adequately regenerate, leaving the tendon susceptible to further injury. Recently, the Indian hedgehog (Ihh) signaling pathway has emerged as an attractive therapeutic target for tendon-to-bone repair because of its critical role in enthesis development. Thus, the objective of this study was to test whether delivery of recombinant Ihh protein could improve tendon-to-bone healing outcomes in a murine patellar tendon (PT) injury model. To control spatial and temporal delivery of Ihh in vivo, we designed and fabricated electrospun polycaprolactone (PCL) scaffolds and infused them with Ihh using subcritical CO2. In vitro testing confirmed that the scaffolds were capable of releasing bioactive Ihh for at least 3 weeks. To determine the effects of Ihh on tendon-to-bone healing, an acute surgical injury was created in the murine PT enthesis and subsequently treated with either an Ihh-infused scaffold, blank scaffold, bolus injection of Ihh, or bolus injection of saline.
Biomechanical and histological healing outcomes were assessed at 5 weeks post-surgery. We found that tendons treated with an Ihh-infused scaffold were more likely to spontaneously rupture during the healing period. Tensile testing of non-ruptured tendons revealed no significant differences in structural properties, regional strain patterns, or failure location between treatment groups, but the Ihh-infused scaffold group did exhibit a trend towards decreased material properties. Although histologic outcomes were highly variable between specimens, the repair tissue in tendons treated with Ihh-infused scaffolds had high cellularity but contained less extracellular matrix, indicating an attenuation of the normal fibrotic healing response. Based on these results, we conclude that the Ihh-infused scaffolds did not provide a functional benefit and may actually have been deleterious to tendon-to-bone healing in our murine PT injury model.
98
6.2. Introduction
Tendon and ligament injuries represent nearly half of all musculoskeletal injuries in the
US and account for millions of treatment episodes per year237. Rotator cuff tears, flexor tendon ruptures, and traumatic injuries to the Achilles and patellar tendons are common yet remain difficult to surgically repair. Whether using a tendon graft or performing a direct repair, reattaching tendon tissue to bone has proven to be a challenge for orthopaedic surgeons. The limited intrinsic healing capacity of mature tendons coupled with the drastic difference in material properties between tendon and bone impedes the regeneration of a normal, zonal attachment between these two tissues. The distinct fibrocartilage transition region which is normally present at the tendon-bone interface does not re-form following injury. Instead, deposition of fibrous scar tissue results in a morphologically abnormal and biomechanically inferior junction88, 91, 192. Despite advances in surgical technique and rehabilitation protocols, re- rupture of repaired tendons remains a serious complication for many patients. Thus, there exists a critical need for innovative strategies to improve tendon-to-bone healing.
Functional tissue engineering (FTE) has been proposed as a promising alternative to traditional tendon repair strategies12, 14, 22. Tissue engineers seek to use some combination of cells, scaffolds, biochemical signals, and mechanical stimulation to create replacement tissues in vitro or to augment repair of damaged tissues in vivo. The FTE paradigm places a special emphasis on restoring not only native structure, but also normal mechanical function.
Unfortunately, the field of tendon tissue engineering has been limited by a lack of understanding of basic tendon biology and consequently a dearth of validated biological targets to modulate during healing. In our laboratory, we believe that strategically reintroducing and/or manipulating the critical signals which govern normal tendon development is an innovative approach that
99 could be used to improve functional outcomes of tendon healing. This type of biologic augmentation could also be combined with traditional surgical repair, thus allowing orthopaedic surgeons to treat tendon injuries with a more effective, multi-faceted approach.
Recently, the Indian hedgehog (Ihh) signaling pathway has emerged as an attractive therapeutic target for tendon-to-bone repair because of its newly discovered role in enthesis development. Gain- and loss-of function studies have implicated hedgehog signaling as a key regulator of enthesis fibrocartilage differentiation, growth, and mineralization43, 171, 181. Forced activation of the hedgehog signaling pathway in Scx-positive tenocytes caused proteins usually confined to the fibrocartilaginous insertion site (e.g. tenascin-C, biglycan, and collagen type 2) to be ectopically expressed throughout the tendon midsubstance43. Similarly, tendon explants maintained in an ex vivo organ culture system and stimulated with recombinant Ihh protein (500 ng/ml) for three days demonstrated ectopic upregulation of insertion site genes and increased expression of insertion site proteins throughout the tendon43. Conversely, ablation of the hedgehog-responsive cell population or Cre-mediated deletion of Smo in Scx-expressing cells both result in less differentiated fibrochondrocytes, reduced glycosaminoglycan content, and significant reductions in mineralized fibrocartilage area at the enthesis43, 171, 181. Analysis of the patellar tendons from Smo conditional knockout mice at 12 weeks of age revealed a 32% reduction in mineralized fibrocartilage area and ~2.3 fold greater mechanical strains in the insertion region181. Together, these findings indicate that Ihh signaling plays a critical role in chondrogenic differentiation and postnatal mineralization of the fibrocartilaginous enthesis and is necessary for the maturation of a functional tendon-bone attachment site.
Given its ability to stimulate chondrocyte proliferation, differentiation, and mineralization, Ihh is intriguing as a potential therapeutic molecule for augmenting fibrocartilage
100 healing during adult tendon-to-bone repair. Thus, the objective of this study was to test whether delivery of recombinant Ihh protein to the healing enthesis via custom-designed polymeric drug delivery scaffolds could improve functional tendon-to-bone repair outcomes in a murine PT injury model. We hypothesized that Ihh delivery to the healing insertion site would stimulate the proliferation, differentiation, and mineralization of neo-fibrocartilage tissue, thereby resulting in restoration of a zonal fibrocartilaginous enthesis with improved biomechanical properties compared to controls.
6.3. Materials and methods
6.3.1. Scaffold design and fabrication
The drug delivery scaffolds used in this study were custom-designed for translational tendon/ligament repair applications and were fabricated in the laboratory of one of the study authors (H. Powell, The Ohio State University). At the outset of the project, we established a set of design goals with input from an interdisciplinary team of experts, including bioengineers, biologists, materials scientists, orthopaedic surgeons, and industry representatives. Namely, we wanted the scaffolds to have the following properties:
1. Strong and stiff enough to permit handling and implantation by surgeons, but
compliant enough to be molded to fit the natural anatomic contours of the injury site.
2. Biodegradable, to eliminate the need for retrieval surgery.
3. Non-cytotoxic to tendon and surrounding tissues.
4. Non-immunogenic, to prevent an immune-mediated foreign body response or fibrotic
encapsulation of the scaffold.
101
5. Non-adherent to the overlying skin or the underlying tendon tissue, to ensure the
preservation of full range of motion throughout the healing process.
6. Able to be infused with controlled amounts of drugs or growth factors.
7. Able to be sterilized prior to implantation, to prevent infection.
8. Able to deliver drugs/growth factors locally and in a unidirectional manner (i.e.
toward the wound site), to prevent widespread diffusion and off-target effects.
9. Able to deliver drugs/growth factors at a predictable, relatively steady rate for
sustained time periods (>2 weeks), to maximize their biologic effect.
10. Composed of readily available, cost-effective, and FDA approved materials, to
minimize fabrication costs and maximize commercialization potential.
With these criteria in mind, we began by fabricating and testing different electrospun polymeric scaffolds in vitro. Electrospinning has been used extensively to create functional nanofiber scaffolds for biomedical applications, including drug/growth factor delivery238, 239. The inherent high porosity of electrospun scaffolds allows for a more precisely controlled degradation profile which can be tuned by altering polymer composition and fiber morphology, thereby leading to sustained drug release. We chose to evaluate three different scaffold compositions: pure poly--caprolactone (PCL), PCL plus collagen (9:1 blend), and PCL plus poly lactic-co-glycolic acid (PLGA) (4:1 blend). Owing to its biocompatibility and biodegradability, PCL is an ideal drug carrier for sustainable, targeted delivery and has been studied extensively for both controlled drug release and tissue engineering applications in vivo240, 241. Adding collagen or PLGA to the PCL can predictably alter the mechanical properties and degradation rate, which can be beneficial for achieving the desired characteristics for a given application242-245. Electrospinning was performed according to published methods246 and the
102 resulting nanofiber scaffolds were systematically characterized to determine their mechanical properties, degradation profile, fiber morphology, and biocompatibility (see flowchart in Fig.
6.1). Due to its high mechanical strength, slow degradation, excellent biocompatibility, and low cost, we selected the pure PCL material for our final scaffold design. To facilitate handling and add structural support to the electrospun portion of the scaffold, a high molecular weight PCL film was used as a backer. This biphasic design allows the electrospun surface of the scaffold to be in close proximity to the healing tendon-bone insertion site while the dense PCL film is oriented towards the overlying skin, thereby acting as a diffusion barrier and promoting unidirectional drug release toward the site of injury.
Our final scaffold design consisted of a porous body of electrospun PCL nanofibers (200-
400 um thick) firmly bonded to a dense PCL film backer (~50 um thick) (Fig. 6.2A,B). The scaffolds were initially fabricated as large sheets (75 mm x 25 mm). To impregnate the scaffolds with drug, droplets containing 2 ug of recombinant Ihh protein (R&D Systems, catalog number:
1705-HH-025/CF) and rhodamine dye were deposited onto the electrospun sheets using ink jet printing technology. To ensure long lasting release, the Ihh droplets were then infused into the electrospun PCL fibers using subcritical CO2 (900 psi for 60 min), a technique which has been described in detail elsewhere247. Due to the presence of the rhodamine dye, the location of the droplets could be easily visualized both before and after the infusion (Fig. 6.2C,D). Prior to in vitro testing or in vivo implantation, a 4 mm diameter sterile biopsy punch was used to cut the
Ihh-containing regions out of the larger scaffold sheet. Of note, we also experimented with larger size scaffolds (5 mm and 8 mm diameter, Fig. 6.2E), but quickly realized that they were more difficult to surgically implant and were simply too large to be used in the murine model.
103
Figure 6.1: Flowchart for selecting final scaffold material and design. Three different materials (pure PCL, PCL + collagen, and PCL + PLGA) were used to create electrospun scaffolds, which were then characterized in vitro in order to select the most appropriate material. In our final design, the electrospun portion of the scaffold was combined with a PCL film backer.
104
Figure 6.2: Final Ihh-infused scaffold design. (A,B) SEM imaging of a cross section (A) and top view (B) of the final scaffold prototype shows the differences in structure between the dense PCL backer and the electrospun PCL body. The electrospun PCL fibers form an interconnected mesh that is highly porous (B, inset), while the backer acts as a diffusion barrier. (C-E) Droplets containing 2 ug of recombinant Ihh protein and rhodamine dye were deposited on the surface of the scaffold using ink jet printing technology (C) and then infused into the PCL fibers using subcritical CO2 (D). The Ihh-containing regions of the scaffold were then cut to size using biopsy punches of varying diameter (E). Pilot studies revealed that the 4 mm diameter scaffolds were the most appropriate size for use in our murine PT injury model.
6.3.2. In vitro studies
6.3.2.1. Ihh release kinetics
Drug release kinetics for the Ihh-infused scaffolds were determined in vitro in the absence of cells. Scaffolds (n=4) were placed in a 12-well plate and incubated in 1 mL PBS per well at 37°C. The releasate was collected from each well at pre-determined time points (from 0
105 to 21 days) and replaced with 1 ml of fresh PBS. The concentration of Ihh protein in the releasate at each time point was quantified using a commercially available murine Indian hedgehog
ELISA kit (ABIN426300, Antibodies-Online).
6.3.2.2. Verification of bioactivity of released Ihh
In order to verify that the Ihh released from the scaffolds was bioactive and had not been degraded during the infusion process, we tested the ability of Ihh-infused scaffolds to activate the hedgehog signaling pathway in cultured cells. Tenocytes were isolated and cultured from the patellar tendons of 12 week old Scx-GFP;Gli1lacZ/+ mice as previously described (see chapter
5.3.3). The cells were then incubated under standard conditions (37°C, 95% relative humidity,
5% CO2) in normal cell culture media (MEM alpha basal media plus 10% fetal bovine serum and
1% antibiotic/antimycotic) in the presence of either an Ihh-infused scaffold or a blank scaffold.
As a positive control, a separate group of cells was incubated in media that had been directly supplemented with 2 ug/mL recombinant Ihh. Media was changed every 48 hours. After either 3 or 14 days of treatment, the cells were fixed and stained with X-Gal to detect expression of the
LacZ reporter as previously described (see chapter 5.3.4). The cells were counterstained with
0.1% nuclear fast red and imaged on an Olympus IX81 inverted microscope equipped with a color CCD camera.
6.3.3. In vivo studies
6.3.3.1. Experimental design
Skeletally mature, male Scx-GFP;Gli1lacZ/+ mice and Scx-GFP;Gli1+/+ littermates (12-16 weeks old, 36.3 ± 3.8 g, n=73) were used for in vivo studies and are described in detail elsewhere
106
(see chapter 5.3.1). Since this was a biomechanical study, we first verified that the LacZ-positive mice did not display baseline biomechanical deficits compared to their LacZ-negative littermates
(Fig. 6.3). Acute, full-length, full-thickness surgical injuries involving the tibial enthesis were created bilaterally in the central-third of each PT as described below. Immediately following surgery, each injured tendon was randomized into one of five possible treatment groups (Fig.
6.4): (1) no treatment (“natural healing”); (2) implantation of a 4 mm diameter Ihh-infused scaffold (“Ihh scaffold”); (3) implantation of a blank scaffold (“blank scaffold”); (4) bolus injection of Ihh (“Ihh bolus”); or (5) bolus injection of saline (“saline bolus”). A separate group of uninjured tendons (“normal”) were included as controls. Mice were euthanized at 5 weeks post-surgery and the PTs were harvested for biomechanical and morphological analysis. Tendons that had spontaneously ruptured during the healing period were excluded from these analyses.
Figure 6.3: Mouse breeding scheme and baseline biomechanical properties. (A) Scx-GFP females were crossed with Gli1lacZ/+ males to generate the mice used in this study. (B) Presence of the LacZ reporter did not affect mouse size or baseline PT biomechanical properties compared to LacZ-negative littermates at 12 weeks. n=10 male mice per group; error bars are SEM.
107
Figure 6.4: Experimental design, treatment groups, and response measures. Following creation of bilateral surgical injuries, each PT was randomized into one of five treatment groups (natural healing, Ihh scaffold, blank scaffold, Ihh bolus, or saline bolus). At 5 weeks post-surgery, mice were euthanized and the PTs harvested for biomechanical and histological analysis. Ruptured tendons identified via gross examination were excluded from these analyses. A separate group of uninjured mice served as normal controls.
6.3.3.2. Murine patellar tendon injury model
All murine surgical procedures were approved by the University of Cincinnati
Institutional Animal Care and Use Committee prior to study initiation. Mice were anesthetized with 4% isoflurane, subcutaneously injected with 1 mg/kg buprenorphine, and both hind limbs were shaved and aseptically prepped. Longitudinal skin incisions were made to expose the PT in each limb. An acute surgical injury was then created in each PT using a previously described
108 technique88. Briefly, two full-length longitudinal incisions were made to isolate the central-third portion of the PT from adjacent medial and lateral struts. The central-third of the PT was then transected as close as possible to its distal insertion into the tibia (Fig. 6.5A). The proximal patellar insertion was left intact. The transected central-third was laid back in its normal anatomic position between the struts with the distal end in close proximity to its original insertion site on the tibia (Fig. 6.5B). However, no attempt was made to physically reattach the tendon tissue to the bone. Thus, this model allows natural PT-to-bone healing to proceed in a load-protected environment without compromising the function of the knee extensor mechanism.
On the other hand, the absence of load on the transected tendon immediately after injury may have influenced the ability of the tendon to heal back to bone due to elimination of mechanical signals that drive the healing process.
6.3.3.3. Treatment groups
After creation of the acute surgical injury, each injured tendon was randomized into one of five possible treatment groups: natural healing, Ihh scaffold, blank scaffold, Ihh bolus, or saline bolus (n=20-28 tendons per group, Fig. 6.4). For the natural healing group, no treatment was provided and the injured tendon-to-bone insertion site was allowed to heal via natural mechanisms. For the Ihh scaffold group, a 4 mm diameter scaffold infused with 2 ug of recombinant Ihh was inserted through the skin incision with jeweler’s forceps and centered directly over the damaged enthesis (Fig. 6.5C). Blank scaffolds (0 ug Ihh per scaffold) were implanted in an equivalent manner. The scaffolds were gently manipulated to take on the contours of the murine knee in an attempt to keep them stationary during knee flexion/extension and eliminate any areas of high friction between the scaffold and surrounding tissues. However,
109 no rigid attachment methods (such as sutures or glue) were used to fix the scaffolds in place, largely because these methods require additional disruption to surrounding structures which can affect the healing process and confound experimental results. For the Ihh bolus group, 1 uL of
PBS containing 200 ng of recombinant Ihh was injected directly into the damaged enthesis using a microsyringe and a 26 gauge beveled needle. The saline bolus was delivered in an equivalent manner but did not contain any Ihh protein. Following treatment, skin incisions were closed with
5-0 prolene suture (Fig. 6.5D). Mice were allowed full range of motion and unlimited cage activity immediately following surgery.
Figure 6.5: PT surgical injury and scaffold implantation. (A) The central-third of the murine PT was isolated from medial and lateral struts then transected at the tibial insertion site. (B) The central-third flap was laid back in its anatomic position with the transected end in close proximity to its original insertion on the tibia. (C) A 4 mm diameter PCL scaffold, either blank or infused with 2 ug recombinant Ihh, was surgically implanted adjacent to the anterior surface of the damaged PT and centered over the tibial insertion site. (D) Skin incisions were closed with 5-0 prolene suture. Correct scaffold positioning was verified by visualizing or palpating the scaffold through the skin.
110
6.3.3.4. Dissection and gross examination
At 5 weeks post-surgery, mice were euthanized by carbon dioxide asphyxiation. Limbs assigned for histology were freshly dissected while limbs assigned for biomechanics were frozen at -20 °C until the date of tensile testing. On the day of dissection, the skin and subcutaneous tissue were removed to expose the PT and surrounding structures. In scaffold-treated limbs, the position of the implanted scaffold relative to the PT enthesis was recorded. The degrading scaffolds were then carefully peeled away from the underlying repair tissue using forceps. In cases where the healing PT had formed adhesions with the scaffold, a scalpel was used to carefully cut away any excess fibrotic tissue without damaging the underlying body of the tendon. High resolution digital photographs were taken throughout the dissection process to document the gross appearance of each specimen.
6.3.3.5. Histological sectioning and staining
Both decalcified and non-decalcified fixed frozen sections were made by fixing the dissected limbs in 4% paraformaldehyde for 1-2 days at 4°C, decalcifying the samples with 0.5M
EDTA for 7 days at 4°C if indicated, then submersing them in 30% sucrose overnight prior to embedding in Tissue-Tek O.C.T. compound (Sakura Finetek). Thin (8-12 um) frozen sections of the PT were cut in the sagittal plane on a cryostat using Kawamoto’s cryofilm technique
196, 197 (Section-Lab, Hiroshima, Japan) . Decalcified sections were hydrated in dH2O for 15 minutes and then stained with 0.025% toluidine blue O (T161-25; Fisher Scientific) for 30-60 seconds. Non-decalcified sections were stained with 2% alizarin red S at pH 4.2 for 2 minutes.
Sections were rinsed and mounted in 30% glycerol prior to imaging. All imaging was performed on an Olympus IX81 inverted microscope equipped with a color CCD camera. Images were
111 acquired using the same settings across all treatment groups and were subsequently refined using
Fiji image analysis software199.
6.3.3.6. Biomechanical testing and analysis
Biomechanical testing was performed as described previously88 (see chapter 4.3.3).
Briefly, on the day of biomechanical testing, mice were thawed and the hind limbs were dissected to expose the PT. The central-third of each PT was isolated by dissecting away the medial and lateral struts. Tendons were marked with three horizontal stain lines, mounted in a materials testing system (100R, Test Resources), preloaded to 0.02 N, preconditioned for 25 cycles, and then failed in tension at 0.003 mm/s. High resolution digital images were captured at
15 second intervals throughout the failure test in order to optically measure regional tissue strains and assess failure location. Ultimate load, displacement at failure, ultimate stress, and strain at failure were recorded for each PT specimen. Using an automated linear regression algorithm, stiffness and modulus were calculated from the linear region of the load-displacement and stress- strain curves, respectively. Regional tissue strains were calculated by optically tracking the applied stain lines, as described previously88. Failure location was also noted for each specimen.
6.3.4. Statistical analysis
Data is reported as mean ± SD unless otherwise noted. All statistical testing was performed using IBM SPSS Statistics 21.0. Significance was set at p < 0.05. Pearson’s chi- square test was used to determine if a relationship existed between treatment group and the frequency of tendon ruptures at 5 weeks post-surgery. This omnibus test was followed by four individual pairwise comparisons (Ihh scaffold vs natural healing, Ihh scaffold vs blank scaffold,
112
Ihh scaffold vs Ihh bolus, Ihh scaffold vs saline bolus) using the chi-square test with a
Bonferroni correction for multiple comparisons (p < 0.05/4). Any tendons which were found to be ruptured at the time of dissection were excluded from further biomechanical analysis. One- way ANOVA followed by Tukey’s HSD post-hoc analysis was used to evaluate differences between treatment groups in terms of structural properties (cross sectional area, ultimate load, stiffness), material properties (ultimate stress, modulus) properties, and regional strain patterns.
6.4. Results
6.4.1. Scaffold characterization and in vitro testing
The PCL scaffolds used in this study were custom-designed to provide sustained delivery of Ihh to the healing tendon enthesis. In vitro testing in the absence of cells showed that the scaffolds were capable of releasing Ihh continuously for at least 21 days. The in vitro Ihh release profile was biphasic in nature, with a prominent initial burst release in the first 48 hours followed by sustained linear release for at least 3 weeks (Fig. 6.6). In the initial burst phase, 324 ng
(16.2%) of the total infused Ihh was released from the scaffolds over a two day period. Over the next 19 days, an additional 306 ng (15.3%) of the total infused Ihh was released in a linear fashion. This correlates to a release rate of approximately 16 ng (0.8%) of the total infused Ihh per day. This data is useful because it allows us calculate how much total Ihh needs to be infused into a scaffold in order to achieve a specific threshold level of daily Ihh release. However, the
Ihh release rate in vivo will likely exceed what was observed in vitro due to the presence of cells and enzymes which accelerate scaffold degradation and drug release from the polymer fibers.
113
Figure 6.6: Ihh release kinetics in vitro. ELISA was used to quantify the amount of Ihh released from the scaffolds at each of the indicated time points. Each scaffold was initially infused with 2 ug of Ihh. When incubated in PBS in the absence of cells, the scaffolds continuously release Ihh for at least 21 days. The release profile is biphasic, with an initial burst phase followed by a sustained linear phase. During the linear phase (red dotted line), about 0.8% of the initial drug load is released per day. Error bars represent SD.
In order to verify that the Ihh released from the scaffolds was still bioactive following the infusion procedure, we compared Gli1-LacZ reporter expression in cultured murine tenocytes that were grown in the presence of either a blank scaffold, an Ihh-infused scaffold, or in media supplemented directly with Ihh. X-gal staining was predominantly absent from tenocytes grown for 3 days in the presence of a blank scaffold, although some cells did stain weakly positive, possibly indicating a low baseline level of endogenous hedgehog signaling (Fig. 6.7A). In contrast, tenocytes grown in the presence of an Ihh-infused scaffold or in media supplemented with 2 ug/mL Ihh exhibited strongly positive X-gal staining, indicating a high degree of active hedgehog signaling under these conditions (Fig. 6.7B,C). After 14 days in culture, the cells grown in the presence of the Ihh-infused scaffold remained strongly X-gal positive and looked
114 morphologically indistinguishable from the cells stimulated directly with Ihh in the media (Fig.
6.7D-F).
Figure 6.7: Bioactivity of Ihh does not change as a result of infusion into scaffolds. Tenocytes grown in the presence of an Ihh-infused scaffold for 3 or 14 days stain strongly positive for X- gal (blue), indicating active hedgehog signaling. The cells stimulated with an Ihh-infused scaffold appear morphologically indistinguishable from cells grown in media supplemented directly with Ihh. In contrast, cells grown in the presence of a blank scaffold have similar morphology as shown by nuclear fast red staining (pink), but X-gal staining is weak or absent.
6.4.2. Surgical outcomes and gross examination
At the time of surgery, successful scaffold implantation or bolus injection was achieved in over 90% of limbs. Following surgery, mice did not display any noticeable gait alterations or changes in behavior. However, dissection and gross examination at 5 weeks post-surgery revealed an unexpectedly large number of spontaneously ruptured tendons, especially in the Ihh scaffold group (Fig. 6.8). Omnibus chi-square analysis revealed a statistically significant
115 relationship between treatment group and the frequency of spontaneous tendon rupture (p =
0.025). In the Ihh scaffold group, 64% (18 of 28) of the PTs were found to be ruptured at dissection, compared to only 27% (6 of 22) in the natural healing group (p = 0.009) and 38.5%
(10 of 26) in the blank scaffold group (p = 0.058).
In both the Ihh scaffold and blank scaffold groups, the ruptured tendons were often extensively adhered to the scaffold material (Fig. 6.9C,D,G,H). This was in stark contrast to non- ruptured limbs, in which the scaffold remained mobile within a thin layer of No spontaneous ruptures were found in normal, unoperated tendons. All ruptured tendons were excluded from further biomechanical analysis.
Figure 6.8: Frequency of spontaneous PT rupture at 5 weeks post-surgery. PTs treated with an Ihh scaffold were significantly more likely to rupture compared to natural healing (p = 0.009) or saline bolus (p = 0.007) tendons. There was no significant difference in rupture rate between tendons treated with an Ihh scaffold compared to a blank scaffold or Ihh bolus.
116
Figure 6.9: Gross appearance of intact and ruptured tendons at 5 weeks post-surgery. In both examples shown here, the ruptured tendons have formed extensive fibrotic adhesions to the scaffolds, making it difficult to separate the tendon from the scaffold during dissection. In limbs where the PT was intact, the scaffolds were coated with a thin layer of subcutaneous connective tissue but were freely mobile and could be easily retrieved without damaging the underlying PT.
117
6.4.3. Structural and material properties
As expected, normal PT exhibited significantly lower cross-sectional area and significantly higher ultimate load, stiffness, ultimate stress, and modulus compared to all other experimental groups (Fig. 6.10 and 6.11). However, once the normal group was excluded from the analysis, no significant differences were detected between any of the remaining groups for any of the structural or material properties measured in this study. Contrary to our hypothesis, the tendons treated with an Ihh scaffold for 5 weeks actually had the lowest ultimate load, stiffness, ultimate stress, and modulus of any of the treatment groups, but none of these differences reached statistical significance (Fig. 6.10 and 6.11).
Figure 6.10: Average stress-strain curves for all treatment groups at 5 weeks post-surgery. Although normal tendons exhibited superior biomechanical properties, no significant biomechanical differences were detected between the natural healing, Ihh scaffold, blank scaffold, Ihh bolus, or saline bolus groups. n=7-15 per group (see Fig. 6.4 for exact sample sizes). Error bars represent SEM.
118
Figure 6.11: Structural and material properties at 5 weeks post-surgery. Normal PT exhibited significantly lower cross-sectional area and significantly higher ultimate load, stiffness, ultimate stress, and modulus compared to all other experimental groups (indicated by an asterisk). No other differences between groups were detected. The Ihh scaffold group actually had the lowest ultimate load, stiffness, ultimate stress, and modulus of any of the treatment groups, but none of these differences reached statistical significance. Error bars represent SD.
119
6.4.4. Regional strain patterns
No significant differences in strain patterns were detected between any of the treatment groups (data not shown). Strain analysis was of limited utility due to a very large amount of inter-specimen variability in the regional strain measurements, especially within the Ihh scaffold and blank scaffold groups, where the coefficients of variation often exceeded 100%.
6.4.5. Failure mode and location
During biomechanical testing, the healthy, unoperated PTs most commonly failed via an anterior-posterior delamination mechanism (67%; see Fig. 4.8B for an example). In contrast, tendons subjected to a surgical injury and allowed to heal for 5 weeks predominantly failed at the distal insertion site, usually via a transverse rupture of the repair tissue at the enthesis (>60% in all treatment groups; see Fig. 4.8C for an example) or sometimes via an avulsion of soft tissue off of the tibia (20-30%) (Fig. 6.12). Treatment with an Ihh-infused scaffold or an Ihh bolus did not change the failure location or failure mode.
120
Figure 6.12: Treatment with Ihh did not change PT failure mechanisms. Normal, unoperated tendons usually failed via a delamination mechanism, while healing tendons failed at the tibial insertion either via a transverse rupture at the tendon-bone junction or via a soft tissue avulsion in which a large swath of periosteum and/or scar tissue was pulled off the tibia during failure.
6.4.6. Histologic outcomes
Histologic analysis revealed no appreciable differences in repair tissue morphology between the natural healing group, the blank scaffold group, the Ihh bolus group, and the saline bolus group at 5 weeks post-surgery. However, implantation of an Ihh-infused scaffold did produce morphological changes at the healing enthesis that were evident to varying degrees in all samples examined in this study. In general, the repair tissue in tendons treated with Ihh-infused scaffolds appeared very hypercellular and contained numerous large, rounded, chondrocyte-like
121 cells (Fig. 6.13C,F). Cells with this type of morphology were rare in naturally healing entheses, which instead contained repair tissue dominated by smaller, elongated, fibroblast-like cells (Fig.
6.13B,E). The repair tissue in tendons treated with an Ihh-infused scaffold also had a paucity of extracellular matrix compared to the extensive collagen deposition and fibrous scar formation in naturally healing tendons (Fig. 6.13).
Figure 6.13: Repair tissue morphology differs between natural healing and Ihh-treated tendons. (A,D) Normal tendons display a prominent tibial tuberosity with a well-defined tidemark. (B,E) Following surgical injury to the enthesis, naturally healing tendons produce scar tissue containing many small, fibroblast-like cells and large amounts of collagenous matrix. (C,F) In contrast, the repair tissue in tendons treated with an Ihh-infused scaffold contains numerous large, round, chondrocyte-like cells but a relative paucity of collagenous matrix. Toluidine blue O stain; decalcified tissue sections.
122
Detailed histologic analysis of the interface between unmineralized and mineralized tissue at the tibial enthesis again revealed the presence of large, rounded, chondrocyte-like cells in tendons that had been treated with an Ihh-infused scaffold but not in normal tendons or those treated with a blank scaffold (Fig. 6.14). These chondrocyte-like cells were surrounded by an extracellular matrix that exhibited deep purple metachromasia when stained with toluidine blue
O, suggesting the presence of proteoglycans. Finally, although we hypothesized that delivery of
Ihh to the healing tendon might induce new mineralization, alizarin red staining did not show any evidence of ossification or ectopic mineralization in the Ihh-treated tendons (Fig. 6.15).
Figure 6.14: Morphology of the tendon-bone interface in control and Ihh-treated PTs. Detailed examination of the tibial enthesis (see yellow box in A for representative region of interest) in tendons treated with an Ihh-infused scaffold revealed the presence of large, round, chondrocyte- like cells (D, yellow arrowheads) surrounded by a proteoglycan rich matrix. These findings are characteristic of fibrocartilage tissue. Toluidine blue O stain; non-decalcified tissue sections.
123
Figure 6.15: Ihh-infused scaffolds do not induce ectopic mineralization in the PT. Alizarin red staining of non-decalcified tissue sections at 5 weeks post-surgery did not show any areas of heterotopic ossification in the tendon midsubstance. Red = mineralized tissue; yellow/brown = unmineralized tissue.
6.5. Discussion
Our knowledge of the role of Indian hedgehog signaling in tendon development and healing has grown steadily in recent years but still remains far from complete. In addition to its role in enthesis development and maturation, Ihh may also be important during adult tendon homeostasis, normal tendon aging, and in the etiopathogenesis of degenerative tendinopathy (see chapter 5). As one of the major molecular regulators of chondrocyte phenotype, the Indian hedgehog pathway represents an intriguing therapeutic target for augmenting fibrocartilage healing during adult tendon-to-bone repair. Thus, the main objective of this study was to test whether delivery of recombinant Ihh protein to the healing PT enthesis could improve functional outcomes in a murine PT injury model. We hypothesized that Ihh delivery to the healing
124 insertion site would stimulate the proliferation, differentiation, and mineralization of neo- fibrocartilage tissue, thereby resulting in restoration of a zonal fibrocartilaginous enthesis with improved biomechanical properties compared to controls.
In order to test this hypothesis, we first had to devise an effective method for delivering recombinant Ihh to the healing enthesis in vivo. Direct injection of recombinant Ihh protein at the site of injury is the most straightforward approach, but this method had no detectable effect on histologic or biomechanical healing outcomes at 5 weeks post-surgery in this study. The serum half-life of hedgehog ligands is on the order of hours, and their rapid degradation when exposed to the in vivo environment likely limits the therapeutic efficacy of single bolus injections248. To combat this, researchers have tried modifying the structure of hedgehog proteins (e.g. via
PEGylation) in order to increase the half-life and improve pharmacokinetic and pharmacodynamic properties248. Rapid diffusion of the Ihh ligands away from the site of injection also decreases efficacy and increases the likelihood of unwanted effects on neighboring tissues. Notably, upregulated Ihh signaling is a mechanism of tumorigenesis in certain cancers, especially basal cell carcinoma of the skin249. Dysregulated Ihh signaling has also been implicated in the development of osteoarthritis, and the amount of Ihh protein in synovial fluid appears to be a sensitive biomarker for disease progression, with higher levels corresponding to more severe articular cartilage damage250-256. Thus, therapeutic strategies aimed at modulating hedgehog signaling must be able to precisely target the correct tissues and/or cells in order to eliminate the potential for serious side effects.
In this study, we attempted to achieve localized and sustained Ihh delivery to the healing enthesis by creating biodegradable Ihh-infused scaffolds that could be implanted directly at the site of injury. After testing a variety of polymeric scaffold materials, we settled on electrospun
125
PCL due to its high mechanical strength, biocompatibility, low cost, and current FDA approval for many different medical uses in human patients, including as a component of drug delivery devices. Once prototype scaffolds had been fabricated, in vitro testing confirmed that they were able to steadily release Ihh for at least 3 weeks and activate hedgehog signaling in cultured cells.
Despite these promising in vitro results, implantation of the Ihh-infused scaffolds into our acute murine PT injury model did not improve biomechanical healing outcomes at 5 weeks post- surgery. In fact, treatment with Ihh-infused scaffolds led to a significantly increased frequency of spontaneous PT ruptures during the 5 week healing period compared to natural healing alone.
When comparing only the remaining non-ruptured tendons, treatment with an Ihh-infused scaffold did not result in any statistically significant differences in terms of tendon structural properties, material properties, or regional strain patterns after 5 weeks of healing compared to a blank scaffold or no scaffold at all. Contrary to our hypothesis, these biomechanical results suggest that delivery of Ihh to the healing enthesis does not improve functional outcomes of tendon-to-bone healing and may even have a deleterious effect.
Histological analysis revealed several intriguing findings that may help explain the high rupture rate and poor biomechanical outcomes in the Ihh-infused scaffold group. The repair tissue at the healing enthesis in tendons treated with an Ihh-infused scaffold was found to be highly cellular but contained a paucity of extracellular matrix. The cells residing within this repair tissue were large, round, and had a chondrocytic appearance. In contrast, tendons allowed to heal naturally or treated with a blank scaffold formed scar tissue at the insertion site that contained abundant extracellular matrix which appeared well-integrated with both the transected tendon end and the underlying bone. The cells in this scar tissue were smaller, more elongated, and fibroblastic in appearance. Large, round, chondrocytic cells surrounded by a proteoglycan-
126 rich matrix were also abundant at the mineralization front in tendon entheses treated with an Ihh- infused scaffold but were not present in the blank scaffold group. We did not detect any ectopic mineralization as a result of the Ihh-infused scaffolds, nor did we see any evidence of inflammation or necrosis. These results are consistent with a mechanism in which sustained treatment with exogenous Ihh during enthesis healing attenuates the normal fibrotic healing response and instead promotes cell proliferation and chondrogenic differentiation at the insertion site. These findings are also consistent with the data from our in vitro studies on the effects of
Ihh stimulation on cultured tendon fibroblasts, which suggested that stimulation with Ihh inhibited Scx expression and induced cells to transdifferentiate from a fibroblastic phenotype into a more chondrocytic phenotype (see chapter 5.4.5). Although reducing fibrous scar formation may seem desirable in the context of tendon-to-bone healing, we hypothesize that the relative lack of fibrotic repair tissue at the insertion site of tendons treated with an Ihh-infused scaffold actually made the tendon-bone unit considerably weaker and more susceptible to rupture in this study.
In addition to the biological effects of Ihh, it is also likely that the presence of a relatively large polymeric scaffold in close proximity to the damaged insertion site acts as a physical impediment to healing. Prior studies have shown that central-third defects in the murine PT heal primarily via an extrinsic mechanism in which cells from the paratenon and retinaculum become activated following injury, begin expressing Scleraxis, and migrate from the periphery of the tendon into the wound site where they begin to produce conditional matrix47. By implanting a scaffold directly over the wound site following surgery, cells may be unable to efficiently migrate into the defect, effectively limiting the healing process. In cases where tendons were treated with an Ihh-infused scaffold, the physical presence of the scaffold coupled with the
127 biological effects of the Ihh may have resulted in repair tissue that was inferior in terms of its structural and material properties when compared to the other treatment groups. Since mice do not limit their activity following surgery, these weakened tendons would have been at high risk for spontaneous rupture during the healing period.
This study did have a number of limitations. We tested our scaffolds in a previously characterized murine PT injury model, but this model lacks direct clinical relevance because the transected central-third of the PT is not surgically reattached to its tibial insertion following injury. In addition, since no surgical repair is performed, the transected tendon remains unloaded during early healing. This is important because Ihh signaling has been shown to be mechanosensitive, and so responsiveness to exogenous Ihh stimulation may depend in part on the mechanical environment. Also, all our analyses in this study were performed at a single time point (5 weeks post-surgery), which limits the conclusions we can draw about the temporal nature of tendon-to-bone healing. We are in the process of examining earlier time points (1 week and 2 weeks post-surgery) in order to better understand the natural progression of tendon-to-bone healing in this model and to assess the involvement of Ihh signaling at different stages of this process. Finally, although we tested our scaffolds in vitro before moving on to in vivo studies, it is often difficult to predict in vivo parameters (such as scaffold degradation rate or drug release kinetics) based solely on in vitro data because the two environments are very different. Thus, it is possible that our Ihh-infused scaffolds were not performing as expected in vivo.
This study also revealed some of the major experimental challenges that arise when attempting to move from the realm of basic research towards more applied or translational research. Our data showed that the Ihh-infused scaffolds used in this study did not improve biomechanical healing outcomes in our murine injury model. However, this certainly does not
128 mean that recombinant Ihh has no future utility as a therapeutic molecule for augmenting tendon- to-bone repair. More likely, it simply means that we did not infuse an appropriate dose of Ihh into each scaffold, or that the scaffold degradation rate was too slow, or that we used scaffolds that were too large and disruptive to the healing process, or that we used an ineffective delivery method altogether, etc. There are myriad new variables that must be considered when attempting to translate a new biological finding into a clinically-relevant product, yet often it is unclear which of these variables are most important. In the case of Indian hedgehog, it is still unclear what role Ihh plays during normal tendon-to-bone healing, so attempting to design an effective biological therapy for tendon repair by targeting the hedgehog pathway may still be premature at the current time. Further mechanistic studies are certainly needed in order to more fully characterize the effects of Ihh on enthesis healing.
As our knowledge of tendon biology continues to expand, the ongoing challenge for clinicians and bioengineers will be to translate this growing knowledge into effective treatments for tendon disorders. In general, efficacious biological therapies for tendon-to-bone repair will likely remain elusive unless more precise methods for controlling the spatiotemporal delivery of biologic factors to the site of injury are developed. Nonetheless, this study paves the way for future tissue engineering studies in which the Ihh signaling pathway is targeted during tendon repair. Ultimately, we hope our work will eventually lead to new therapeutic strategies for tendon disorders via modulation of hedgehog signaling.
6.6. Acknowledgements
We gratefully acknowledge the National Institutes of Health for providing research support (R01 AR056943) as well as student funding via the University of Cincinnati MSTP
129 training grant (T32 GM063483). Student tuition and stipend support was also provided by the
University of Cincinnati Graduate School Dean’s Fellowship (to S.D. Gilday). We thank Dr.
Andrew Breidenbach, Dr. Andrea Lalley, Dr. Nathaniel Dyment, and Dr. Han Liu for technical assistance. We would like to acknowledge Dr. Daria Narmoneva for providing microscopy and imaging support and Cindi Gooch for helping with mouse husbandry, breeding, and surgery.
130
CHAPTER 7
Conclusions and Recommendations
7.1. Summary and discussion of principal findings
Tendon injuries are common, debilitating, and often difficult to treat. Reattaching ruptured tendons to their bony insertions has been a fundamental challenge in orthopaedics for decades, yet effective solutions that restore normal fibrocartilaginous enthesis architecture and mechanical function are still lacking. To address this problem, our biomedical engineering research group at the University of Cincinnati established a formal partnership with developmental biologists at Cincinnati Children’s Hospital Medical Center via an NIH
Bioengineering Research Partnership grant titled “A developmentally-based tissue engineering approach to improve tendon repair”. We believed that in order to successfully engineer a replacement tissue or develop effective treatments for tendon-to-bone healing, it would be necessary to first understand the biological signals that govern how the natural interface between tendon and bone develops. Thus began our initial investigations into tendon development and specifically the role of Indian hedgehog signaling in this process.
As discussed in Chapter 2, we found that Ihh is a critical regulator of fibrocartilaginous enthesis differentiation, growth, and mineralization. Ihh is critical for the development and maturation of a morphologically normal and biomechanically functional enthesis. It promotes chondrogenic differentiation and growth of entheseal fibrocartilage during the perinatal period and is also a positive regulator of post-natal mineralization at the tendon-bone interface. These discoveries were a fundamental breakthrough that laid the groundwork for the studies presented
131 in this dissertation. In this work, our goal was to build upon our developmental studies by evaluating the activity of Ihh signaling in adult tendons and investigating its therapeutic potential for adult tendon-to-bone repair. Below, we summarize our principal findings.
1. Restoration of the zonal fibrocartilaginous enthesis following injury is an important biological success criterion for functional tissue engineering of tendon. The long term goal of our laboratory is to design new therapeutic strategies that can return function to damaged musculoskeletal tissues. Building on the FTE paradigm (see chapter 1.3), in this dissertation we established biological design criteria for our tissue-engineered tendon repairs.
Our conceptual strategy for establishing these design criteria was to investigate the normal biological processes responsible for tendon development and maturation, identify and categorize the biological parameters that are most important, and finally, assess how these chosen biological parameters affect the mechanical outcomes of tissue repair. Given its importance in anchoring the tendon to the bone and dissipating stress concentrations that build up at the tendon-bone interface, we feel that the restoration of a zonal fibrocartilaginous enthesis is one of the critical design criteria that must be met in order to achieve a successful, functional tendon repair.
2. Once disrupted, the murine PT enthesis does not recover normal biomechanical properties even after 8 weeks of natural healing. Before we could begin designing strategies to augment tendon-to-bone healing, we first needed an appropriate animal model. We developed a murine model of acute PT injury and characterized the biomechanical outcomes of natural healing in this model (see chapter 4). We found that once the PT insertion site had been surgically damaged, 8 weeks of natural healing was still insufficient to restore full function to the tendon-bone unit. Thus, this study not only introduced a new murine model of tendon-to-bone
132 healing into the literature, but it also established the natural healing baseline upon which future therapeutic studies will attempt to improve.
3. Normal murine PT exhibits regional variations in tissue strains. In order to fully characterize healing biomechanics in our murine PT injury model, we developed an optical technique to measure microscale local surface strains in mouse tendons during tensile loading
(see chapter 4). Similar to human tendons, we found that the murine PT exhibited 2-3x higher strain in the insertion region compared to the midsubstance during sub-failure loading. This finding reinforces the idea that regional variation in material properties along the length of a tendon is necessary for optimal biomechanical function. Our regional strain measurement technique has now also been successfully used to detect phenotypic differences between normal and mutant (e.g. SmoKO) tendons43, 181.
4. The adult murine PT enthesis contains a population of hedgehog-responsive cells in the unmineralized entheseal fibrocartilage. Using a Scx-GFP;Gli1-LacZ double reporter mouse, we were able to identify discrete populations of cells in and around the PT that were actively responding to hedgehog signaling. We found that the murine PT enthesis (and all other entheses examined as well) contained a population of hedgehog-responsive (Gli1-positive) cells that were restricted to the unmineralized entheseal fibrocartilage (see chapter 5). The strikingly consistent spatial distribution of these cells along the unmineralized side of the tidemark strongly suggests that they play a role in the regulation of mineralization. Furthermore, we examined
Gli1-LacZ reporter mice ranging from 12 to 46 weeks of age and found that the patterns of activated hedgehog signaling at the PT enthesis did not appreciably change as a function of aging, even long after the slowing of tendon growth and decline in mineral apposition rate (see chapter 5). The mechanisms by which these hedgehog-responsive cells in the entheseal
133 fibrocartilage regulate mineralization, homeostasis, and/or healing are still unknown, thus providing an exciting and potentially very impactful opportunity for future work.
5. Hedgehog signaling is active in regions of tendon undergoing fibrocartilaginous metaplasia, likely in response to compressive loading. One of the most interesting findings from this dissertation was the observation that hedgehog signaling activity was present in regions of the tendon midsubstance where compressive forces would be expected to occur, such as where tendons wrap around or become pressed against bone. We observed that in these regions, the cells have a more chondrocytic phenotype, consistent with the known phenomenon of fibrocartilaginous metaplasia (see chapter 5). We did not explicitly quantify or experimentally vary the amount of compressive force on the tendons in vivo in order to establish a causal relationship between compression, hedgehog signaling, and fibrocartilage formation, but our observational data has led us to hypothesize that Scx-positive tenocytes have the ability to directly transdifferentiate into chondrocyte-like cells in response to compressive force. The role of hedgehog signaling in this process is unknown, but it could serve as a biological transducer which links the mechanical environment with cell phenotype. The mechanobiological relationships between loading, hedgehog signaling, and fibrocartilaginous metaplasia in tendons should be explored mechanistically in future work.
6. Stimulation with Ihh causes tenocytes to downregulate Scx expression and adopt a more chondrocytic phenotype. In vitro experiments have also supported a role for Ihh in mediating the phenotypic transition from tendon fibroblast to fibrochondrocyte. Tenocytes cultured in monolayer and exposed to recombinant Ihh protein downregulate Scx expression and their morphology becomes more chondrocyte-like (see chapter 5). Treating cells with cyclopamine, an inhibitor of hedgehog signaling, has the opposite effect. These results lend
134 further credence to the idea that hedgehog signaling may well have therapeutic utility for stimulating fibrocartilage formation and growth, as long as it can be delivered effectively.
7. In vivo delivery of Ihh to the healing PT enthesis via a controlled release scaffold produces histologic changes consistent with increased chondrogenic differentiation at 5 weeks post-surgery. In vitro testing showed that the novel drug delivery scaffolds we developed in this work were able to successfully release bioactive Ihh at a predictable rate for a duration of at least 2 weeks. In healing tendons treated with our Ihh-infused scaffolds in vivo, we observed histologic changes consistent with increased cell proliferation and chondrogenic differentiation at the insertion site at 5 weeks post-surgery (see Chapter 6). Although encouraging, more work is needed in order to better understand how exogenously delivered Ihh affects the healing process in vivo. Earlier post-surgical time points and additional response measures (such as qPCR to assess gene expression changes or immunohistochemistry to assess specific fibrocartilage markers) should be utilized to provide a more complete assessment of the effects of recombinant
Ihh on fibrocartilage formation during tendon-to-bone healing.
8. Our use of Ihh-infused scaffolds to augment tendon-to-bone healing had a deleterious effect on tendon biomechanics at 5 weeks post-surgery. In this work, when the
Ihh-infused scaffolds were tested in vivo, they caused healing tendons to rupture 64% of the time. Tendons treated with an Ihh-infused scaffold displayed a lack of fibrotic scar formation at the enthesis after the surgical injury compared to controls, and this could have contributed to the inferior biomechanical properties and increased rupture rate (see Chapter 6). It is unclear if these effects were primarily attributable to the biological effects of the Ihh released from the scaffold or if they were simply due to the physical presence of the scaffold itself, which may have interfered with healing. Although the biomechanical results were not as we expected, the
135 therapeutic potential of Ihh for tendon-to-bone repair should not be ruled out based on the outcomes of this initial study. Creative new strategies to deliver Ihh to the appropriate cells at the healing insertion are needed and may yield more promising biomechanical results.
7.2. Unanswered questions and recommendations for future studies
In this dissertation, we have only started investigating the many possibilities by which the
Ihh pathway could be modulated in an attempt to improve musculoskeletal tissue repair. Our experiments have produced some interesting data, but many important questions still remain unanswered. Our work has brought to light some interesting new directions that warrant further investigation. These include both basic science questions about the mechanistic role of Indian hedgehog signaling in tendons as well as translational questions about how to therapeutically target the hedgehog pathway in tendon disease, healing, and repair.
1. What is the role of Ihh signaling in entheseal fibrocartilage in mature adult tendons? In addition to its established role in enthesis development, our identification of Gli1- positive cells in the unmineralized entheseal fibrocartilage in adult murine tendons implies a role for hedgehog signaling at the mature enthesis. The distinct localization of active hedgehog signaling to the unmineralized fibrocartilage region and its consistent association with the tidemark most likely signifies a role in fibrocartilage homeostasis and/or the regulation of enthesis mineralization. This hypothesis is consistent with our developmental data, which shows deficits in fibrocartilage growth and mineralized fibrocartilage area if Ihh signaling is ablated during development. In further support of this hypothesis, data from work in articular cartilage by other groups has shown that Ihh signaling is a chondrogenic inducer220 and a recent study by
136
Dyment and colleagues257 showed that fibrocartilage cells in growing tendons mineralize via hedgehog signaling to form the zonal enthesis.
However, if Ihh signaling does indeed stimulate chondrogenic differentiation and mineralization as the data suggest, then the fact that hedgehog signaling remains active at the enthesis even in mature tendons raises an interesting question: Why doesn’t the enthesis fibrocartilage continue to expand and mineralize throughout life? One possible explanation is that there are other developmental signals in addition to Ihh (such as BMPs) that dictate the temporal decline in fibrocartilage growth and mineralization. Another possibility is that inhibitory signals, produced either in the maturing tendon itself or from neighboring tissues such as muscle, limit fibrocartilage growth in adult animals. The changing mechanical environment as the tendon grows and is exposed to increased forces likely regulates fibrocartilage formation and mineralization at the enthesis as well.
Our current understanding of the role of Ihh signaling at the enthesis is limited in part because we do not know what upstream genes or signals activate the hedgehog pathway in tendon cells. Similarly, we do not fully understand all the downstream targets and effectors of activated hedgehog signaling. Thus, fundamental mechanistic studies are needed in order to elucidate the details of these molecular pathways. These studies could be conducted using inducible gain- and loss-of-function transgenic mice in which critical elements of the hedgehog signaling pathway are either constitutively active or knocked out within the tendons of adult animals. With the availability of CRISPR-Cas genome editing technology, it is now possible to overcome many of the technical limitations for creating the necessary transgenic mouse models at a reasonable cost and within a shorter time frame. After inducible overexpression or elimination of hedgehog signaling activity at a designated time point (e.g. at 20 weeks of age),
137 we could assess the resulting morphologic and biomechanical effects on the enthesis fibrocartilage and thus gain further insight into the role of hedgehog signaling in mature tendons.
2. What is the role of endogenous Indian hedgehog signaling during natural tendon- to-bone healing? Although our research group has studied hedgehog in the context of tendon development and attempted to use recombinant Ihh protein as a biologic therapy for tendon repair, we have not systematically studied the role of hedgehog signaling during natural healing of a tendon injury. By creating surgical injuries and assessing the spatiotemporal patterns of endogenous hedgehog pathway activation, we could assess whether or not this pathway is upregulated or downregulated during healing. Furthermore, genetically inhibiting or activating this pathway during healing and assessing the resulting biomechanical outcomes at 2, 5, and 8 weeks could help us understand how Ihh signaling might be contributing to functional outcomes.
3. What is the relationship between compressive force, Ihh signaling, and fibrocartilaginous metaplasia in tendons? Despite its prevalence, the molecular mechanisms governing fibrocartilaginous metaplasia in tendons have not been fully elucidated. In this work, we observed that in areas of tendon in which higher compressive forces would be expected, such as where the tendon surface presses against rigid bone, the cells displayed a more chondrocytic morphology and expressed Gli1, indicating active hedgehog signaling was occurring. Based on this observation, we now hypothesize that Ihh signaling may be a key biological intermediary that links mechanical forces, namely compression, to the development of fibrocartilage in tendons (Fig. 7.1). This mechanism may help explain not only fibrocartilaginous metaplasia in the tendon midsubstance, but also the formation of fibrocartilage at the postnatal enthesis.
138
Figure 7.1: Proposed mechanism of fibrocartilage metaplasia in tendons. [1] Tendons are composed of abundant Scx-expressing tenocytes (green) and rare tendon stem cells (TSC, grey). The tenocytes are aligned along the axis of tension and thus have a spindle shape. [2] When subjected to compressive forces, tenocytes begin expressing and secreting Ihh (red), which acts in an autocrine or paracrine manner on nearby cells. [3a] Tenocytes in the region of compression begin responding to the Ihh and begin to transdifferentiate into Gli1-expressing cells, or alternatively, [3b] TSCs begin responding to the Ihh and begin to proliferate and differentiate into Gli1-expressing cells. [4] If compressive forces persist, the tendon undergoes fibrocartilaginous metaplasia, as evidenced by the appearance of mature, rounded, Gli1- expressing fibrochondrocytes (blue) and the production of a proteoglycan-rich matrix which protects the tendon from damage caused by compressive loading. If the compressive forces are removed, then the tendon cells will revert back to their original morphology and phenotype.
139
In this dissertation, we did not explicitly quantify or experimentally vary the amount of compressive force on tendons in vivo in order to establish a causal relationship between compression, hedgehog signaling, and fibrocartilaginous metaplasia. Further mechanistic studies which test this relationship directly are needed in order to truly reveal whether compressive forces and hedgehog signaling are necessary and/or sufficient to induce fibrocartilage formation in tendon. These studies could be conducted by subjecting the murine PT to compression in vivo, although a custom device would need to be created to apply controlled and measurable amounts of compressive force to specific regions of the tendon without harming the animal. An alternative approach would be to isolate murine tenocytes and expose them to compressive forces in vitro using a mechanical stimulator. By assaying for active hedgehog signaling at early time points (e.g. using the Gli1-LacZ reporter) and for markers of fibrocartilage at later time points (e.g. using qPCR or immunohistochemistry) in groups of tendons subjected to differing levels of compressive force, it would be possible to establish a relationship between compression, Ihh activity, and fibrocartilage formation. By performing these same experiments in a Smo-KO animal or in the presence of a hedgehog inhibitor like cyclopamine, it would be possible to determine if Ihh activity is absolutely required for the development of fibrocartilage, or if compressive force in the absence of hedgehog is sufficient. Furthermore, if Ihh signaling is indeed the critical intermediary linking compressive force and fibrocartilage metaplasia, then direct activation of Smo in tenocytes by either chemical or genetic means should faithfully recapitulate the phenotypic changes seen as a result of subjecting the tenocytes to compression.
The mechanobiology of Ihh signaling in tendons will likely be a fruitful area for further research, but it is not without challenges. A major hurdle is how to quantitatively measure the mechanical forces on tendon cells. Cells in the tendon enthesis are subjected to complex loading
140 patterns that include tension, compression, and shear. Even a simple maneuver like stretching a tendon uniaxially causes a cell at the insertion site to experience tensile forces, compressive forces (due to the Poisson effect), and shear forces (due to sliding of the collagen fibers with respect to one another). These different forces may have opposing effects on hedgehog signaling.
Another challenge is how to accurately quantify hedgehog signaling activity. The Gli1-LacZ reporter mouse used extensively in this dissertation allows for visualization of cells responding to hedgehog ligands, but this is qualitative instead of quantitative. Finally, it is unclear if the phenomenon of fibrocartilage metaplasia is due to transdifferentiation of tenocytes into fibrochondrocytes or proliferation and fibrochondrogenic differentiation of resident tendon stem cells (Fig. 7.1 panels 3a and 3b). This is an interesting question that also has therapeutic implications, but rigorously determining the answer will require lineage tracing studies to assess whether the Gli1-expressing fibrochondrocytes within tendon are the clonal progeny of resident tendon stem cells or whether they are derived from the Scx-expressing lineage.
4. Is Ihh signaling involved in the etiopathogenesis of chronic tendinopathy? In this work, we observed the age-related formation and expansion of new foci of active hedgehog signaling in the midsubstance of the murine Achilles tendon but not the patellar tendon. It is unclear whether this phenomenon represents a normal age-related change, possibly due to a changing mechanical environment as the Achilles tendon stiffens with age, or if it is due to some type of pathology, such as degenerative tendinopathy, that causes progressive phenotypic changes within the tendon cells. Additional studies making use of human tissues from patients with tendinosis and animal models of tendinopathy are warranted in order to try and discern the role of hedgehog signaling, if any, in the etiology and pathogenesis of age-related tendon degeneration. If Ihh signaling is involved in this process, then the door will be opened for future
141 therapeutic studies that attempt to modulate hedgehog signaling as a means of preventing or reversing chronic tendon degeneration.
5. What is the most appropriate animal injury model for the continued study of tendon-to-bone healing? In this dissertation, we relied heavily on a modified murine PT injury model based off of our laboratory’s previous central-third patellar tendon defect injury models.
Although very familiar and thus easy to use, this model lacks direct clinical relevance because no surgical repair is performed following the creation of the acute injury. Thus, the transected tendon remains unsecured and unloaded during early healing. As has already been discussed, the loading environment following surgery is likely very important because Ihh signaling has been shown to be highly mechanosensitive. In support of this hypothesis, a recent study of tendon-to- bone healing in a rat ACL reconstruction model showed that pre-tensioning of the graft at the time of surgery resulted in increased Ihh signaling at 3 weeks258. Thus, the responsiveness of healing tendons to Ihh may depend in part on how much load is placed on the tissue. Other injury models of tendon-to-bone healing are available, including a recently developed murine model of supraspinatus tendon repair259. In the future, the choice of model will be very important, especially for translational studies. Although murine models are well suited for basic research due to their genetic tractability and low cost, a larger animal model such as a rat or rabbit would allow for more clinically realistic tendon repair surgeries. This may be required in order to accurately assess the performance of future drug delivery scaffolds or other therapeutic devices.
6. Which cell types are the primary contributors to enthesis fibrocartilage healing following injury? One of the major reasons why biologic therapies aimed at augmenting tendon- to-bone healing have been largely unsuccessful is that it is often unclear which cell types are the primary mediators of repair, and thus the chosen therapeutic intervention cannot be targeted or
142 localized to the correct cells. Injuries to the tendon enthesis often involve not only tendon tissue, but also fibrocartilage, bone, fascia, paratenon, periosteum, muscle, vascular tissue, and adipose tissue, all of which could serve as a source of cells during the healing process. Which of these cell types contribute to entheseal fibrocartilage healing after a tendon injury? What is the relative contribution of stem/progenitor cells versus differentiated cells? Does this differ widely based on the anatomic location of the tendon, the local environment, and the mechanism of the injury?
In major load bearing tendons such as the Achilles, PT, and rotator cuff, extrinsic healing predominates following injury. Studies using an empty central-third murine PT defect model have shown that the main contributors to the healing response are paratenon cells which migrate into the wound site, where they differentiate into Scx-expressing cells and begin to produce collagenous matrix47. It is likely that these same cells are the primary mediators of repair in our
PT injury model as well. Nonetheless, further studies are needed to better define the phenotype and behavior of these cells at different stages of healing. In addition, lineage tracing and genetic labeling techniques should be employed to map the expansion of specific pools of cells (such as the Gli1-expressing fibrochondrocytes at the enthesis) following injury to determine if they contribute to the final repair tissue.
7. How could we make our Ihh-infused scaffolds more efficacious? What other experimental approaches could be used to modulate Ihh signaling in healing tendons?
During the course of our work, we invested considerable time and effort in order to design, fabricate, and test different polymeric drug delivery scaffolds. One argument for continuing work with the scaffolds is that we have already characterized them in vitro and shown that they release bioactive Ihh for sustained time periods and at a predictable rate. We can easily tailor the size, shape, Ihh concentration, or scaffold composition in future studies to control the drug release
143 profile. We could even infuse the scaffolds with something other than Ihh, such as BMP, and then test their effects in vitro and in vivo using similar methods as before. In essence, we now have a platform technology that can be easily customized and adjusted depending on our experimental plans. In retrospect, implanting a relatively large (4 mm) scaffold into a relatively small murine injury model was probably not the most effective method for localizing Ihh delivery to the healing enthesis. In addition, the physical presence of the scaffolds likely altered the healing process. If these scaffolds are to be used in future therapeutic studies in the mouse, they should be made considerably smaller (<1 mm) and affixed to the healing enthesis after the tendon has been surgically reattached to the bone. This would lessen the probability of the scaffolds interfering with the extrinsic healing process and would help ensure that Ihh is released locally at the junction between tendon and bone.
Future work should also investigate cell-based therapies in addition to Ihh-infused scaffolds. Mesenchymal stem cells or tendon fibroblasts could be harvested from mice, genetically engineered to express Ihh, and then implanted into an enthesis defect. In theory, these implanted cells would produce and secrete Ihh, which could then act in an autocrine or paracrine manner to stimulate proliferation and chondrogenic differentiation of cells at the healing enthesis, ultimately resulting in increased fibrocartilage formation compared to controls. As long as the implanted cells remained viable and did not migrate away from the defect site, this approach would ensure a relatively constant and long lasting supply of Ihh ligands in the immediate vicinity of the healing enthesis. Although not as readily translatable to larger animal models or humans, the simplicity of this cell-based approach makes it an attractive alternative to scaffold-based drug delivery.
144
7.3. Overall impact on the field of tendon tissue engineering
The work presented in this dissertation represents our laboratory’s initial attempts to understand the role of hedgehog signaling in adult tendon and to utilize Ihh as a therapeutic for tendon-to-bone repair. Prior to our studies, Ihh had never before been investigated in the context of normal tendon biology or as a potential mediator of tendon healing following injury. Over the last five years, we have made significant progress towards elucidating the mechanism by which
Ihh signaling regulates fibrocartilage development and mineralization at the enthesis. We have also been proactive regarding the translation of our basic science findings into biologic therapies for tendon-to-bone repair. Our unique multidisciplinary approach, which involves the close collaboration between developmental biologists, bioengineers, materials scientists, and clinicians, and our strategy of ‘developmentally inspired tissue engineering’ can serve as a paradigm that can be applied to other areas of regenerative medicine.
We chose to focus solely on the hedgehog signaling pathway in this dissertation, but many other pathways are involved in tendon healing and thus deserve further research as well.
One of the grand challenges of tissue engineering is the integration of the many individual pathways and mechanisms involved in tissue formation and repair into a single unified framework. This is particularly difficult due to the many interactions between pathways and the inherent redundancy of biological systems. In addition to the biological aspects of tendon repair, the mechanical signals acting on the healing enthesis must also be considered. Our work has provided hints that a profound relationship between mechanical force, hedgehog signaling, and cell phenotype may underlie the formation, maintenance, and repair of fibrocartilage in tendons.
More work is certainly needed in this area, but the mechanobiology principles at work in tendon will likely be applicable to other musculoskeletal tissues as well.
145
Tendon tissue engineering is still an attractive alternative to traditional tendon repair strategies, but our work highlights some of the challenges of translating basic research into efficacious tissue engineering solutions with potential for commercialization. Targeting the Ihh pathway during tendon-to-bone healing is straightforward in principle, but difficult in practice.
We hope our experiences with Ihh-infused scaffolds and our suggestions for alternative approaches to modulate Ihh signaling in healing tendons, such as cell- based therapies, will prove instructive to future researchers and serve as a starting point for additional studies. Furthermore, we believe that investigation of the hedgehog pathway as a therapeutic target should not be restricted to cases of acute tendon-to-bone healing following tendon rupture, but should be expanded to include chronic tendon disorders such as degenerative tendinopathy. The opportunities for continued study of hedgehog signaling in tendon development, homeostasis, disease, and repair are vast, and we hope those who continue to pursue this line of inquiry will be met with positive results that will eventually lead to improvements in the treatment of tendon disorders.
146
Bibliography
1. Delale F. 1984. Stress singularities in bonded anisotropic materials. Int J Solids Structures 20:
31-40.
2. Mittelstedt C, Becker W. 2007. Free-edge effects in composite laminates. Applied Mechanics
Reviews 60: 217-245.
3. Liu Y, Birman V, Chen C, Thomopoulos S, Genin GM. 2011. Mechanisms of bimaterial attachment at the interface of tendon to bone. J Eng Mater Technol 133: 011006.
4. United States Bone and Joint Initiative. 2011. The burden of musculoskeletal diseases in the united states, 2nd ed. Rosemont, IL: American Academy of Orthopaedic Surgeons; 129 p.
5. Yamamoto A, Takagishi K, Osawa T, et al. 2010. Prevalence and risk factors of a rotator cuff tear in the general population. J Shoulder Elbow Surg 19: 116-120.
6. Reilly P, Macleod I, Macfarlane R, Windley J, Emery RJ. 2006. Dead men and radiologists don't lie: A review of cadaveric and radiological studies of rotator cuff tear prevalence. Ann R
Coll Surg Engl 88: 116-121.
7. Randelli P, Spennacchio P, Ragone V, et al. 2012. Complications associated with arthroscopic rotator cuff repair: A literature review. Musculoskelet Surg 96: 9-16.
8. Charousset C, Bellaiche L, Kalra K, Petrover D. 2010. Arthroscopic repair of full-thickness rotator cuff tears: Is there tendon healing in patients aged 65 years or older? Arthroscopy 26:
302-309.
147
9. Dy CJ, Hernandez-Soria A, Ma Y, Roberts TR, Daluiski A. 2012. Complications after flexor tendon repair: A systematic review and meta-analysis. J Hand Surg Am 37: 543-551.e1.
10. Maffulli N, Wong J. 2003. Rupture of the achilles and patellar tendons. Clin Sports Med 22:
761-776.
11. Reddy SS, Pedowitz DI, Parekh SG, Omar IM, Wapner KL. 2009. Surgical treatment for chronic disease and disorders of the achilles tendon. J Am Acad Orthop Surg 17: 3-14.
12. Breidenbach AP, Gilday SD, Lalley AL, et al. 2014. Functional tissue engineering of tendon:
Establishing biological success criteria for improving tendon repair. J Biomech 47: 1941-1948.
13. Rodrigues MT, Reis RL, Gomes ME. 2013. Engineering tendon and ligament tissues: Present developments towards successful clinical products. J Tissue Eng Regen Med 7: 673-686.
14. Butler DL, Goldstein SA, Guilak F. 2000. Functional tissue engineering: The role of biomechanics. J Biomech Eng 122: 570-575.
15. Guilak F, Butler DL, Goldstein SA, Mooney D editors. 2003. Functional tissue engineering,
New York: Springer; 426 p.
16. Benjamin M, Kaiser E, Milz S. 2008. Structure-function relationships in tendons: A review. J
Anat 212: 211-228.
17. Kjaer M. 2004. Role of extracellular matrix in adaptation of tendon and skeletal muscle to mechanical loading. Physiol Rev 84: 649-698.
148
18. Rettig AC, Liotta FJ, Klootwyk TE, Porter DA, Mieling P. 2005. Potential risk of rerupture in primary achilles tendon repair in athletes younger than 30 years of age. Am J Sports Med 33:
119-123.
19. Amin NH, Old AB, Tabb LP, et al. 2013. Performance outcomes after repair of complete achilles tendon ruptures in national basketball association players. Am J Sports Med 41: 1864-
1868.
20. Kato H, Minami A, Suenaga N, Iwasaki N, Kimura T. 2002. Long-term results after primary repairs of zone 2 flexor tendon lacerations in children younger than age 6 years. J Pediatr Orthop
22: 732-735.
21. Silfverskiold KL, May EJ, Oden A. 1993. Factors affecting results after flexor tendon repair in zone II: A multivariate prospective analysis. J Hand Surg Am 18: 654-662.
22. Butler DL, Juncosa-Melvin N, Boivin GP, et al. 2008. Functional tissue engineering for tendon repair: A multidisciplinary strategy using mesenchymal stem cells, bioscaffolds, and mechanical stimulation. J Orthop Res 26: 1-9.
23. Juncosa N, West JR, Galloway MT, Boivin GP, Butler DL. 2003. In vivo forces used to develop design parameters for tissue engineered implants for rabbit patellar tendon repair. J
Biomech 36: 483-488.
24. Malaviya P, Butler DL, Korvick DL, Proch FS. 1998. In vivo tendon forces correlate with activity level and remain bounded: Evidence in a rabbit flexor tendon model. J Biomech 31:
1043-1049.
149
25. West JR, Juncosa N, Galloway MT, Boivin GP, Butler DL. 2004. Characterization of in vivo achilles tendon forces in rabbits during treadmill locomotion at varying speeds and inclinations. J
Biomech 37: 1647-1653.
26. Korvick DL, Cummings JF, Grood ES, et al. 1996. The use of an implantable force transducer to measure patellar tendon forces in goats. J Biomech 29: 557-561.
27. Benjamin M, Ralphs JR. 1998. Fibrocartilage in tendons and ligaments--an adaptation to compressive load. J Anat 193 ( Pt 4): 481-494.
28. Thomopoulos S, Williams GR, Gimbel JA, Favata M, Soslowsky LJ. 2003. Variation of biomechanical, structural, and compositional properties along the tendon to bone insertion site. J
Orthop Res 21: 413-419.
29. Arruda EM, Calve S, Dennis RG, Mundy K, Baar K. 2006. Regional variation of tibialis anterior tendon mechanics is lost following denervation. J Appl Physiol (1985) 101: 1113-1117.
30. Butler DL, Sheh MY, Stouffer DC, Samaranayake VA, Levy MS. 1990. Surface strain variation in human patellar tendon and knee cruciate ligaments. J Biomech Eng 112: 38-45.
31. Rigozzi S, Muller R, Snedeker JG. 2009. Local strain measurement reveals a varied regional dependence of tensile tendon mechanics on glycosaminoglycan content. J Biomech 42: 1547-
1552.
32. Brent AE, Schweitzer R, Tabin CJ. 2003. A somitic compartment of tendon progenitors. Cell
113: 235-248.
150
33. Murchison ND, Price BA, Conner DA, et al. 2007. Regulation of tendon differentiation by scleraxis distinguishes force-transmitting tendons from muscle-anchoring tendons. Development
134: 2697-2708.
34. Pryce BA, Watson SS, Murchison ND, et al. 2009. Recruitment and maintenance of tendon progenitors by TGFbeta signaling are essential for tendon formation. Development 136: 1351-
1361.
35. Schweitzer R, Chyung JH, Murtaugh LC, et al. 2001. Analysis of the tendon cell fate using scleraxis, a specific marker for tendons and ligaments. Development 128: 3855-3866.
36. Ito Y, Toriuchi N, Yoshitaka T, et al. 2010. The mohawk homeobox gene is a critical regulator of tendon differentiation. Proc Natl Acad Sci U S A 107: 10538-10542.
37. Liu W, Watson SS, Lan Y, et al. 2010. The atypical homeodomain transcription factor mohawk controls tendon morphogenesis. Mol Cell Biol 30: 4797-4807.
38. Guerquin MJ, Charvet B, Nourissat G, et al. 2013. Transcription factor EGR1 directs tendon differentiation and promotes tendon repair. J Clin Invest 123: 3564-3576.
39. Lejard V, Blais F, Guerquin MJ, et al. 2011. EGR1 and EGR2 involvement in vertebrate tendon differentiation. J Biol Chem 286: 5855-5867.
40. Blitz E, Sharir A, Akiyama H, Zelzer E. 2013. Tendon-bone attachment unit is formed modularly by a distinct pool of scx- and Sox9-positive progenitors. Development 140: 2680-
2690.
151
41. Sugimoto Y, Takimoto A, Akiyama H, et al. 2013. Scx+/Sox9+ progenitors contribute to the establishment of the junction between cartilage and tendon/ligament. Development 140: 2280-
2288.
42. Blitz E, Viukov S, Sharir A, et al. 2009. Bone ridge patterning during musculoskeletal assembly is mediated through SCX regulation of Bmp4 at the tendon-skeleton junction. Dev Cell
17: 861-873.
43. Liu CF, Breidenbach A, Aschbacher-Smith L, Butler D, Wylie C. 2013. A role for hedgehog signaling in the differentiation of the insertion site of the patellar tendon in the mouse. PLoS One
8: e65411.
44. Lejard V, Brideau G, Blais F, et al. 2007. Scleraxis and NFATc regulate the expression of the pro-alpha1(I) collagen gene in tendon fibroblasts. J Biol Chem 282: 17665-17675.
45. Edom-Vovard F, Schuler B, Bonnin MA, Teillet MA, Duprez D. 2002. Fgf4 positively regulates scleraxis and tenascin expression in chick limb tendons. Dev Biol 247: 351-366.
46. Shukunami C, Takimoto A, Oro M, Hiraki Y. 2006. Scleraxis positively regulates the expression of tenomodulin, a differentiation marker of tenocytes. Dev Biol 298: 234-247.
47. Dyment NA, Liu CF, Kazemi N, et al. 2013. The paratenon contributes to scleraxis- expressing cells during patellar tendon healing. PLoS One 8: e59944.
48. Eliasson P, Andersson T, Aspenberg P. 2009. Rat achilles tendon healing: Mechanical loading and gene expression. J Appl Physiol (1985) 107: 399-407.
152
49. Scott A, Sampaio A, Abraham T, Duronio C, Underhill TM. 2011. Scleraxis expression is coordinately regulated in a murine model of patellar tendon injury. J Orthop Res 29: 289-296.
50. Dyment NA, Kazemi N, Aschbacher-Smith LE, et al. 2012. The relationships among spatiotemporal collagen gene expression, histology, and biomechanics following full-length injury in the murine patellar tendon. J Orthop Res 30: 28-36.
51. Bagchi RA, Czubryt MP. 2012. Synergistic roles of scleraxis and smads in the regulation of collagen 1alpha2 gene expression. Biochim Biophys Acta 1823: 1936-1944.
52. Espira L, Lamoureux L, Jones SC, et al. 2009. The basic helix-loop-helix transcription factor scleraxis regulates fibroblast collagen synthesis. J Mol Cell Cardiol 47: 188-195.
53. Mendias CL, Gumucio JP, Davis ME, et al. 2012. Transforming growth factor-beta induces skeletal muscle atrophy and fibrosis through the induction of atrogin-1 and scleraxis. Muscle
Nerve 45: 55-59.
54. von der Mark K. 1981. Localization of collagen types in tissues. Int Rev Connect Tissue Res
9: 265-324.
55. Lapiere CM, Nusgens B, Pierard GE. 1977. Interaction between collagen type I and type III in conditioning bundles organization. Connect Tissue Res 5: 21-29.
56. Dressler MR, Butler DL, Wenstrup R, et al. 2002. A potential mechanism for age-related declines in patellar tendon biomechanics. J Orthop Res 20: 1315-1322.
153
57. Wenstrup RJ, Smith SM, Florer JB, et al. 2011. Regulation of collagen fibril nucleation and initial fibril assembly involves coordinate interactions with collagens V and XI in developing tendon. J Biol Chem 286: 20455-20465.
58. Izu Y, Ansorge HL, Zhang G, et al. 2011. Dysfunctional tendon collagen fibrillogenesis in collagen VI null mice. Matrix Biol 30: 53-61.
59. Ritty TM, Roth R, Heuser JE. 2003. Tendon cell array isolation reveals a previously unknown fibrillin-2-containing macromolecular assembly. Structure 11: 1179-1188.
60. Fujioka H, Wang GJ, Mizuno K, Balian G, Hurwitz SR. 1997. Changes in the expression of type-X collagen in the fibrocartilage of rat achilles tendon attachment during development. J
Orthop Res 15: 675-681.
61. Zhang G, Ezura Y, Chervoneva I, et al. 2006. Decorin regulates assembly of collagen fibrils and acquisition of biomechanical properties during tendon development. J Cell Biochem 98:
1436-1449.
62. Ezura Y, Chakravarti S, Oldberg A, Chervoneva I, Birk DE. 2000. Differential expression of lumican and fibromodulin regulate collagen fibrillogenesis in developing mouse tendons. J Cell
Biol 151: 779-788.
63. Galloway MT, Lalley AL, Shearn JT. 2013. The role of mechanical loading in tendon development, maintenance, injury, and repair. J Bone Joint Surg Am 95: 1620-1628.
64. Wren TA, Yerby SA, Beaupre GS, Carter DR. 2001. Mechanical properties of the human achilles tendon. Clin Biomech (Bristol, Avon) 16: 245-251.
154
65. Blissett AR, Garbellini D, Calomeni EP, et al. 2009. Regulation of collagen fibrillogenesis by cell-surface expression of kinase dead DDR2. J Mol Biol 385: 902-911.
66. Zhang G, Young BB, Ezura Y, et al. 2005. Development of tendon structure and function:
Regulation of collagen fibrillogenesis. J Musculoskelet Neuronal Interact 5: 5-21.
67. Kadler KE, Hill A, Canty-Laird EG. 2008. Collagen fibrillogenesis: Fibronectin, integrins, and minor collagens as organizers and nucleators. Curr Opin Cell Biol 20: 495-501.
68. Goh KL, Holmes DF, Lu Y, et al. 2012. Bimodal collagen fibril diameter distributions direct age-related variations in tendon resilience and resistance to rupture. J Appl Physiol (1985) 113:
878-888.
69. Parry DA, Barnes GR, Craig AS. 1978. A comparison of the size distribution of collagen fibrils in connective tissues as a function of age and a possible relation between fibril size distribution and mechanical properties. Proc R Soc Lond B Biol Sci 203: 305-321.
70. Lechner BE, Lim JH, Mercado ML, Fallon JR. 2006. Developmental regulation of biglycan expression in muscle and tendon. Muscle Nerve 34: 347-355.
71. Hedlund H, Mengarelli-Widholm S, Heinegard D, Reinholt FP, Svensson O. 1994.
Fibromodulin distribution and association with collagen. Matrix Biol 14: 227-232.
72. Oryan A, Moshiri A, Meimandi-Parizi AH. 2012. Short and long terms healing of the experimentally transverse sectioned tendon in rabbits. Sports Med Arthrosc Rehabil Ther
Technol 4: 14.
155
73. Rigozzi S, Muller R, Snedeker JG. 2010. Collagen fibril morphology and mechanical properties of the achilles tendon in two inbred mouse strains. J Anat 216: 724-731.
74. Shino K, Oakes BW, Horibe S, Nakata K, Nakamura N. 1995. Collagen fibril populations in human anterior cruciate ligament allografts. electron microscopic analysis. Am J Sports Med 23:
203-8; discussion 209.
75. Battaglia TC, Clark RT, Chhabra A, et al. 2003. Ultrastructural determinants of murine achilles tendon strength during healing. Connect Tissue Res 44: 218-224.
76. Frank C, McDonald D, Bray D, et al. 1992. Collagen fibril diameters in the healing adult rabbit medial collateral ligament. Connect Tissue Res 27: 251-263.
77. Khanna A, Friel M, Gougoulias N, Longo UG, Maffulli N. 2009. Prevention of adhesions in surgery of the flexor tendons of the hand: What is the evidence? Br Med Bull 90: 85-109.
78. Wong JK, Lui YH, Kapacee Z, et al. 2009. The cellular biology of flexor tendon adhesion formation: An old problem in a new paradigm. Am J Pathol 175: 1938-1951.
79. Thomopoulos S, Marquez JP, Weinberger B, Birman V, Genin GM. 2006. Collagen fiber orientation at the tendon to bone insertion and its influence on stress concentrations. J Biomech
39: 1842-1851.
80. Genin GM, Kent A, Birman V, et al. 2009. Functional grading of mineral and collagen in the attachment of tendon to bone. Biophys J 97: 976-985.
156
81. Ralphs JR, Benjamin M, Waggett AD, et al. 1998. Regional differences in cell shape and gap junction expression in rat achilles tendon: Relation to fibrocartilage differentiation. J Anat 193 (
Pt 2): 215-222.
82. de Palma L, Marinelli M, Meme L, Pavan M. 2004. Immunohistochemistry of the enthesis organ of the human achilles tendon. Foot Ankle Int 25: 414-418.
83. Waggett AD, Ralphs JR, Kwan AP, Woodnutt D, Benjamin M. 1998. Characterization of collagens and proteoglycans at the insertion of the human achilles tendon. Matrix Biol 16: 457-
470.
84. Wang IE, Mitroo S, Chen FH, Lu HH, Doty SB. 2006. Age-dependent changes in matrix composition and organization at the ligament-to-bone insertion. J Orthop Res 24: 1745-1755.
85. Lake SP, Miller KS, Elliott DM, Soslowsky LJ. 2010. Tensile properties and fiber alignment of human supraspinatus tendon in the transverse direction demonstrate inhomogeneity, nonlinearity, and regional isotropy. J Biomech 43: 727-732.
86. Miller KS, Connizzo BK, Feeney E, Soslowsky LJ. 2012. Characterizing local collagen fiber re-alignment and crimp behavior throughout mechanical testing in a mature mouse supraspinatus tendon model. J Biomech 45: 2061-2065.
87. Stouffer DC, Butler DL, Hosny D. 1985. The relationship between crimp pattern and mechanical response of human patellar tendon-bone units. J Biomech Eng 107: 158-165.
157
88. Gilday SD, Casstevens EC, Kenter K, Shearn JT, Butler DL. 2014. Murine patellar tendon biomechanical properties and regional strain patterns during natural tendon-to-bone healing after acute injury. J Biomech 47: 2035-2042.
89. Liu CF, Aschbacher-Smith L, Barthelery NJ, et al. 2012. Spatial and temporal expression of molecular markers and cell signals during normal development of the mouse patellar tendon.
Tissue Eng Part A 18: 598-608.
90. Schwartz AG, Lipner JH, Pasteris JD, Genin GM, Thomopoulos S. 2013. Muscle loading is necessary for the formation of a functional tendon enthesis. Bone 55: 44-51.
91. Galatz LM, Sandell LJ, Rothermich SY, et al. 2006. Characteristics of the rat supraspinatus tendon during tendon-to-bone healing after acute injury. J Orthop Res 24: 541-550.
92. Kinneberg KR, Galloway MT, Butler DL, Shearn JT. 2011. Effect of implanting a soft tissue autograft in a central-third patellar tendon defect: Biomechanical and histological comparisons. J
Biomech Eng 133: 091002.
93. Rodeo SA, Arnoczky SP, Torzilli PA, Hidaka C, Warren RF. 1993. Tendon-healing in a bone tunnel. A biomechanical and histological study in the dog. J Bone Joint Surg Am 75: 1795-1803.
94. Newsham-West R, Nicholson H, Walton M, Milburn P. 2007. Long-term morphology of a healing bone-tendon interface: A histological observation in the sheep model. J Anat 210: 318-
327.
158
95. Functional Tissue Engineering Conference Group. 2008. Evaluation criteria for musculoskeletal and craniofacial tissue engineering constructs: A conference report. Tissue Eng
Part A 14: 2089-2104.
96. Nawroth JC, Parker KK. 2013. Design standards for engineered tissues. Biotechnol Adv 31:
632-637.
97. Birch HL. 2007. Tendon matrix composition and turnover in relation to functional requirements. Int J Exp Pathol 88: 241-248.
98. Mackey AL, Heinemeier KM, Koskinen SO, Kjaer M. 2008. Dynamic adaptation of tendon and muscle connective tissue to mechanical loading. Connect Tissue Res 49: 165-168.
99. Klatte-Schulz F, Pauly S, Scheibel M, et al. 2012. Influence of age on the cell biological characteristics and the stimulation potential of male human tenocyte-like cells. Eur Cell Mater
24: 74-89.
100. Yu TY, Pang JH, Wu KP, et al. 2013. Aging is associated with increased activities of matrix metalloproteinase-2 and -9 in tenocytes. BMC Musculoskelet Disord 14: 2-2474-14-2.
101. Magnusson SP, Hansen M, Langberg H, et al. 2007. The adaptability of tendon to loading differs in men and women. Int J Exp Pathol 88: 237-240.
102. Miller BF, Hansen M, Olesen JL, et al. 2007. Tendon collagen synthesis at rest and after exercise in women. J Appl Physiol (1985) 102: 541-546.
159
103. Abate M, Schiavone C, Salini V, Andia I. 2013. Occurrence of tendon pathologies in metabolic disorders. Rheumatology (Oxford) 52: 599-608.
104. de Oliveira RR, Lemos A, de Castro Silveira PV, da Silva RJ, de Moraes SR. 2011.
Alterations of tendons in patients with diabetes mellitus: A systematic review. Diabet Med 28:
886-895.
105. Obst SJ, Barrett RS, Newsham-West R. 2013. Immediate effect of exercise on achilles tendon properties: Systematic review. Med Sci Sports Exerc 45: 1534-1544.
106. Reeves ND. 2006. Adaptation of the tendon to mechanical usage. J Musculoskelet Neuronal
Interact 6: 174-180.
107. Clark LD, Clark RK, Heber-Katz E. 1998. A new murine model for mammalian wound repair and regeneration. Clin Immunol Immunopathol 88: 35-45.
108. Lalley AL, Dyment NA, Kazemi N, et al. 2015. Improved biomechanical and biological outcomes in the MRL/MpJ murine strain following a full-length patellar tendon injury. J Orthop
Res 33: 1693-1703.
109. Benjamin M, Kumai T, Milz S, et al. 2002. The skeletal attachment of tendons--tendon
"entheses". Comp Biochem Physiol A Mol Integr Physiol 133: 931-945.
110. Thomopoulos S, Williams GR, Gimbel JA, Favata M, Soslowsky LJ. 2003. Variation of biomechanical, structural, and compositional properties along the tendon to bone insertion site. J
Orthop Res 21: 413-419.
160
111. Thomopoulos S, Genin GM, Galatz LM. 2010. The development and morphogenesis of the tendon-to-bone insertion - what development can teach us about healing -. J Musculoskelet
Neuronal Interact 10: 35-45.
112. Thomopoulos S. 2011. The role of mechanobiology in the attachment of tendon to bone.
IBMS BoneKEy 8: 271-285.
113. Astrom M, Rausing A. 1995. Chronic achilles tendinopathy. A survey of surgical and histopathologic findings. Clin Orthop Relat Res (316): 151-164.
114. Streit JJ, Shishani Y, Rodgers M, Gobezie R. 2015. Tendinopathy of the long head of the biceps tendon: Histopathologic analysis of the extra-articular biceps tendon and tenosynovium.
Open Access J Sports Med 6: 63-70.
115. Thomopoulos S, Parks WC, Rifkin DB, Derwin KA. 2015. Mechanisms of tendon injury and repair. J Orthop Res 33: 832-839.
116. Sharma P, Maffulli N. 2006. Biology of tendon injury: Healing, modeling and remodeling. J
Musculoskelet Neuronal Interact 6: 181-190.
117. Koh TJ, DiPietro LA. 2011. Inflammation and wound healing: The role of the macrophage.
Expert Rev Mol Med 13: e23.
118. Chan KM, Fu SC, Wong YP, et al. 2008. Expression of transforming growth factor beta isoforms and their roles in tendon healing. Wound Repair Regen 16: 399-407.
161
119. St Pierre P, Olson EJ, Elliott JJ, et al. 1995. Tendon-healing to cortical bone compared with healing to a cancellous trough. A biomechanical and histological evaluation in goats. J Bone
Joint Surg Am 77: 1858-1866.
120. Aoki M, Oguma H, Fukushima S, et al. 2001. Fibrous connection to bone after immediate repair of the canine infraspinatus: The most effective bony surface for tendon attachment. J
Shoulder Elbow Surg 10: 123-128.
121. Silva MJ, Thomopoulos S, Kusano N, et al. 2006. Early healing of flexor tendon insertion site injuries: Tunnel repair is mechanically and histologically inferior to surface repair in a canine model. J Orthop Res 24: 990-1000.
122. Liu SH, Panossian V, al-Shaikh R, et al. 1997. Morphology and matrix composition during early tendon to bone healing. Clin Orthop Relat Res (339): 253-260.
123. Fujioka H, Thakur R, Wang GJ, et al. 1998. Comparison of surgically attached and non- attached repair of the rat achilles tendon-bone interface. cellular organization and type X collagen expression. Connect Tissue Res 37: 205-218.
124. Galatz LM, Rothermich SY, Zaegel M, et al. 2005. Delayed repair of tendon to bone injuries leads to decreased biomechanical properties and bone loss. J Orthop Res 23: 1441-1447.
125. Thomopoulos S, Hattersley G, Rosen V, et al. 2002. The localized expression of extracellular matrix components in healing tendon insertion sites: An in situ hybridization study.
J Orthop Res 20: 454-463.
162
126. Thomopoulos S, Williams GR, Soslowsky LJ. 2003. Tendon to bone healing: Differences in biomechanical, structural, and compositional properties due to a range of activity levels. J
Biomech Eng 125: 106-113.
127. Rodeo SA. 2007. Biologic augmentation of rotator cuff tendon repair. Journal of Shoulder and Elbow Surgery 16: S191-S197.
128. Longo UG, Lamberti A, Maffulli N, Denaro V. 2010. Tendon augmentation grafts: A systematic review. British Medical Bulletin 94: 165-188.
129. Smith L, Xia Y, Galatz LM, Genin GM, Thomopoulos S. 2012. Tissue-engineering strategies for the tendon/ligament-to-bone insertion. Connect Tissue Res 53: 95-105.
130. Oliva F, Gatti S, Porcellini G, Forsyth NR, Maffulli N. 2012. Growth factors and tendon healing. Med Sport Sci 57: 53-64.
131. Gulotta LV, Rodeo SA. 2009. Growth factors for rotator cuff repair. Clin Sports Med 28:
13-23.
132. Bedi A, Maak T, Walsh C, et al. 2012. Cytokines in rotator cuff degeneration and repair. J
Shoulder Elbow Surg 21: 218-227.
133. Nixon AJ, Watts AE, Schnabel LV. 2012. Cell- and gene-based approaches to tendon regeneration. J Shoulder Elbow Surg 21: 278-294.
163
134. Liu CF, Aschbacher-Smith L, Barthelery NJ, et al. 2011. What we should know before using tissue engineering techniques to repair injured tendons: A developmental biology perspective. Tissue Eng Part B Rev 17: 165-176.
135. Ingham PW, Nakano Y, Seger C. 2011. Mechanisms and functions of hedgehog signalling across the metazoa. Nat Rev Genet 12: 393-406.
136. Mullor JL, Sanchez P, Ruiz i Altaba A. 2002. Pathways and consequences: Hedgehog signaling in human disease. Trends Cell Biol 12: 562-569.
137. Ruiz i Altaba A. 1999. Gli proteins encode context-dependent positive and negative functions: Implications for development and disease. Development 126: 3205-3216.
138. Matise MP, Joyner AL. 1999. Gli genes in development and cancer. Oncogene 18: 7852-
7859.
139. Bai CB, Auerbach W, Lee JS, Stephen D, Joyner AL. 2002. Gli2, but not Gli1, is required for initial shh signaling and ectopic activation of the shh pathway. Development 129: 4753-4761.
140. Weitzman JB. 2002. Agonizing hedgehog. J Biol 1: 7.
141. Kronenberg HM. 2003. Developmental regulation of the growth plate. Nature 423: 332-336.
142. Lai LP, Mitchell J. 2005. Indian hedgehog: Its roles and regulation in endochondral bone development. J Cell Biochem 96: 1163-1173.
143. Pan A, Chang L, Nguyen A, James AW. 2013. A review of hedgehog signaling in cranial bone development. Front Physiol 4: 61.
164
144. Ma G, Yu J, Xiao Y, et al. 2011. Indian hedgehog mutations causing brachydactyly type A1 impair hedgehog signal transduction at multiple levels. Cell Res 21: 1343-1357.
145. Gao B, Guo J, She C, et al. 2001. Mutations in IHH, encoding indian hedgehog, cause brachydactyly type A-1. Nat Genet 28: 386-388.
146. Hellemans J, Coucke PJ, Giedion A, et al. 2003. Homozygous mutations in IHH cause acrocapitofemoral dysplasia, an autosomal recessive disorder with cone-shaped epiphyses in hands and hips. Am J Hum Genet 72: 1040-1046.
147. St-Jacques B, Hammerschmidt M, McMahon AP. 1999. Indian hedgehog signaling regulates proliferation and differentiation of chondrocytes and is essential for bone formation.
Genes Dev 13: 2072-2086.
148. Vortkamp A, Lee K, Lanske B, et al. 1996. Regulation of rate of cartilage differentiation by indian hedgehog and PTH-related protein. Science 273: 613-622.
149. Lanske B, Karaplis AC, Lee K, et al. 1996. PTH/PTHrP receptor in early development and indian hedgehog-regulated bone growth. Science 273: 663-666.
150. Long F, Zhang XM, Karp S, Yang Y, McMahon AP. 2001. Genetic manipulation of hedgehog signaling in the endochondral skeleton reveals a direct role in the regulation of chondrocyte proliferation. Development 128: 5099-5108.
151. Kobayashi T, Soegiarto DW, Yang Y, et al. 2005. Indian hedgehog stimulates periarticular chondrocyte differentiation to regulate growth plate length independently of PTHrP. J Clin
Invest 115: 1734-1742.
165
152. Koziel L, Wuelling M, Schneider S, Vortkamp A. 2005. Gli3 acts as a repressor downstream of ihh in regulating two distinct steps of chondrocyte differentiation. Development
132: 5249-5260.
153. Long F, Chung UI, Ohba S, et al. 2004. Ihh signaling is directly required for the osteoblast lineage in the endochondral skeleton. Development 131: 1309-1318.
154. Shimoyama A, Wada M, Ikeda F, et al. 2007. Ihh/Gli2 signaling promotes osteoblast differentiation by regulating Runx2 expression and function. Mol Biol Cell 18: 2411-2418.
155. Kesper DA, Didt-Koziel L, Vortkamp A. 2010. Gli2 activator function in preosteoblasts is sufficient to mediate ihh-dependent osteoblast differentiation, whereas the repressor function of
Gli2 is dispensable for endochondral ossification. Dev Dyn 239: 1818-1826.
156. Nakamura T, Aikawa T, Iwamoto-Enomoto M, et al. 1997. Induction of osteogenic differentiation by hedgehog proteins. Biochem Biophys Res Commun 237: 465-469.
157. Enomoto-Iwamoto M, Nakamura T, Aikawa T, et al. 2000. Hedgehog proteins stimulate chondrogenic cell differentiation and cartilage formation. J Bone Miner Res 15: 1659-1668.
158. Lenton K, James AW, Manu A, et al. 2011. Indian hedgehog positively regulates calvarial ossification and modulates bone morphogenetic protein signaling. Genesis 49: 784-796.
159. Hu H, Hilton MJ, Tu X, et al. 2005. Sequential roles of hedgehog and wnt signaling in osteoblast development. Development 132: 49-60.
166
160. Mak KK, Chen MH, Day TF, Chuang PT, Yang Y. 2006. Wnt/beta-catenin signaling interacts differentially with ihh signaling in controlling endochondral bone and synovial joint formation. Development 133: 3695-3707.
161. Hilton MJ, Tu X, Cook J, Hu H, Long F. 2005. Ihh controls cartilage development by antagonizing Gli3, but requires additional effectors to regulate osteoblast and vascular development. Development 132: 4339-4351.
162. Chung UI, Schipani E, McMahon AP, Kronenberg HM. 2001. Indian hedgehog couples chondrogenesis to osteogenesis in endochondral bone development. J Clin Invest 107: 295-304.
163. Karp SJ, Schipani E, St-Jacques B, et al. 2000. Indian hedgehog coordinates endochondral bone growth and morphogenesis via parathyroid hormone related-protein-dependent and - independent pathways. Development 127: 543-548.
164. Kobayashi T, Chung UI, Schipani E, et al. 2002. PTHrP and indian hedgehog control differentiation of growth plate chondrocytes at multiple steps. Development 129: 2977-2986.
165. Murakami S, Noda M. 2000. Expression of indian hedgehog during fracture healing in adult rat femora. Calcif Tissue Int 66: 272-276.
166. Le AX, Miclau T, Hu D, Helms JA. 2001. Molecular aspects of healing in stabilized and non-stabilized fractures. J Orthop Res 19: 78-84.
167. Meyer RA,Jr, Meyer MH, Tenholder M, et al. 2003. Gene expression in older rats with delayed union of femoral fractures. J Bone Joint Surg Am 85-A: 1243-1254.
167
168. Desai BJ, Meyer MH, Porter S, Kellam JF, Meyer RA,Jr. 2003. The effect of age on gene expression in adult and juvenile rats following femoral fracture. J Orthop Trauma 17: 689-698.
169. Ferguson C, Alpern E, Miclau T, Helms JA. 1999. Does adult fracture repair recapitulate embryonic skeletal formation? Mech Dev 87: 57-66.
170. Vortkamp A, Pathi S, Peretti GM, et al. 1998. Recapitulation of signals regulating embryonic bone formation during postnatal growth and in fracture repair. Mech Dev 71: 65-76.
171. Schwartz AG, Long F, Thomopoulos S. 2015. Enthesis fibrocartilage cells originate from a population of hedgehog-responsive cells modulated by the loading environment. Development
142: 196-206.
172. Thomopoulos S, Kim HM, Rothermich SY, et al. 2007. Decreased muscle loading delays maturation of the tendon enthesis during postnatal development. J Orthop Res 25: 1154-1163.
173. Liu Y, Schwartz AG, Birman V, Thomopoulos S, Genin GM. 2014. Stress amplification during development of the tendon-to-bone attachment. Biomech Model Mechanobiol 13: 973-
983.
174. Benjamin M, Toumi H, Ralphs JR, et al. 2006. Where tendons and ligaments meet bone:
Attachment sites ('entheses') in relation to exercise and/or mechanical load. J Anat 208: 471-490.
175. Kim HM, Galatz LM, Patel N, Das R, Thomopoulos S. 2009. Recovery potential after postnatal shoulder paralysis. an animal model of neonatal brachial plexus palsy. J Bone Joint
Surg Am 91: 879-891.
168
176. Han X, Guo L, Wang F, Zhu Q, Yang L. 2014. Contribution of PTHrP to mechanical strain- induced fibrochondrogenic differentiation in entheses of achilles tendon of miniature pigs. J
Biomech 47: 2406-2414.
177. Chen J, Sorensen KP, Gupta T, et al. 2009. Altered functional loading causes differential effects in the subchondral bone and condylar cartilage in the temporomandibular joint from young mice. Osteoarthritis Cartilage 17: 354-361.
178. Ng TC, Chiu KW, Rabie AB, Hagg U. 2006. Repeated mechanical loading enhances the expression of indian hedgehog in condylar cartilage. Front Biosci 11: 943-948.
179. Shao YY, Wang L, Welter JF, Ballock RT. 2012. Primary cilia modulate ihh signal transduction in response to hydrostatic loading of growth plate chondrocytes. Bone 50: 79-84.
180. Wu Q, Zhang Y, Chen Q. 2001. Indian hedgehog is an essential component of mechanotransduction complex to stimulate chondrocyte proliferation. J Biol Chem 276: 35290-
35296.
181. Breidenbach AP, Aschbacher-Smith L, Lu Y, et al. 2015. Ablating hedgehog signaling in tenocytes during development impairs biomechanics and matrix organization of the adult murine patellar tendon enthesis. J Orthop Res 33: 1142-1151.
182. Clayton RA, Court-Brown CM. 2008. The epidemiology of musculoskeletal tendinous and ligamentous injuries. Injury 39: 1338-1344.
183. Shaw HM, Benjamin M. 2007. Structure-function relationships of entheses in relation to mechanical load and exercise. Scand J Med Sci Sports 17: 303-315.
169
184. Haraldsson BT, Aagaard P, Krogsgaard M, et al. 2005. Region-specific mechanical properties of the human patella tendon. J Appl Physiol 98: 1006-1012.
185. Huang CY, Wang VM, Pawluk RJ, et al. 2005. Inhomogeneous mechanical behavior of the human supraspinatus tendon under uniaxial loading. J Orthop Res 23: 924-930.
186. Wu JZ, Brumfield A, Miller GR, Metheny R, Cutlip RG. 2004. Comparison of mechanical properties of rat tibialis anterior tendon evaluated using two different approaches. Biomed Mater
Eng 14: 13-22.
187. Lieber RL, Leonard ME, Brown CG, Trestik CL. 1991. Frog semitendinosis tendon load- strain and stress-strain properties during passive loading. Am J Physiol 261: C86-92.
188. Doschak MR, Zernicke RF. 2005. Structure, function and adaptation of bone-tendon and bone-ligament complexes. J Musculoskelet Neuronal Interact 5: 35-40.
189. Malliaras P, Kamal B, Nowell A, et al. 2013. Patellar tendon adaptation in relation to load- intensity and contraction type. J Biomech 46: 1893-1899.
190. Butler DL, Grood ES, Noyes FR, Zernicke RF, Brackett K. 1984. Effects of structure and strain measurement technique on the material properties of young human tendons and fascia. J
Biomech 17: 579-596.
191. Schwartz AG, Pasteris JD, Genin GM, Daulton TL, Thomopoulos S. 2012. Mineral distributions at the developing tendon enthesis. PLoS One 7: e48630.
170
192. Galatz LM, Gerstenfeld L, Heber-Katz E, Rodeo SA. 2015. Tendon regeneration and scar formation: The concept of scarless healing. J Orthop Res 33: 823-831.
193. Ahn S, Joyner AL. 2004. Dynamic changes in the response of cells to positive hedgehog signaling during mouse limb patterning. Cell 118: 505-516.
194. Pryce BA, Brent AE, Murchison ND, Tabin CJ, Schweitzer R. 2007. Generation of transgenic tendon reporters, ScxGFP and ScxAP, using regulatory elements of the scleraxis gene.
Dev Dyn 236: 1677-1682.
195. Brownell I, Guevara E, Bai CB, Loomis CA, Joyner AL. 2011. Nerve-derived sonic hedgehog defines a niche for hair follicle stem cells capable of becoming epidermal stem cells.
Cell Stem Cell 8: 552-565.
196. Kawamoto T. 2003. Use of a new adhesive film for the preparation of multi-purpose fresh- frozen sections from hard tissues, whole-animals, insects and plants. Arch Histol Cytol 66: 123-
143.
197. Kawamoto T, Kawamoto K. 2014. Preparation of thin frozen sections from nonfixed and undecalcified hard tissues using kawamot's film method (2012). Methods Mol Biol 1130: 149-
164.
198. Chen JK, Taipale J, Cooper MK, Beachy PA. 2002. Inhibition of hedgehog signaling by direct binding of cyclopamine to smoothened. Genes Dev 16: 2743-2748.
199. Schindelin J, Arganda-Carreras I, Frise E, et al. 2012. Fiji: An open-source platform for biological-image analysis. Nat Methods 9: 676-682.
171
200. Wren TA, Beaupre GS, Carter DR. 2000. Mechanobiology of tendon adaptation to compressive loading through fibrocartilaginous metaplasia. J Rehabil Res Dev 37: 135-143.
201. Malaviya P, Butler DL, Boivin GP, et al. 2000. An in vivo model for load-modulated remodeling in the rabbit flexor tendon. J Orthop Res 18: 116-125.
202. Yang J, Andre P, Ye L, Yang YZ. 2015. The hedgehog signalling pathway in bone formation. Int J Oral Sci 7:73-9.
203. Dyment NA, Breidenbach AP, Schwartz AG, et al. 2015. The cellular and matrix dynamics of enthesis growth. Transactions of the annual meeting of the Orthopaedic Research Society
Paper No. 0086.
204. Ingham PW, McMahon AP. 2001. Hedgehog signaling in animal development: Paradigms and principles. Genes Dev 15: 3059-3087.
205. Petrova R, Joyner AL. 2014. Roles for hedgehog signaling in adult organ homeostasis and repair. Development 141: 3445-3457.
206. Dashty M. 2014. Hedgehog signaling pathway is linked with age-related diseases. J
Diabetes Metab 5: 2.
207. Pazin DE, Gamer LW, Rosen V. 2012. Gene signature of the developing meniscus.
Transactions of the annual meeting of the Orthopaedic Research Society Paper No. 0221.
208. Kim EJ, Cho SW, Shin JO, et al. 2013. Ihh and Runx2/Runx3 signaling interact to coordinate early chondrogenesis: A mouse model. PLoS One 8: e55296.
172
209. Amano K, Densmore M, Nishimura R, Lanske B. 2014. Indian hedgehog signaling regulates transcription and expression of collagen type X via Runx2/smads interactions. J Biol
Chem 289: 24898-24910.
210. Yoshida CA, Yamamoto H, Fujita T, et al. 2004. Runx2 and Runx3 are essential for chondrocyte maturation, and Runx2 regulates limb growth through induction of indian hedgehog. Genes Dev 18: 952-963.
211. Tu X, Joeng KS, Long F. 2012. Indian hedgehog requires additional effectors besides
Runx2 to induce osteoblast differentiation. Dev Biol 362: 76-82.
212. Iwasaki M, Le AX, Helms JA. 1997. Expression of indian hedgehog, bone morphogenetic protein 6 and gli during skeletal morphogenesis. Mech Dev 69: 197-202.
213. Bitgood MJ, McMahon AP. 1995. Hedgehog and bmp genes are coexpressed at many diverse sites of cell-cell interaction in the mouse embryo. Dev Biol 172: 126-138.
214. Grimsrud CD, Romano PR, D'Souza M, et al. 2001. BMP signaling stimulates chondrocyte maturation and the expression of indian hedgehog. J Orthop Res 19: 18-25.
215. Kameda T, Koike C, Saitoh K, Kuroiwa A, Iba H. 1999. Developmental patterning in chondrocytic cultures by morphogenic gradients: BMP induces expression of indian hedgehog and noggin. Genes Cells 4: 175-184.
216. Seki K, Hata A. 2004. Indian hedgehog gene is a target of the bone morphogenetic protein signaling pathway. J Biol Chem 279: 18544-18549.
173
217. Minina E, Wenzel HM, Kreschel C, et al. 2001. BMP and ihh/PTHrP signaling interact to coordinate chondrocyte proliferation and differentiation. Development 128: 4523-4534.
218. Minina E, Kreschel C, Naski MC, Ornitz DM, Vortkamp A. 2002. Interaction of FGF, ihh/pthlh, and BMP signaling integrates chondrocyte proliferation and hypertrophic differentiation. Dev Cell 3: 439-449.
219. Pathi S, Rutenberg JB, Johnson RL, Vortkamp A. 1999. Interaction of ihh and BMP/noggin signaling during cartilage differentiation. Dev Biol 209: 239-253.
220. Steinert AF, Weissenberger M, Kunz M, et al. 2012. Indian hedgehog gene transfer is a chondrogenic inducer of human mesenchymal stem cells. Arthritis Res Ther 14: R168.
221. Reichert JC, Schmalzl J, Prager P, et al. 2013. Synergistic effect of indian hedgehog and bone morphogenetic protein-2 gene transfer to increase the osteogenic potential of human mesenchymal stem cells. Stem Cell Res Ther 4: 105.
222. Sieker JT, Kunz M, Weissenberger M, et al. 2015. Direct bone morphogenetic protein 2 and indian hedgehog gene transfer for articular cartilage repair using bone marrow coagulates.
Osteoarthritis Cartilage 23: 433-442.
223. He J, Rosen CJ, Adams DJ, Kream BE. 2006. Postnatal growth and bone mass in mice with
IGF-I haploinsufficiency. Bone 38: 826-835.
224. Glatt V, Canalis E, Stadmeyer L, Bouxsein ML. 2007. Age-related changes in trabecular architecture differ in female and male C57BL/6J mice. J Bone Miner Res 22: 1197-1207.
174
225. Ralphs JR, Benjamin M, Thornett A. 1991. Cell and matrix biology of the suprapatella in the rat: A structural and immunocytochemical study of fibrocartilage in a tendon subject to compression. Anat Rec 231: 167-177.
226. Milz S, Putz R, Ralphs JR, Benjamin M. 1999. Fibrocartilage in the extensor tendons of the human metacarpophalangeal joints. Anat Rec 256: 139-145.
227. Koob TJ, Clark PE, Hernandez DJ, Thurmond FA, Vogel KG. 1992. Compression loading in vitro regulates proteoglycan synthesis by tendon fibrocartilage. Arch Biochem Biophys 298:
303-312.
228. Benjamin M, Qin S, Ralphs JR. 1995. Fibrocartilage associated with human tendons and their pulleys. J Anat 187 ( Pt 3): 625-633.
229. Wang Y, Wang J, Bai D, et al. 2013. Cell proliferation is promoted by compressive stress during early stage of chondrogenic differentiation of rat BMSCs. J Cell Physiol 228: 1935-1942.
230. Nowlan NC, Prendergast PJ, Murphy P. 2008. Identification of mechanosensitive genes during embryonic bone formation. PLoS Comput Biol 4: e1000250.
231. Morrow D, Sweeney C, Birney YA, et al. 2007. Biomechanical regulation of hedgehog signaling in vascular smooth muscle cells in vitro and in vivo. Am J Physiol Cell Physiol 292:
C488-96.
232. Tang GH, Rabie AB, Hagg U. 2004. Indian hedgehog: A mechanotransduction mediator in condylar cartilage. J Dent Res 83: 434-438.
175
233. Burssens A, Forsyth R, Bongaerts W, et al. 2013. Arguments for an increasing differentiation towards fibrocartilaginous components in midportion achilles tendinopathy. Knee
Surg Sports Traumatol Arthrosc 21: 1459-1467.
234. Archambault JM, Jelinsky SA, Lake SP, et al. 2007. Rat supraspinatus tendon expresses cartilage markers with overuse. J Orthop Res 25: 617-624.
235. Gigante A, Marinelli M, Chillemi C, Greco F. 2004. Fibrous cartilage in the rotator cuff: A pathogenetic mechanism of tendon tear? J Shoulder Elbow Surg 13: 328-332.
236. Bi Y, Ehirchiou D, Kilts TM, et al. 2007. Identification of tendon stem/progenitor cells and the role of the extracellular matrix in their niche. Nat Med 13: 1219-1227.
237. Praemer A, Furner S, Rice DP. 1999. Musculoskeletal conditions in the united states,
Rosemont, IL: American Academy of Orthopaedic Surgeons; 182 p.
238. Liang D, Hsiao BS, Chu B. 2007. Functional electrospun nanofibrous scaffolds for biomedical applications. Adv Drug Deliv Rev 59: 1392-1412.
239. Hadjiargyrou M, Chiu JB. 2008. Enhanced composite electrospun nanofiber scaffolds for use in drug delivery. Expert Opin Drug Deliv 5: 1093-1106.
240. Dash TK, Konkimalla VB. 2012. Poly-E-caprolactone based formulations for drug delivery and tissue engineering: A review. J Control Release 158: 15-33.
176
241. Xiao Y, Yuan M, Zhang J, Yan J, Lang M. 2014. Functional poly(epsilon-caprolactone) based materials: Preparation, self-assembly and application in drug delivery. Curr Top Med
Chem 14: 781-818.
242. Idaszek J, Bruinink A, Swieszkowski W. 2015. Ternary composite scaffolds with tailorable degradation rate and highly improved colonization by human bone marrow stromal cells. J
Biomed Mater Res A 103: 2394-2404.
243. Samavedi S, Vaidya P, Gaddam P, Whittington AR, Goldstein AS. 2014. Electrospun meshes possessing region-wise differences in fiber orientation, diameter, chemistry and mechanical properties for engineering bone-ligament-bone tissues. Biotechnol Bioeng 111:
2549-2559.
244. Choi da J, Choi SM, Kang HY, et al. 2015. Bioactive fish collagen/polycaprolactone composite nanofibrous scaffolds fabricated by electrospinning for 3D cell culture. J Biotechnol
205: 47-58.
245. Xu H, Su J, Sun J, Ren T. 2012. Preparation and characterization of new nano-composite scaffolds loaded with vascular stents. Int J Mol Sci 13: 3366-3381.
246. Powell HM, Boyce ST. 2009. Engineered human skin fabricated using electrospun collagen-PCL blends: Morphogenesis and mechanical properties. Tissue Eng Part A 15: 2177-
2187.
247. Powell HM, Ayodeji O, Summerfield TL, et al. 2007. Chemotherapeutic implants via subcritical CO2 modification. Biomaterials 28: 5562-5569.
177
248. Pepinsky RB, Shapiro RI, Wang S, et al. 2002. Long-acting forms of sonic hedgehog with improved pharmacokinetic and pharmacodynamic properties are efficacious in a nerve injury model. J Pharm Sci 91: 371-387.
249. Otsuka A, Levesque MP, Dummer R, Kabashima K. 2015. Hedgehog signaling in basal cell carcinoma. J Dermatol Sci 78: 95-100.
250. Zhang C, Wei X, Chen C, et al. 2014. Indian hedgehog in synovial fluid is a novel marker for early cartilage lesions in human knee joint. Int J Mol Sci 15: 7250-7265.
251. Yang Z, Song L. 2014. Correlation between indian hedgehog and damage degree of knee osteoarthritis cartilage. Zhonghua Yi Xue Za Zhi 94: 2434-2437.
252. Shuang F, Zhou Y, Hou SX, et al. 2015. Indian hedgehog signaling pathway members are associated with magnetic resonance imaging manifestations and pathological scores in lumbar facet joint osteoarthritis. Sci Rep 5: 10290.
253. Wei F, Zhou J, Wei X, et al. 2012. Activation of indian hedgehog promotes chondrocyte hypertrophy and upregulation of MMP-13 in human osteoarthritic cartilage. Osteoarthritis
Cartilage 20: 755-763.
254. Zhou J, Chen Q, Lanske B, et al. 2014. Disrupting the indian hedgehog signaling pathway in vivo attenuates surgically induced osteoarthritis progression in Col2a1-CreERT2; ihhfl/fl mice. Arthritis Res Ther 16: R11.
178
255. Zhou J, Wei X, Wei L. 2014. Indian hedgehog, a critical modulator in osteoarthritis, could be a potential therapeutic target for attenuating cartilage degeneration disease. Connect Tissue
Res 55: 257-261.
256. Ishizuka Y, Shibukawa Y, Nagayama M, et al. 2014. TMJ degeneration in SAMP8 mice is accompanied by deranged ihh signaling. J Dent Res 93: 281-287.
257. Dyment NA, Breidenbach AP, Schwartz AG, et al. 2015. Gdf5 progenitors give rise to fibrocartilage cells that mineralize via hedgehog signaling to form the zonal enthesis. Dev Biol
405: 96-107.
258. Carbone A, Carballo C, Ma R, et al. 2015. Indian hedgehog signaling and the role of graft tension in tendon-to-bone healing: Evaluation in a rat ACL reconstruction model. J Orthop Res. doi: 10.1002/jor.23066 [Epub ahead of print].
259. Bell R, Taub P, Cagle P, Flatow EL, Andarawis-Puri N. 2015. Development of a mouse model of supraspinatus tendon insertion site healing. J Orthop Res 33: 25-32.
179