pathology of calcific aortic valve disease: the role of … · 2013. 12. 18. · calcific aortic...

Pathology of Calcific Aortic Valve Disease: The Role of Mechanical and Biochemical Stimuli in Modulating the

Phenotype of and Calcification by Valvular Interstitial Cells

by

Cindy Ying Yin Yip

A thesis submitted in conformity with the requirements for the degree of Doctor of Philosophy

Graduate Department of the Institute of Biomaterials and Biomedical Engineering and Cardiovascular Collaborative Sciences Program

University of Toronto

© Copyright by Cindy Ying Yin Yip 2010

ii

Pathology of Calcific Aortic Valve Disease: The Role of Mechanical

and Biochemical Stimuli in Modulating the Phenotype of and

Calcification by Valvular Interstitial Cells

Cindy Ying Yin Yip

The Degree of Doctor of Philosophy

Institute of Biomaterials and Biomedical Engineering and

Cardiovascular Sciences Collaborative Program

University of Toronto

2010

Abstract

Calcific aortic valve disease (CAVD) occurs through multiple mutually non-exclusive

mechanisms that are mediated by valvular interstitial cells (VICs). VICs undergo pathological

differentiation during the progression of valve calcification; however the factors that regulate

cellular differentiation are not well defined. Most commonly recognized are biochemical factors

that induce pathological differentiation, but little is known regarding the biochemical factors that

may suppress this process. Further, the contribution of matrix mechanics in valve pathology has

been overlooked, despite increasing evidence of close relationships between changes in tissue

mechanics, disease progression and the regulation of cellular response. In this thesis, the effect of

matrix stiffness on the differentiation of and calcification by VICs in response to pro-calcific and

anti-calcific biochemical factors was investigated. Matrix stiffness modulated the response of

VICs to pro-calcific factors, leading to two distinct calcification processes. VICs cultured on the

more compliant matrices underwent calcification via osteoblast differentiation, whereas those

cultured on the stiffer matrices were prone to myofibroblast differentiation. The transition of

iii

fibroblastic VICs to myofibroblasts increased cellular contractility, which led to contraction-

mediated, apoptosis-dependent calcification. In addition, C-type natriuretic peptide (CNP), a

putative protective molecule against CAVD, was identified. CNP supressed myofibroblast and

osteoblast differentiation of VICs, and thereby inhibited calcification in vitro. Matrix stiffness

modulated the expression of CNP-regulated transcripts, with only a small number of CNP-

regulated transcripts not being sensitive to matrix mechanics. These data demonstrate the

combined effects of mechanical and biochemical cues in defining VIC phenotype and responses,

with implications for the interpretation of in vitro models of VIC calcification and possibly

disease devleopment. The findings from this thesis emphasize the necessity to consider both

biochemical and mechanical factors in order to improve fundamental understanding of VIC

biology.

iv

Acknowledgments

I would not have made it this far in academia without years of encouragement from two excellent

high school teachers: Dr. Doug Edward Burt – a dedicated and patient chemistry teacher – and

Mrs. Linda Willey – an enthusiastic biology teacher who supervised me on my first experiment

ever! I would also like to thank the members of the Cvitkovitch laboratory, especially Elena

Voronejskaia, Kristen Krastel, Richard Mair, Prahsath Suntharalingam and Celine Levesque, for

teaching me molecular biology techniques and many laboratory “tricks” during my graduate

study at the University of Toronto. In particular, Dr. Dennis Cvitkovitch, Dr. Celine Levesque

and Dr. Richard Ellen provided endless support during my Master’s degree, and they continue to

be amazing mentors throughout my Ph.D. study.

Most importantly, this Ph.D. work was made possible with the support and guidance of my

supervisor, Dr. Craig Simmons. I sincerely thank Dr. Simmons for providing the opportunity,

and with many risks (and presumably some faith), taking me on as a Ph.D. student. His

optimistic attitude kept me going during the difficult times and motivated me when exciting

research problems were encountered. I really appreciate the level of trust and respect he has

given me over the years. His patience to listen and willingness to consider countless scientific

(and non-scientific) ideas made my graduate experience particularly enjoyable. I am honored to

have had the opportunity to facilitate the establishment of and to exercise my “bossiness” in

managing the Simmons laboratory. I am grateful we shared this working experience and I feel

privileged to have had you as my supervisor.

I would also like to acknowledge the members of the Simmons group, who have continuously

worked together as a dynamic team to build the Simmons laboratory from an empty room on the

fourth floor of Rosebrugh Building to what it is today. I especially thank Jan-Hung Chen,

Christopher Moraes, Derek Watt, Kristine Wyss and Edmond Young, who were a part of the

Simmons group from the very beginning and contributed hours in setting up our research facility.

Please remember to take the “Simmons lab time capsule” if you move to another research facility

in the future! In addition, I would like to thank: Ruogang Zhao, for offering an endless supply of

pig hearts from the slaughterhouse regardless of the weather conditions – without Ruogang, I

would not be able to conduct all my experiments with freshly isolated primary cells; Jan-Hung

Chen, for organizing birthday lunches, the annual Centre Island BBQ, the Christmas party and

v

most importantly, for providing all the intense, but valuable scientific discussions, including the

endless arguments regarding the definition of “bone nodules”; Kelly Chen, for yelling and

screaming in the laboratory all of the time to make sure everyone did their laboratory chores;

Christopher Moraes, for organizing lab meetings and getting CO2 tanks for the incubators –

without you my “cells” will not go on; Mark Blaser, for help with the siRNA transfection

experiments and for putting up with my lousy supervision; Zahra Mirzaei, for keeping the lab “in

one piece” and being the world’s best “lab mom”; Edmond Young, for always providing

alternative perspectives to life and hours of fun during our trip in New York; Kristine Wyss, for

being a great collaborator, a reliable scuba diving buddy and an adventurous travel buddy. To

Krista Sider, Morakot Likhitpanichkul, Bogdan Beca, Suthan Srigunapalan and Wing-Yee

Cheung – thank you for working together to create the Simmons Group. Also I would like to

thank the high school, undergraduate, summer and work-study students who worked with me. In

particular, Susie Ferrante, Melissa Filice and Xiao Zhong for their help with the C-type

natriuretic peptide project and Stephanie Ting for her participation in the matrix stiffness study.

Thank you all and thank you for putting up with me! I wish you all the best with much success

and happiness in the future.

I was also fortunate to receive technical assistance from other laboratories. I thank Brent Steer

for his technical advice on molecular biology techniques. I thank Justin Parreno from the Kandel

laboratory for teaching me the hydroxyproline assay. I thank Jian Wang and Robert Chernecky

for their help with scanning electron microscopy. In addition, Jian Wang helped with tensile

testing of aortic valve leaflets and I always enjoyed our conversations about his garden. I thank

the Stanford laboratory, especially Wing Yan Chang and Tammy Reid for their help with

Western blots. I also thank Kelly Jackson at the University Health Network Microarray Center

for answering all my questions regarding microarray experiments. I thank Dr. Mete Civelek at

the University of California Los Angeles for his advice on microarray data analysis. And I thank

Dr. Michelle Bendeck, Dr. Christopher McCulloch, Dr. Christopher Yip, Dr. Lidan You and Dr.

Linda Demer for serving on my committee and providing insightful feedback.

I wish to thank all my friends for their support and lending an ear during the tough times. Special

thanks to: Seema Nagaraj, for her support and for lending me her quiet apartment to write my

thesis; Scott Brehm, for always pushing me to see life through the eyes of the optimist; Frederick

Suen, for providing a shoulder to lean on when needed; Alan Wong, for all the endless

vi

discussions about calcium – I hope we will have a chance to work on the calcium solid state

NMR project in the future; and many others who provided endless laughter. Without you all, I

would have been a complete “lab rat” and not have had a life outside of the lab.

Finally, I am thankful that my parents did not force me into another career path and allowed me

to pursue a Ph.D. degree.

vii

Table of Contents

ACKNOWLEDGMENTS .......................................................................................................... IV

TABLE OF CONTENTS ......................................................................................................... VII

LIST OF TABLES .................................................................................................................... XII

LIST OF FIGURES .................................................................................................................XIII

LIST OF ABBREVIATIONS ................................................................................................. XVI

CHAPTER 1.................................................................................................................................. 1

1. INTRODUCTION ................................................................................................................ 1

1.1. MOTIVATION ..................................................................................................................... 1

1.2. CURRENT RESEARCH PROBLEM......................................................................................... 1

1.3. OBJECTIVES....................................................................................................................... 3

1.4. THESIS ORGANIZATION ..................................................................................................... 3

CHAPTER 2.................................................................................................................................. 5

2. LITERATURE REVIEW.................................................................................................... 5

2.1. INTRODUCTION.................................................................................................................. 5

2.2. AORTIC VALVE FUNCTION AND STRUCTURE ..................................................................... 5

2.3. AORTIC VALVE PATHOLOGY: SCLEROSIS AND CALCIFICATION ....................................... 6

2.3.1. Economic and Clinical Burden of Calcific Aortic Valve Disease ............................. 6

2.3.2. Pathogenesis of CAVD............................................................................................... 7

2.3.3. Cellular and Molecular Mechanisms of CAVD ......................................................... 9

2.3.3.1. Valvular Endothelial Cells and Side-Dependent Susceptibility ............................ 9

2.3.3.2. Valvular Interstitial Cells and Phenotypes............................................................. 9

2.3.3.3. Calcification by Valvular Interstitial Cells .......................................................... 11

2.3.3.4. Pathological Extracellular Matrix Remodeling ................................................... 13

2.3.3.5. Extracellular Signals: Cytokines and Growth Factors ......................................... 15

2.3.3.6. Intracellular Signal: Transcription Factors .......................................................... 16

2.3.3.7. Natriuretic Peptides and Cardiovascular Disorders ............................................. 17

2.3.3.8. Progress in Therapeutic Development ................................................................. 19

2.4. MECHANOBIOLOGY ......................................................................................................... 20

2.4.1. Definition of stiffness ................................................................................................. 21

viii

2.4.2. Stiffness Sensing ........................................................................................................ 22

2.4.3. Test Systems: Engineering the Stiffness of Culture Substrata ................................... 24

2.4.4. Effect of Matrix Stiffness on Cell Response .............................................................. 25

2.4.4.1. Cell Shape and Spreading .................................................................................... 25

2.4.4.2. Cell Growth and Death ........................................................................................ 26

2.4.4.3. Cell Phenotype and Differentiation ..................................................................... 29

2.4.5. Matrix Stiffness and Pathologies................................................................................ 31

CHAPTER 3................................................................................................................................ 34

3. HYPOTHESES, OBJECTIVES AND CONTRIBUTIONS........................................... 34

3.1. RATIONALE ..................................................................................................................... 34

3.2. THESIS HYPOTHESES ....................................................................................................... 34

3.3. OBJECTIVES AND SPECIFIC AIMS ..................................................................................... 35

3.4. OVERVIEW OF CONTRIBUTIONS....................................................................................... 35

CHAPTER 4................................................................................................................................ 37

4. IMPLEMENTATION AND CHARACTERIZATION OF THE CELL CULTURE

SYSTEM………………………………………………………………………………………...37

4.1. MATERIALS AND METHODS............................................................................................. 38

4.1.1. Fabrication of Collagen Matrices ............................................................................. 38

4.1.2. Scanning Electron Microscopy for Topographic Evaluation ................................... 38

4.1.3. Determination of Matrix Mechanics ........................................................................ 39

4.1.4. Measurement of Collagen Content ........................................................................... 39

4.1.5. Statistical Analysis ................................................................................................... 40

4.2. RESULTS.......................................................................................................................... 40

4.2.1. Collagen Matrices with Tunable Stiffness ............................................................... 40

4.2.2. Substrate Topography............................................................................................... 43

4.2.3. Collagen Content and Stiffness of Matrices Over Culture Duration........................ 43

4.3. DISCUSSION..................................................................................................................... 44

CHAPTER 5................................................................................................................................ 48

5. EFFECT OF SUBSTRATE STIFFNESS ON CALCIFICATION BY VICS............... 48


5.1.1. Valve Interstitial Cell Culture .................................................................................. 49

ix

5.1.2. Measurement of Cellular Proliferation..................................................................... 49

5.1.3. Determination of Cell Shape and Spreading ............................................................ 50

5.1.4. Staining of Viable, Dead and Apoptotic Cells ......................................................... 50

5.1.5. Polymerase Chain Reaction for Expression of Osteogenic Markers........................ 51

5.1.6. Measurement of Runt-Related Transcription Factor 2 (Runx2) Protein .................. 52

5.1.7. Alkaline Phosphatase and Alizarin Red S Staining.................................................. 53

5.1.8. Osteocalcin Immunohistochemical Staining ............................................................ 53

5.1.9. Immunofluorescent Staining of Cytoskeletal Proteins ............................................. 53

5.1.10. Disruption of Cytoskeleton Assembly...................................................................... 54

5.1.11. Response to TGF-1................................................................................................. 54

5.1.12. Expression of TGF-1.............................................................................................. 54

5.1.13. Contraction-Dependent Apoptosis and Akt Activation............................................ 55


5.2. RESULTS.......................................................................................................................... 56

5.2.1. Morphological Changes, Proliferation and Cell Spreading...................................... 56

5.2.2. More Compliant Matrices Promote Osteogenic Differentiation of VICs ................ 56

5.2.3. Stiffer Matrices Promote Calcification Through Apoptosis..................................... 62

5.2.4. Aggregate Formation on Stiffer Matrices is Mediated by Cytoskeletal Tension..... 62

5.2.5. Response to TGF-β and the Expression of its Receptors Are Matrix Stiffness

Dependent................................................................................................................. 66

5.3. DISCUSSION..................................................................................................................... 68

CHAPTER 6................................................................................................................................ 73

6. EFFECT OF CNP ON PATHOLOGICAL DIFFERENTIATION OF VICS.............. 73


6.1.1. Animal Model and Valve Interstitial Cell Isolation ................................................. 74

6.1.2. Cell Culture .............................................................................................................. 74

6.1.3. Histological Analysis................................................................................................ 75

6.1.4. Dose Response of CNP ............................................................................................ 75

6.1.5. Cellular Proliferation ................................................................................................ 76

6.1.6. Evaluation of Osteogenic Differentiation................................................................. 76

6.1.7. Evaluation of Myofibroblast Differentiation............................................................ 77

x


6.2. RESULTS.......................................................................................................................... 78

6.2.1. Expression of Pathological Markers and CNP in Normal and Sclerotic Aortic

Valves ....................................................................................................................... 78

6.2.2. Molecular Components of CNP Signaling ............................................................... 78

6.2.3. Dose Response of CNP ............................................................................................ 78

6.2.4. Cellular Proliferation and Morphological Changes.................................................. 80

6.2.5. CNP Inhibits Calcification and Osteogenic Differentiation of VICs ....................... 81

6.2.6. Inhibition of Myofibroblast Differentiation by CNP................................................ 83

6.3. DISCUSSION..................................................................................................................... 87

CHAPTER 7................................................................................................................................ 93

7. THE COMBINED EFFECTS OF MECHANICAL AND BIOCHEMICAL CUES ON

THE TRANSCRIPTIONAL REGULATION OF VICS ........................................................ 93


7.1.1. Cell Culture .............................................................................................................. 94

7.1.2. Sample Preparation................................................................................................... 94

7.1.3. Microarray Experiments ........................................................................................... 95

7.1.4. Data Analysis............................................................................................................ 95

7.1.5. Partial Annotation Mapping and Identification of Biological Processes ................. 95

7.1.6. Venn Diagram Analysis ........................................................................................... 96

7.2. RESULTS.......................................................................................................................... 96

7.2.1. Sample Characterization........................................................................................... 96

7.2.2. Differential Gene Expression by Matrix Stiffness ................................................... 97

7.2.3. Differential Gene Expression by CNP ................................................................... 100

7.2.4. The Combined Effect of Matrix Stiffness and CNP on Transcriptional Regulation

103

7.3. DISCUSSION................................................................................................................... 110

CHAPTER 8.............................................................................................................................. 115

8. CONCLUSIONS AND RECOMMENDATIONS ......................................................... 115

8.1. CONCLUSIONS ............................................................................................................... 115

8.2. FUTURE WORK .............................................................................................................. 116

xi

8.2.1. Determination of Changes in Valve Matrix Mechanics in vivo ............................. 116

8.2.2. Improvement of the Cell Culture System............................................................... 117

8.2.3. Effect of CNP treatment at Different Stages of Disease Progression .................... 118

8.2.4. Identification of Transcriptional Pathways that Regulate Pathological

Differentiation of VICs........................................................................................... 118

REFERENCES.......................................................................................................................... 120

APPENDIX A............................................................................................................................ 142

A. PROTOCOLS................................................................................................................... 142

A.1. FABRICATION OF TYPE I COLLAGEN MATRICES............................................................... 142

A.2. SCANNING ELECTRON MICROSCOPY................................................................................ 143

A.3. HYDROXYPROLINE ASSAY ............................................................................................... 144

A.4. VALVULAR INTERSTITIAL CELL ISOLATION ..................................................................... 147

A.5. CRYOPRESEVATION OF VICS ........................................................................................... 149

A.6. RELEASING CELLS FROM COLLAGEN MATRICES.............................................................. 150

A.7. CELLULAR PROLIFERATION ASSAY.................................................................................. 151

A.7. CELL VIABILITY ASSAY ................................................................................................... 152

A.8. ALKALINE PHOSPHATASE STAINING ................................................................................ 153

A.9. INDIRECT IMMUNOSTAINING PROTOCOL .......................................................................... 155

A.10. WESTERN BLOT ............................................................................................................. 157

A.11. PRIMER SEQUENCES FOR PCR AND QRT-PCR............................................................... 166

APPENDIX B ............................................................................................................................ 167

B. PRELIMINARY DATA ...................................................................................................... 167

B.1. THE EFFECT OF STATINS ON THE EXPRESSION OF CNP BY VICS ..................................... 167

B.2. CULTURING PRIMARY VICS ON POLYACRYLAMIDE SUBSTRATES.................................... 169

B.3. ISOLATION OF VICS FROM MOUSE AORTIC VALVE.......................................................... 172

xii

List of Tables

Table 2.1. VIC phenotypes and functions in normal and diseased aortic valves.......................... 10

Table 2.2. A comparison of the cellular characteristics between calcification by VICs in vivo and

in vitro .......................................................................................................................... 11

Table 2.3. Biochemical factors in valve sclerosis and calcification ............................................. 16

Table 2.4. Compliance of culture materials and tissues................................................................ 22

Table 2.5. Matrix stiffness mediated cellular responses ............................................................... 27

Table 2.6. Advantages and disadvantages of various culture systems for studying the effect of

substrate stiffness ......................................................................................................... 28

Table 7.1. A subset of transcripts with higher expression in VICs cultured on compliant matrices

relative to those cultured on stiff matrices ................................................................. 100

Table 7.2. A subset of transcripts with lower expression in VICs cultured on compliant matrices

relative to those cultured on stiff matrices ................................................................. 101

Table 7.3. A subset of transcripts with higher expression in CNP-treated VICs relative to

untreated cells when cultured on compliant matrices. ............................................... 103

Table 7.4. A subset of transcripts with lower expression in CNP-treated VICs relative to

untreated cells when cultured on compliant matrices. ............................................... 104


untreated cells when cultured on stiff matrices.......................................................... 105


untreated VICs cells when cultured on stiff matrices ................................................ 106

Table 7.7. A partial list of CNP-regulated, mechanically-insensitive genes .............................. 109

xiii

List of Figures

Figure 2.1. Porcine aortic valve. ..................................................................................................... 7

Figure 2.2. Cellular and extracellular matrix of the aortic valve. ................................................... 8

Figure 2.3. Morphology and molecular content of multicellular aggregates formed by cultured

VICs ............................................................................................................................ 13

Figure 2.4. The interplay of physical and biochemical signals in a dynamic feedback system

involving the interactions of ECM, surface receptors and cytoskeleton (adapted

from154) ....................................................................................................................... 24

Figure 2.5. Change in arterial wall stiffness in rabbit fed a high cholesterol diet to induce

atherosclerosis............................................................................................................. 33

Figure 4.1. Type I collagen matrices ............................................................................................ 41

Figure 4.2. Mechanical properties of thick and thin collagen matrices ........................................ 42

Figure 4.3. Microstructure of collagen matrices........................................................................... 43

Figure 4.4. Collagen content and effective stiffness of the two matrices over the ....................... 45

Figure 5.1. Proliferation and morphology of VICs cultured in DMEM ....................................... 57

Figure 5.2. Proliferation and morphology of VICs cultured in calcifying media......................... 58

Figure 5.3. Cell shape and spreading 48 hours after initial seeding ............................................. 59

Figure 5.4. Calcification by VICs on the two matrices................................................................. 60

Figure 5.5. Compliant matrices promote osteogenic phenotypes. ................................................ 61

Figure 5.6. Stiffer matrices promote dystrophic calcification associated with VIC apoptosis..... 63

Figure 5.7. Cytoskeleton expression............................................................................................. 64

Figure 5.8. Aggregate formation by VICs cultured on stiff matrices ........................................... 65

xiv

Figure 5.9. Effect of cytoskeleton disruption on aggregate formation ......................................... 66

Figure 5.10. Molecular factors involved in contraction-induced apoptosis on stiff substrates .... 67

Figure 5.11. Expression of TGF- receptors I and II ................................................................... 68

Figure 6.1. Expression of CNP, -SMA and Runx2/Cbfa-1 in normal and sclerotic porcine aortic

valves .......................................................................................................................... 79

Figure 6.2. Expression of NPR-B and activation of cGMP by CNP ............................................ 80

Figure 6.3. Dose-dependent -SMA expression by VICs ............................................................ 81

Figure 6.4. Proliferation and morphology of cells with or without CNP treatment ..................... 82

Figure 6.5. CNP modulates calcification by VICs........................................................................ 84

Figure 6.6. Expression of bone-related transcripts ....................................................................... 85

Figure 6.7. Expression of bone-related proteins ........................................................................... 86

Figure 6.8. Effect of CNP on osteoprogenitor subpopulation ...................................................... 87

Figure 6.9. CNP inhibits expression of myofibroblast marker ..................................................... 88

Figure 6.10. Quantification of -SMA expression ....................................................................... 89

Figure 6.11. Mutually exclusive expression of CNP and -SMA in cultured VICs. ................... 90

Figure 6.12. CNP affects function associated with activated myofibroblasts .............................. 91

Figure 6.13. Contractility of VICs with or without CNP treatment.............................................. 92

Figure 7.1. CNP inhibites aggregate formation on VICs cultured on compliant and stiff matrices

..................................................................................................................................... 97

Figure 7.2. CNP significantly reduces the total number of aggregates formed by VICs on

compliant and stiff matrices........................................................................................ 98

xv

Figure 7.3. The distribution of sequences differentially expressed with matrix stiffness ............ 98

Figure 7.4. The distribution of sequences differentially expressed with CNP treatment in cultures

on compliant matrices ............................................................................................... 102

Figure 7.5. The distribution of sequences differentially expressed with CNP treatment in cultures

on stiff matrices......................................................................................................... 102

Figure 7.6. Transcript expression modulated by matrix stiffness and/or by CNP in cultures on

compliant matrices .................................................................................................... 107

Figure 7.7. Transcript expression modulated by matrix stiffness and/or by CNP in cultures on

stiff matrices.............................................................................................................. 108

Figure 7.8. A three-way Venn diagram showing the commonly and exclusively modulated genes

by matrix stiffness and by CNP ................................................................................ 111

Figure A.1. Assembly of the protein transfer tank...................................................................... 163

Figure B.1. Expression of CNP transcript after three days of simvastatin treatment ................. 168

Figure B.2. Inhibition of aggregate formation by simavastatin treatment…………………....170

Figure B.3. Calcification by primary VICs on PA substrates with stiffness of 11 kPa, 22 kPa, 50

kPa and 144 kPa ………………………………………...…...……………………..172

Figure B.4. Isolation of mouse aortic valve………………………………………………..…173

Figure B.5. Mouse aortic valve and VICs…………………………………………………….174

xvi

List of Abbreviations

Full name Abbreviation

2-dimensional 2D

3-dimensional 3D

Alkaline phosphatase ALP

Angiotensin converting enzyme ACE

A-type natriuretic peptide ANP

Alizarin red S ARS

Alpha-smooth muscle actin -SMA

Aortic valve AV

Bone morphogenetic protein BMP

B-type natriuretic peptide BNP

Calcific aortic valve disease CAVD

Core-binding factor alpha-1 Cbfa-1

Cyclin-depenent kinases Cdk

Colony forming unit-alkaline phosphatase CFU-ALP

C-type natriuretic peptide CNP

Carbon dioxide CO2

Cardiovascular disease CVD

Cyclic guanosine monophosphate cGMP

Dulbecco's Modified Eagle Medium DMEM

Discoidin domain receptor DDR

Extracellular matrix ECM

Extracellular signal-regulated kinase ERK

Focal adhesion kinase FAK

Fetal bovine serum FBS

False discovery rate FDR

Finite element FE

Fibroblast growth factor FGF

xvii

Full name Abbreviation

Fluoroscein isothiocyanate FITC

Glyceraldehyde-3-phosphate dehydrogenase GAPDH

Gene ontology GO

Guanosine tri-phopshate GTP

Hydrochloride acid HCl

Low-density lipoprotein LDL

Matrix metalloproteinase MMP

Sodium bicarbonate NaHCO3

Sodium hydroxide NaOH

Notch homolog 1 Notch1

Natriuretic peptide receptor B NPR-B

Neutral buffered formalin NBF

Nuclear factor of kappa light polypeptide gene enhancer in B-cells

inhibitor-

NFKBIA

Osteocalcin OC

Osteonectin ON

Osteoprotegerin OPG

Phosphate buffered saline PBS

Receptor activator of nuclear factor κB RANK

Ligand of receptor activator of nuclear factor κB RANKL

RNA integrity number RIN

Runt-related transcription factor 2 Runx2

Scanning electron microscopy SEM

Swinholide A SWA

Tissue cultured treated polystyrene TCPS

Transforming growth factor-beta TGF-

Inhibitor of metalloproteinases TIMP

Valvular interstitial cell VIC

1

Chapter 1 1. Introduction

1.1. Motivation

It is estimated that 17 million people around the world die of cardiovascular disease each year1.

Calcific aortic valve disease (CAVD) is one of the most common cardiovascular diseases,

afflicting more than 25% of the population over age 652 and resulting in a 50% increase risk of

other cardiovascular events3. Surgical valve replacement at the end stage of the disease remains

the only treatment, because to date, there are no effective therapeutics. A strong fundamental

knowledge of valve cell biology will improve our understanding of the etiology of valve disease,

perhaps leading to new treatment options.

1.2. Current Research Problem

Progress in basic research on atherosclerosis, a disease that shares many features with CAVD,

and in vivo and in vitro valve studies have advanced fundamental knowledge of CAVD.

Although in vivo animal experimental models of CAVD are informative and often display

disease characteristics similar to those found in humans4-6, in vitro studies provide fundamental

mechanistic insights that are difficult to dissect otherwise. However the majority of in vitro

valve studies have focused on biochemically regulated cellular responses and overlooked the

potential contributions of mechanical cues, despite increasing evidence of close relationships

between changes in tissue mechanics, disease progression7-9 and the mechanical regulation of

cellular responses in other cell types (reviewed in 10). Moreover, mechanical cues can modulate

cell response to exogenous biochemical factors11, 12, and therefore evaluation of cell response to

biochemical cues in the context of the cellular mechanical environment will provide a more

complete understanding of valve cell biology.

Beyond its relevance for interpreting in vitro models of valve calcification, matrix stiffness may

also contribute to CAVD pathology in vivo. Calcified valve leaflets demonstrate significant

remodeling of the valve extracellular matrix (ECM), which alters matrix protein composition13, 14

and stiffness15. Although the contribution of alterations in the micromechanical properties of the

2

valve matrix to disease development has yet to be determined, changes in local matrix stiffness

of other tissues have been shown to occur prior to abnormal biological alterations. For example,

liver stiffness increases prior to the activation of liver fibroblasts and pathological matrix

deposition7, and changes in the local stiffness of atherosclerotic lesions occur prior to substantial

histological changes in the matrix16. These findings suggest an association between the

biophysical properties of the tissue matrix and disease progression. As atherosclerosis and

CAVD share many disease characteristics, similar early dynamic changes in matrix mechanics

are postulated to occur in sclerotic valves. Thus, the determination of whether VICs sense and

respond to substrate stiffness will not only provide fundamental knowledge of the

mechanobiology of VICs, but also may be relevant to understanding the progression of CAVD.

In addition to the lack of fundamental knowledge regarding valve cell mechanoregulation, there

are limited in vivo and in vitro studies evaluating biochemical factors that protect against the

development of disease-related VIC phenotypes. Clinical trials of statins, one of the most

studied potential therapeutics for CAVD, have so far generated inconclusive and controversial

results (reviewed in 17, 18), suggesting the necessity for an improved understanding of the cellular

and/or molecular mechanisms of statins or alternatively, the identification of other novel

therapeutic targets. Clues to alternative targets to prevent and treat CAVD may originate from

the tendency of calcification to occur more readily in the fibrosa (disease-prone aortic side) of

the leaflets than the ventricularis (disease-protected ventricular side)14. Profiling of gene

expression by normal porcine aortic valve endothelial cells revealed statistically significant side-

dependent differential expression of 584 genes6. Of these, C-type natriuretic peptide (CNP) was

among the most differentially expressed, with higher expression on the disease-protected side of

the leaflet in comparison to the disease-prone side. CNP and its activator, furin, were recently

shown to be distinctly downregulated in human sclerotic valves19, suggesting a putative

protective role of CNP against CAVD. Although CNP has been extensively studied in the

skeletal and reproductive systems, its effect on valvular cells is completely unknown. The study

of CNP-mediated regulation of VIC phenotypic changes will identify the cellular basis

responsible for the putative protective effect of CNP against CAVD, and will provide the

fundamental knowledge essential for mechanistic investigations at the molecular level in the

future. Further, matrix stiffness is known to regulate cellular response to soluble factors13. It is

possible that matrix stiffness may regulate the response of VICs to CNP. If VICs are

3

mechanically regulated, it will be important to investigate the effect of CNP on VICs when

cultured on various substrate stiffnesses. The investigation of substrate stiffness-mediated VIC

responses to anti-sclerotic or anti-calcific factors has yet to be conducted, despite the fact that

VICs populate the valve matrix, which changes its mechanical properties over the progression of

the disease. An in vitro study that incorporates both biochemical and mechanical cues will

provide a concrete foundation for subsequent in vivo testing of CNP, if it is proven to elucidate a

protective effect against VIC calcification in vitro.

1.3. Objectives

The overall goal of this project was to investigate the role of substrate stiffness on modulating

VIC responses to both pro-calcific and anti-calcific biochemical factors. The specific objectives

of this study are as follows:

1. To implement and characterize a cell culture system with tunable stiffness;

2. To investigate if substrate stiffness modulates the response of VICs to pro-calcific

biochemical factors;

3. To identify the cellular basis of CNP, a putative disease-protective agent, in protecting

against VIC sclerosis and calcification; and

4. To evaluate the combined effect of substrate stiffness and CNP on the transcriptional

regulation of VICs.

1.4. Thesis Organization

Chapter One presents the clinical motivation, an overview of the research problem and the thesis

objectives. Chapter Two provides a thorough review of topics specifically related to this

research, including aortic valve structure, CAVD pathology, C-type natriuretic peptide and

mechanobiology. Chapter Three states the hypotheses, objectives and overall approach. A

summary of contributions from my collaborators in completing this Ph.D. thesis is included at

the end of Chapter Three. Chapter Four describes the characterization of the cell culture system

with adjustable stiffness. Chapter Five describes the effect of substrate stiffness on the response

of VICs to pro-calcific biochemical factors. Chapter Six describes the cellular basis responsible

for the protective properties of CNP against sclerosis and calcification. Chapter Seven evaluates

the combined effect of both substrate stiffness and CNP on transcriptional regulation of VICs.

4

Chapter Eight summarizes the results and provides recommendations for future studies. A list of

references, detailed experimental protocols and preliminary data from most recent work are

appended to the end of this document.

5

Chapter 2 2. Literature Review

2.1. Introduction

The interdisciplinary nature of this thesis requires a review of both the biochemical and the

mechanical factors that regulate CAVD and cell behaviour. This chapter is divided into two main

portions. The first part describes the aortic valve structure and functions, the valve extracellular

matrix and the valvular cells, with an in-depth discussion of VICs. A brief summary of the

pathogenic processes associated with valve sclerosis and calcification is included, followed by a

discussion of current pharmacological inventions. The second portion summarizes the recent

findings in mechanobiology, mainly focused on the role of matrix stiffness in modulating cell

phenotype and its association with various diseases. A brief discussion of several technologies

with adjustable substrate stiffness is included.

2.2. Aortic Valve Function and Structure

The aortic valve leaflets open and close more than three billion times within one lifespan to

regulate unidirectional blood flow from the left ventricle to the aorta20. The dynamic operation of

the aortic valve is driven by the pressure gradient between the aortic and the ventricular sides of

the leaflets. Higher pressure in the left ventricle during systole forces the valve to open, thereby

pumping oxygenated blood through the valve to the aorta. The pressure in the left ventricle

decreases as systole ends, leading to a higher pressure on the aortic side during diastole which

forces the aortic valve to close. The closure of the aortic valve prevents retrograde blood flow

back into the left ventricle.

The aortic valve consists of three half moon shaped leaflets (Figure 2.1) each with similar matrix

structure and composition as illustrated in Figure 2.2.A. The aortic and ventricular surfaces of

the aortic valve are lined with valve endothelial cells (VECs, Figure 2.2: B and C), whereas

valve interstitial cells (VICs, Figure 2.2: B and C) can be found throughout the ECM. The ECM

of a healthy aortic valve is arranged into three highly organized layers each with distinct matrix

protein composition13 and mechanical properties21-23. The fibrosa contains primarily of a dense

6

network of type I collagen; the spongiosa consists of loosely arranged proteoglycans; and the

ventricularis is composed of a dense network of collagen and elastin fibers20, 24. The

composition and the organization of matrix proteins likely govern the mechanical properties of

each layer. For example, collagen fiber orientation determines the ability of the tissue to

withstand tensile stresses. This is evident in biaxial mechanical testing of the ventricularis layer

in which circumferentially oriented collagen fibers dominated its stress-strain response21. Of

note, restoration of collagen fiber orientation between successive loading cycle requires elastin25,

suggesting the role of ECM proteins in maintaining proper valve matrix mechanics to serve its

function. Presumably, the organization and the composition of the ECM regulate VICs

biochemically and mechanically, enabling these cells to withstand continuous physical stresses

and to maintain valve homeostasis through physiologic matrix remodeling. VIC responsiveness

to ECM composition is evident in cell-matrix interactions via cell surface receptors. For instance,

components of the ECM such as fibronectin, fibrin, laminin and collagen have been showed to

modulate VIC calcification26. However, the responsiveness of VICs to ECM mechanics has yet

been evaluated thoroughly. Further discussion of valve cells and ECM in relation to CAVD and

calcification follows in the next section.

2.3. Aortic Valve Pathology: Sclerosis and Calcification

2.3.1. Economic and Clinical Burden of Calcific Aortic Valve Disease

Cardiovascular disease (CVD) costs the Canadian economy about $18 billion27 each year and is

the leading economic burden of illness nationally28. An estimated eight million Canadians suffer

from some form of CVD29 and CVD contributes to one third of all deaths in Canada30. Calcific

aortic valve disease (CAVD) is one of the most common CVDs2. Risk factors of CAVD are

similar to those of atherosclerosis including: increasing age, male gender, hypertension and

elevated serum level of low-density lipoprotein (LDL)31. CAVD increases in prevalence with

age: from 35 percent between 75 and 84 years of age to 50 percent of those over age 802, 32. Life

expectancy is projected to increase globally in the next half century, with an increase of 19

percent and 27 percent of the population at age 60 and 80, respectively33, suggesting that CAVD

will become even more prevalent. Further, CAVD is associated with a 40 percent increase risk of

myocardial infarction34 as well as a 50 percent increase risk of cardiovascular death35. About 50

percent of patients with severe aortic sclerosis that progresses to stenosis (i.e., clinically-

significant valve narrowing) also suffer from coronary artery disease36. Despite the economic

7

and clinical consequences, there is no treatment other than surgical replacement of severely

stenotic valves, in large part because of the poor understanding of the biochemical and

mechanical factors contributing to the disease as well as a lack of fundamental knowledge in

VIC biology.

Figure 2.1. Porcine aortic valve.

(A) The left coronary cusp (L), right coronary cusp (R) and non-coronary (N) cusp of the porcine

aortic valve are similar in shape. (B) The aortic side of a porcine valve leaflet.

2.3.2. Pathogenesis of CAVD

CAVD is characterized by thickening of the leaflets and significant pathological matrix

remodeling, which often leads to calcification that eventually impairs leaflet motion and causes

significant hemodynamic obstruction due to valve stiffening14. CAVD was once considered as a

degenerative process resulted from “wear and tear” with aging. However, findings from recent

studies have identified multiple cell-driven processes contributing to the development of CAVD,

some of which are similar to those found in atherosclerosis37, 38 such as chronic inflammation39,

lipid deposition40 and pathological matrix remodeling41, 42.

L RN

A B

L RN

A B

8

Figure 2.2. Cellular and extracellular matrix of the aortic valve.

(A) Valve extracellular matrix consists of three distinct tissue layers: fibrosa (F), spongiosa (S)

and ventricularis (V). (B and C) show the lining of valve endothelial cells (VECs) on the surface

of the valve leaflet and valve interstitial cells (VICs, nuclei are stained blue) permeates the entire

valve extracellular matrix. (Source: Adapted from 6).

Spatial characteristics of CAVD are evident at both the macro- and micro-scale. Lipoprotein

deposition43 and the formation of fibrocalcific masses6 (also known as “calcific nodules”) are

particularly prominent in the fibrosa, on the aortic side of the leaflets. Macrophage and T-

lymphocyte infiltration, subendothelial thickening, basement membrane disruption, accumulation

of collagen and elastin, and focal calcium deposition are observed on the aortic side of the

leaflets or within the fibrosa layer at the early stage of lesion formation14. The susceptibility of

the aortic side to disease development may result from the differences in the local

microenvironment (e.g., infiltration of inflammatory cells, lipoprotein deposition, hemodynamic

forces44, mechanical stress45, 46, or extracellular matrix composition13) and heterogeneity between

cells from the aortic and the ventricular side of the valve6.

Aortic valves undergo substantial matrix composition and cellular phenotypic changes as CAVD

progresses, including pathological remodeling of the tri-layered extracellular matrix13, an

increase in apoptotic cells5, an increase in myofibroblast content47 and calcium deposition14.

Strikingly, ectopic bone and cartilage are also found in diseased aortic valves48, 49. The

upregulation of several bone-related markers including osteopontin, bone sialoprotein,

osteocalcin, alkaline phosphatase (ALP) and osteoblast-specific transcription factor core binding

9

factor -1/runt-related transcription factor 2 (Cbfa-1/Runx2) has been reported in human

calcified valves49, suggesting that cell-mediated processes are of bone development. The cell

source and the molecular mechanisms leading to these phenotypic alterations have not yet been

fully defined, although various non-mutually exclusive mechanisms have been identified some of

which are described in the next section.

2.3.3. Cellular and Molecular Mechanisms of CAVD

2.3.3.1. Valvular Endothelial Cells and Side-Dependent Susceptibility

VECs are anchored to the basal lamina, lining the two surfaces of the leaflets. The VEC

monolayer acts as a semi-permeable boundary that regulates transport of soluble factors from the

circulating blood into the valve interstitium and shields VICs from hemodynamic effects.

Endothelial dysfunction is often found in sclerotic leaflets50 and is associated with the

accumulation of inflammatory cells. Although it has been thought that VECs share similar

characteristics with those in the vasculature, tissue-specific functions have been suggested based

on the observed phenotypic and functional differences between the two EC populations in

culture51. Further, VECs from opposite sides of normal leaflets have been shown to display

distinct gene expression profiles that correlate with side-specific susceptibility of aortic valves to

sclerosis and calcification6, indicating local spatial phenotypic heterogeneity of the EC

population within the aortic valve. Of note, among the 584 differentially expressed genes on the

two sides of normal porcine aortic valves, disease-protective transcripts such as osteoprotegerin

(OPG) and C-type natriuretic peptide (CNP) were expressed at a significantly higher level on the

ventricular (disease-protected) side in comparison to the aortic (disease-prone) side. It has been

postulated that this spatial heterogeneity may be important in regulating autocrine signaling

within the VEC population and paracrine signaling to VICs locally. Biochemical factors

synthesized by VECs may also regulate VICs locally, contributing to the side-dependent

susceptibility of aortic valves to sclerosis and calcification. In this thesis, I was particularly

interested in investigating the response of VICs to CNP.

2.3.3.2. Valvular Interstitial Cells and Phenotypes

The VIC population is not localized to any one region of the leaflet, but resides in all three layers

of the ECM52, 53. VICs regulate valve homeostasis and are primarily responsible for valve

sclerosis and calcification54. The population of VICs in healthy aortic valves is heterogeneous

10

(Table 1), consisting mainly of quiescent fibroblasts, a small population (~1-5 percent) of

myofibroblasts (activated VICs) and smooth muscle cells47, 55-57, and a subpopulation of

progenitor cells with multipotent differentiation potential58. VICs undergo phenotypic changes

in response to the microenvironmental cues, including the extracellular matrix26, biochemical

soluble factors (e.g., cytokines)59 and mechanical forces55 (reviewed in60). In some cases, these

phenotypic alterations can contribute to pathogenesis. For example, activation of quiescent VICs

into myofibroblasts increases the myofibroblast content of sclerotic leaflets to 30 percent of the

total VIC population47, 61. Activated fibroblasts express prominent stress fibres containing -

smooth muscle actin (-SMA)57, 62-64, and are associated with increased collagen synthesis and

cellular contractility65. Pathological myofibroblast activation may lead to an unbalanced matrix

remodeling that subsequently alters the biochemical and mechanical properties of the

microenvironment within the valves. Additionally, osteoblasts-like cells are often found in

calcified aortic valves49 and likely originate from osteogenic differentiation of resident

progenitor VICs58 or bone marrow-derived hematopoietic stem cells66. VICs comprise a

phenotypically dynamic cell population (Table 2.1) and their differentiation into various

phenotypes is closely associated with defined sets of cellular functions, which presumably

modulate the progression of CAVD.

Table 2.1. VIC phenotypes and functions in normal and diseased aortic valves

Phenotype Normal aortic valves

Sclerotic/ calcified AVs

Function(s)

Quiescent VICs Abundant Less abundant than in normal valves.

Maintain valve homeostasis.

Progenitor VICs (Likely a subpopulation of quiescent VICs)

~ 50%58 Data not available

Multipotent differentiation potential. Can differentiate into multiple lineages including chondrocytes, adiopocytes and osteoblasts. If differentiated into osteoblasts, these cells can secret alkaline phosphatase and osteocalcin, and deposit calcium58.

Activated VICs (Myofibroblasts)

~1-5%67 ~ 30%61 -SMA positive contractile cells that participate in active cellular repair processes such as matrix remodeling55.

11

2.3.3.3. Calcification by Valvular Interstitial Cells

The cellular mechanisms by which VICs contribute to calcification are not well defined, largely

because in vivo and ex vivo studies are limited to evaluating the end-stage of the disease.

Although large animal models such as porcine and ovine are excellent disease models with

similar lipoprotein serum levels and hemodynamic profiles to humans, it is difficult to track the

changes of individual molecular signaling pathways over the course of disease progression partly

due to the lack of imaging modalities that can monitor valve pathological development in a time-

dependent manner. Alternatively, ex vivo studies with human aortic valves are often limited to

those at the end-stage rather than at the onset of the disease, and therefore these studies are not

capable of identifing molecular mechanisms responsible for disease initiation. Hence in vitro cell

culture systems are often used to study disease-related molecular and cellular events in hopes of

deciphering the underlying mechanisms of valve calcification, as these cell culture models often

display in vivo characteristics of the disease (Table 2.2).

Table 2.2. A comparison of the cellular characteristics between calcification by VICs in vivo

and in vitro

Characteristics Calcified aortic valves In vitro calcification by VICs

Upregulated49 Upregulated54, 58 Upregulated49 Upregulated58 Upregulated49 Upregulated58

Expression of bone-related markers: Alkaline phosphatase Cbfa-1/Runx2 Osteocalcin Osteopontin Upregulated49 Upregulated68

Hydroxyapatite mineral Present49 Present58

Apoptotic cell content Increased5 Increased69

Expression of - SMA and myofibroblast content

Upregulated47 Upregulated26

MMP activities Altered47, 70, 71 Altered72

Calcification by VICs can be induced in vitro by cytokines (e.g., transforming growth factor-1

(TGF-1)69 and bone morphogenetic proteins (BMPs))54, 73 or by addition of a calcifying media

supplement that consists of a combination of -glycerophosphate, ascorbic acid and/or

dexamathesone58, 73. Calcification can be achieved without additional biochemical stimuli only if

12

VICs are seeded at high (> 50,000 cells/cm2) cell-seeding density26 or cultured for a long

duration (> 21 days)54. Under appropriate culture conditions, VICs form multicellular aggregates

of various shapes, sizes, and transcript and protein expression (Figure 2.3). These aggregates are

associated with the expression of bone-related transcripts and proteins26, 54, 58, 74, the expression

of myofibroblast markers26, and/or apoptosis69, all of which are accompanied by localized

calcium deposition within the aggregates. Cellular calcification has been reported to occur

through various non-mutually exclusive processes. Cells such as VICs54, pericytes75 and

calcifying vascular smooth muscle cells76 can differentiate into osteoblast-like cells that secrete

bone matrix proteins (e.g., ALP, osteonectin and osteocalcin) and form hydroxyapatite (a bone

mineral)26, 54. Apoptosis is another calcification process that is evident in calcified tissue in

cartilage, arteries and aortic valves, and it is apparent that apoptosis is involved in calcification

by valve cells69, 77 and vascular cells77. Apoptotic bodies have been proposed to serve as a

nucleation site for calcium crystal formation77. Other processes such as active uptake of calcium

ions and phosphate by matrix vesicles, hydrolysis of pyrophosphate by alkaline phosphatase,

hydrolysis of ATP by ATPase leading to the release of pyrophosphate, and accumulation of

phospholipids as a potential apatite nucleation site may also contribute to physiologic

calcification (reviewed in78). Because of the complexity of these calcification processes, it is

unclear whether some of the in vitro calcification reported in VIC studies (Table 2.2) represents

one or many of these calcification processes, confounding the interpretation of cell culture data

and limiting our understanding of the mechanisms underlying calcification by VICs.

13

Figure 2.3. Morphology and molecular content of multicellular aggregates formed by

cultured VICs

In different studies, cell aggregates have been reported to vary in cellular and molecular content

from different studies. These cells can express (A) localized alkaline phosphatase (stained violet,

adapted from58), (B) osteocalcin (stained brown, adapted from 79), (C) -SMA (stained green,

adapted from26) and (D) annexin V indicative of apoptosis (stained green, adapted from69). (E)

Calcium deposition is often observed within the aggregates, closely apposed to intact, viable

cells. M: mineral; C: cells (adapted from58). (F) Electron diffraction patterns of mineral deposits

from calcific aggregates matched the reference pattern for hydroxyapatite (JCPDS 9-0432)

(adapted from58).

2.3.3.4. Pathological Extracellular Matrix Remodeling

The ECM provides biochemical and mechanical cues to adherent cells. ECM composition13 and

mechanical7, 16 alterations are characteristics of sclerotic diseases. Disorganization of collagen

bundles42, fragmentation and stratification of the elastic fibers13, infiltration of loose connective

tissue within the collagen matrix and an increase in proteoglycan matrix13 are commonly found

in the diseased valve leaflets. In addition, the expression of bone-related matrix proteins such as

osteocalcin, osteopontin, osteonectin, and bone morphogenetic proteins (BMPs) is often

upregulated in calcified valves48. These bone proteins are known to regulate hydroxyapatite

formation, therefore they are important mediators of cell-mediated calcification80, 81.

A B C

D E F

A B C

D E F

14

The matrix architecture can be remodeled by zinc- and calcium-dependent endopeptidases

known as matrix metalloproteinases (MMPs). Activated MMPs can degrade ECM components

such as elastin and collagen fibers; therefore a fine balance between the activities of MMPs and

their inhibitors (i.e., tissue inhibitors of metalloproteinases, TIMPs) dictates the rate of ECM

turnover. The results of MMP studies with respect to valve pathology have been inconsistent,

with some but not all studies reporting an upregulation of the expression and/or activity of

MMP-1, -2, -3, and -9 and TIMP-1 and TIMP-2 in calcified valves42, 82. Presumably, changes in

the relative activity of MMPs and TIMPs in diseased leaflets contribute to pathological matrix

remodeling, which alters cell-matrix interactions and matrix mechanics. Interestingly, calcium

deposition is spatially associated with damaged basal membrane83 and it has been postulated that

MMPs may facilitate calcification process(es) by degrading components of damaged basal

membrane84.

The function of MMPs goes beyond matrix remodeling. For example, MMP-2 has been

identified as an intracellular signaling molecule capable of modulating fibroblast proliferation

and fibrotic process in non-valve tissues85-87. The expression of MMP-2 is often found in -SMA

positive VICs70, however whether MMP-2 has similar proliferative and fibrotic effects on VICs

as observed in other cell types is unknown.

Although the involvement of MMPs and TIMPs in the pathogenesis of CAVD is evident, the

mechanisms leading to changes in the expression and/or activity of these endopeptidases is not

known. VICs have been reported to produce some of the MMPs and TIMPs. Studies have

demonstrated stimulation of the synthesis and activation of MMP-1 and MMP-2 in VICs by

proinflammatory cytokines such as interleukin-171 and tumour necrosis factor-72. Other ECM

proteins such as tenascin C have been shown to stimulate MMP-2 expression and gelatinolytic

activity in cultured VIC culture88. Tenascin-C is a multifunctional ECM glycoprotein known to

regulate cell proliferation, migration, differentiation and apoptosis89. The expression of

tenascin-C increases proportionally with the severity of valve calcification90 and is co-localized

with -SMA positive valve cells88. Further, the expression and activation of MMPs can be self-

regulated. For example, the activation of MMP-3 can trigger the activation of MMP-1 and

MMP-9. In addition to VICs, inflammatory cells that often infiltrate diseased leaflets can also

produce MMPs82, which may potentiate the development of CAVD.

15

2.3.3.5. Extracellular Signals: Cytokines and Growth Factors

Cytokines and growth factors are known to modulate AV pathogenesis (Table 2.3). The TGF-β

superfamily consists of a wide range of regulatory proteins, some of which are upregulated in

calcified AVs. The most frequently studied is TFG-1, a multifunctional cytokine that regulates

proliferation, phenotype, differentiation and fibrosis, with diverse functions in the vascular

system. Its ability to induce -SMA expression in myofibroblast precursors is indicative of its

role in myofibroblast differentiation65. Of note, TGF-1 accelerated the differentiation of

cultured VICs into -SMA positive myofibroblasts via Smad signaling and its myofibrogenic

effect was repressed by fibroblast growth factor (FGF-2)64. Jian et al identified abundant TFG-1

expression and a moderate decrease in TFG-1 receptors (RI and RII) in calcified human AV in

comparison to normal AV69. The effect of TGF-1 in culture is multifaceted, including

stimulation of MMP-9 transcript expression, MMP-2 activity74, ALP activity74, and apoptosis-

dependent calcification by VICs69. Other members of the TGF- superfamily, the bone

morphogenetic proteins (BMPs), are also upregulated in calcified AVs91. Cell culture with

BMP-2, BMP-4 or BMP-7 was found to promote calcification by VICs92. These data suggest the

involvement of proteins from the TGF- superfamily in CAVD pathogenesis.

Receptor activator of nuclear factor B (RANK), its ligand (RANKL), and the soluble receptor

osteoprotegerin (OPG) are cytokines of the tumour necrosis factor superfamily known to regulate

bone turnover, and vascular and valvular calcification. The activation of the RANK-RANKL

pathway promotes osteogenic differentiation of vascular smooth muscle cells (reviewed in98).

OPG, a soluble decoy receptor, binds to RANKL to inhibit its interaction with RANK. RANKL

and OPG are differentially expressed in calcified AVs, with RANKL expression co-localized

with areas of calcification and OPG downregulated93. RANKL treatment on cultured VICs

elevated MMP-1and MMP-2 activities72, DNA binding activity of the transcription factor

Cbfa-1/Runx2, bone-related matrix protein expression (e.g., ALP and osteocalcin) and

calcification84. These data suggest that RANKL promotes valvular calcification whereas OPG

elicits anti-calcific effects.

16

Table 2.3. Biochemical factors in valve sclerosis and calcification

Factor General functions Expression in sclerotic/calcified AV*

Effect on AV

Cytokines and growth factors TFG-1 Proliferation, differentiation of

myofibroblast 69 Pathologic

BMP Promote bone and cartilage formation

91 Pathologic

FGF Wound healing response Data not available Pathologic RANKL Bone development,

Connective tissue remodeling 93 Pathologic

OPG Bone resorption 84 Protective Transcription factors

Sox 9 Chondrocyte differentiation 94 Pathologic Cbfa-1/Runx2

Osteoblast differentiation 49 Pathologic

Smad-3 Mediate multiple signaling pathways

(Unpublished data) Pathologic

Egr-1 Proliferation, differentiation and engagement in cell death pathways

95 Pathologic

Other physiological factors ACE Lipoprotein oxidization,

Lipoprotein retention 96 Pathologic

BNP Vasodilation, natriuretic activity (In circulation) 97 Pathologic CNP Vasodilation 19 Protective

* Upregulated expression; Downregulated expression

2.3.3.6. Intracellular Signal: Transcription Factors

An increase in the expression of various transcription factors in aortic valves has been linked to

CAVD (Table 2.3). These factors are often associated with VIC differentiation into pathologic

osteoblasts and myofibroblasts. The transcript expression level of Sox9 and Cbfa1/Runx2, which

are transcription factors for chondrocyte and osteoblast differentiation respectively, were

significantly increased in calcified AVs49, 94. Mechanistically, it has been identified that Notch

homolog 1 (Notch1) regulates osteogenic differentiation by repressing Cbfa1/Runx2

transcriptional activation99 and BMP2 signaling100. Genetic research discovered a Notch1

mutation in a family of individuals who had bicuspid aortic valves and calcific aortic valve

disease. Individuals in the family with calcified tricuspid aortic valves also had a Notch1

mutation. Additionally, a Notch1 frameshift mutation was identified in an unrelated family with

17

similar aortic valve diseases. Subsequent in vitro and in vivo studies confirmed that Notch1

signalling regulates the development of aortic valves and calcification by VICs100. Inhibition of

Notch signaling in cultured VICs led to an increase in valve calcification and upregulated

expression of bone-related transcripts. Additionally, heterozygous Notch1-null

(Notch1+/-) mice had greater than five-fold more aortic valve calcification than age- and sex-

matched wildtype littermates100. These data suggest a signaling pathway that regulates activity of

transcription factors and calcification in CAVD. Egr-1 is another transcription factor that may

play a role in valve calcification. It is a nuclear phosphoprotein that regulates the transcription of

a diverse group of genes such as TGF-1, TNF-, basic fibroblast growth factor, MMPs, and

tenascin-C. Its expression was found upregulated in calcified AVs95. In vitro, the expression of

Egr-1 was upregulated in VICs cultured on non-calcifying substrates, but downregulated in

parallel with an upregulation of osteopontin and tenascin-C in VICs cultured on pro-calcific

substrates95. These confounding in vitro and in vivo results suggest further investigations are

required to determine the precise regulatory function of Egr-1 in cell-mediated calcification.

Transcription factors that regulate the differentiation of fibroblastic VICs to pathological

myofibroblasts may also play a role in CAVD. The Smad3 transcription factor was reported to

mediate myofibroblast activation of cultured porcine VICs, leading to an increase in -SMA

expression64. Similarly, our laboratory observed co-localized expression of Smad3 and -SMA

in sclerotic porcine AVs (unpublished data).

2.3.3.7. Natriuretic Peptides and Cardiovascular Disorders

Recently, increasing attention has been drawn to the role of natriuretic peptides in CAVD

pathogenesis. The family of natriuretic peptides consists of A-type natriuretic peptide (ANP), B-

type natriuretic peptide (BNP), C-type natriuretic peptide (CNP) and Dendroaspis-type

natriuretic peptide (DNP). ANP, BNP, CNP and DNP differ in their structures and functions.

Traditionally, these peptides are known to maintain blood pressure and volume, but other

physiological functions including their role in pathoregulation of cardiovascular diseases have

been recognized in recent years. ANP, BNP and CNP mediate their physiological effects by

binding with membrane receptors: natriuretic peptide receptor type A (NPR-A), natriuretic

peptide receptor type B (NPR-B) and natriuretic peptide receptor type C (NPR-C). NPR-A and

NPR-B are guanyly cyclase-coupled receptors that potentiate the catalytic conversion of

18

guanosine tri-phopshate (GTP) to cyclic guanosine monophosphate (cGMP). NPR-A has high

affinity to ANP and BNP101, whereas NPR-B is activated by CNP102. NPR-C binds to all three

natriuretic peptides. It is known as the “clearance receptor” because it lacks enzymatic activity

and it can remove natriuretic peptides from the extracellular environment through receptor-

mediated internalization and degradation103.

The study of natriuretic peptides in maintaining valve homeostasis is an emerging research field.

There have been an increasing number of studies that demonstrate the relation between BNP,

CNP and valve pathogenesis. The expression of circulating BNP, a 32 amino acid polypeptide

secreted by cardiac atria and ventricles, and its N-terminal fragment (NT-pro BNP) were

elevated in patients with aortic stenosis and the levels correlated proportionally with disease

severity97. BNP and NT-pro BNP have therefore been considered as potential diagnostic

biomarkers of valve stenosis104.

CNP is expressed and stored as a prepropeptide that consists of 103 amino acids. The prepro-

CNP becomes biologically active upon cleavage by furin into a 53 amino acid peptide or by an

unknown enzyme into a 22 amino acid peptide105 CNP binds with NPR-B receptors and induces

cGMP synthesis, which subsequently activates downstream signaling molecules to mediate its

diverse biological effects. For example, cGMP-dependent protein kinase I (PKG I) is activated

by CNP/NPR-B/cGMP signaling, which regulates cardiac contractility106. In the skeletal system,

CNP/NPR-B signaling regulates endochondral ossification via mitogen-activated protein kinase

(MAPK) pathways107, 108.

In calcified human AVs, the expression of CNP and furin were downregulated19. As mentioned

previously in this chapter, microarray studies of healthy porcine aortic valve endothelial cells

found higher expression of CNP on the disease-protected ventricular side6 in comparison to the

ECs on the disease-prone aortic side. In cultured VICs, CNP was able to suppress the formation

of TGF-1 induced calcific aggregates109. These data suggest a plausible protective role of CNP

against valve sclerosis and calcification and is one of the motivations of the current thesis work.

CNP is widely recognized to regulate fibrosis. For example, administration of CNP in animal

models reduced fibrosis associated with vascular intimal thickening110, pulmonary fibrosis111,

and myocardial infarction112. Its efficacy in preventing cardiac fibrosis after myocardial

19

infarction in rats has led to the postulation that CNP might have a potent inhibitory effect on

proliferation of myofibroblast-like cells112. However, it is unclear whether VICs respond to CNP

in a similar manner as observed in other cell types, and more importantly if CNP modulates any

of the cell-mediated pathological processes in CAVD.

2.3.3.8. Progress in Therapeutic Development

Although surgical replacement of the stenotic valve is an established procedure with relatively

high success rate (i.e., operative mortality of less than 10%)113, the operative mortality risk

increases markedly with the presence of other medical complications114 and age115, 116. Further,

many cardiovascular complications may occur in association with the usage of general

anesthesia, thoracotomy and a heart-lung machine. Non-invasive medical treatment, which can

interfere with pathological processes to either halt disease initiation in high-risk patients or slow

disease progression into stenosis, will minimize the need for surgery. The development of

pharmacological therapies for CAVD is ongoing and a number of these compounds have been

tested in animal models or clinical trials. However, the outcome of these clinical studies are

controversial, inconclusive and in some cases, unsuccessful.

LDL accumulation is present at the early stage of CAVD and continues as lesions progress. The

co-localization of angiotensin converting enzyme (ACE) with oxidized LDL in diseased leaflets

suggests a possible role of the renin-angiotensin system in retention and oxidization modification

of LDL96. Retrospective study of patients with mild to moderate aortic stenosis demonstrated no

improvement in these patients upon treatment with an ACE inhibitor117. Consistent with this,

O’Brien et al also found that ACE inhibitors failed to abolish the occurrence of aortic stenosis,

but treated patients experienced a slower rate of disease progression118.

Lipoprotein deposition may be controlled by interfering with the rate-limiting step of cholesterol

biosynthesis with 3-hydroxy-3-methyl-glutaryl-coenzyme A (HMG-CoA) reductase inhibitors

(statins), which inhibit the conversion of HMG-CoA into mevalonate. In cultured VICs, statin

treatment reduced the expression of inflammatory cytokines92, ALP activity73, osteocalcin

expression92 and calcific aggregate formation73, 119, 120, however its mechanism of action remains

unclear. Wu et al. showed the inhibitory effect of statins on aggregate formation by VICs

occured via inhibition of the mevalonate-dependent cholesterol biosynthesis pathway

20

independent of protein prenylation119, whereas Monzack et al found that the decrease in

aggregate number was achieved via modulation of ROCK signaling pathway, and did not rely on

the mevalonate-dependent cholesterol biosynthesis120. Retrospective clinical trials with statins

identified a reduction in calcium accumulation and progression of aortic stenosis117. However,

data from prospective clinical trials are contradictory, with the Scottish Aortic Stenosis and Lipid

Lowering Therapy (SALTIRE) trial indicating statins fail to halt or reverse the progression of

calcific aortic stenosis121 in patients with advanced aortic stenosis, whereas the Rosuvastatin

Affecting Aortic Valve Endothelium (RAAVE) study demonstrated that in patients with

moderate to severe aortic stenosis and hyperlipidemia, statin therapy slowed the progression of

aortic stenosis122. The patient conditions and the assessement criteria of the effectiveness of

statins on CAVD were significantly different in those two clinical trials, which might have led to

the discrepancies in the clinical outcomes. Further, the most recent clinical trial, Simvastatin and

Ezetimibe in Aortic Stenosis (SEAS), demonstrated no effect in mediating CAVD123.

Nonetheless, ongoing prospective clinical trials such as Aortic Stenosis Progression Observation:

Measuring the Effects of Rosuvastatin (ASTRONOMER) and Stop Aortic Stenosis (STOP-AS)

will help delineate the patient populations with aortic stenosis that might benefit from statin

therapy. Other agents that have been considered as potential therapeutics are neutral

endopeptidase124, angiotensin type 1 anatogonists125 and MMP inhibitors (reviewed in126).

However, further investigations are necessary to verify the protective effect of these agents

against CAVD. The molecular mechanisms targeted by these potential agents, including those

that have been tested clinically, remained poorly understood. In vitro VIC studies may facilitate

the identification of cellular mechanisms of these therapeutic interventions prior to animal and

clinical tests.

2.4. Mechanobiology Although it is generally recognized that biochemical factors influence functions and phenotypes

of valve cells, recent studies have shown that these cells, similar to others, are also responsive to

mechanical cues from their microenvironment. Valve cells experience an array of mechanical

forces including shear stress, compression, tension and flexure (reviewed in127), some of which

have been shown to regulate VIC behaviour. For instance, cyclic strain reduces the expression of

pro-inflammatory genes by VICs128, clearly demonstrating the mechanosensing ability of this

21

cell type. Presumably, VICs are similar to other adherent cells in the body, which require

interactions with the ECM to maintain proper physiologic function. The regulation of cell

functions by the ECM has traditionally been recognized purely as a series of biochemical

processes involving interactions between matrix proteins (e.g., collagen, elastin, and

proteoglycans) and cell surface receptors. It is increasingly evident that adherent cells are

responsive to the inherent mechanical properties of the ECM in vivo as well as the adhesion

substrate in vitro. As reviewed in this section, the stiffness or rigidity of the adhesion surface

(i.e., matrix stiffness) is a potent regulator of cell morphology129, differentiation130, 131 and

mineralization132 in various non-valve cell types. Matrix stiffness has important clinical

implications in disease (reviewed in133) such as osteoporosis, atherosclerosis16 and liver fibrosis7,

in which progression of these diseases often involve alterations of the physical properties of the

ECM. However, the role of matrix stiffness in valve pathogenesis has yet to be identified, partly

due to the lack of study of mechanobiology of VICs as well as suitable culture systems that can

decouple biochemical cues from mechanical cues. In this section, the mechanisms by which cells

sense the stiffness of their microenvironment are described. A summary of experimental

approaches and findings related to matrix mechanics-mediated cell responses are presented. The

association of matrix stiffness with various pathologies is discussed.

2.4.1. Definition of stiffness

Stiffness is the resistance of an elastic material to deformation by an applied stress and is

typically measured in Newtons per metre (N/m). The inherent stiffness of an isotropic elastic

material (independent of specimen geometry) can be described by the elastic modulus (E), which

is the ratio of stress over strain within the linear range (Hooke’s Law); its unit is N/m2 or Pascal

(Pa). Elastic modulus can be calculated from the slope of the linear region of a stress-strain curve

that is experimentally determined from tensile tests or compression tests. Alternatively, the

inherent stiffness of a material can be represented by the shear modulus (G). The shear modulus

can be determined by measuring the deformation of an object when a force is applied parallel to

one face of the object while the opposite face is held fixed by another equal force. The unit of

shear modulus is Pascal (Pa). Most cells and tissue have elastic moduli on the order of 101 to 106

Pa (Table 2.4), which is much more compliant than glass or tissue-cultured treated polystyrene

(TCPS) commonly used in cell culture. Therefore, studies of the effects of substrate stiffness on

22

cell function may not only improve our understanding of cell mechanobiology in general, but

also may clarify the interpretation of in vitro data and better relate these data to in vivo settings.

Table 2.4. Compliance of culture materials and tissues

Culture materials/ tissues Compliance (Pa) TCPS ~ 1012 Pa134 Glass ~ 1012 Pa135 Early atherosclerotic lesion ~ 3 x 104 – 5 x 104 Pa16 Mature collagenous bone > 105 Pa130 Brain ~ 102 – 103 Pa130 Striated muscle ~ 8 x 103 – 17 x 103 Pa136 Tendon and cartilage ~ 106 Pa137 Liver ~ 3 x 102 – 6 x 102 Pa137 Fibrotic liver ~ 2 x 104 Pa137 Mammary glands ~ 1.5 x 102 Pa138 Breast tumours ~ 4 x 103 Pa138

2.4.2. Stiffness Sensing

While it is clear that cells response to matrix mechanics, the molecular mechanisms by which

load-sensitive cells process mechanical cues into intracellular signals are still open to

investigation. Traditionally mechanotransduction was recognized as the balance of external-

internal force leading to the sequential processes of “force-signaling-response”; however, it is

increasingly apparent that mechanotransduction requires the interplay of physical and

biochemical signals in a dynamic feedback system involving the interactions of the ECM,

surface receptors and the cytoskeleton (Figure 2.4). The formation of contacts between cells and

the ECM through surface receptors is a crucial component of mechanosensing. Integrins, a

family of transmembrane heterodimeric glycoproteins with - and - subunits, are one of the

most studied mechanosensing surface receptors. Integrin activation begins with conformational

changes of the integrin - and - subunits, which is essential for the initiation of focal

complexes. The binding of integrins to ECM proteins induces the recruitment of scaffolding

proteins inside the cell to physically link the ECM with the actin cytoskeleton at the site of focal

adhesions139. Some of the proteins known to facilitate integrin-cytoskeleton linkages include

talin140, integrin-linked kinases141 and tensin142. The maturation of focal complexes into focal

adhesions depends on the force applied to the ECM-integrin-cytoskeleton connections either

externally (e.g., ECM motion) or internally (e.g., actin polymerization). Mature focal adhesions

23

are localized at the termini of stress fibers, consisting of bundles of actin filaments. The

participation of focal adhesions in sensing matrix stiffness is evident with the suppression of

focal adhesion maturation and reduced phosphorylation of focal adhesion kinase in cells cultured

on compliant substrates143. Further, myofibroblasts lacking the cytoskeleton protein, -SMA, fail

to form mature focal adhesion144, emphasizing the importance of internal actin-mediated force

generation in this process. Of note, -SMA is an abundant cytoskeleton protein found in smooth

muscle cells and myofibroblasts. Polymerization of monomeric -SMA into filamentous actin is

associated with increased cellular contractility. Interestingly, the expression of filamentous

-SMA is dependent on matrix stiffness, suggesting a role of cell-generated tension in regulating

filament formation (reviewed in145) as well as a bi-directional communication between integrin-

ECM adhesion and the actin cytoskeleton.

Matrix stiffness influences integrin-ECM contacts as well as the formation and maturation of

focal adhesions. Formation of focal adhesions and reorganization of the cytoskeleton in

mechanically loaded cells are associated with biochemical signaling pathways known to

modulate cell differentiation, shape, proliferation and motility (summarized in Table 2.5; also see

detailed reviews 146-148). Many of these biochemical pathways are differentially activated on

substrates of different stiffnesses. For instance, RhoA has been widely implicated in integrin-

mediated signaling and mechanotransduction, and was found to modulate substrate stiffness-

dependent osteogenic differentiation in pre-osteoblasts131, 132. Notably, the signaling pathways

involved in stiffness-mediated responses depend on the chemical nature149, 150, the spatial

distribution151, 152, and the topography153 of the adhesive surface, which affect ECM-integrin

interactions and the subsequent activation of specific biochemical pathways. Therefore, substrate

stiffness and surface chemistry responses are often coupled, which can confound interpretation

of experimental data obtained from some of the culture systems.

24

Figure 2.4. The interplay of physical and biochemical signals in a dynamic feedback system

involving the interactions of ECM, surface receptors and cytoskeleton (adapted from154)

Mechanosensing is not a sequential process of “force-signaling-response”; rather it requires the

cell to continually probe and respond to the physical properties of their substrate.

2.4.3. Test Systems: Engineering the Stiffness of Culture Substrata

Various culture systems have been developed to study the effect of substrate mechanics on cell

function (Table 2.6), with hydrogels being the most commonly used. Hydrogels are networks of

natural and/or synthetic polymer chains that are water-soluble. Natural polymers made of ECM

proteins such as collagen, polysaccharide and MatrigelTM are often used because of their

physiologic relevance and their ability to undergo polymerization without the addition of toxic

cross-linking agents. Changing the stiffness of natural polymer-based systems is accomplished

by either altering the concentration of the proteins, the physical constraints, the geometry of the

system or by thermal/chemical treatment of the polymers. However, each of these approaches

has drawbacks. For example, alteration of protein concentration simultaneously alters the

stiffness, the density and the distribution of ECM adhesion ligands, which complicates

interpretation of cellular responses as cell processes are regulated by both stiffness and ligand

density155. The use of thermal or chemical treatment of polymers to modify the extent of cross-

linking often denatures ECM proteins, again affecting surface chemistry. Alteration of the

physical constraints or the geometry of the system lacks the ability to fine-tune substrate

stiffness. Nonetheless, natural polymers remain an excellent culture platform because of their

25

physiologic relevance, potential use as implantable biomaterials and their high adhesiveness to

cells making it an easy platform to use with minimal experimental optimization (reviewed in10).

Alternatively, synthetic polymeric gels offer several advantages over natural polymer-based

systems as tools to study matrix regulation of cell function. Synthetic polymeric gels are inert

and stiffness can be varied over a wide range (e.g., shear moduli of polyacrylamide gels range

from 10 Pa to 5 x 103 Pa) by changing the cross-linker concentration129. The surface can then be

functionalized by coating with physiologically- relevant ECM proteins for cell adhesion. This

allows one to control both the density and the spatial distribution of the ligands while

independently varying the stiffness, therefore decoupling surface chemistry from matrix

mechanics. However, our laboratory has observed that the standard surface functionalization

procedure for polyacrylamide gels often leads to poor adhesion of primary VICs. Additional

efforts to improve surface functionalization are necessary to provide better substrate-cell

interactions.

2.4.4. Effect of Matrix Stiffness on Cell Response

The effect of substrate mechanics on cell phenotype and function has been demonstrated in two-

dimensional cultures (i.e., cells grown on top of the substrate) and in three-dimensional cultures

(i.e., cells embedded within the matrix). As reviewed below, several recent studies have reported

that regulation of cell shape, spreading, migration, growth and differentiation by matrix stiffness

is cell-type specific.

2.4.4.1. Cell Shape and Spreading

The size and the spatial distribution of cell-ECM contacts are determined by the extent of cell

spreading. The cell arranges its cytoskeleton to exert force against its adhesive contacts, defining

the amount of force that can be exerted against the ECM. It is believed that the extent of cell

spreading determines the reorganization of the cytoskeleton and the magnitude of force that the

cytoskeleton can generate156. This in turn regulates a series of cellular processes including cell

migration157, ECM remodeling158 and apoptosis152. Cells cultured on stiff substrates generally

spread more. 129. When cells are cultured on substrates with discrete changes in stiffness159, cells

migrate toward stiffer regions of the substrate. Presumably, reduced cell spreading on compliant

substrate is related to the formation of fewer and smaller focal adhesions and reduced

26

cytoskeletal tension, resulting in smaller tractional forces generated by the cell160. Notably, the

effect of matrix stiffness on cell spreading and migration is cell type specific: for example, the

stiffness at which migration rate is maximal differs for smooth muscle cells161, preosteoblasts132

and neutrophils129, suggesting the necessity to study mechanoregulation of each cell type

individually.

Although it may seem that mechanically-regulated cellular responses are mediated through

changes in cytoskeletal structure, leading to deformation of the plasma membrane and changes of

cell shape, mechanotransduction can occur without large-scale changes in cell shape. Direct

application of mechanical stress to cell surface receptors with a magnetic twisting device was

able to induce focal adhesion formation and cytoskeleton stiffening without changes in global

cell shape162. In addition, studies that constrained cell shape still displayed different focal

adhesion dynamics on compliant versus stiff substrates163, suggesting the relation between

stiffness sensing and cell shape is more complex than originally anticipated.

2.4.4.2. Cell Growth and Death

Cells generally proliferate faster on stiffer substrates than on more compliant substrates166, 167,

however the optimal stiffness for proliferation is cell type specific168. It is increasingly clear that

substrate stiffness mediates proliferative effects by manipulating the sensitivity of cells to soluble

growth factors169, 170 and their ability to progress through the cell cycle (reviewed in171).

Engagement of integrins with ECM proteins not only contributes to the formation of focal

adhesions, actin cytoskeleton reorganization and alteration in cell tension, but also to clustering

of integrins that promotes the activation of signaling components such as growth factor receptors

and their downstream targets such as phosphoinositide-3 kinase/Akt survival signaling172. For

example, binding to αvβ3 integrin is associated with the activation of platelet-derived growth

factor (PDGF) receptor, the insulin and insulin-like growth factor-I (IGF-I) receptors, and the

vascular endothelial growth factor (VEGF) receptor-2, supporting the role of integrin-based

matrix adhesion in cell growth-related intracellular signaling173.

27

Table 2.5. Matrix stiffness mediated cellular responses

Response Matrix stiffness Cell type Reference Osteogenic differentiation increases with substrate stiffness involving activation of MAPK pathway downstream of RhoA and ROCK.

Soft substrate ~ 7 kPa Stiff substrate ~ 160 kPa

MC3T3-E1 Khatiwala et al. 2009164

Greater cell spreading with higher expression of actin stress fiber and focal adhesion on stiff substrate.


Smooth muscle cells

Engler et al. 2004136,

155

Cell shape and spreading unaffected by substrate stiffness.

Soft substrate ~ 2 kPa Stiff substrate ~ 1012 Pa

Neutrophils Yeung et al. 2005129

Cells differentiated to neuronal, myogenic and osteogenic lineages when cultured on soft, immediate and stiff substrate, respectively.

Soft substrate ~ 0.1 – 1 kPa Immediate stiffness ~ 8 –17 kPa Stiff substrate ~ 25 – 40 kPa

Marrow-derived mesenchymal stem cells

Engler et al. 2006130

Formation of focal adhesion and stress fibers depends on substrate stiffness.


Smooth muscle cells

Peyton et al. 2005161

Soft substrate downregulated cyclin D1 and upregulated p27Kip1, leading to reduced cell proliferation.

Free floating gels (soft) Constrained gels (stiff)

Fibroblasts Fringer et al. 2001165

Formed striations only when cultured on intermediate (~ 11 kPa) stiffness.

Substrates at 1, 8, 11 and 17 kPa

Myoblasts Engler et al. 2004136

28

Table 2.6. Advantages and disadvantages of various culture systems for studying the effect

of substrate stiffness

Substrate Advantages Disadvantages Collagen - Easy to handle

- Physiologic relevant - Readily adhesive to cells - Can support 3D culture - Can decouple substrate chemistry from mechanics if stiffness is adjusted by changing the geometry in 2D culture

- Cannot support a wide range of stiffness, unless concentration of collagen is adjusted

MatrigelTM - Can tune the stiffness by cross-linking the gels using glutaraldehyde

- Heterogeneous composition and batch-to-batch variability - Cannot decouple substrate chemistry from mechanics as stiffness is adjusted by altering the concentration of MatrigelTM

Fibrin - Can support 3D culture - Cannot decouple substrate

chemistry from mechanics

Agarose - Easy to handle - Not readily adhesive to cells, requires surface modification - Cannot decouple substrate chemistry from mechanics

Polyacrylamide - Can adjust to a wide range of stiffness

- Limited to 2D culture - Cross-linker is toxic to cells - Requires surface functionalization

Poly(ethylene glycol) - Can adjust to a wide range of stiffness - Can generate 3D culture - Has physical characteristics similar to those of soft tissues, e.g., permeable to oxygen, nutrients and water-soluble metabolites

- Requires surface functionalization - When used in 3D culture, substrate chemistry and mechanics cannot be decoupled

Polydimethylsiloxane - Can adjust to wide range of stiffness - Surface chemistry and substrate stiffness are decoupled

- Surface functionalization is difficult - Can only support short culture periods of a few days

29

Substrate stiffness modulates proliferation by interfering with cell cycle progression (i.e., the

cycle of G1 phase, S phase, G2 phase and M phase). Focal adhesion kinase (FAK) and actin

polymerization were found to regulate the expression of the cell cycle regulatory proteins and

cylcin-dependent kinases (cdk) through extracellular signal-regulated kinase (ERK), Ras and

Rho signaling pathways depending on the level of compliance. For example, cells cultured on

free-floating collagen gels exhibited reduced cell proliferation due to G1 phase arrest, which was

accompanied by reduced FAK autophosphorylation, ERK activity, and absence of cyclin

expression and upregulation of cdk inhibitor expression. Stiffer substrates favour the formation

of mature focal adhesions and FAK activation, which is associated with ERK activation and

ERK-dependent induction of cyclin expression, hence permitting the cells to enter the S phase.

Of note, ERK- and FAK-independent induction of cyclin has also been reported and further work

is required to identify the interplay among the different mechanisms by which substrate stiffness

alters cyclin and cdk expression171.

Without doubt, cell-ECM contacts are an important aspect of stiffness sensing. It is also known

that the lack of a firm adhesive substrate contributes to anoikis (a type of apoptosis which is

induced by the detachment of cells from their adherent surfaces). Presumably, the higher number

of mature focal adhesions in cells cultured on stiffer substrates provides a firm adhesive substrate

and protects the cells from apoptosis. This notion is supported by the reduction in apoptosis of

pre-osteoblasts when cultured in stiffer substrates174. It is known that apoptosis is associated

with integrin signaling and cytoskeleton content175, but the direct relation between stiffness-

mediated cytoskeleton reorganization and apoptosis remains unknown.

2.4.4.3. Cell Phenotype and Differentiation

While the potential of mesenchymal stem cells to differentiate into a diverse range of lineages is

well-known, other cell types in the body such as VICs can also undergo phenotypic changes

during pathologic differentiation47, 49. Recent studies demonstrate that cellular differentiation is

not only regulated by soluble factors and matrix composition, but also depends on substrate

stiffness. Notably, cells appear to express their differentiated phenotype in vitro on substrates

that are similar to the stiffness of their native ECM130, 136. It has therefore been postulated that

changes in tissue mechanics during pathologic development may partly contribute to cellular

pathological differentiation16, 176. Further, matrix stiffness-induced phenotypic alterations may

30

influence the interpretation of in vitro studies, which are often conducted with glass or TCPS that

are orders of magnitude stiffer than any tissue. Lastly, the fundamental understanding of matrix

mechanics in regulating differentiation has substantial influence on the development of

biomaterials for tissue engineering as material stiffness may influence cell differentiation and

functions.

The ability to differentiate stem cells into a desired lineage is important in tissue engineering and

regenerative medicine. Interestingly, Engler’s study showed that mesenchymal stem cells

differentiate into different lineages as a function of substrate stiffness130. Mesenchymal stem

cells grown on substrates with brain-like compliance (E ~ 0.1-1 kPa) underwent neuronal

differentiation. Myogenic differenation was observed in cells cultured on substrates of

intermediate stiffness (E ~ 8-17 kPa) and osteogenic differentiation was found in cells grown on

relatively stiff substrates (E ~ 25-40 kPa). These data suggest that cells are able to recognize

physiologically relevant substrate stiffnesses and differentiate accordingly. Engler et al further

demonstrated that soluble factors and matrix stiffness synergistically guide MSC commitment to

particular lineages130.

The effect of substrate stiffness on differentiation is not limited to progenitor cells, but also

impacts on various cell lines and committed mammalian cells (e.g., pre-osteoblastic cells,

hepatocytes, myofibroblasts, VICs). One of the most frequently studied cell lines is the pre-

osteoblastic MC3T3-E1 cell line, which displays substrate stiffness-dependent osteogenic

differentiation and calcium deposition. When MC3T3-E1 cells were cultured on substrates

coated with RGD rather than full-length type I collagen, osteogenic differentiation was optimal

on more compliant substrates (E = 20 kPa)177. However, the same cells cultured on type I

collagen-coated substrates deposited more bone mineral on stiff substrates (E > 20 kPa)132.

These differences may result from the unique engagement of integrins to particular ECM

proteins (i.e., typically αvβ3 and α5β1 integrins interact with RGD and α2β1 interacts with type I

collagen), which together with matrix stiffness differentially regulate cellular differentiation.

This emphasizes the importance of decoupling surface chemistry from substrate mechanics.

Similar matrix stiffness effects on cell-mediated calcification were also found in VICs. An

increase in calcium deposition by VICs was identified when cultured on fibrin-modified TCPS

(i.e., a stiff substrate) in comparison to those cultured on fibrin-modified soft polyethylene

31

glycol178. These results demonstrate that matrix stiffness regulates cell-mediated calcification.

Calcification by cultured cells is associated with integrin binding and FAK activation149.

Presumably, substrate stiffness may regulate calcification by influencing integrin binding, focal

adhesion signaling, and possibly cytoskeleton reorganization, although further studies are

necessary to decipher the precise molecular mechanisms.

Matrix stiffness also regulates and maintains the differentiation of committed cell types. This has

implications in disease development, where ECM composition and tissue compliance are often

altered and phenotypic changes of cells are observed. For example, hepatocytes cultured on soft

substrates (e.g., collagen gels and MatrigelTM of G′~ 34 Pa) remain differentiated179, whereas

those cultured on stiff substrates (e.g., collagen film and cross-linked MatrigelTM of G′~118 Pa)

adopt a dedifferentiated phenotype179, 180. Another example is the differentiation of fibroblasts to

contractile myofibroblasts in wound repair, in which matrix stiffness, cell tension, and TGF-β

release are all required to promote and maintain myofibroblast differentiation that is responsible

for wound closure (reviewed in137, 181). In vitro studies of myofibroblasts cultured on stiff

constrained collagen gels (E ~ 20 kPa) responded to TGF-β and expressed a high level of

-SMA and abundant stress fibers. In contrast, cells cultured on free floating (more compliant)

collagen gels (E ~ 8 kPa) had reduced -SMA expression and stress fibers regardless of the

presence or absence of TGF-β, suggesting that matrix stiffness regulates cellular sensitivity to

soluble factors related to myofibroblast differentiation12.

2.4.5. Matrix Stiffness and Pathologies

Regulation of cellular responses by substrate stiffness is not just an in vitro phenomenon, but

also has relevant clinical implications. Studies have linked matrix stiffness to various pathologies

such as cancer, osteoporosis and atherosclerosis. Significant stiffening of tumour tissue has been

correlated with an increase in tumour cell migration and proteolysis182, which has been suggested

to partially explain the failure of protease inhibitors as cancer therapies183, 184. This result

emphasizes the value of understanding the effect of matrix stiffness on the response of cells to

biochemical factors including therapeutics.

Fibrosis and tissue stiffening are often found as liver disease progresses. The differentiation of

portal fibroblasts and hepatic stellate cells into hepatic myofibroblasts is the key pathological

32

mediator. Interestingly, studies with rat models found that liver stiffening occurs prior to

fibrosis7, 185, suggesting that matrix mechanics may induce myofibroblast differentiation in early

liver disease. Matrix stiffness also impacts vascular diseases. Vascular calcification correlates

with arterial stiffening. Dynamic tissue stiffening was found in atherosclerotic lesions, in which

the initial soft lesion was mainly composed of foam cells, and subsequent stiffening of the lesion

was accompanied by the presence of smooth muscle cells. Further stiffening of the artery occurs

when there is marked calcification within the tissue (Figure 2.5). These data suggest a close

correlation between tissue stiffness and its pathology16. Although atherosclerosis and CAVD

shared many risk factors and disease features, it has yet to be established if similar mechanical

changes occur as the disease progresses from lipid deposition to fibrosis to eventually the

formation of calcific nodules. Studies with normal and abnormal mitral valves support a tight

association between mechanical changes and valve pathogenesis186, however such data for the

AV will have to be determined in the future.

33

Figure 2.5. Change in arterial wall stiffness in rabbit fed a high cholesterol diet to induce

atherosclerosis

Foam cells accumulated at the site of lesions after eight weeks on high cholesterol diet. The

stiffness of these early lesions was significantly lower than that of the normal tissue. The

stiffness of lesions increased substantially by week 28, in which marked calcification was found

in the tissue16.

34

Chapter 3 3. Hypotheses, Objectives and Contributions

3.1. Rationale

Myofibroblasts and osteoblasts form in sclerotic, calcified valve leaflets through processes that

are not well-defined47, 49. Although the aortic valve undergoes significant matrix remodeling and

changes in stiffness during CAVD, the role of the mechanical properties of the extracellular

matrix in regulating VIC differentiation have yet to be investigated. Matrix stiffness influences

the differentiation of and mineralization by other non-valve cell types132, 185, and may have

similar effects on VICs. Biochemical factors also regulate the pathological differentiation of

VICs69. One particularly interesting secreted protein is CNP, as it has been identified to

putatively protect against CAVD. CNP regulates calcification and differentiation of other non-

valve cell types187, 188 and its anti-fibrotic effect has been demonstrated in animal model

studies110-112, suggesting it may suppress differentiation of VICs into pathological myofibroblasts

and osteoblasts to prevent calcific aortic sclerosis. Furthermore, as matrix stiffness modulates the

response of cells to biochemical factors12, it is expected that CNP-mediated cellular responses

are affected by matrix stiffness.

3.2. Thesis Hypotheses

It is hypothesized that the response of VICs to biochemical stimuli is regulated by matrix

stiffness, which in turn modulates the pathological differentiation of and calcification by VICs.

In particular, I hypothesized that:

I. Matrix stiffness regulates the differentiation of VICs into myofibroblasts and

osteoblasts, phenotypes associated with calcification;

II. CNP, a putative protective agent, suppresses myofibroblast and osteoblast

differentiation of VICs, thereby inhibiting calcification by VICs; and

III. Matrix stiffness, CNP and their combination differentially regulate mRNA expression

by VICs.

35

3.3. Objectives and Specific Aims

The objectives (Roman numerals) and specific aims (Arabic numerals) of this thesis are:

I. To implement and characterize a cell culture system with tunable stiffness

1. Design and fabricate collagen matrices with tunable stiffness

2. Determine the mechanical properties of the matrices

3. Evaluate the surface topography of the collagen matrices

4. Evaluate the integrity of the collagen matrices over culture duration

II. To evaluate the response of VICs to pro-calcific biochemical factors and substrate

stiffness

1. Determine proliferation and morphological responses of VICs cultured on the

collagen-based culture system

2. Utilize the culture system to evaluate the effect of matrix stiffness on calcification by

VICs induced by pro-calcific media

III. To identify the cellular basis of CNP in protecting against valve calcification

1. Investigate the ability of CNP to regulate pathological differentiation of VICs

2. Identify the effect of CNP on calcification by VICs

IV. To identify the transcriptional changes in VICs when exposed to mechanical

stimuli and biochemical stimuli

1. Investigate morphological changes of VICs cultured on compliant and stiff

substrates with and without CNP treatment

2. Investigate if matrix stiffness, CNP and their combination differentially regulate

transcript expression of VICs

3.4. Overview of Contributions

The work presented in this thesis was made possible with the collaborative efforts from various

students and staff. Included in this section is a record of their contributions in implementing and

executing the proposed research ideas.

36

Technical support for the work on matrix stiffness presented in Chapters Four and Five was

provided by: Stephanie Ting, an undergraduate summer research student under my supervision,

who performed the Swinholide A dose-dependent experiment and the cyotoxcity test; Zahra

Mirzaei who did the TGF-1 ELISA and cGMP synthesis assay; Jian Wang and Robert

Chernecky from the Faculty of Dentistry who assisted with scanning electron microscopy; and

Ruogang Zhao who did the finite element models and analyses. I would like to also acknowledge

numerous insightful discussions with Jan-Hung Chen throughout the project.

The study of the effect of CNP on the pathological differentiation of VICs is described in

Chapter Six. The in vitro study of CNP began with the help from Susie Ferrante and Melissa

Filice, who were Grade 12 students under my supervision. Xiao Zhong did the clonal assay

under my supervision with help from Jan-Hung Chen. Zahra Mirzaei did the immunostaining of

CNP, -SMA and Cbfa-1/ Runx2 in valve leaflet sections. Also, I would like to acknowledge

help from Mark Blaser with siRNA transfection experiments. Lastly, the microarray experiments

in Chapter Seven were done by Kelly Jackson and partial data analysis was performed by Zhibin

Lu at the Microarray Centre of the University Health Network.

37

Chapter 4 4. Implementation and Characterization of the Cell Culture

System

A variety of hydrogel-based matrices have been developed to study the effects of substrate

stiffness on cell function. As reviewed in Chapter Two, a common feature of these systems is the

ability to tune the substrate stiffness by either altering the structural stiffness of the matrices or

by altering the elastic modulus of the material that makes up the matrices. Substrate chemistry

and mechanics are not decoupled in some of these culture systems, because adjustment of the

substrate stiffness often requires changes in ECM protein concentration, which alters ligand

availability and density. In the case where synthetic polymers are used to form an inert surface

with defined stiffness, surface modification is needed to increase the adhesiveness of the surface

to cells. Fine-tuning of the substrate stiffness can be achieved in such systems, however surface

modification can be time consuming and costly, and sometimes does not provide the desire

adhesiveness to cells. As shown in our laboratory, primary VICs adhere poorly on

polyacrylamide substrates coated with monomeric collagen. Additional optimization is often

necessary to improve the adhesiveness of these inert polymer-based platforms to specific cell

types189-191. The relatively complex fabrication process and increased cost associated with

surface modification to some extent limits the widespread appeal of these systems for cell

biology studies. To address these issues, we used a collagen gel-based culture system described

previously11 to test the sensitivity of VIC calcification to substrate stiffness. The culture system

was chosen with the following criteria:

Tunable stiffness while maintaining identical surface chemistry

Maintenance of the substrate stiffness for at least five days, as this is the shortest

culture duration necessary for VICs to form calcific aggregates120, 192

Optical transparency for visualization of cells by bright field and fluorescent

microscopy

Composed of ECM proteins that mimic those of native heart valve

The work presented here was published in Arterioscler Thromb Vasc Biol. 2009 Jun; 9(6): 36-42.

38

Ease of handling

Readily supports VIC adhesion and proliferation without any surface modification of

the substrate

The system used in this work satisfied all of these selection criteria. Matrices were composed of

bovine type I collagen, constrained to wells of microtiter plates. Two different stiffnesses were

achieved by changing the thickness of the matrices, while maintaining the same surface area and

chemistry for cell adhesion. We further characterized the culture system by evaluating the

biochemical and mechanical changes that occur in the collagen matrices over the culture period

to better define the relative contribution of substrate stiffness to the observed cellular response.

4.1. Materials and Methods

Unless otherwise stated, all reagents were purchased from Sigma-Aldrich (Oakville, ON,

Canada). Detailed protocols can be found in Appendix A.

4.1.1. Fabrication of Collagen Matrices

Fresh collagen solutions were prepared for each experiment as described previously193. Briefly, a

mixture with the following chemicals was prepared on ice: (1) 0.3 mL 10X concentrated

Dulbecco's Modified Eagle Medium (DMEM); (2) 0.3 mL 0.25 M sodium bicarbonate

(NaHCO3) buffer; (3) 0.3 mL fetal bovine serum (FBS, Hyclone, Logan, UT, lot # KRA25425);

(4) 0.3 mL penicillin/streptomycin mixture; (5) 0.12 mL 0.1 M sodium hydroxide (NaOH)

buffer; and (6) 2.5 mL 3 mg/mL bovine collagen (PureCol, Advanced BioMatrix, San Diego,

California). For thick collagen matrices, 500 L of collagen mixture was pipetted into each well

of a 24-well microtiter plate lined with sterile coverslips. Polymerization of collagen was

achieved by incubating the collagen mixture at 37 oC in a 5% CO2 incubator overnight. To make

thin collagen matrices, the same volume of collagen mixture was applied to the surface of the

well and the coverslips for one minute at room temperature. Excess collagen mixture was then

removed by aspiration, leaving a thin uniform collagen coating in the well that was polymerized

overnight.

4.1.2. Scanning Electron Microscopy for Topographic Evaluation

Thick and thin collagen matrices with or without cells were evaluated by scanning electron

microscopy (SEM). Samples were fixed with 4% formaldehyde, followed by dehydration in a

39

series of ethanol washes at 30%, 50%, 70%, 95% and 100% ethanol for 30 minutes each.

Samples were then critical point dried with liquid carbon dioxide in a Polaron CPD7501,

mounted on SEM aluminum stubs and sputter coated with gold using a Polaron SC 515 SEM

Coating System. The samples were examined at 1,000 X to 5,000 X magnifications using a

scanning electron microscope (Model S-2500, Hitachi Instrument). For collagen matrices

without cells, images were used for estimating the collagen fibril diameters with ImageJ software

(NIH, Bethesda, MD).

4.1.3. Determination of Matrix Mechanics

The effective stiffness of the initial collagen matrices (hydrated) as well as those cultured for

three and eight days was measured in compression using a Biosyntech Mach-1 mechanical test

system (Laval, QC). Constrained thick and thin matrices (n = 4 of each) were compressed within

24-well culture plates using an eight mm diameter loading platen. Load-displacement curves

were recorded, from which the effective stiffness was determined as the initial tangential slope.

To demonstrate the relative difference in the stiffness of the thick and thin collagen matrices

under shear loading that better mimics the tractional forces applied by cells on the matrices, we

performed finite element (FE) analysis using ANSYS (Canonsburg, PA). Two-dimensional,

plane strain FE models of the matrices, representing a vertical slice through the thick and thin

matrices, were constructed. The thickness of the thick and thin collagen matrices was set to the

initial thickness of 2.5 mm and 10 μm, respectively, to mimic the experimental conditions. To

demonstrate the relative difference in shear deformation between the thick and thin matrices, an

arbitrary constant shear force of 0.48 μN/mm was applied to the top surface of both matrices

over the entire width of the model. Eight-node elements were used to discretize the models.

Collagen was modeled as a hyperelastic neo-Hookean material with shear modulus of 30 kPa.

The gels were constrained in both the horizontal and vertical directions on the left and bottom

edges. The right edge of the models was left unconstrained to allow for measurement of the shear

deformation.

4.1.4. Measurement of Collagen Content

Collagen content of the thick and thin matrices was measured by colormetric hydroxyproline

assay after zero, three and eights days in culture. Briefly, collagen matrices were papain digested,

40

followed by release of hydroxyproline with acid hydrolysis using 6 N hydrochloride acid (HCl).

The hydroxlate was then neutralized with 5.7 N NaOH. The extracted hydroxyproline was

oxidized into a pyrrole with 0.05 N chloramine T, followed by treatment with

4-dimethylaminobenaldehyde to develop a colour change. The amount of hydroxyproline was

quantified by measuring the absorbance of the solution at 560 nm.

4.1.5. Statistical Analysis

Results are presented as mean standard error. Samples sizes were at least three in all cases,

and experiments were repeated at least three times. Unpaired Student’s t-test was used for

comparisons between two groups. ANOVA and Fisher’s least significant difference test were

used to evaluate statistically significant differences in multiple group comparisons.

4.2. Results

4.2.1. Collagen Matrices with Tunable Stiffness

Fibrillar collagen matrices were made in wells lined with coverslips. The coverslip facilitates

matrix removal from the well for immunostaining and SEM. Using a similar methodology as

previously described11, collagen matrices of two different stiffnesses were achieved by altering

only the thickness of constrained collagen matrices. This allows the same culture surface area

(~ 2 cm2) and collagen concentration to be used for all culture conditions, hence keeping the

ligand density consistent when cells are seeded on the surface only. Thick collagen matrices

were made with 500 L of collagen mixture, which gave rise to matrices with thickness

approximately 2.5 mm (Figure 4.1). Thin collagen matrices were achieved by coating the

coverslips with collagen mixture of the same concentration (1.97 mg/mL). The approximated

thickness of the thin matrices was 10 m (estimated previously11 (Figure 4.1)).

41

Figure 4.1. Type I collagen matrices

(A) A schematic of the experimental set up and (B) Wells of a 24-well plate with a thick collagen

matrix (top) and a thin collagen matrix (bottom). Cells were subsequently cultured on the surface

of the matrices.

The effective stiffness in this system is dictated in part by the geometry and boundary constraints

of the matrices in the wells, and therefore tensile stiffness could not be measured in situ and

compression tests were performed on constrained collagen matrices instead. Because of the

differences in the matrix thickness, the thick matrices were significantly more compliant than the

thin matrices in compression (P < 0.05, Figure 4.2 A). We performed FE analysis to estimate the

shear loading that mimics the tractional forces applied by cells on the matrices. The model

estimated a maximum shear displacement of 10.16 μm and 0.36 μm for the thick and thin

matrices, respectively (Figure 4.2 B). These results indicate that the effective stiffness of thick

matrices under shear forces was significantly less than that of the thin matrices, as expected and

consistent with the results from the compression tests.

Coverglass

Collagen matrix

Thick (2.5 mm) Thin (10 m)

A B

Thick matrix

Thin matrix

42

Figure 4.2. Mechanical properties of thick and thin collagen matrices

(A) Thin matrices were three times stiffer than the thick matrices in compression. * P < 0.05. (B

and C) FEA model of the thick and thin collagen matrices. The shear displacement was larger on

the thick matrices in comparison to the thin matrices (colour scale bar represents the magnitude

of displacement in mm)

A. B.

0

2000

4000

6000

8000

1 2

*

Co

mp

ress

ive

Stif

fne

ss(N

/m)

Thick Thin

C.

1.88 x 10-4 1.02 x 10-2

0 0.30 x 10-3

4.11 x 10-3

1.31 x 10-4

Displacement (mm)

Displacement (mm)

A. B.

0

2000

4000

6000

8000

1 2

*

Co

mp

ress

ive

Stif

fne

ss(N

/m)

Thick Thin

C.

1.88 x 10-4 1.02 x 10-2

0 0.30 x 10-3

4.11 x 10-3

1.31 x 10-4

Displacement (mm)

Displacement (mm)

43

4.2.2. Substrate Topography

Substrate topographical features, such as fiber diameter, can significantly influence cell

adhesion, migration and proliferation194. Thick and thin matrices displayed uniform fibrillar

collagen microstructure (Figure 4.3: A, B). Collagen fiber diameters ranged from 0.10 to 0.15

m and were not significantly different between the two matrices (P = 0.27, Figure 4.3: C).

These data indicate that alteration of only the thickness of these matrices did not influence the

microstructure of the culture surface.

Figure 4.3. Microstructure of collagen matrices

(A and B) SEM images of thick and thin collagen matrices. (C) Comparison of fibril diameters.

Fibril diameters were not significantly different between the two collagen matrices, P = 0.27.

4.2.3. Collagen Content and Stiffness of Matrices Over Culture Duration

Cells can enzymatically degrade collagen matrices over the culture duration, altering the

thickness of the matrices, which in turn may change the substrate stiffness. To ensure that the

mechanical properties were maintained over the culture duration, we estimated by FE analysis

that a substrate thickness of less than 100 m is required for the cells to “feel” the underlying

A. Thick matrix

B. Thin matrix

0.00

0.05

0.10

0.15

0.20

1 2

Fib

er

dia

me

ter

(m

)Thick Thin

CA. Thick matrix

B. Thin matrix

0.00

0.05

0.10

0.15

0.20

1 2

Fib

er

dia

me

ter

(m

)Thick Thin

0.00

0.05

0.10

0.15

0.20

1 2

Fib

er

dia

me

ter

(m

)Thick Thin

C

44

glass substrate, thereby increasing the perceived matrix stiffness. Experimentally, 30% collagen

degradation was detected over eight days of culturing VICs on the thick substrates (Figure 4.4:

A), which reduced the overall thickness of the thick matrices to approximately 1.75 mm.

Collagen content did not change significantly on the thin substrates (Figure 4.4: B). Consistent

with our FE analysis, this amount of collagen degradation did not affect the effective stiffness of

both the thick and thin matrices (Figure 4.4: C and D).

4.3. Discussion

A variety of hydrogel-based matrices have been developed to study the effect of substrate

stiffness on cell function, many of which lack the ability to decouple surface chemistry with

substrate mechanics. This complicates interpretation of in vitro cellular responses as cell

processes are regulated by multiple factors in the microenvironment including ligand density,

ligand type, substrate stiffness and the microstructure of culture surface (e.g., fiber diameter). In

this study, matrix composition and polymerization conditions were kept constant, while the

volume of neutralized collagen placed in each well was altered. This resulted in matrices of

different thicknesses, but similar microtopographical characteristics.

Matrix stiffness regulation of cell function is often studied using synthetic hydrogels coated with

monomeric collagen or other ECM proteins (reviewed in Yip et al.10). We chose to use fibrillar

type I collagen instead of monomeric collagen to better mimic the native ECM composition in

heart valves, as fibrillar type I collagen is the most abundant ECM protein in aortic valves24.

This may be important as different intracellular signaling pathways are activated when cells bind

fibrillar versus monomeric collagen195, 196. For example, transcript expression197 and

proliferation195 of smooth muscle cells are different when cultured on fibrillar collagen versus

monomeric collagen. It is plausible that a similar effect may apply to VICs, and therefore it was

crucial to keep the polymerization condition consistent to ensure the formation of collagen fibrils

on both thick and thin matrices. In addition, consistent polyermization should also minimize

difference in fiber density between the two matrices. However, further tests are necessary to

confirm if the fiber density is identical in the compliant and stiff matrices.

45

Figure 4.4. Collagen content and effective stiffness of the two matrices over the

course of cell culture

Data were expressed as percentage of collagen remaining relative to day 0. (A) Significant

collagen degradation was observed in thick matrices after eight days of culture (N = 3). * P <

0.05 indicates statistically significant difference in collagen content between matrices from day

zero and day eight. (B) Content of collagen remained relatively constant for the thin matrices.

Stiffness of thick matrices (C) and thin matrices (D) did not change significantly over the culture

duration.

It has been reported in various cell types that integrin binding to polymerized collagen fibrils can

induce the expression of collagenases such as MMP-1198, 199 and MMP-13200. This effect is

particularly important to our culture system because changes in collagen content within the

matrices would alter the overall matrix thickness: if an excessive amount of collagen was

degraded, the matrix could become thin enough that the cells would be able to “feel” the

underlying stiff glass substrate. By selecting the initial thickness (2.5 mm) of the thick matrices

to be substantially larger than that needed to maintain its compliance (estimated to be 100 m in

thickness by FE analysis), we prevented the effects of degradation from impacting matrix

mechanics over the culture duration.

Culture duration

020406080

100120

Day 0 Day 3 Day 8

Pe

rce

nt o

f co

llag

enA B

020406080

100120

Day 0 Day 3 Day 8

0.00.20.40.60.81.01.21.4

Day 0 Day 3 Day 8

Re

lativ

e s

tiffn

ess

n

orm

aliz

ed

to D

ay

0

C D

0.00.20.40.60.81.01.21.4

Day 0 Day 3 Day 8

*

Culture duration

020406080

100120

Day 0 Day 3 Day 8

Pe

rce

nt o

f co

llag

enA B

020406080

100120

Day 0 Day 3 Day 8

0.00.20.40.60.81.01.21.4

Day 0 Day 3 Day 8

Re

lativ

e s

tiffn

ess

n

orm

aliz

ed

to D

ay

0

C D

0.00.20.40.60.81.01.21.4

Day 0 Day 3 Day 8

*

46

The collagen used in this study was commercially available and extracted from bovine hides with

purity between 97 % to 99.9 % (determined using SDS-PAGE by the manufacturer). As it was

not possible to obtain 100 % pure type I collagen extract, because type I collagen naturally

associates with type III collagen in tissues201-203, there might be a small percentage (< 3 %) of

type III collagen present in the mixture. Both type I and type III collagens are fibrillar. Type III

collagen is a potent regulator of type I collagen fibrollogenesis, which modulates the fibril

diameter of type I collagen203. As previously mentioned, fibril diameter influences cell adhesion,

migration and proliferation194. We examined the fibril diameter in the collagen matrices made

from different batches of type I collagen extract, and found that fibril diameter was similar

among these matrices despite the potential of minor type III collagen contamination203. These

data suggest that the small percentage of type III collagen had little impact on cell function

regulated by fibril diameter. In addition, interaction with collagen affects cell adhesion,

migration, proliferation and differentiation mediated through integrin, adhesion molecules from

the immunoglobulin superfamily and discoidin domain receptors (DDR 1 and 2) (reviewed in 204,

205). Of the integrins, 11, 21, 101, 111 are known to bind to various types of collagen206-

208. Type I collagen is thought to primarily interact with 21 integrin, and type III collagen

interacts with integrin 11. However, the I domains from the 1 and 2 subunits demonstrate

similar binding affinities to type I and III collagen209. Further, some members of the

immunoglobulin superfamily have been shown to facilitate cell-integrin interaction; for example

syndecan-1 has been shown to mediate the binding of integrin 21 to collagen210. DDRs can

bind to type I and type III collagen to control cellular responses to the extracellular matrix211,212.

Aortic VICs express predominately 1, 2, 3, 4, 5 and 1 integrin subunits, but little or no

expression of adhesion molecules of the immunoglobulin superfamily (e.g. syndecan-4,

syndecan-1, E-selectin)67. DDRs have been studied widely in non-valve cell types, but the

expression and the function of DDRs in VICs has yet to be identified. As type I and III collagen

both bind to 11 and 21 integrins as well as DDRs, various batches of type I collagen matrices

with small differences in type III collagen contamination would likely elicit cellular responses

via similar receptor binding.

While the initial composition of the collagen matrices was well controlled (i.e., same

concentration and polymerization condition), compositional changes over the culture duration

were probable because of cellular remodeling. Previous studies have demonstrated that

47

calcification by valvular and vascular cells is similar on collagen and fibronectin-coated

substrates26, 213, suggesting that changes in certain ECM components during remodeling would

not impact calcification by VICs. However, characterization of the compositional changes that

occurred in the matrices would better define the relative contributions of matrix stiffness versus

composition.

To manipulate only matrix stiffness, we were limited to a two-dimensional (2D) system, as

decoupling of matrix mechanics and chemistry is not possible with three-dimensional (3D)

fibrillar collagen matrices10. Based on previous studies120, 213, some VIC responses, including

calcification92, are similar in 3D matrices as on 2D surfaces. Data from these previous studies

support the feasibility of using a 2D culture system as a “proof-of-principle” test platform, to

evaluate the effect of matrix stiffness on the calcification by VICs.

In summary, a collagen-based cell culture system with tunable stiffness was achieved by

changing the thickness of the matrices without altering the surface chemistry. Despite the

degradation of the compliant matrices over the culture period, which reduced the overall

thickness of these matrices by 20%, the effective stiffness of both the thick and thin matrices

remained constant. Although fine-tuning of stiffness was not possible in this culture system,

using an ECM protein that is commonly found in native aortic valves provided a culture

environment that was physiologically relevant and avoided the needs of costly surface

modification, which is often required in systems with finely adjustable stiffness. The selection of

a suitable culture system largely depends on the application and the research objectives. Our

current approach provides a simple means to achieve two culture surfaces with significantly

different stiffnesses. This system is sufficient to investigate whether VICs, if at all, sense the

mechanical cues from their microenvironment and how such mechanical cues may influence

their phenotype and calcification.

48

Chapter 5 5. Effect of Substrate Stiffness on Calcification by VICs

Dysregulation of normal cellular processes214, 215 leads to CAVD that often involves chronic

inflammation, fibrosis, and calcification48, 216. The progression of sclerosis and calcification is

mediated primarily by VICs that populate the interstitial matrix54, 214. As in the vasculature217,

calcification of the aortic valve occurs through multiple mechanisms218, including apoptosis-

related calcification typically associated with myofibrogenic activation of VICs4, 5, calcium

deposition associated with necrotic cells219, and bone formation by resident VICs49 or bone

marrow-derived cells5. However, details of the cellular mechanisms by which VICs contribute to

calcification are not well understood, largely due to the limited number of studies in vitro and

difficulties with their interpretation. For example, when VICs are induced to form calcified

multicellular aggregates in vitro, the aggregates are associated with the expression of bone-

related transcripts and proteins54, 74, 92, 220, the expression of myofibroblast markers26, and/or

apoptosis187, 220. It is unclear if these features represent a single or multiple calcification

process(es).

The factors that contribute to the dysregulation of VICs leading to calcification are also not fully

defined. While a variety of biochemical cues, such as TGF-β, BMP-2 and BMP-4, have been

implicated in valve calcification216, mechanical cues from the extracellular matrix may also

regulate cell function, both in vivo and in vitro. Notably, cells are able to “sense” the local

mechanical properties of their extracellular matrix, and matrix stiffness is known to regulate

motility, proliferation, and differentiation in various cell types221. The differentiation of VICs to

myofibroblasts was recently shown to be influenced by matrix stiffness55, but the role of matrix

mechanics in regulating calcification by VICs has yet to be determined. Presumably, the

combined effects of matrix stiffness and biochemical cues ultimately define VIC phenotype. An

improved understanding of VIC-matrix interactions is required to aid in interpretation of VIC

calcification studies in vitro; to guide the selection of biomaterials with appropriate mechanical

The work presented here was published in Arterioscler Thromb Vasc Biol. 2009 Jun; 9(6): 36-42.

49

properties for valve tissue engineering; and to assess if alterations in extracellular matrix

mechanics that occur with disease16, 186 modulate pathologic changes in VIC phenotypes and

calcification processes.

To gain a better understanding of calcification by VICs and its regulation by mechanical cues,

we used the fibrillar collagen-based system with tunable substrate stiffness described in Chapter

Four to study the influence of matrix stiffness on primary porcine aortic VICs in vitro. We found

that the response of VICs to pro-calcific soluble factors is sensitive to matrix stiffness. VICs

grown in pro-calcific conditions preferentially differentiate to osteoblast-like cells on compliant

substrates that mimic the stiffness of normal or early sclerotic tissue, but differentiate to

myofibroblasts on stiffer substrates that mimic the stiffness of stenotic tissue. Calcified

aggregates form in both cases, but through distinct processes that are differentially mediated by

cytoskeletal tension.



Canada). Detailed protocols are described in Appendix A. Assays that followed the protocols

from the manufacturers without any modification are not listed in Appendix A; these protocols

can be found on the websites of the respective suppliers.

5.1.1. Valve Interstitial Cell Culture

Primary VICs were isolated from aortic valves of eight-month-old pigs by collagenase digestion.

Compliant and stiff collagen matrices were constructed following procedures described in

Chapter 4. VICs were seeded on collagen matrices at 10,000 cells/cm2 in either complete

medium consisting of DMEM with 10% FBS (, Hyclone, Logan, UT, lot# KRA25425), 10,000

units/mL penicillin, and 10 mg/mL streptomycin, or in calcifying medium consisting of complete

medium supplemented with 10 mM β-glycerophosphate, 10 g/mL ascorbic acid and 10 nM

dexamathesone.

5.1.2. Measurement of Cellular Proliferation

Proliferation was determined at various time points based on measurement of DNA content of

cell pellets via binding with fluorescent dye from the CyQUANT® NF cell proliferation assay

50

kit (Invitrogen, Burlington, ON). Cells cultured on compliant or stiff matrices were rinsed with

sterile phosphate buffered saline (PBS) without calcium and magnesium. VICs were released

from collagen matrices with collagenase digestion (300 units/mL) for one hour at 37 oC. Culture

media, degraded collagen and collagenase solution were centrifuged at 4 oC at 16200 × g for five

minutes, followed by aspiration of the supernatant. Cell pellets were rinsed with ice-cold PBS,

and resuspended and incubated with the CyQUANT® dye for one hour at 37 oC in a 96-well

microtiter plate. The fluorescence intensity of each sample was measured using a microplate

reader with excitation at 485 nm and emission detection at 530 nm. A standard curve consisting

of 100 to 20,000 cells was generated with primary VICs, which was used to quantify the actual

number of cells in the test samples.

5.1.3. Determination of Cell Shape and Spreading

Cell shape and spreading of VICs cultured on compliant and stiff matrices were determined after

48 hours in culture. Cells were fixed with 10% neutral buffered formalin (NBF), followed by

permeabilization with 0.1% Triton X. Cells were stained with 5 g/mL of fluoroscein

isothiocyanate (FITC)-conjugated phalloidin (excitation /emission wavelengths: 490 nm/525

nm). Samples were subsequently mounted on microscope slides with PermaFluor mounting

medium and examined by fluorescence microscopy immediately (Olympus Model IX71,

Olympus, Center Valley, PA). The cell contours were identified based on the images of

phalloidin-stained cells. Cell spreading was estimated by tracing the cell contours and measuring

the cell area with ImageJ (NIH, Bethesda, MD).

5.1.4. Staining of Viable, Dead and Apoptotic Cells

VICs cultured on compliant or stiff matrices were quickly rinsed with sterile PBS. Viable cells

were determined by fluorescent labeling with 4 M Calcein AM (excitation/emission

wavelengths: 494 nm/517 nm) and dead cells were labeled with 2 M Ethidium Homodimer-1

(excitation/emission wavelengths: 528 nm/617 nm; LIVE/DEAD® Viability/Cytotoxicity Kit for

mammalian cells, Invitrogen, Burlington, ON). Cells were incubated with fluorescent dye for one

hour at 37 oC and then the nuclei were counterstained with Hoechst 33242 dye

(excitation/emission wavelengths: 350 nm/461 nm) for five minutes. Samples were subsequently

mounted on microscope slides with PermaFluor mounting medium and examined by

fluorescence microscopy immediately. As a negative control, cells were killed by formalin

51

fixation prior to Calcein AM staining and nuclear counterstaining to confirm that the Calcein

AM staining was specific to viable cells and not simply binding to calcium.

Apoptotic cells were identified by cellular uptake of APOPercentageTM dye (Biocolor Ltd,

Carrickfergus, UK) as a result of apoptosis-induced membrane phosphatidylserine and

phosphatidylcholine translocation. In brief, samples were quickly rinsed with sterile PBS with

calcium and magnesium prior to incubation with APOPercentageTM dye diluted 1:20 in

supplemented DMEM at 37 °C for 30 minutes. Samples were then mounted on microscope slides

with PermaMount mounting medium and images were captured under the microscope. Positive

controls were achieved by chemically-induced apoptosis of cells using 5 mM hydrogen peroxide

for three hours at 37 °C prior to staining. Intense staining, typically bright pink or purple

depending on the culture substrate and cell density, was observed in the positive controls,

ensuring the validity of the apoptosis detection method in VICs. Negative controls were achieved

by incubating samples without the APOPercentageTM dye.

5.1.5. Polymerase Chain Reaction for Expression of Osteogenic Markers

VICs were released from collagen matrices via collagenase digestion. Cell pellets were obtained

by centrifugation, followed by aspiration of the supernatant. Total RNA was isolated from cell

pellets following standard protocols of the Micro RNeasy System (Qiagen, Mississauga, ON).

Subsequently, total RNA was reverse transcribed with oligo-(dT)12-18 primers (Invitrogen,

Burlington, ON) and Superscript III reverse transcriptase (Invitrogen). cDNA was quantified

with a NanoDrop Spectrophotometer (ND-1000, NanoDrop Technologies, Wilmington, DE), and

then used as the template for real-time PCR using SYBR Green, an annealing temperature of 60 oC, and 35 cycles. Two osteoblast-related transcripts, osteonectin (Accession number:

AW436132, forward primers: 5’-tccggatctttcctttgctttcta-3’ and reverse primer 5’-

ccttcacatcgtggcaagagtttg-3’) and osteocalcin (Accession number: AW346755, forward primers:

5’-tcaaccccgactgcgacgag-3’ and reverse primer 5’-ttggagcagctgggatgatgg-3’) were tested222.

Glyceraldehyde-3-phosphate dehydrogenase (GAPDH, Accession number: AF017079, forward

primers: 5’-tgtaccaccaactgcttggc-3’ and reverse primer 5’-ggcatggactgtggtcatgag-3’) was used as

the housekeeping gene223. Transcriptional expression was quantified by the comparative Ct

(Cycle threshold) method (2-Ct method) with the following equations:

Ct = Ct of target gene – Ct of housekeeping gene (1)

52

Ct = Ct Compliant matrices - Ct Stiff matrices (2)

Fold increase between two matrices = 2-Ct (3)

5.1.6. Measurement of Runt-Related Transcription Factor 2 (Runx2) Protein

Runx2/Cbfa-1 protein expression was measured using a commercially available ELISA-based

immunoassay (TransAM kit, ActiveMotif, Carlsbad, CA). Nuclear extracts were prepared from

VICs following the manufacturer’s recommendations. Briefly, VICs were released from collagen

matrices with collagenase digestion (as described above) and cell pellets were obtained. Samples

were then rinsed with ice-cold PBS with phosphatase inhibitor buffer (125 mM sodium fluoride,

250 mM -glycerophosphate, 250 mM para-nitrophenyl phosphate and 25 Mm sodium vanadate)

to prevent inactivation of Runx2/Cbfa-1. Pellets were resuspended in ice-cold hypotonic buffer

(20 mM HEPES, 5 mM sodium fluoride, 10 M sodium thioglycolate and 0.1 M EDTA) and

allowed to swell on ice for 15 minutes. Cell membranes were disrupted by gentle mixing with

10% Igepal CA-630, followed by centrifugation. Nuclear pellets were resuspended in Complete

Lysis Buffer (1 M DTT, protease inhibitor cocktail, lysis buffer AM4) and rocked gently on ice

for 30 minutes. Nuclear extract was obtained by collecting the supernatant upon centrifugation.

The protein concentration in the nuclear extract was determined using a micro BCA assay

(Pierce, Rockford, IL). For each sample, 20 g of nuclear extract was used to measure the

abundance of Runx2/Cbfa-1 by an ELISA-based immunoassay following the manufacturer’s

protocol. Briefly, nuclear extracts containing unknown amount of activated transcription factor

were incubated for one hour in 96-well microtiter plates pre-coated with oligonucleotides

containing an AML-3/Runx2/Cbfa-1 consensus binding site. Primary antibody for AML-

3/Runx2/Cbfa-1 was added to the samples for another one hour of incubation, followed by one

hour incubation with horseradish peroxidase-conjugated anti-rabbit IgG. The colorimetric

reaction was initiated by a five-minute incubation with the Developing Solution, followed by the

Stop Solution. Absorbance of each sample was read immediately with a spectrophotometer at

450 nm with a reference wavelength of 655 nm. Absorbance readings were normalized by total

cell number per sample. A positive control for AML-3/Runx2/Cbfa-1 activation was performed

with 5 g/well of Saos-2 nuclear extract. Negative controls were achieved with blank compliant

and stiff collagen matrices without cells.

53

5.1.7. Alkaline Phosphatase and Alizarin Red S Staining

To detect alkaline phosphatase activity (ALP), VICs on collagen matrices were fixed in 10%

NBF and rinsed in distilled water. Samples were stained using Napthol AS MX-PO4 (Fisher

Scientific, Ottawa, ON) as the substrate, N, N-Dimethylformaide, Trizma-hydrochloride acid,

and Fast Red Violet LB salt. The stained samples were rinsed with distilled water three times and

examined under a light microscope. Positively stained cells display a reddish/purple color. To

detect the presence of calcium salts, formalin fixed samples were washed with distilled water,

followed by staining with 0.02 mg/mL Alizarin red S (ARS) solution. Cells with calcium

deposition were stained bright red. Subsequently, ARS dye was released from the stained

samples using 0.6 N HCl, followed by neutralization with 10% (vol/vol) ammonium hydroxide.

The total amount of ARS released from the culture was quantified by measuring the absorbance

of the solublized dye in solution at 405 nm.

5.1.8. Osteocalcin Immunohistochemical Staining

Immunohistochemical staining for osteocalcin was performed using Vectastain Universal Elite

ABC Kit (Vector Laboratories, Burlingame, CA). Samples were fixed in 10% NBF and washed

with 0.05% Tween 20 diluted in PBS. Following fixation, samples were treated with 3%

hydrogen peroxide in methanol at room temperature for 10 minutes, blocked with horse serum

for 20 minutes at room temperature, and then incubated for three hours at room temperature with

20 μg/mL mouse anti-bovine osteocalcin antibody (clone OCG4; Affinity BioReagents, Golden,

CO) diluted in 0.3% Triton X-100 in PBS. Secondary biotinylated antibody and 3,3'-

diaminobenzidine (DAB) substrate were then applied. The stained samples were dehydrated in

an ethanol gradient. Negative controls were achieved by omitting the primary antibody.

5.1.9. Immunofluorescent Staining of Cytoskeletal Proteins

VICs on collagen matrices were fixed in 10% NBF, followed by permeabilization with 0.1%

Triton-X and rinsing with PBS. Samples were blocked with 3% bovine serum albumin to

minimize non-specific binding. To stain for -smooth muscle actin (-SMA), monoclonal

mouse anti--SMA antibody (20 g/mL clone 1A4 mouse anti-human monoclonal primary

antibody ) and 20 g/mL Alexa Fluor® 568 goat anti-mouse antibody were used. Filamentous

(F)-actin was stained by fluoroscein isothiocyanate (FITC)-conjugated phalloidin (5 g/mL).

54

Nuclei were stained with Hoechst 33242 dye (1 g/mL). Samples were mounted on microscope

slides with PermaFluor and were examined under a fluorescence microscope.

5.1.10. Disruption of Cytoskeleton Assembly

VICs were treated with 0.4 nM of Swinholide A (SWA) after six days of culture on compliant or

stiff matrices in calcifying media. Medium with fresh SWA was exchanged every 24 hours for

two days. After two days of treatment with SWA, expression of -SMA and F-actin were

analyzed with immunostaining and the number of aggregates was recorded.

5.1.11. Response to TGF-1

VICs were cultured on compliant and stiff matrices and were immediately treated with calcifying

media (containing 10% FBS) and 5 ng/mL of TGF-1 for five days to induce -SMA

expression. The FBS used in these experiments was reported by the manufacturer to contain 10-

22 ng/mL TGF- (equivalent to 1- 2.2 ng/mL in the calcifying media). We attempted to reduce

the baseline TGF-1 concentration in the media by using 1% FBS, but this resulted in a

significant reduction in proliferation and no formation of calcified aggregates after up to 10 days

on both compliant and stiff matrices, even with the addition of exogenous TGF-1. Extended

long term culture (> 10 days) was not feasible due to collagen degradation. Responsiveness of

VICs to TGF-1 was also determined by measuring the transcript expression of TGF-1 receptor

I and II by RT-PCR using the following primers sequence: (1) TGF-1 receptor I (Accession

number: AB182260.1, forward primers: 5’-gacggcattccagtgtttct-3’ and reverse primer 5’-

tgcacatacaaatggcctgt-3’) and (2) TGF-1 receptor II (Accession number: EF396957.1, forward

primers: 5’-cagggaagaacgttcatggt-3’ and reverse primer 5’-ccaaccaaagctgagtccat-3’).

5.1.12. Expression of TGF-1

VICs were grown on compliant or stiff matrices for eight days, with media changes every two

days. Conditioned media was collected on the last day of culture. TGF-1 in the conditioned

media was measured using the TGF-1 Emax® ImmunoAssay System (Promega, Nepean, ON)

according to the manufacturer’s directions. Briefly, TGF-1 coat monoclonal antibody, which

binds to soluble TGF-1, was adhered to the surface of a 96-well microtiter plate. Samples for

generating the standard curve and the test samples were applied to each pre-coated well. Samples

55

were incubated with anti-TGF-1 polyclonal antibody, followed by incubation with a species-

specific antibody conjugated to horseradish peroxidase. Colorimetric development was achieved

by the addition of TMB One solution. The reaction was stopped by 1 N HCl. Absorbance was

read at 450 nm on a plate reader and then normalized by total cell number per sample.

5.1.13. Contraction-Dependent Apoptosis and Akt Activation

A stress-relaxation collagen gel model224 was used to evaluate the relationship between cell

contraction, apoptosis, and Akt activity. Briefly, VICs were cultured on constrained collagen gels

for six days with osteogenic media at a cell density of 10,000 cells/cm2. Cellular contraction was

induced by releasing the gels from the walls of the culture wells. Cells were stained with

APOPercetageTM dye after 0, 0.5, and 3 hours of gel release to detect the presence of apoptosis.

The level of apoptosis was quantified colorimetrically (absorbance at 550 nm) upon dye release.

In a separate experiment using the same culture methodology, proteins were extracted from cells

after 0 and 1 hour of gel release, followed by western blot for the detection of total Akt and

phosphorylated Akt. Briefly, cell cultures were washed with ice-cold PBS followed by the

addition of 1× lysis buffer. Cells were scraped and transferred to pre-chilled tubes, followed by

30-minute incubation on ice. Cell lysates were obtained by centrifugation at 16200 × g for 15

minutes at 4 oC. The protein concentrations of cell lysates were determined by micro BCA

protein assay (Pierce, Rockford, USA). 10 g of protein extract from compliant and stiff

matrices was loaded on to two identical 10% SDS-polyacrylamide gels for electrophoresis.

Samples were then transferred to two separate polyvinylidene fluoride membranes. The first

membrane was used for detection of Akt (1:1000 dilution, 60 kDa, Akt rabbit polycloncal

antibody, Cell Signaling Technology, Danvers, MA) and glyceraldehyde-3-phosphate

dehydrogenase (1:3000 dilution, 40 kDa, GAPDH mouse monoclonal antibody, Stressgen, Ann

Arbor, Michigan). The second membrane was used for detection of phospho-Akt (1:1000

dilution, 60 kDa, phospho-Akt rabbit polycloncal antibody, Cell Signaling Technology, Danvers,

MA) and GAPDH immunoblot. Expression of Akt and phospho-Akt were quantified by

densitometry using ImageJ Software (NIH, Bethesda, MD) and normalized to GAPDH

expression. Full details of western blot procedure including incubation duration, concentration of

horseradish peroxidase conjugated secondary antibody and detection of proteins with x-ray films

are described in Appendix A.

56



and experiments were repeated at least three times. Unpaired Student’s t-tests were used for



5.2. Results

5.2.1. Morphological Changes, Proliferation and Cell Spreading

VICs were cultured on the thick (relatively more compliant) matrices or the thin (stiffer)

matrices. When cultured in complete medium without pro-calcific supplements, VICs

proliferated more rapidly on compliant matrices (P < 0.05; Figure 5.1: A), but the morphology

was similar on the two matrices (Figure 5.1: B and C). In contrast, VIC proliferation rate was not

significantly different on stiff and compliant matrices when cultured in calcifying media (Figure

5.2: C), but morphological differences were substantial. In calcifying media, VICs on the more

compliant matrices formed multicellular aggregates after eight to ten days in culture (Figure 5.2:

A). In contrast, VICs on the stiffer matrices formed fewer aggregates (P < 0.05, Figure 5.2: D)

and instead tended to form ridges (Figure 5.2: B, ridges indicated by the red arrows). In medium

without pro-calcific supplements, there was no aggregation on either substrate over the culture

period.

Because VICs cultured in calcifying media displayed substantial morphological differences on

the two matrices after eight days in culture, we further investigated if initial cell-matrix contacts

altered individual cell morphology (cell shape and spreading) prior to the formation of

aggregates or ridges. Cell shape was visualized by FITC-conjugated phalloidin labeling 48 hours

after initial cell seeding. A wide variety of cell shapes were found on both matrices (Figure 5.3:

A and B), indicative of preservation of VIC heterogeneity. Although there was no striking

difference in cell shape on the two matrices, cells on the stiffer matrices spread significantly

more than those grown on compliant matrices (Figure 5.3: C).

5.2.2. More Compliant Matrices Promote Osteogenic Differentiation of VICs

VICs can form aggregates in vitro that contain calcium deposits and osteoblast-related proteins54,

so we investigated if this was the case for the cell aggregates on compliant and stiff collagen

57

matrices. There was a trend for greater calcification on the compliant matrices (P < 0.06; Figure

5.4: A). Calcium deposition was localized within the aggregates formed on both matrices (Figure

5.4: B and C).

Figure 5.1. Proliferation and morphology of VICs cultured in DMEM

supplemented with 10% FBS

(A) Cellular proliferation rate was significantly higher on compliant matrices in comparison to

those grown on stiff matrices after six and eight days in culture (* P < 0.05). (B) Bright field

images showing similar morphology of VICs on the two matrices after eight days in culture.

C. Stiff matricesB. Compliant matrices

B0

100

200

300

400

4 days 6 days 8 days

A.


*

Compliant matrices

Nu

mb

er

of

c ells

(x

10

-3)

Stiff matrices

*

Culture duration


B0

100

200

300

400


A.


*

Compliant matrices

Nu

mb

er

of

c ells

(x

10

-3)

Stiff matricesCompliant matrices

Nu

mb

er

of

c ells

(x

10

-3)

Stiff matrices

*

Culture duration

58

Figure 5.2. Proliferation and morphology of VICs cultured in calcifying media

(A and C) Bright field images showing aggregates (indicated by black arrows) formation on

compliant and stiff matrices. Ridge (indicated by red arrowheads) formation can be observed

only on the stiff matrices. (B) Cellular proliferation rate was similar between the two matrices.

(C) VICs grown on compliant matrices formed more aggregates in comparison to those cultured

on stiff matrices. * P < 0.05.

B. Stiff matricesA. Compliant matrices

B0

50

100

150

200

Compliant matrices Stiff matrices

Nu

mb

er

of

ag

gre

ga

tes

*

C.

Compliantmatrices

D.

0

200

400

600

800

1000

4 days 6 days 10 days4 days 6 days 8 days

Compliant matrices

Nu

mb

er

of

cells

(x

10

-3) Stiff matrices

StiffmatricesCulture duration


B0

50

100

150

200


Nu

mb

er

of

ag

gre

ga

tes

*

C.

Compliantmatrices

D.

0

200

400

600

800

1000


Compliant matrices

Nu

mb

er

of

cells

(x

10

-3) Stiff matrices

Stiffmatrices


B0

50

100

150

200


Nu

mb

er

of

ag

gre

ga

tes

*

C.

Compliantmatrices

D.

0

200

400

600

800

1000


Compliant matrices

Nu

mb

er

of

cells

(x

10

-3) Stiff matrices


B0

50

100

150

200


Nu

mb

er

of

ag

gre

ga

tes

*

C.

Compliantmatrices

D.

0

200

400

600

800

1000


Compliant matrices

Nu

mb

er

of

cells

(x

10

-3) Stiff matrices

Compliant matrices

Nu

mb

er

of

cells

(x

10

-3) Stiff matrices

StiffmatricesCulture duration

59

Figure 5.3. Cell shape and spreading 48 hours after initial seeding

Heterogeneous cell shape found on both the more compliant matrices (A) and the stiffer matrices

(B), scale bar represents 50 m. (C) Comparison of cell spreading on the two matrices.

* P < 0.05.

A. Compliant matrices B. Stiff matrices

C.

*

0

100

200

300

400

500

Compliant Stiff Type of collagen matrices

Ce

ll sp

rea

din

g a

rea

(m

2 )

Ce

ll sp

rea

din

g a

rea

(m

2)

*


C.

*

0

100

200

300

400

500


Ce

ll sp

rea

din

g a

rea

(m

2 )

Ce

ll sp

rea

din

g a

rea

(m

2)


C.

*


C.


C.

*

0

100

200

300

400

500


Ce

ll sp

rea

din

g a

rea

(m

2 )

Ce

ll sp

rea

din

g a

rea

(m

2)

0

100

200

300

400

500


Ce

ll sp

rea

din

g a

rea

(m

2 )

Ce

ll sp

rea

din

g a

rea

(m

2)

*

60

Figure 5.4. Calcification by VICs on the two matrices.

(A) Relative amount of calcium on the two matrices as measured by releasing the ARS dye.

# P = 0.06. (B and C) ARS staining for calcium expression showing localization of calcium

deposition within cell aggregates.

Transcriptional expression of osteonectin and osteocalcin were significantly higher in VICs

cultured on compliant matrices (Figure 5.5: A), as were the protein level of Runx2/Cbfa-1

(Figure 5.5: B), ALP activity (Figure 5.5: C), and osteocalcin protein expression assessed by

immunostaining (Figure 5.5: D). ALP activity and osteocalcin expression were localized within

the aggregates on compliant matrices. On stiff substrates, ALP activity was weak and dispersed

throughout the cell layer (Figure 5.5: C) and minimal osteocalcin expression was observed, even

in aggregates (Figure 5.5: D).

B. Compliant matrices

A.

0.0

0.5

1.0

1.5

2.0


Re

lati

ve a

mo

unt

o

f ca

lciu

m (

AU

)

#

Compliant Stiff


A.

0.0

0.5

1.0

1.5

2.0


Re

lati

ve a

mo

unt

o

f ca

lciu

m (

AU

)

#

Compliant Stiff

C. Stiff matrices

61

Figure 5.5. Compliant matrices promote osteogenic phenotypes.

(A) Relative mRNA levels of osteocalcin and osteonectin by VICs, * P < 0.05 and # P < 0.06.

(B) Runx2/Cbfa-1 expression, * P < 0.05. (C) ALP staining. (D) Osteocalcin expression, inset

showing negative control.

Am

oun

t of

Run

x2/C

bfa

-1

(no

rma

lize

d, A

U)

02468

1012

Compliant matrices Stif f matrices

A.

*F

old

incr

ea

se re

lati

ve

to

stif

f ma

tric

es

02468

10

Osteonectin Osteocalcin

*

#

Compliant Stiff

Compliant Stiff

B.

C.

Compliant StiffD.

Am

oun

t of

Run

x2/C

bfa

-1

(no

rma

lize

d, A

U)

02468

1012


A.

*F

old

incr

ea

se re

lati

ve

to

stif

f ma

tric

es

02468

10


*

#

Compliant Stiff

Compliant Stiff

B.

C.

Compliant StiffD.

02468

1012


A.

*F

old

incr

ea

se re

lati

ve

to

stif

f ma

tric

es

02468

10


*

#

Compliant Stiff

Compliant Stiff

B.

C.

Compliant StiffD.

62

5.2.3. Stiffer Matrices Promote Calcification Through Apoptosis

Although VICs on stiff matrices expressed osteoblast-related markers at low levels, detectable

calcium deposition was observed within the few aggregates that formed. Morphological analysis

by SEM revealed significant differences in the spreading and shape of these cells on the surface

of and around the multicellular aggregates on the two matrices (Figure 5.6: A and B). Aggregates

on stiff matrices were rounded and symmetrical with fewer cells adjacent to the aggregates,

suggesting that these aggregates might have resulted from contraction of the confluent cell

sheets. In contrast, cells covered the entire surface of the compliant matrices. Some cells on the

aggregates were cuboidal as opposed to the elongated morphology that are often observed with

fibroblasts, suggesting a change in cell phenotype and potentially the presence of clonal growth

within these cell aggregates on the compliant matrices. Such differences in cell organization and

morphology suggested that the calcific aggregates might have occurred through different

mechanisms on the two matrices. Both in vivo and in vitro, calcification can occur through a

process involving apoptosis4, 187 (discussed in Chapter 2), so we examined whether cell death and

apoptosis was associated with localized calcium deposition in these multicellular aggregates. On

compliant matrices, the aggregates contained viable cells with little evidence of apoptosis

(Figure 5.6: C and E). Positive Calcein AM staining was not due to the presence of calcium, as

formalin-fixed calcified aggregates stained negatively (Figure 5.6: C inset). In contrast,

aggregates formed on stiff matrices contained dead and apoptotic cells (Figure 5.6: D, F).

5.2.4. Aggregate Formation on Stiffer Matrices is Mediated by Cytoskeletal Tension

The striking differences in VIC phenotypes, aggregate morphology, and calcifiation process on

compliant versus stiff matrices in otherwise identical culture conditions suggested that VICs

sense and respond to matrix stiffness. Stiff culture surfaces, such as TCPS, are known to promote

myofibroblast differentiation of VICs and increase expression of filamentous -SMA55, a

cytoskeletal protein that contributes to the contractility of activated VICs. We found that VICs

displayed F-actin fibers regardless of matrix stiffness (Figure 5.7: A and B). However, VICs on

compliant matrices expressed predominantly monomeric -SMA (Figure 5.7: E), whereas

abundant expression of filamentous -SMA was observed only in cells cultured on stiff matrices

(Figure 5.7: F) consistent with the emergence of a myofibroblast phenotype. On the stiff

matrices, ridge formation by VICs was associated with the expression of filamentous -SMA

(Figure 5.8: A, B, D, E), which eventually led to the formation of aggregates (Figure 5.8: C, F).

63

We investigated the role of actin assembly in matrix stiffness-dependent aggregate formation by

disrupting actin filaments in VICs. Actin depolymerization was observed in VICs after 48 hours

of SWA treatment on both matrices (Figure 5.7: C, D, G, H).

Figure 5.6. Stiffer matrices promote dystrophic calcification associated with VIC apoptosis.

(A and B) SEM of cell aggregates showing distinct morphological differences. (C and D)

Calcein AM staining for live cells (green) and Ethidium homodimer-1 staining for dead cells

(red) staining, inset showing the negative control with nuclear counterstaining (blue). (E and F)

APOPercentageTM staining for apoptosis, inset showing positive control (purple).

StiffCompliantS

EM

Ca

lciu

m A

MA

po

pto

sis

A B

C D

E F

StiffCompliantS

EM

Ca

lciu

m A

MA

po

pto

sis

A B

C D

E F

64

Figure 5.7. Cytoskeleton expression.

Comparison of the expression of F-actin (green, A -D) and -SMA (red, E-F) by VICs with

nuclei counterstained blue) with and without swinholide A treatment on cells cultured on the

compliant and stiff matrices.

Upon treatment with SWA, cells remained attached to the matrices, but some were rounded with

limited extension of cytoplasmic processes (Figure 5.9: A - D). On compliant matrices, actin

disruption had no effect on the formation of cell aggregates that displayed osteogenic phenotypes

(Figure 5.9: C). In contrast, on stiff matrices disruption of actin assembly significantly reduced

the formation of aggregates (Figure 5.9: B, D, E).

Without SWA With SWA

Compliant matrix Stiff matrix

A B C D


-S

MA

E F G

F-a

ctin

H



A B C D


-S

MA

E F G

F-a

ctin

H

65

Figure 5.8. Aggregate formation by VICs cultured on stiff matrices

(A and D) Confluent cells on stiff matrices expressed abundant filamentous -SMA (red). (B and

E) Prior to the formation of aggregate cells expressing filamentous -SMA (red) form ridges. (C

and F) Expression of -SMA was also found in cell aggregates.

These data, along with the “contracted” appearance of the aggregates formed on the stiff

matrices (Figure 5.8: F), suggested that apoptosis leading to calcification on stiff substrates may

be due to local contraction of the cell layer resulting from increased cytoskeletal tension, which

is then released upon aggregation. To test this, we released constrained collagen gels seeded with

VICs and observed a significant increase in the number of apoptotic cells (Figure 5.10: A).

Previous studies have identified the Akt signaling pathway as a mediator of contraction-

dependent apoptosis225. We found that Akt activation was downregulated upon gel contraction

by VICs (Figure 5.10: B), prior to apoptosis, suggesting mechanically-regulated Akt activity

influences apoptosis in VICs.

Low magnification

High magnification

Overlapping cell layers Ridges Aggregates

A B C

D E F

Low magnification

High magnification


Low magnification

High magnification


A B C

D E F

66

Figure 5.9. Effect of cytoskeleton disruption on aggregate formation

(A – D) Swinholide A reduced the formation of ridges on stiff matrices. (E) Number of

aggregates was significantly reduced on stiff, but not compliant matrices (* P < 0.05).

5.2.5. Response to TGF-β and the Expression of its Receptors Are Matrix Stiffness

Dependent

TGF-1 is a potent inducer of -SMA expression and myofibroblast differentiation. It is also

expressed in calcified aortic valves69 and promotes VIC apoptosis and calcified aggregate

formation in vitro54, 69. We evaluated if matrix stiffness-dependent aggregation was influenced by

TGF-1. No differences were detected in endogenous total TGF-1 production by VICs on

compliant versus stiff matrices (data not shown). However, when treated with 5 ng/mL of

exogenous TGF-1, aggregate formation was accelerated on stiff matrices as early as after five

days in culture (28 ± 4 aggregates). No aggregates were found in the absence of TGF-1 after the

same culture duration (zero aggregates; P < 0.05). In contrast, no aggregates were observed after

Compliant Stiff


Compliant Stiff

A B C D

E

0

50

100

150

200


Without SWA. 8 days in cultureWith SWA, 8 days in culture

Nu

mb

er

of

ag

gre

ga

tes

*

Compliant Stiff


Compliant Stiff

A B C D

E

0

50

100

150

200


Without SWA. 8 days in cultureWith SWA, 8 days in culture

Nu

mb

er

of

ag

gre

ga

tes

*

67

five days on compliant matrices in the absence or presence of TGF-1. VICs grown in calcifying

media for eight days on either substrate had significantly lower expression of both TGF-1

receptor I and II compared with freshly isolated VICs. Whereas TGF- receptor II expression

was not different on compliant versus stiff matrices, TGF- receptor I expression was

significantly lower in VICs on compliant matrices (Figure 5.11). These observations suggest that

the preferential responsiveness to TGF-1 on stiff matrices is mediated in part through matrix

stiffness-dependent expression of TGF- receptor I.

Figure 5.10. Molecular factors involved in contraction-induced apoptosis on stiff substrates

(A) Apoptosis determined by APOPercentage dye upon release of constrained collagen gels,

* P < 0.05. (B), Western blot of total Akt (60 kDa) and phosphorylated Akt (p-Akt, 60 kDa) in

cells on gels before release and one hour after release. Densitometric quantification of Western

blot expressed as the ratio of phosphorylated Akt to total Akt. # P = 0.06

Pho

spho

ryla

ted

Akt

/To

tal A

kt

0.0

0.1

0.2

0.3

0.4

0.5

0 0.5 3.5Time after gel release (hours)

A B

*

0.00

0.05

0.10

0.15

0.20

0.25

0.30

0.35

0 1

#

Total Akt

p-Akt

0 1(hr)

Re

lati

ve a

po

pto

sis

(A

U)

Time after gel release (hours)

Pho

spho

ryla

ted

Akt

/To

tal A

kt

0.0

0.1

0.2

0.3

0.4

0.5


Pho

spho

ryla

ted

Akt

/To

tal A

kt

0.0

0.1

0.2

0.3

0.4

0.5


A B

*

0.00

0.05

0.10

0.15

0.20

0.25

0.30

0.35

0 1

#

Total Akt

p-Akt

0 1(hr)

Re

lati

ve a

po

pto

sis

(A

U)


A B

*

0.00

0.05

0.10

0.15

0.20

0.25

0.30

0.35

0 1

#

Total Akt

p-Akt

0 1(hr)

Re

lati

ve a

po

pto

sis

(A

U)

A B

*

0.00

0.05

0.10

0.15

0.20

0.25

0.30

0.35

0 1

#

Total Akt

p-Akt

0 1(hr)

Re

lati

ve a

po

pto

sis

(A

U)


68

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

TGF-beta Receptor I TGF-beta Receptor II

Freshly isolated VICs Compliant matricesStiff matrices

*

**

*

No

rma

lize

dm

RN

Ae

xpre

ssio

n (

AU

)

*

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

TGF-beta Receptor I TGF-beta Receptor II

Freshly isolated VICs Compliant matricesStiff matrices

*

**

*

No

rma

lize

dm

RN

Ae

xpre

ssio

n (

AU

)

*

Figure 5.11. Expression of TGF- receptors I and II

Transcript expression of TGF- receptors in freshly isolated VICs from normal porcine AV as

well as in VICs cultured on compliant and stiff matrices, * P < 0.05.

5.3. Discussion

VICs often form calcified nodules in vitro54, however the mechanisms by which VICs form

calcified aggregates and the factors that regulate these processes are not well defined. Here, we

demonstrated two distinct calcification processes that are mechanically regulated and are

associated with the differentiation of VIC into two distinct cell phenotypes. When cultured in

calcifying media, VICs grown on more compliant matrices were viable and formed calcified

bone-like nodules identified as such by the localized expression bone-related transcripts and

proteins. In contrast, VICs cultured in the same media but on the stiffer matrices had minimal

osteoblast marker expression, differentiated to contractile myofibroblasts, and formed calcified

aggregates containing apoptotic cells. Importantly, calcification on either matrix occurred within

the culture duration only when the VICs were exposed to biochemical factors that promote

calficiation. Thus, matrix stiffness alone was insufficient to cause calcification, but worked in

conjunction with soluble factors to regulate VIC differentiation and calcification. This is of

particular interest because biochemical factors are never decoupled with mechanical factors in

vivo.

69

The experimental system used here permitted specific investigation of distinct calcification

mechanisms. This has not been possible to date and has been largely ignored, confounding

interpretation of cell culture data and limiting our understanding of the mechanisms underlying

calcification by VICs. We used primary VICs rather than passaged VICs to capture the cellular

heterogeneity in intact valves 215, 226 and to avoid the phenotypic changes that occur with

subculture, including myofibroblast differentiation55 and loss of osteoprogenitors58. Cell shape

data further confirmed the preservation of cellular heterogeneity of our cultures. Although it was

difficult to determine the effect of matrix stiffness on cell shape due to heterogeneity of the cell

source, cell spreading was significantly different between the two matrices. Cell spreading is

driven by forces generated from actin polymerization227. VICs cultured on the stiffer matrices

expressed abundant actin filaments and spread more than those cultured on the more compliant

matrices. It is likely that VICs sense matrix stiffness via integrin-cytoskeleton linkages, which in

turn regulates cytoskeleton organization and cell spreading. This view is supported by studies

with epithelial keratocytes lacking adaptor proteins such as talin, which normally facilitates the

integrin-cytoskelton linkage, fail to spread228. Further investigations are necessary to identify the

molecular components necessary for stiffness sensing in VICs.

To manipulate only matrix stiffness, we were limited to a 2D system, as decoupling of matrix

mechanics and chemistry is not possible with 3D fibrillar collagen matrices10.

Although efforts were made to achieve similar surface chemistry on the compliant and sitff

matrices, some limiations remained with our 2D system. For instance, changes in matrix

composition due to cell remodeling may have occurred with time in culture. The effect of other

matrix components produced by VICs on calcification is unknown, and thus characterization of

the compositional changes that occur in the gels due to remodeling would better define the

relative contributions of matrix stiffness versus composition. Further, cells might have invaded

the collagen matrices as a result of matrix remodeling and migration, which would affect cell

shape and spreading as well as stiffness sensing by the cells. Although SEM images showed that

layers of overlapping cells occur locally at the site of aggregation, additional experiments are

necessary to evaluate the spatial distribution of cells over the given culture duration to fully

determine if cells remained on the surface of the collagen matrices.

70

Based on Hertz contact analysis of microindentation data of collagen gels identical to those used

here229, the apparent elastic moduli of the thick and thin collagen matrices are estimated to be 27

kPa and 113 kPa, respectively. Of note, VICs underwent osteogenic differentiation in the same

stiffness range (25-40 kPa) as bone marrow-derived MSCs130, consistent with recent evidence

that the aortic valve also contains a subpopulation of MSCs with robust osteogenic calcification

potential58.

The differentiation of VICs to contractile myofibroblasts that express filamentous -SMA on

stiff substrates has been reported previously55. The primary inducers of myofibroblast

differentiation are mechanical tension and TGF-β181. Cytoskeletal tension is generated

intrinsically by cells as they exert tractional forces on the surrounding extracellular matrix; stiff

matrices provide greater resistance to deformation, resulting in greater tractional forces221. The

incorporation of -SMA into stress fibers aids in force generation230. We found that -SMA

stress fibers were critical to aggregation on stiffer matrices, as this process was inhibited by

treatment with SWA, which disrupts polymerization of -SMA221, and was promoted by TGF-

1. The dependency of aggregation on cytoskeletal tension, along with the appearance of ridges

and the final symmetrical morphology of the aggregates suggested that the aggregates formed by

local contraction of the cell layer. Release of mechanical tension in VICs, as would occur with

contraction-induced aggregation, reduced Akt activity and subsequently triggered apoptosis as it

does in other myofibroblasts224, 225, 231. Apoptosis is associated with calcification of vascular and

valvular cells in vitro77 and in vivo4, 232, and is required for TGF-β1-induced calcification by

VICs69. Our observations suggest a mechanically-based mechanism with which to interpret in

vitro models of apoptosis-associated VIC calcification, particularly those performed on stiff

polystyrene tissue culture plates that induce myofibroblast differentiation55. This finding is

consistent with recent study by Anseth’s group, in which overexpression of -SMA increased

VIC calcific nodule formation, whereas knockdown of -SMA with siRNA transfection reduced

this process233. This mechanism is also likely to be important in vivo where increases in

myofibroblasts47, 55, 234, apoptotic cells4, and TGF-69 are observed in sclerotic leaflets and

alterations in matrix tension are believed to be a trigger for myofibroblast to undergo apoptosis

during wound repair224, 231.

71

In contrast to stiff matrices, addition of exogenous TGF-β1 did not accelerate calcification on

compliant substrates. The relative insensitivity of the cells on the more compliant matrices to

TGF-1 likely resulted from lower expression of the TGF- receptor I. Fibroblasts are less

sensitive to TGF- when grown in 3D spheroids than when grown as 2D monolayers on glass235.

The differential responsiveness was reported to correlate with downregulation of TGF- receptor

expression in 3D culture, which in light of our findings, may reflect differences in effective

matrix mechanical properties between 2D and 3D culture systems. We also observed significant

downregulation of both TGF- receptor I and II transcripts in VICs grown under calcifying

conditions on either substrate compared to those freshly isolated from normal valves. This is

consistent with observations from explanted human aortic valves, in which these receptors were

moderately downregulated in calcified leaflets relative to non-calcified leaflets69.

While our findings have clear implications for the interpretation of VIC calcification in vitro and

for the selection of biomaterials for valve regeneration, the relevance to valve calcification in

vivo remains to be determined. Similar to atherosclerosis, CAVD is an active pathobiological

process that involves extensive matrix remodeling13, 42, 216. The extracellular matrix provides

biochemical and mechanical cues to adherent cells, and alterations in the composition13, 236 and

mechanical properties16 of the ECM are characteristic of sclerotic diseases. The effects of matrix

composition on calcification have been reported26, 237, 238, but the influence of matrix stiffness on

vascular or valvular calcification has not been investigated. Notably, changes in the local

stiffness of atherosclerotic lesions occur early, prior to substantial histological changes in the

matrix16. Similar early dynamic changes in matrix mechanics are expected in sclerotic valves,

but the alterations in the micromechanical stiffness of the valve matrix that occur with disease

progression and the factors that contribute to these mechanical changes have yet to be

determined. While the collagen matrices used here are far less complex than valve tissue in

composition and structure, the modulus of the stiffer gels was comparable to that of sclerotic

valve tissue (based on relative changes from normal tissue186) and the modulus of the more

compliant gels was approximately two- to three-fold greater than the micromechanical tensile

modulus of normal aortic valve tissue23, but similar to that of early atherosclerotic lesions16. The

matrices that mimicked the normal or early disease stiffness promoted osteogenic differentiation

when the cells were exposed to calcific soluble signals, consistent with the appearance of

osteoblast-like VICs early in CAVD234 prior to the substantial matrix changes and calcification

72

that ultimately stiffen the matrix. While these findings are intriguing, further investigation is

required to determine the role of matrix stiffness in modulating osteogenic and non-osteogenic

calcification processes in vivo. In particular, translation of these findings to valve disease

requires additional studies of the dynamic temporal and spatial changes that occur in matrix

structure and composition during disease development and their relationships to the

micromechanical properties of the valve matrix and cell phenotypes.

In summary, by using a simple collagen-based culture platform, we were able to demonstrate the

ability of VICs to “sense” their local mechanical microenvironment. Our data demonstrate that

the differentiation of and calcification by VICs in response to biochemical factors are modulated

by the mechanical properties of the matrix. These data suggest an important regulatory role for

matrix mechanics in valve cell biology, with implications for the interpretation of in vitro models

of VIC calcification, the selection of biomaterials for tissue engineered heart valves, and possibly

disease development. While we observed that either osteoblast or myofibroblast differentiation

of VICs can result in calcification in vitro, the two processes are distinct and respectively mimic

aspects of either bone formation or apoptosis-associated calcification in vivo. The identification

of distinct calcification processes suggests the need for therapies that are specific, yet capable of

targeting multiple pathways involved in VIC pathological differentiation and valve calcification.

73

Chapter 6

6. Effect of CNP on Pathological Differentiation of VICs

The heterogeneous population of VICs has been shown to differentiate into pathologic

myofibroblasts and osteoblasts in various cell culture studies55, 58, 239. Myofibroblasts and

osteoblasts are often found in calcified AV47, 49. Myofibroblasts are responsible for pathological

matrix remodeling and fibrosis, whereas osteoblasts contribute to bone-matrix protein deposition

and calcification. In culture, myofibroblast differentiation of VICs is associated with apoptosis-

dependent calcification69. Clearly, these phenotypic changes resulting from pathological

differentiation of VICs contribute to valve dysfunction, and therefore pharmacological inhibition

of VIC pathological differentiation may prevent the occurrence of valve sclerosis and

calcification.

Clues to the molecular determinants of CAVD may come from the tendency for calcified lesions

to form more readily in the fibrosa of aortic valve leaflets than on the ventricular (disease-

protected) side14. Profiling of gene expression by endothelial cells from opposite sides of normal

porcine aortic valves revealed statistically significant side-dependent differential expression of

584 genes6. One of the most highly differentially expressed genes with higher expression on the

disease-protected side of the leaflet was CNP. It was postulated that CNP regulates valve

homeostatsis via paracrine signaling between VECs and VICs, and therefore the higher

expression of CNP by the endothelium on the ventricular side may contribute to the side-specific

disease protection. Further, CNP was found to be expressed in healthy valve leaflets, but its

expression and that of its activator, furin, were downregulated in VICs and VECs of human

sclerotic valves19, indicating that the changes in CNP expression during the progression of

CAVD are not limited to VECs. Although these data suggest that CNP may act in both autocrine

and paracrine manners to protect against sclerosis, the direct influence of CNP on VIC response

has yet to be studied. Further, it is not even clear that VICs can respond to CNP, as the

expression of components of the CNP signaling pathway, such as the NPR-B receptor and cGMP

activity, has yet to be verified. Motivated by findings from the microarray study with VECs and

the ex vivo study with human sclerotic valves, we hypothesized that VICs are equipped with

74

CNP signaling components and CNP elicits protective effects against CAVD by regulating the

pathological differentiation of VICs into myofibroblasts and osteoblasts. To test the hypothesis,

we first tested correlations between CNP expression and pathological differentiation of VICs in

normal and sclerotic valve leaflets. We then determined the expression of NPR-B, the primary

receptor for CNP, and the ability of CNP to induce the activation of cGMP signaling in vitro.

Myofibroblast and osteoblast differentiation of VICs in the presence or absence of CNP was

determined in vitro. As demonstrated previously, matrix stiffness modulates VICs response to

exogenous biochemical factors, and therefore we kept surface stiffness constant in this current

study in hopes to first identify if CNP has any direct cellular effect on VICs.






6.1.1. Animal Model and Valve Interstitial Cell Isolation

For histological analysis, sclerotic and healthy leaflets were kindly provided by Dr. Peter Davies’

laboratory at the University of Pennsylvania. Briefly, sclerotic leaflets were obtained from pigs

fed an atherogenic diet240 for five months. This hypercholesterolemic diet induces focal

calcification preferentially in the fibrosa of the valves, similar to those observed in human

sclerotic valves6. Control pigs were fed normal chow. Leaflets were fixed in 10% NBF, paraffin-

embedded and serial sectioned for immunofluorescent staining.

For cell culture, normal aortic valves were obtained from eight month old pigs immediately after

death (Quality Meat Packers, Toronto, ON). VICs were isolated by collagenase digestion as

described in Chapter Five. Only primary VICs were used to preserve the heterogeneity of the cell

population as explained in Chapter Five.

6.1.2. Cell Culture

Cells were plated on uncoated TCPS at a seeding density of 10,000 cells/cm2, and medium was

changed every two days unless otherwise stated. To evaluate myofibroblast differentiation, cells

75

were cultured in complete medium (DMEM with 10% FBS (Hyclone, Logan, UT, Lot #

KRA25425), 10,000 units/mL penicillin, and 10 mg/mL streptomycin) for seven days at 37 °C

and 5% CO2. To determine the myofibroblast content of freshly isolated VICs, an aliquot of

isolated cells was cytospun onto microscope glass slides, fixed and stained for α-SMA with

monoclonal mouse anti--SMA antibody (20 g/mL clone 1A4 mouse anti-human monoclonal

primary antibody) and 20 g/mL Alexa Fluor® 568 goat anti-mouse antibody, and for nuclei

with Hoechst 33242 dye. The dose response of CNP (CNP-22, Bachem, Torrance, CA) was

determined by measuring cGMP activity with concentrations of CNP of 0 nM, 1 nM and 100nM,

and -SMA expression of VICs with concentrations of CNP of 0 nM, 0.1 nM, 1 nM, 10 nM and

100 nM. The concentration of CNP with the most detectable activation of cGMP activity and

reduction in -SMA expression was used in subsequent experiments. To evaluate osteogenic

differentiation, cells were cultured in the presence or absence of CNP with calcifying medium

consisting of complete medium supplemented with 10 mM β-glycerophosphate, 10 g/mL

ascorbic acid and 10 nM dexamathesone for up to 21 days at 37 °C and 5%CO2.

6.1.3. Histological Analysis

Paraffin-embedded leaflets from normal and sclerotic leaflets were sectioned transversely to the

valve long axis. Serial paraffin sections were immunostained for α-SMA, Runx2/Cbfa-1 (20

g/mL Runx2 polyclonal rabbit anti-mouse antibody, Santa Cruz Biotechnology, Santa Cruz,

CA and 20 g/mL Alexa Fluor ® 568 goat anti-rabbit secondary antibody) and CNP (2 g/mL ,

goat polyclonal C-19 primary antibody, Santa Cruz Biotechnology, Santa Cruz, CA and

20 g/mL Alexa Fluor ® 568 rabbit anti-goat secondary antibody). The sections were also

stained for nuclei with Hoechst 33242 dye.

6.1.4. Dose Response of CNP

VICs were cultured in complete DMEM for 48 hours and RNA was extracted. Transcript

expression of NPR-B was determined by PCR, followed by gel electrophoresis (Accession

number: DQ487044.1, forward primer: 5’-agcattaccgtaccctgggtg-3’ and reverse primer: 5’-

tagtgaggccggtcatcatgt -3’).

76

In a separate series of experiments, VICs were treated with 0 nM, 1 nM and 100 nM of CNP for

10 minutes in DMEM. The level of cGMP activity of the cultures was measured using cGMP

Direct Biotrak EIA assay kit (Amersham/GE Healthcare, Baie d’Urfe, Quebec).

To determine the dose-dependent response of -SMA expression by VICs, cells were cultured in

complete DMEM with CNP of 0 nM, 0.1 nM, 1 nM, 10 nM and 100 nM for seven days,

followed by western blotting. . Briefly, cell cultures were washed with ice-cold PBS followed by

the addition of 1× lysis buffer. Cells were scraped and transferred to pre-chilled tubes, followed

by incubation on ice for 30 minutes. Cell lysates were obtained by centrifugation at 16200 × g

for 15 minutes at 4 oC. The protein concentrations of cell lysates were determined by the micro

BCA protein assay (Pierce, Rockford, USA). 1.5 g of protein extract from each treatment group

was loaded on to 10% SDS-polyacrylamide gels for electrophoresis. Samples were then

transferred to polyvinylidene fluoride membranes, followed by immunoblotting for α-SMA (20

ng/mL clone 1A4 mouse anti-human monoclonal primary antibody) and glyceraldehyde-3-

phosphate dehydrogenase (83 ng/mL GAPDH mouse monoclonal antibody, Stressgen, Ann

Arbor, Michigan). Expression of α-SMA was quantified by densitometry using ImageJ Software

(NIH, Bethesda, MD) and normalized to GAPDH expression.

6.1.5. Cellular Proliferation

Proliferation of VICs up to 15 days in culture was determined based on measurement of cellular

DNA content (CyQuant® NF cell proliferation assay kit, Invitrogen, Burlington, ON).

6.1.6. Evaluation of Osteogenic Differentiation

VICs were cultured in calcifying media to promote osteogenic differentiation. Transcript levels

of Cbfa-1/Runx2, osteonectin and osteocalcin after three, eight and sixteen days of culture in

calcifying media with or without CNP were measured using qRT-PCR. ALP and osteocalcin

staining was performed after 21 days of culture in calcifying media. Calcium deposition was

determined by ARS staining after 14 days of culture. In addition, a colony forming unit-ALP

(CFU-ALP) assay was used to determine the frequency of VIC osteoprogenitor differentiation

under the influence of CNP58. Briefly, viable primary VICs were seeded at 0.2 cells/well into

96-well plates. The cells were cultured for three days in complete media to permit cell adhesion,

at which point the media were replaced with calcifying media with or without CNP and changed

77

every two days for three weeks. The cells were subsequently fixed with 10% NBF and stained

for ALP. For each plate, the number of wells without ALP-positive aggregates was recorded.

The ratio of wells without ALP-positive aggregates to the total number of wells was calculated.

Based on Poisson’s distribution, the negative natural logarithm of the ratio of wells without

aggregates to the total number of wells is the CFU-ALP frequency or equivalently, the expected

number of osteoprogenitors per plate under the specific culture conditions.

6.1.7. Evaluation of Myofibroblast Differentiation

After seven days of culture in complete DMEM with or without CNP, cells were fixed with 10%

NBF, permeablized and immunostained for α-SMA, followed by nuclear counterstaining.

Expression of α-SMA was quantified by densitometry of the immunoblots. In addition, CNP

expression in cultured VICs was evaluated by immunoflurescence staining after plating and after

five days of treatment with 1 ng/mL TGF-1 to induce myofibroblast differentiation.

To investigate changes in myofibroblast function, collagen deposition and cell contractility were

analyzed. Collagen content was measured with Sirius Red dye release method as described

previously241. Briefly, cultured cells were fixed with 10% NBF, followed by one hour incubation

at room temperature with 0.1% Sirius Red F3BA reconstituted in saturated picric acid. Stained

samples were washed five times with 10 mM HCl and then rinsed with distilled water. For

quantification, Sirius Red dye was released by 0.1 M NaOH for five minutes. The absorbance of

the supernatant containing the released dye was measured at 540 nm. Absorbance was

normalized by total cell number determined by DNA content, which was measured using

PicoGreen (Molecular Probes, Eugene, OR). Contractility was measured using standard stress-

relaxation collagen gel model224. In brief, VICs were cultured on constrained, compliant collagen

gels for four days in complete media with (100 nM) or without CNP at a cell density of 10,000

cell/cm2. Myofibroblast differentiation of VICs was induced by treating the cells with 1 ng/mL

of TGF-1 for 48 hours, after which the gels were released to allow for contraction. Images of

the collagen gels were taken every 30 minutes and the gel areas were determined using ImageJ.



and experiments were repeated at least three times. Unpaired Student’s t-test was used for

78



6.2. Results

6.2.1. Expression of Pathological Markers and CNP in Normal and Sclerotic Aortic Valves

Immunofluorescent staining of normal and sclerotic aortic valve leaflets revealed that the

expression of CNP is spatially mutually exclusive with that of -SMA and Runx2/Cbfa-1

(Figure 6.1). In normal valves, CNP expression was abundant throughout the interstitium, with

slightly higher expression in the ventricularis (Figure 6.1: A). There were few -SMA positive

cells in normal leaflets (Figure 6.1: D). Numerous cells in sclerotic valves stained strongly

positive for -SMA, particularly near lesions (Figures 6.1: E and F), indicative of VIC

myofibroblast differentiation. In contrast, weak or absent CNP staining was observed in these

regions in sclerotic valves (Figures 6.1: B and C). Cbfa-1/ Runx2, an osteochondral transcription

factor, was not detected in normal leaflets (Figure 6.1: G), but was abundant in sclerotic valves

(Figure 6.1: H and I) accompanied by low expression of CNP.

6.2.2. Molecular Components of CNP Signaling

Natriuretic receptors (NPR-A, NPR-B, NPR-C) are expressed in cardiac atria and ventricles, but

their expression in VICs is unknown. Of the three receptors, CNP has the highest binding affinity

with NPR-B. NPR-B is linked to a guanylyl cyclase domain, and CNP binding with NPR-B

receptor induces cGMP synthesis, which mediates downstream cellular responses (reviewed in 242). To determine if VICs are equipped with molecular components to detect and to process CNP

signaling, we identified the expression of NPR-B mRNA by VICs as well as the induction of

cGMP synthesis by CNP. Primary VICs indeed express transcripts of NPR-B (Figure 6.2: A) and

cGMP synthesis by VICs was observed within 10 minutes of treatment with 100 nM of CNP

(Figure 6.2: B).

6.2.3. Dose Response of CNP

Induction of cGMP synthesis by CNP showed a threshold effect. VICs appeared to be highly

responsive to 100 nM of CNP (Figure 6.2: B). We further investigated if there was a dose-

dependent response by measuring the expression level of -SMA with Western blotting. By

79

treating the cells with CNP concentrations of 0, 0.1, 1, 10 and 100 nM, it was found that

suppression of -SMA was most prominent with 100 nM of CNP (Figure 6.3: A and B).

Together with the cGMP synthesis data, we decided that in this in vitro culture model, 100 nM of

CNP was the most effective in mediating detectable cellular responses and all subsequent

experiments were conducted with 100 nM of CNP.

Figure 6.1. Expression of CNP, -SMA and Runx2/Cbfa-1 in normal and sclerotic porcine

aortic valves

(A-C) Immunofluorescent staining of CNP (red) and nucleus (blue). (D-F) Immunostaining of -

SMA (red) and nucleus (blue). (G-I) Immunostaining of Runx2/Cbfa-1 (red) and nucleus (blue).

“AO” denotes the aortic side and “V” denotes the ventricular side of the normal leaflets. Only

the aortic side of the sclerotic leaflets was shown.

Normal Sclerotic

CN

P

-SM

AR

un

x2

/Cb

fa1

Boxed Area

Boxed Area

G

Boxed Area

A B C

D E F

H I

100 m

AO

V

AO

V

V

AO

Normal Sclerotic

CN

P

-SM

AR

un

x2

/Cb

fa1

Boxed Area

Boxed Area

G

Boxed Area

A B C

D E F

H I

100 m

AO

V

AO

V

V

AO

CN

P

-SM

AR

un

x2

/Cb

fa1

Boxed Area

Boxed Area

G

Boxed Area

A B C

D E F

H I

100 m

AO

V

AO

V

V

AO

80

NPR-B (142 \bp)

0 1 100

CNP concentration (nM)

cGM

P c

onc

en

tra

tion

(fm

ol/w

ell)

No

t de

tec

ted

No

t de

tec

ted

0

3

6

9

12*

B.

A.

NPR-B (142 \bp)

0 1 100


cGM

P c

onc

en

tra

tion

(fm

ol/w

ell)

No

t de

tec

ted

No

t de

tec

ted

0

3

6

9

12*

B.

A.

NPR-B (142 \bp)

0 1 100


cGM

P c

onc

en

tra

tion

(fm

ol/w

ell)

No

t de

tec

ted

No

t de

tec

ted

0

3

6

9

12*

B.

0

3

6

9

12*

B.

A.

Figure 6.2. Expression of NPR-B and activation of cGMP by CNP

(A) Primary VICs from three different isolations expressed transcript for NPR-B (lane 1,2,3).

(B) Treatment of VICs with 100 nM of CNP significantly induced cGMP synthesis in

comparison to those treated with 1 nM of CNP and the untreated culture, * P < 0.05

6.2.4. Cellular Proliferation and Morphological Changes

CNP had little effect on proliferation whether cells were cultured in complete media

(Figure 6.4: A) or calcifying media (Figure 6.4: B). In complete media, VICs displayed typical

fibroblastic morphology with elongated processes, and no substantial morphological differences

were observed between CNP-treated and untreated cells (data not shown). However, when cells

were cultured in calcifying media for more than ten days, formation of multicellular aggregates

was prominent only in the absence of CNP (Figures 6.4: C and D). Since CNP had no effect on

cellular proliferation in all culture conditions (P ≥ 0.4 between untreated and CNP treated

samples at each time point), cell density was similar in all cases and was unlikely to contribute to

the observed differences in aggregation or the phenotypic differences described below.

81

Figure 6.3. Dose-dependent -SMA expression by VICs

Immunoblots of -SMA and GAPDH and the corresponding densitometric quantification for

culture after seven days with CNP concentrations of 0, 0.1, 1, 10 and 100 nM. ** P < 0.001 and

* P < 0.05 in comparison to no CNP treatment.

6.2.5. CNP Inhibits Calcification and Osteogenic Differentiation of VICs

Consistent with previous studies58, VICs formed multicellular aggregates when cultured in

calcifying media (Figure 6.4: D). In untreated cultures, the aggregates contained calcium as

shown with the intense positive staining of ARS (Figure 6.5: A and B). CNP-treated VICs

stained diffusely for calcium with minimal aggregate formation (Figure 6.5: C). The few

aggregates that did form in the CNP-treated culture stained weakly for ARS (Figure 6.5: D).

Quantification of the number of ARS-positive aggregates confirmed that CNP treatment

inhibited calcification (Figure 6.5: E).

SMA (42 kDa)

GAPDH (36 kDa)

0 0.1 1 10 100

Concentration of CNP (nM)

0

20

40

60

80

100

120

0 0.1 1 10 100


Pe

rce

nta

ge

of S

MA

e

xpre

ssio

n r

ela

tive

t o n

o tr

ea

tme

nt *

** **

**

SMA (42 kDa)

GAPDH (36 kDa)

0 0.1 1 10 100


0

20

40

60

80

100

120

0 0.1 1 10 100


Pe

rce

nta

ge

of S

MA

e

xpre

ssio

n r

ela

tive

t o n

o tr

ea

tme

nt *

** **

**

82

Figure 6.4. Proliferation and morphology of cells with or without CNP treatment

(A) Proliferation of VICs from four hours to 15 days in complete media and (B) calcifying

media. (C) In calcifying media, few aggregates formed in CNP-treated cultures, (D) whereas

abundant aggregate formation was observed in untreated cultures.

Formation of calcified aggregates occurs in vitro through at least two possible processes, one

associated with myofibroblast apoptosis and the other associated with osteoblast

differentiation79. Both types of aggregates were observed in the untreated cultures, indicating

calcification in the untreated VICs was due in part to osteoblast differentiation. Notably,

transcript expression of Cbfa-1/Runx2, osteonectin and osteocalcin in CNP-treated cells was

lower than that of untreated samples over the culture duration, with a significant reduction in

osteonectin expression after as little as eight days of culture (Figures 6.6: A, B and C).

Consistent with the transcriptional profile, expression of bone-related proteins was also reduced

with CNP treatment. ALP activity (Figure 6.7: A and B) and osteocalcin expression were low

Ce

ll nu

mb

er(

1x1

0-3

)

0

100

200

300

400

500

600

700

4 hrs 5 days 10 days 15 daysCulture duration

Complete media + 100nM CNPComplete media

A. B.

0

100

200

300

400

500

600

700

4 hrs 5 days 10 days 15 days

Culture duration

OS media + 100nM CNPOS media

C. Calcifying media + 100 nM CNP D. Calcifying media

Culture duration

Calcifying media + 100 nM CNPCalcifying media

Complete media + 100 nM CNPComplete media

Ce

ll nu

mb

er(

1x1

0-3

)

0

100

200

300

400

500

600

700

4 hrs 5 days 10 days 15 daysCulture duration

Complete media + 100nM CNPComplete media

A. B.

0

100

200

300

400

500

600

700

4 hrs 5 days 10 days 15 days

Culture duration

OS media + 100nM CNPOS media

C. Calcifying media + 100 nM CNP D. Calcifying media

Culture duration

Calcifying media + 100 nM CNPCalcifying media

Complete media + 100 nM CNPComplete media

83

within CNP-treated cultures (Figure 6.7: C), but high within multicellular aggregates in the

untreated cultures (Figure 6.7: D).

To further understand the cellular target of the anti-osteogenic effect of CNP, we determined the

CFU-ALP frequency as a measure of the differentiation of single osteoprogenitor cells58. Under

CNP treatment, the CFU-ALP frequency was significantly reduced in comparison to the

untreated culture (Figure 6.8). These data suggest that the anti-osteogenic effect of CNP was

mediated by suppressing the osteogenic differentiation of progenitor cells in the VIC population.

6.2.6. Inhibition of Myofibroblast Differentiation by CNP

Myofibroblast differentiation can be induced by biochemical (e.g., cytokines such as TGF-165)

and mechanical stimuli (e.g., a rigid culture surface55, 79). To evaluate the myofibroblast

differentiation of VICs, we cultured freshly isolated VICs on stiff TCPS with complete medium.

As in Chapter Five, we aimed to achieve a cell population that reflected that of native valves, and

therefore the effect of CNP was only tested on primary VICs, as subculturing induces

myofibroblast differentiation55. Freshly isolated VICs did not express -SMA, indicative of an

undifferentiated cell population (Figure 6.9: A). After seven days of culture, cells treated with

CNP (Figure 6.9: B) expressed little -SMA compared to untreated cells, which had prominent

-SMA stress fibers (Figure 6.9: C). Western blotting of CNP-treated and untreated cultures

further confirmed significantly lower -SMA expression with CNP treatment (Figures 6.10: A

and B).

84

Figure 6.5. CNP modulates calcification by VICs

(A and B) Untreated cultures stained intensely with ARS, indicating high and localized

concentration of calcium deposition. (C) CNP-treated cultures displayed diffuse ARS staining

and (D) only weak ARS staining even in the few aggregates that formed. (E) CNP inhibited

formation of calcified aggregates after fourteen days in osteogenic medium, * P < 0.05.

A. B.

0

20

40

60

80

100

120

100nM CNP No CNP

Nu

mb

er

of A

RS

-po

sitiv

ea

gg

reg

ate

s

*

E.

C. D.

A. B.

0

20

40

60

80

100

120

100nM CNP No CNP

Nu

mb

er

of A

RS

-po

sitiv

ea

gg

reg

ate

s

*

E.

C. D.

85

Figure 6.6. Expression of bone-related transcripts

(A) Transcript expression of Runx2/Cbfa-1, (B) osteonectin and (C) osteocalcin in CNP-treated

cultures relative to that of untreated samples. * P < 0.05. (A relative gene expression level of less

than one indicates lower expression with CNP treatment relative to that without CNP treatment).

A.

B.

Re

lati

ve R

un

x2/C

bfa

-1

exp

ress

ion

(AU

) *

0.00.20.40.60.81.01.21.41.6

Day 3 Day 8 Day 16

*

0.00.20.40.60.81.01.2

Day 3 Day 8 Day 16

**

Re

lati

ve o

ste

one

ctin

exp

ress

ion

(AU

)

C.

Re

lati

ve o

ste

oca

lcin

exp

ress

ion

(AU

)

0.0

1.0

2.0

3.0

4.0

5.0

Day 8 Day 16

*

A.

B.

Re

lati

ve R

un

x2/C

bfa

-1

exp

ress

ion

(AU

) *

0.00.20.40.60.81.01.21.41.6

Day 3 Day 8 Day 16

*

0.00.20.40.60.81.01.2

Day 3 Day 8 Day 16

**

Re

lati

ve o

ste

one

ctin

exp

ress

ion

(AU

)

C.

Re

lati

ve o

ste

oca

lcin

exp

ress

ion

(AU

)

0.0

1.0

2.0

3.0

4.0

5.0

Day 8 Day 16

*

86

Figure 6.7. Expression of bone-related proteins

(A) At the protein level, CNP-treated cells had low levels of ALP activity and (C) weak diffuse

staining for osteocalcin. (B) Untreated cultures displayed high levels of localized ALP activity as

well as (D) osteocalcin within the aggregates (inset represents no primary antibody control; black

arrow indicates aggregate).

To investigate if CNP regulates biochemically-induced myofibroblast differentiation of VICs, we

compared the difference in CNP expression by VICs before and after five days of culture with

TGF-1. VICs expressed CNP but not -SMA after one day in culture (Figure 6.11: A). After

five days of induction with TGF-β1, the majority of VICs differentiated into -SMA positive

myofibroblasts as expected (Figure 6.11: B). Similar to the mutually exclusive expression of

-SMA and CNP observed in the histological analysis, -SMA positive myofibroblasts did not

express CNP. A few cells that did not differentiate into myofibroblasts with TGF-1 induction

stained positive for CNP (Figure 6.11: B, arrow). Co-expression of -SMA and CNP was

therefore rarely observed in vivo or in vitro.

A. B.

C. D.

A. B.

C. D.

87

Figure 6.8. Effect of CNP on osteoprogenitor subpopulation

The CFU-ALP frequency was reduced in CNP-treated cultures, suggesting that CNP inhibits

osteogenic differentiation of the valve progenitor subpopulation (* P < 0.05).

Myofibroblasts typically have increased collagen synthesis and cellular contractility55, 59, 65. CNP

treatment suppressed collagen synthesis compared with untreated cells (Figure 6.12). To

evaluate the effect of CNP on myofibroblast-induced contractility, we cultured TGF-1-treated

VICs on constrained collagen gels and then measured gel contraction by VICs upon gel release.

Gels treated with CNP contracted significantly less than untreated gels, suggesting that CNP

suppressed TGF-1-induced myofibroblast differentiation (Figure 6.13: A and B).

6.3. Discussion

CNP is expressed in disease-protected regions of normal porcine valves6 and in normal human

valves19, but its expression is down-regulated in calcified aortic valves19 and in advanced

atherosclerotic lesions243. In the current study, we identified the expression of NPR-B receptor

and cGMP activity in VICs, which are components of CNP signaling. We found mutually

exclusive spatial expression of CNP and disease-related VIC phenotypes in vivo, and confirmed

in vitro that CNP inhibits differentiation of VICs to myofibroblasts and osteoblasts, phenotypes

associated with CAVD. Our current findings provide a cellular basis responsible for the

protective of CNP against valve calcification.

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

100 nM CNP No CNP

CF

U-A

LP

fre

que

ncy

*

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

100 nM CNP No CNP

CF

U-A

LP

fre

que

ncy

*

88

Figure 6.9. CNP inhibits expression of myofibroblast marker

(A) Immunostaining of freshly isolated VICs, (B) CNP-treated VICs and (C) untreated VICs

for -SMA (red) and nucleus (blue).

The influence of CNP on osteogenic differentiation and calcification appear to be tissue and

cell-type specific. Several animal models with either targeted disruption of NPPC244, the gene

for natriuretic peptide precursor C, or loss-of-function mutation in NPR-B receptor107, 245

display skeletal defects due to disturbed chondrogenesis during endochondral ossification.

CNP-dependent skeletal growth was also demonstrated in cell culture studies with pre-

osteoblastic cells188, 246, 247 and calvaria cells248. Treatment of pre-osteoblastic cells with CNP

has been reported to increase calcium deposition and the expression of ALP and osteocalcin

via NPR-B/cGMP signaling, indicative of CNP-induced osteoblast differentiation188, 246, 247.

Although CNP promotes ossification in bone cells, the reciprocal effect was found in

vascular cells, suggesting its response is cell-type specific. Vascular smooth muscle cells

treated with CNP displayed reduced calcium deposition and ALP expression187. Here, by

manipulating the culture conditions to promote osteogenic differentiation, CNP inhibited the

differentiation of VICs into osteoblasts, as demonstrated by reduced calcium deposition and

A.

B. C.

A.

B. C.

89

No

rma

lize

d

-SM

A e

xpre

ssio

n-SMA (42 kDa)

GAPDH (36 kDa)

100nM CNP no CNP

0.0

0.5

1.0

1.5

2.0

2.5

CNP + CNP -100nM CNP no CNP

*

No

rma

lize

d

-SM

A e

xpre

ssio

n-SMA (42 kDa)

GAPDH (36 kDa)

100nM CNP no CNP

0.0

0.5

1.0

1.5

2.0

2.5


*

-SMA (42 kDa)

GAPDH (36 kDa)

100nM CNP no CNP

0.0

0.5

1.0

1.5

2.0

2.5


*

lowered expression of bone-related transcripts and proteins in whole cell populations,

providing additional evidence to support the paradoxical effect of CNP in the vascular

system and the skeletal system.

Figure 6.10. Quantification of -SMA expression

Western blot of -SMA and GAPDH and the corresponding densitometric quantification for

culture after seven days with or without CNP treatment. * P < 0.05.

We have previously identified a large subpopulation of progenitor cells in VICs with multipotent

differentiation potential58, and therefore investigated their responsiveness to CNP using single-

cell clonal assays. By treating cells at the start of the experiments, we tested the response of

uncommitted, undifferentiated valve progenitors to CNP. We found that CNP attenuated

osteogenic differentiation of the valve progenitor subpopulation significantly. Although the

effect of CNP on valve progenitors at different stages of committment will require further

investigation, our initial findings suggest the ability of CNP to prevent osteogenic differentiation

of at least the undifferentiated valve progenitor cells. Commitment of cells to specific lineages

has been shown to influence the responsiveness of cells to CNP treatment. For example, ROB-

C26 cells induced by BMP-2 to undergo osteoblast differentiation displayed high levels of CNP-

mediated cGMP activity, whereas the same cells committed to the adipogenic lineage with

dexamathesone treatment exhibited marked reduction of CNP-mediated cGMP activity249.

90

Presumably VICs, including the subpopulation of osteoprogenitors, undergo pathological

differentiation during CAVD pathogenesis, leading to their commitment to myofibroblast or

osteoblast cell lineages, which may ultimately alter their response to CNP. Hence, the

effectiveness of CNP against CAVD in vivo may depend on the stage of the disease. Therefore,

future work on the therapeutic application of CNP for CAVD should explore the stage-related

effect of CNP treatment as a function of the disease progression.

Figure 6.11. Mutually exclusive expression of CNP and -SMA in cultured VICs.

(A) After one day in culture, VICs expressed CNP (red). (B) After five days of growth in media

containing TGF-β1, the majority of VICs differentiated into myofibroblasts that expressed -

SMA (green); however a few cells that did not express -SMA stained positive for CNP (white

arrow).

VICs can undergo myofibroblast differentiation, which is closely associated with apoptosis-

dependent calcification in vitro as described in Chapter Five. CNP is widely recognized to

regulate fibrosis in other tissues. For example, administration of CNP in animal models reduced

fibrosis associated with vascular intimal thickening110, pulmonary fibrosis111, and myocardial

infarction112. We observed that CNP attenuated myofibroblast differentiation of quiescent VICs

as reflected by the down-regulation of -SMA and loss of myofibroblast-related functions. Co-

expression of -SMA and CNP was rarely observed in vivo or in vitro. Although the influence

of CNP on apoptosis varies with cell type250-252, an anti-apoptotic effect of CNP has been

reported in some cell types such as pulmonary endothelial cells252. In our culture, notable cell

death was not observed as majority of the cells were well-spread on TCPS and plasma membrane

A. B.A. B.

91

blebbing resulting from cleavage of cytoskeleton proteins by caspases during apoptosis253, 254

was not evident. These observations clearly demonstrate that CNP is not pro-apoptotic in VICs.

Figure 6.12. CNP affects function associated with activated myofibroblasts

Collagen production of CNP-treated cells was significantly less than the

untreated culture. * P < 0.05.

It has been well documented that the actions of CNP are modulated through membrane-bound

receptors, mainly NPR-B and NPR-C, of which only NPR-B is linked to the cGMP-dependent

signaling cascade. We found that VICs express the transcript of NPR-B and synthesize cGMP in

response to CNP treatment. Additional studies are required to determine if the observed CNP-

mediated effects on VIC differentiation involve NPR-B/cGMP pathway. Because of the lack of

an NPR-B antagonist, siRNA-based knockdown of NPR-B would be one approach to test its

role.

Co

llag

en

pro

duc

tion

(A

54

0nm

no

rma

lize

d t

o t

ota

l DN

A)

0.0

1.0

2.0

3.0

4.0

5.0

6.0

7.0

DMEM + CNP DMEM 100nM CNP no CNP

*

Co

llag

en

pro

duc

tion

(A

54

0nm

no

rma

lize

d t

o t

ota

l DN

A)

0.0

1.0

2.0

3.0

4.0

5.0

6.0

7.0

DMEM + CNP DMEM 100nM CNP no CNP

*

92

Figure 6.13. Contractility of VICs with or without CNP treatment.

Cells were seeded on the surface of constrained collagen gels. Contractility was recorded every

half an hour after gel release. Untreated VICs were more contractile than those treated with CNP

(* P < 0.05, ** P < 0.06).

In summary, the results of this study demonstrate that VICs express components of CNP

signaling. CNP inhibits myofibroblast and osteoblast differentiation of VICs, which may prevent

calcification. These findings provide a cellular mechanism by which CNP maintains valve

homeostasis and protects against aortic valve calcification in vivo. This fundamental knowledge

regarding CNP enables future studies aimed at the identification of the molecular mechanisms of

its putative protective actions, both in vitro and in vivo.

0 hr 0.5 hr 1 hr 1.5 hrs 2.0 hrs

No

CN

P1

00

nMC

NP

0.0

0.2

0.4

0.6

0.8

1.0

1.2

0.5 1 1.5 2

100nM CNP No CNP

Ge

l are

a (

cm2) * **

Time after gel release (hr)

0 hr 0.5 hr 1 hr 1.5 hrs 2.0 hrs

No

CN

P1

00

nMC

NP

0.0

0.2

0.4

0.6

0.8

1.0

1.2

0.5 1 1.5 2

100nM CNP No CNP

Ge

l are

a (

cm2) * **

Time after gel release (hr)

93

Chapter 7

7. The Combined Effects of Mechanical and Biochemical

Cues on the Transcriptional Regulation of VICs

It is well accepted that matrix stiffness regulates phenotypic drift and functions of a wide range

of cell types55, 221. Recent studies further suggest that substrate stiffness may modulate the release

of and the response to biochemical factors by cells12. The correlative link between cellular

response, mechanical cues and biochemical cues from the microenvironment has yet to be

studied thoroughly. Little is known regarding the impact of the combined effects of mechanical

and biochemical cues on VIC biology, despite increasing evidence of close relationships between

changes in tissue mechanics, soluble factors and disease progression7-9, 176.

Matrix stiffness regulates the responses of cells to biochemical factors to ultimately define cell

behaviour. The first demonstration of this was the differential effects of TGF-β on

myofibroblasts by matrix stiffness11. Myofibroblasts cultured on more compliant substrates were

insensitive to TGF-β, whereas those cultured on stiffer substrates were highly responsive to

TGF-β. Similar differential effects of TGF-β were also observed in VICs cultured on compliant

and stiff matrices as described in Chapter Five, suggesting that matrix stiffness may also

modulate the response of VICs to soluble factors. Others have also reported growth factors

mediate cellular response in a matrix stiffness-dependent manner169, 170. In addition, soluble

factors and matrix stiffness have been shown to synergistically guide stem cell commitment to

particular lineages130.

We have demonstrated the ability of matrix stiffness or CNP alone to modulate the pathological

differentiation of VICs into myofibroblasts and osteoblasts. In this chapter, we investigated the

combined effect of matrix stiffness and CNP on the transcriptional regulation of VICs. We

identified the impact of matrix stiffness on CNP-dependent transcript expression. The evaluation

of cell response to biochemical cues in the context of the cellular mechanical environment will

provide a more complete understanding of valve cell biology.

94


7.1.1. Cell Culture





7.1.2. Sample Preparation

Primary VICs were isolated from porcine aortic valves by collagenase digestion as described in

Chapter Five. Compliant and stiff collagen matrices were constructed following procedures

described in Chapter Four. VICs were seeded on collagen matrices at 10,000 cells/cm2 in

calcifying media with (100 nM) or without CNP. A total of four different experimental

conditions were tested:

I. VICs cultured on compliant collagen matrices with calcifying media

II. VICs cultured on stiff collagen matrices with calcifying media

III. VICs cultured on compliant collagen matrices with calcifying media and 100 nM

CNP

IV. VICs cultured on stiff collagen matrices with calcifying media and100 nM CNP

After nine days in culture, VICs were released from collagen matrices by collagenase digestion.

Cell pellets were obtained by centrifugation, followed by aspiration of the supernatant. Total

RNA was isolated from cell pellets following standard protocols of the Micro RNeasy System

(Qiagen, Mississauga, ON). A total of 16 RNA samples from the four culture conditions using

cells from four separate VIC isolations were collected (N = 4). Universal reference RNA was

obtained by extracting RNA directly from freshly isolated VICs. RNA samples were quantified

with a NanoDrop Spectrophotometer (ND-1000, NanoDrop Technologies, Wilmington, DE).

Sample integrity based on the 28S:18S ribosomal RNA ratio was determined using Agilent 2100

Bioanalyzer (Agilent Technologies Canada, Mississauga, ON). All microarray samples had RNA

integrity number (RIN) of at least nine.

95

7.1.3. Microarray Experiments

RNA samples from the four culture conditions were labeled with Cy5 and the universal reference

RNA labeled with Cy3. Labeled RNA samples were competitively hybridized onto 44k 60-mer

Porcine Gene Expression Microarrays (G2519F, Design ID: 020109, Agilent Technologies

Canada, Mississauga, ON) at 65 oC and 20 rpm for 17 hours. Arrays were scanned using the

Agilent dual-laser DNA microarray scanner (G2565CA) and images were analyzed with Agilent

Feature Extraction Software (Version 10.5.1.1.).

7.1.4. Data Analysis

Intensity files were loaded into R statistical analysis software (Version 2.91, http://www.r-

project.org/) for preprocessing. Quality assessment was performed using Bioconductor package

(arrayQualityMetrics Version 2.2.1). One of the replicates for the RNA sample extracted from

VICs cultured on thin matrices with CNP treatment was identified to be an outlier from the log2

intensity boxplots for the red and green channels. Data generated from this sample were

eliminated from subsequent gene expression analyses. Gene expression analysis was done by

GeneSpring GX 10.0.2. Briefly, a flag filter was applied and those probes being present or

marginal in at least two out of eight samples (i.e. n = 4 for each test condition with the exception

of thin matrices with CNP treatment, for which n = 3) were kept for further analysis. An

expression filter was applied and those with raw intensities greater than 100 in at least two out of

eight samples were kept for further analysis. T-tests were performed between samples from two

different experimental conditions and the Benjamini & Hochberg False Discovery Rate (FDR)

method was used for multiple testing corrections. Transcript expression with fold change of

greater than two and P < 0.05 were recorded as statistically significant changes in fold

expression.

7.1.5. Partial Annotation Mapping and Identification of Biological Processes

The contents of the porcine gene expression microarray sourced from RefSeq (release 27, Jan

2008), UniGene (release 33, Feb 2008) and TIGR (The Institute for Genomic Research, release

12, Jun 2006). As porcine genome sequencing has yet to be completed, sequences unavailable in

the porcine genome database were examined for homology to Homo sapiens, Bos taurus or Mus

musculus genome by BLAST searches. The protein sequences were identified based on the gene

sequences using the search engine of the Entrez Nucleotide database

96

(http://www.ncbi.nlm.nih.gov/sites/entrez?db=nuccore). The Gene Ontology (GO) annotation for

each sequence was determined by searching the UniProtKB (Universal Protein Resource

Knowledgebase) database (http://www.uniprot.org/help/uniprotkb), and the corresponding

biological processes were identified as listed in the GO annotations. For entries without complete

GO annotations, biological functions of the genes were determined by literature searches. Due to

the large number of entries and the lack of complete porcine annotations, an exhaustive analysis

of the complete gene list is beyond the scope of this study. Differentially expressed genes with

putative significance to valve pathology were identified and discussed.

7.1.6. Venn Diagram Analysis

Venn diagram analysis was performed to determine genes that are modulated commonly or

exclusively by matrix stiffness and by CNP. Two-way and three-way Venn diagrams were

generated using the list of differentially expressed entries and VENNY online software

(http://bioinfogp.cnb.csic.es/tools/venny/index.html, BioinfoGP Bioinformatics for Genomics

and Proteomics CNB-CSIC, Madrid, Spain)255.

7.2. Results

7.2.1. Sample Characterization

The cell morphology after nine days of culturing on compliant and stiff matrices without CNP

treatment were the same as reported in Chapter Five. Cells formed aggregates more readily on

compliant matrices in comparison to those cultured on stiff matrices. When treated with CNP,

aggregate formation by VICs on both compliant and stiff matrices was reduced (Figure 7.1 and

Figure 7.2).

97

7.2.2. Differential Gene Expression by Matrix Stiffness

In the absence of CNP, 998 sequences were identified as differentially expressed (i.e., fold

difference greater than two and P < 0.05), with higher expression of 648 sequences and lower

expression of 350 sequences in VICs cultured on the compliant matrices relative to those

cultured on stiff matrices. Seventy-eight percent of the upregulated sequences and 93% of the

downregulated sequences displayed two- to five-fold changes in expression (Figure 7.3), while

only a small fraction of the gene list displayed an expression fold difference of greater than 30-

fold.

Figure 7.1. CNP inhibites aggregate formation on VICs cultured on compliant and stiff

matrices

Relief phase contrast images of VICs cultured on compliant matrices with CNP treatment (A)

and without treatment (B), and on stiff matrices with CNP treatment (C) and without treatment

(D). Formation of aggregates was predominately found in untreated cultures (B and D, white

arrows indicate cell aggregates).

A.

CNP treated Untreated

B.

C. D.

Co

mp

lian

tS

tiff

A.


B.

C. D.

A.


B.

C. D.

A.


B.

C. D.

Co

mp

lian

tS

tiff

98

Figure 7.2. CNP significantly reduces the total number of aggregates formed by VICs on

compliant and stiff matrices

Only a few aggregates were formed in CNP-treated cultures on compliant and stiff matrices

(number indicates the average aggregate count). * P < 0.05 compared to untreated cultured.

Figure 7.3. The distribution of sequences differentially expressed with matrix stiffness

Sequences associated with a wide range of biological processes were altered by matrix stiffness

(Tables 7.1 and 7.2). Multiple upregulated transcripts in VICs cultured on compliant matrices

were associated with immune response, apoptosis and growth (Table 7.1). Several osteoinductive

transcripts were upregulated in cells cultured on the compliant matrices, a culture condition that

favored osteogenic differentiation of VICs. These included nuclear factor of kappa light

polypeptide gene enhancer in B-cells inhibitor- (NFKBIA), bone morphogenetic protein-2

0

50

100

150

200

Compliant matrices Stiff matricesCulture substrate

Nu

mb

er

of a

gg

reg

ate

s With 100 nM CNPWithout CNP

*

*21

0

50

100

150

200

Compliant matrices Stiff matricesCulture substrate

Nu

mb

er

of a

gg

reg

ate

s With 100 nM CNPWithout CNP

*

*21

0

50

100

150

200

250

300

350

2.5 5 10 20 30 40 +Fold difference

Nu

mb

er

of e

ntr

ies

UpregulatedDownregulated

99

(BMP2) and mitogen-activated protein kinase phosphatase-1 (MKP-1). NFKBIA encodes for a

protein that negatively regulates Notch signaling pathway. Notch signaling has been shown to

regulate osteogenesis, including its inhibitory effect on the differentiation of mesenchymal

progenitor cells to osteoblast lineage 256. Notch1, part of the Notch signaling system, represses

the activation of Cbfa-1/Runx2257, which is a transcription factor associated with osteogenic

differentiation. Expression of Cbfa-1/Runx2 was found to be upregulated in mouse and rabbit

models of valvular calcification258, 259 and in cells cultured on compliant collagen matrices

(described in Chapter Five). Negative regulation of Notch signaling by NFKBIA might possibly

promote Runx2 activation. Similarly, BMP2260 and MKP-1261 signaling for osteogenesis require

Cbfa-1/Runx2 activity (see Discussion).

Coupled with the relative high expression of osteoinductive transcripts, the expression of

transforming growth factor beta 3 (TGFB3) was downregulated in cells grown on compliant

matrices. TGFB3 has been shown to inhibit osteogenic differentiation of mesenchymal stem

cells262. The upregulation of osteoinducive transcripts and downregulation of osteorepressive

transcript on cells cultured on compliant matrices presumably contributes to the pro-osteogenic

nature of these matrices. In addition, the expression of transcripts related to cell adhesion and

actin-myosin cytoskeleton system were downregulated in VICs cultured on the more compliant

matrices.

100

Table 7.1. A subset of transcripts with higher expression in VICs cultured on compliant

matrices relative to those cultured on stiff matrices

Gene name* C/S fold change‡

P-value Biological process

CSF2 59.48 5.39 x 10-4 Immune Response IL8 41.88 2.19 x 10-6 Inflammatory response CCL20 27.58 1.21 x 10-3 Immune Response BTG2 15.24 1.91 x 10-5 Negative regulation of apoptosis SELE 9.67 2.2 x 10-3 Cell adhesion SPY2 7.27 2.27 x 10-4 Negative regulation of MAP kinase

activity NFKBIA 4.32 1.14 x 10-3 Negative regulation of Notch signaling

pathway MKP-1 3.45 1.09 x 10-3 Protein amino acid dephosphorylation BMP2 3.25 3.00 x 10-3 Growth C-JUN 2.95 1.05 x 10-4 Regulation of transcription SLN 2.79 4.22 x 10-3 Regulation of calcium ion transport PIAP 2.58 3.76 x 10-3 Regulation of apoptosis

* CSF2, colony stimulating factor 2; IL8, interleukin 8; CCL20, chemokine (C-C motif) ligand

20; BTG2, B-cell translocation gene 2; SELE, selectin; SPY2, sprouty homolog 2; NFKBIA,

nuclear factor of kappa light polypeptide gene enhancer in B-cells inhibitor alpha; MKP-1,

mitogen-activated protein kinase phosphatase-1; BMP2, bone morphogenetic proteins; C-JUN,

C-JUN protein; SLN, sarcolipin; PIAP, inhibitor of apoptosis-like. Sequence mapped to genome of other species (e.g. Homo sapiens, Bos taurus, Mus musculus) ‡ Fold changes of genes expressed by cells cultured on compliant matrices relative to those

cultured on stiff matrices

7.2.3. Differential Gene Expression by CNP

The effect of CNP on altering transcript expression in VICs cultured on compliant and stiff

collagen matrices was evaluated. For VICs cultured on compliant matrices, CNP treatment

influenced the expression of 181 sequences, with higher expression of 73 sequences and lower

expression of 108 sequences. When VICs were cultured on stiff collagen matrices, CNP

treatment altered the expression of 237 transcripts, with higher expression of 139 genes and

lower expression of 98 genes in CNP-treated samples relative to untreated cells. The majority of

the transcripts displayed expression fold differences of less than five (Figures 7.4 and 7.5).

101

CNP treatment affects a wide variety of genes that are associated with various biological

processes. When cells were cultured on compliant matrices, CNP upregulated transcripts related

to metabolic processes such as gluconeogenesis, lipid catabolic process, collagen catabolic

process and ATP biosynthetic process (Table 7.3). CNP downregulated the expression of

transcripts related to ion binding and transport (Table7.4). Similarly when cells were cultured on

stiff matrices, CNP treatment also upregulated several transcripts related to metabolic processes

including glycogen metabolic process and lipid catabolic process (Table 7.5). CNP

downregulated transcripts associated with the actin-myosin cytoskeleton system when cells were

cultured on stiff matrices (Table 7.6).

Table 7.2. A subset of transcripts with lower expression in VICs cultured on compliant

matrices relative to those cultured on stiff matrices

Gene name* C/S fold change‡ P-value Biological classification GP38K 11.75 3.31 x 10-5 Carbohydrate metabolic process TNFSF10 6.58 5.85 x 10-3 Immune response ITIH4 5.10 1.67 x 10-3 Acute phase response CDH5 4.77 2.39 x 10-3 Cell adhesion WNT2B 3.46 2.68 x 10-4 Wnt receptor signaling pathway,

calcium modulating pathway CNN1 3.27 9.89 x 10-5 Actomyosin structure organization TIMP1 2.76 3.12 x 10-3 Erythrocyte maturation TNNC2 2.59 3.06 x 10-3 Calcium ion binding COL5A1 2.52 9.43 x 10-4 Cell adhesion TGFB3 2.37 5.41 x 10-5 Growth, positive regulation of cell

division DDC 2.37 6.07 x 10-3 Metabolic process, catecholamine

biosynthetic process TPM1 2.08 8.61 x 10-4 Actin binding

* GP38K, 38 kDa heparin-binding glycoprotein; TNFSF10, tumour necrosis factor (ligand)

superfamily member 10; ITIH4, inter-alpha (globulin) inhibitor H4 (plasma kallikrein-sensitive

glycoprotein); CHD5, cadherin 5; WNT2B, wingless-type MMTV integration site family,

member 2B;CNN1, calponin 1 basic smooth muscle; TIMP1, TIMP metallopeptidase inhibitor 1;

TNNC2, troponin C type 2; COL5A1, collage type V alpha 1; TGFB3, transforming growth

factor beta 3; DDC, dopa decarboxylase; TPM1, tropomyosin 1. Sequence mapped to genome of other species (e.g. Homo sapiens, Bos taurus, Mus musculus) ‡ Fold changes of genes expressed by cells cultured on compliant matrices relative to those

cultured on stiff matrices

102

0

10

20

30

40

50

60

70

2.5 5 10 20 30 40+Fold difference

Nu

mb

er

of e

ntr

ies

UpregulatedDownregulated

Figure 7.4. The distribution of sequences differentially expressed with CNP treatment in

cultures on compliant matrices

Figure 7.5. The distribution of sequences differentially expressed with CNP treatment in

cultures on stiff matrices

0102030405060708090

2.5 5 10 20 30 40+

Fold difference

Nu

mb

er

of e

ntr

ies Upregulated

Downregulated

103


untreated cells when cultured on compliant matrices.

Gene name T/U fold change‡ P value Biological classification PAH 11.47 2.06 x 10-3 Aromatic amino acid family metabolic

process, L-phenylalanine catabolic process, oxidation reduction

HP 6.98 3.13 x 10-4 Proteolysis MMP1 4.68 7.41 x 10-3 Proteolysis, metabolic process, collagen

catabolic process CCRL1 3.89 6.93 x 10-5 G-protein coupled receptor protein

signaling pathway AMCF-II 2.98 9.21 x 10-3 Immune response, chemotaxis,

inflammatory response HSD11B1 2.62 7.17 x 10-3 Metabolic process ANGPT1 2.62 1.82 x 10-2 Signal transduction, angiogenesis, cell

differentiation C1S 2.22 1.71 x 10-2 Proteolysis, innate immune response,

complement activation classical pathway

VDR 2.22 2.78 x 10-2 Regulation of transcription PTGFR 2.09 4.65 x 10-2 G-protein coupled receptor protein

signaling pathway LPL 2.04 4.40 x 10-2 Lipid catabolic process, lipid metabolic

process ATP9A 2.03 2.97 x 10-2 ATP biosynthetic process, metabolic

process, phospholipid transport PC 2.03 1.10 x 10-2 Gluconeogenesis, metabolic process

*PAH, phenylalanine hydroxylase; HP, haptoglobin; MMP1, matrix metalloproteinase-1;

CCRL1, chemokine receptor-like 1; AMCF-II, alveolar macrophaste-derived chemotactic factor-

II; HSD11B1, 11-beta hydroxysteroid dehydrogenase isoform 1; ANGPT1, angiopoietin 1;C1S,

complement component 1;VDR, vitamin D (1,25-dihydroxyvitamine D3) receptor; PTGFR,

prostaglandin F receptor; LDL, lipoprotein lipase; ATP9A, ATPase class II type 9A; PC,

pyruvate carboxylase. Sequence mapped to genome of other species (e.g. Homo sapiens, Bos taurus, Mus musculus) ‡Fold changes of genes expressed by CNP-treated cells relative to untreated cells

7.2.4. The Combined Effect of Matrix Stiffness and CNP on Transcriptional Regulation

Venn diagrams were constructed to identify the genes commonly and exclusively modulated by

matrix stiffness and by CNP. We first compared the list of differentially expressed sequences by

104

matrix stiffness alone and by CNP on cells cultured on compliant matrices. Matrix stiffness alone

affected five times more transcripts than those regulated by CNP treatment. Of the 998

transcripts modulated by matrix stiffness, the expression of only 18 transcripts was also regulated

by CNP when cells were cultured on compliant matrices (Figure 7.6), suggesting the majority of

the transcripts were exclusively regulated by either matrix stiffnss or by CNP.

A similar trend was observed when comparing the number of sequences affected by matrix

stiffness and by CNP treatment on VICs cultured on stiff matrices. The expression of only a

small portion (i.e., 31 entries) of the sequences was influenced by matrix stiffness as well as by

CNP, further confirming that the majority of the transcripts were exclusively regulated by either

mechanical and biochemical cues (Figure 7.7).


untreated cells when cultured on compliant matrices.

Gene name T/U fold change‡ P value Biological classification PBD-1 7.39 9.65 x 10-4 Defense response PALMD 2.57 2.16 x 10-3 Regulation of cell shape LIM 2.41 1.35 x 10-2 Metal ion binding, zinc ion

binding SRPK3 2.17 3.78 x 10-3 Protein amino acid

phosphorylation KCNN4 2.10 7.40 x 10-5 Potassium ion transport ALDH1A3 2.05 2.56 x 10-2 Positive regulation of apoptosis,

oxidation reduction, metabolic process

GPR183 2.05 1.36 x 10-2 G-protein coupled receptor protein signaling pathway, immune response

UPP1 2.04 1.52 x 10-3 Nucleotide catabolic process *PBD-1, prepro-beta-defensin 1; PALMD, palmdelphin ; LIM, alpha-actinin-2-associated LIM

protein; SRPK3, SFRS protein kinase 3; KCNN4, potassium intermediate/small conductance

calcium-activated channel subfamily N member 4; ALDH1A, aldehyde dehydrogenase family 1

subfamily A3;GPR183, G-protein coupled receptor 183 ; UPP1, uridine phosphorylase 1. Sequence mapped to genome of other species (e.g. Homo sapiens, Bos taurus, Mus musculus) ‡Fold changes of genes expressed by CNP-treated cells relative to untreated cells

105


untreated cells when cultured on stiff matrices

Gene name T/U fold

change‡

P value Biological classification

HP 5.05 7.72 x 10-3 Proteolysis PKHA1 3.34 2.61 x 10-2 Glycogen metabolic process CD36 2.83 9.67 x 10-3 Cell adhesion ANGPTL4 2.25 2.75 x 10-3 Cell differentiation, angiogenesis, signal

transduction CXCL12 2.22 1.55 x 10-2 Immune response ALB 2.17 3.62 x 10-2 Cellular response to starvation,

maintenance of mitochondrion location, negative regulation of apoptosis, transport

PLA2G7 2.16 2.38 x 10-2 Lipid catabolic process FABP4 2.10 9.28 x 10-2 Transport (lipid binding) IGF1 2.06 6.79 x 10-4 Positive regulation of DNA replication ANGPT1 2.04 2.24 x 10-2 Signal transduction, angiogenesis, cell

differentiation SERPINA6 2.03 3.01 x 10-2 Transport NEO 2.02 2.56 x 10-2 Cell adhesion, regulation of

transcription, myoblast fusion RAMP1 2.00 1.79 x 10-2 Regulation of G-protein coupled

receptor protein signaling pathway, intracellular protein transport, transport

*HP, haptoglobin; PHKA1, phosphorylase kinase alpha 1; CD36, thrombospondin receptor;

ANGPTL4, angiopoietin-like 4; CXCL12, chemokine lingand 12 (stromal cell-derived factor 1) ;

ALB, albumin ; PLA2G7, phospholipase A2 group VII; FABP4, fatty acid binding protein 4;

IGF1, insulin-like growth factor 1; ANGPT1, angiopoietin 1; SERPINA6, serpin peptidase

inhibitor, clade A (alpha-1 antiproteinase, antitrypsin); NEO, neogenin; RAMP1, receptor (G-

protein-coupled) activity modifying protein 1. Sequence mapped to genome of other species (e.g. Homo sapiens, Bos taurus, Mus musculus) ‡Fold changes of genes expressed by CNP-treated cells relative to untreated cells

106


untreated VICs cells when cultured on stiff matrices

Gene name T/U fold change‡ P value Biological classification ACTA1 6.69 9.05 x 10-3 Skeletal muscle fiber development,

muscle thin filament assembly CNN1 2.83 5.89 x 10-3 Actomyosin structure organization NPY1R 2.44 3.61 x 10-2 Signal transduction, G-protein coupled

receptor protein signaling pathway GNAO1 2.30 2.00 x 10-3 Locomotory behaviour, regulation of

heart contraction, dopamine receptor signaling pathway, cellular process, G-protein coupled receptor signaling pathway, muscle contraction

APBB1IP 2.23 4.80 x 10-2 Signal transduction NTF3 2.22 3.17 x 10-2 Growth factor activity, neurotrophin

receptor binding *ACTA1, actin alpha skeletal muscle; CNN1, calponin 1 basic smooth muscle; NPY1R ,

neuropeptide Y receptor Y1; GNAO1, guanine nucleotide binding protein (G protein), alpha

activating activity polypeptide O; NTF3, neurotrophin 3; APBB1IP, Amyloid beat (A4) protein-

binding family B member 1 interacting protein. Sequence mapped to genome of other species (e.g. Homo sapiens, Bos taurus, Mus musculus) ‡Fold changes of genes expressed by CNP-treated cells relative to untreated cells

107

Figure 7.6. Transcript expression modulated by matrix stiffness and/or by CNP in cultures

on compliant matrices

Blue region represents transcript expression altered by matrix stiffness. Yellow region represents

transcripts that were changed in VICs cultured on compliant matrices with CNP treatment

relative to the untreated cultures. The union region represents genes that were non-exclusively

regulated by matrix stiffness.

When cells were cultured on compliant or stiff matrices, CNP treatment influenced

approximately 150-200 genes (Figure 7.8). The expression of only 30 CNP-regulated genes was

not affected by matrix stiffness (Figure 7.8, union region), indicating matrix mechanics

significantly modulated the response of VICs to CNP treatment. CNP-modulated, but non-

mechanically regulated genes were related to diverse biological processes including cell division,

replication of DNA, cellular amino acid biosynthetic process (Table 7.7).

980 genesregulated by matrix stiffness

163 genes regulated by CNP (cells cultured on compliant matrices)

18 genes



18 genes

108

Figure 7.7. Transcript expression modulated by matrix stiffness and/or by CNP in cultures

on stiff matrices

Blue region represents sequences regulated by matrix stiffness. Yellow region represents

transcripts that were changed in VICs cultured on stiff matrices with CNP treatment relative to

the untreated cultures. The union region represents genes that were non-exclusively regulated by

matrix stiffness.


208 genes regulated by CNP (cells cultured on stiff matrices)

31 genes967 genesregulated by matrix stiffness


31 genes

109

Table 7.7. A partial list of CNP-regulated, mechanically-insensitive genes

Gene Name Description Biological Processes

STAG2 Stromal antigen 2, transcript variant 4

Cell cycle, cell division, mitosis, meiosis, chromosome segregation

IGF1 Insulin-like growth factor 1 Positive regulation of DNA replication FST Follistatin Undefined† ABCC9 ATP-binding cassette, sub family

C, member 9, transcript variant SUR2B

Potassium ion transport

FUR1 Uracil phosphoribosyltransferase homolog

Transferase activity, transferring glycosyl groups, uracil phosphoribosyltransferase activity

HP Haptoglobin Proteolysis SAA1 Serum amyloid A1, transcript

variant 1 Positive regulation of interleukin-1 secretion, regulation of protein secretion, positive regulation of cell adhesion, acute-phase response, platelet activation, negative regulation of inflammatory response, chemotaxis, elevation of cytosolic calcium ion concentration

PAH Phenylalanine hydroxylase Oxidation reduction, L-phenylalanine catabolic process, metabolic process, aromatic amino acid family metabolic process, cellular amino acid biosynthetic process

SULT1A1 Sulfotransferase family, cytosolic, 1A, phenolpreferring, member 1

Undefined†

HRH1 Histamine receptor H1 Signal transduction, G-protein coupled receptor protein signaling pathway, positive regulation of nitro oxide biosynthetic process, synaptic transmission, inflammatory response

RAMP1 Receptor (G protein-coupled) activity modifying protein 1

Regulation of G-protein coupled recpetor protein signaling pathway, intracellular protein transport, transport

NFE2 Nuclear factor (erythroid-derived 2)

Regulation of transcription (DNA-dependent), Nucleosome disassembly, homeostasis, blood circulation, multicellular organismal development

ANGPT1 Angiopoietin 1 Signal transduction, angiogenesis, multicellular organismal development, cell differentiation

SLC25A25 Solute carrier family 25 (mitochondrial carrier; phosphate carrier), member 25, nuclear gene encoding mitochondrial protein

Transport

CCDC73 Coiled-coil domain containing 73 Undefined†

NCALD Neurocalcin delta, transcript variant 7

Vesicle-mediated transport

Sequence mapped to annotation of other species (e.g. Homo sapiens, Bos taurus, Mus musculus)

† Incomplete GO annotation lacking defined biological process(es). Molecular functions: FST, TGF-

signaling pathway; SULT1A1, transferase activity.

110

A three-way Venn diagram was constructed to further identify any genes that were regulated by

matrix stiffness and/or CNP. The regulation of only one transcript, follistatin, was altered by

matrix stiffness and CNP (Figure 7.8). Interestingly, in the absence of CNP, the expression of

follistatin transcript was 2.4-fold higher on compliant matrices relative to cells cultured on stiff

matrices, indicative of mechanically-regulated follistatin transcript expression in certain

biochemical environments. On the compliant matrices that favored osteogenic differentiation of

VICs, expression of follistatin was 2.0-fold higher in CNP-treated cells relative to the untreated

cells. On the stiff matrices that preferentially promoted myofibroblast differentiation, CNP-

treatment on VICs led to a 3.3-fold upregulation in the expression of follistatin. These data

suggest that the regulation of follistatin transcript expression by VICs in the presence of CNP

was independent of matrix stiffness, while the majority of CNP-regulated transcripts were

sensitive to modulation by matrix mechanics.

7.3. Discussion

Although matrix stiffness is recognized to play critical roles in regulating cell functions and

differentiation, its contribution in regulating cellular response to biochemical factors is poorly

understood. Further, the effect of matrix stiffness on CNP-mediated cellular response has never

been studied. By evaluating the transcriptional profile of VICs cultured on compliant and stiff

collagen matrices in the presence or absence of CNP, we identified a subset of mechanically- and

biochemically-regulated transcripts. The differential gene expression profile suggests that the

majority of CNP-regulated transcripts are sensitive to matrix stiffness. These data demonstrate

the significance of matrix stiffness in modulating the response of VICs to CNP, and the

combined effect of mechanical and biochemical cues in regulating VICs at the transcriptional

level, which ultimately may be important in understanding valve pathology and in determining

VIC response to potential therapeutics, as valve tissues mechanics changes during disease

progression.

111

Figure 7.8. A three-way Venn diagram showing the commonly and exclusively modulated

genes by matrix stiffness and by CNP

Blue region represents transcriptional regulation by matrix stiffness alone. Yellow region

represents CNP-regulated genes in VICs cultured on stiff matrices. Green region represents

CNP-regulated genes in VICs cultured on compliant matrices. The center area in which all three

regions overlap indicates a transcript (follistatin) that was regulated by matrix stiffness and CNP.

In the absence of CNP, higher expression of osteoinductive transcripts (e.g., BMP-2, NFKB1A,

MKP-1) was observed on compliant matrices, substrates that favored osteogenic differentiation

of VICs. The expression of BMP-2 has been observed in VICs differentiated to osteoblast-like

cells in vitro68 and in calcified valvular tissue48, 84. Addition of BMP-2 to VICs in culture

increased their expression of osteoblast-related markers and their rate of calcific aggregate

formation54, 73. BMP-2-induced osteoblastic differentiation has been shown to be mediated in a

MKP-1-dependent manner263. Similarly, we also found an upregulation of MKP-1 in the pro-

osteogenic, compliant matrices. Further, Notch signaling is involved in early stage of valve

formation as well as inhibition of the mediators of osteogenic-dependent valvular calcification257.



30


17

1

29



30


17

1

29



30


17

1

29

112

Presumably, the negative regulator of Notch signaling pathway, NFKB1A, may regulate

osteogenic differentiation indirectly by mitigating the osteorepressive effect of Notch signaling;

however the direct contribution of NFKB1A in CAVD will require further investigation.

When cells were cultured on the stiffer matrices that promoted myofibroblast differentiation,

higher expression of transcripts associated with actin-myosin cytoskeleton system including

calponin 1 (CNN1) and tropomyosin 1 (TPM-1) was observed. These findings are consistent

with a previous study by Chambers et al. in which expression profiling identified upregulation of

genes associated with contractile phenotype and cytoskeletal organization in myofibroblasts264

when compared to quiescent fibroblasts. The transition of fibroblasts to myofibroblasts is

closely related to maturation of focal adhesions265, which depends on the force applied to the

ECM-integrin-cytoskeleton connections either externally (e.g., ECM motion, substrate rigidity)

or internally (e.g., actin polymerization). A stiffer culture surface would presumably permit the

generation of greater traction forces, which facilitates focal adhesion maturation and enables the

transition of fibroblasts to myofibroblasts. The dependency of the transition of VICs into

myofibroblasts on matrix stiffness has previously been reported in the study by Pho et al.55 and

likely plays a role in the phenotypic drift of VICs in stiffened sclerotic valves.

Matrix stiffness alone affected the expression of ~ 1000 transcripts, whereas the expression of

only ~ 200 transcripts was affected by CNP. Notably, CNP upregulated various transcripts

associated with metabolic processes, including lipoprotein lipase (LPL) and phospholipase A2

(PLA2G7). LPL is the rate-limiting enzyme of triglyceride removal from plasma and has been

implicated in atherosclerosis. The expression of LPL transcript was downregulated in

atherosclerotic patients266, whereas statin treatment significantly increased LPL activity in

patients267. Our microarray data suggest a link between lipoprotein catabolic processes and CNP

signaling. Intriguingly, initial work found an upregulation of CNP transcript expression in statin-

treated VICs (Appendix B1); however whether there exists molecular relationships among

lipoprotein, statins and CNP awaits to be determined.

A striking observation from the two-way Venn diagram was the relatively small number of CNP-

dependent transcripts that were insensitive to matrix stiffness. The expression of 97% of all the

differentially expressed CNP-dependent transcripts was regulated by matrix stiffness. These data

emphasize the contribution of matrix mechanics in modulating cellular response to biochemical

113

factors. Also notable in the three-way Venn diagram analysis was the expression of only one

transcript, follistatin, which was non-exclusively regulated by matrix stiffness or CNP. One

possible explanation is that follistatin may regulate both osteoblast and myofibroblast

differentiation. Follistatin is a 34-kDa soluble protein that binds activin with high affinity to

inhibit the activation of TGF- signaling268, 269. In the absence of follistatin, activins bind to the

activin type IIA and type IIB receptors, leading to the recruitment and phosphorylation of type I

receptor, and subsequently the phosphorylation of Smad2/3270. Activation of Smad2/3 signaling

has been shown to increase -SMA expression and myofibroblast differentiation271. In addition

to activin signalling, follistatin has also been demonstrated to form a trimeric complex with BMP

and receptors of BMP to inhibit osteogenic effect mediated by BMP activity272. VICs

preferentially differentiated to pathological osteoblasts when cultured on compliant matrices and

CNP treatment would presumably prevent the osteogenic differentiation of these VICs. The

upregulation of follistatin in CNP-treated cells cultured on compliant matrices may facilitate the

inhibition of osteogenic differentiation by mitigating BMP activity272. CNP repressed

myofibroblast differentiation of VICs on stiff matrices. The higher expression of follistatin on

CNP-treated cells cultured on stiff matrices would likely mitigate TGF- signaling and Smad2/3

phosphorylation, which are required for the activation of fibroblasts to myofibroblasts.

Clearly, matrix stiffness has been shown to modulate various cell behaviours in vitro. Here, we

demonstrated the significance of matrix stiffness not only in modulating VIC phenotype, but also

their response to CNP at the transcriptional level. We identified a subset of mechanically- and

biochemically-regulated transcripts. The differential gene expression profile suggests that the

majority of CNP-regulated transcripts are sensitive to matrix stiffness. These data demonstrate

that matrix stiffness significantly affects the response of VICs to biochemical cues, and the

combined effect of mechanical and biochemical cues may ultimately govern the functions and

phenotypes of VICs. These findings may impact the response of cells to therapeutics in diseases

with substantial tissue matrix remodeling, where changes in tissue mechanics may define cellular

response to soluble factors. For example, significant stiffening of tumour tissue has been

correlated to an increase in proteolysis182, which has been suggested to partially explain the

failure of protease inhibitors as cancer therapies. The aortic valve undergoes significant

pathological matrix remodeling, which may alter local matrix mechanics. It is possible that local

matrix mechanics regulates valve homeostasis and functions of VICs in vivo, which may affect

114

their response to biochemical factors and the effectiveness of therapeutics against CAVD in a

stage-dependent manner. While our findings are suggestive of a correlative link between matrix

mechanics, VIC phenotypes and transcriptional regulation, further investigations are required to

demonstrate causality of this link.

115

Chapter 8 8. Conclusions and Recommendations

8.1. Conclusions

Pathological differentiation of VICs alters cell functions and is closely associated with valve

calcification. It is well accepted that biochemical factors such as TGF-1 induce the pathological

differentiation of VICs69, but little is known regarding factors that can inhibit the differentiation

of VICs into undesirable phenotypes. Further, the role of mechanical stimuli in regulating VIC

phenotype and functions has been overlooked and has yet to be investigated, despite the

observed influence of matrix stiffness on the differentiation of and mineralization by other non-

valve cell types132, 185. Alterations in tissue stiffness have been reported to occur prior to

substantial cellular and histological changes in diseases such as liver fibrosis7 and

atherosclerosis16, suggesting correlative links between tissue mechanics, disease progression and

the regulation of cell response. The aortic valve undergoes significant pathological matrix

remodeling and stiffens when calcified; it is likely that matrix stiffness may modulate VIC

behaviour and response to biochemical factors. Hence, the overall goal of this thesis was to

investigate the effect of matrix stiffness on modulating the response of VICs to pro- and anti-

calcific biochemical factors, which would provide further insights in valve pathology.

The first objective was to implement and characterize a culture system with tunable stiffness.

The morphological and molecular changes in VICs cultured on compliant and stiff matrices were

evaluated. In addition, the ability of CNP, a putative anti-sclerotic and anti-calcific agent, to

suppress pathological differentiation of VICs was tested in vitro. Lastly, the combined effect of

matrix stiffness and CNP on the transcriptional regulation of VICs was investigated.

VICs were found to be highly responsive to matrix stiffness. In conjunction with pro-calcific

biochemical factors, VICs preferentially underwent osteogenic differentiation and calcified when

cultured on the more compliant matrices. In contrast, the stiffer matrix favored myofibroblast

differentiation of VICs, contributing to contraction-mediated calcification that downregulated

Akt activity and was associated with apoptosis. Similarly, microarray study of cells cultured on

compliant and stiff matrices with pro-calcific biochemical factors identified upregulation of

116

osteoinductive transcripts and downregulation of osteorepressive transcripts on the more

compliant matrices relative to those cultured on the stiffer matrices. The ability to distinguish

two calcification processes by simply changing the matrix stiffness provides a useful research

tool to dissect the fundamental mechanisms of cell-mediated calcification.

The protective effect of CNP on VICs was also evident in the cell culture study. CNP inhibited

myofibroblast and osteoblast differentiation of VICs and suppressed in vitro calcification by

VICs. A striking finding was the small number of transcripts that were commonly regulated by

CNP and by matrix stiffness. The microarray results clearly demonstrate that the combined

effects of mechanical and biochemical cues govern transcriptional regulation of VIC, which

further emphasizes the necessity to consider both biochemical and mechanical factors in valve

studies in order to improve our fundamental understanding of VIC biology and valve pathology.

This thesis work contributes to the field of mechanobiology and valve biology. It provides an

improved understanding of VIC-matrix interactions, which is required to aid in interpretation of

VIC calcification studies in vitro; to guide the selection of biomaterials with appropriate

mechanical properties for valve tissue engineering; and to assess if alterations in extracellular

matrix mechanics that occur with disease modulate pathologic changes in VIC phenotypes and

calcification processes. In addition, the current study identifies for the first time the ability of

VICs to respond to CNP and provides a cellular explanation responsible for the protective effect

of CNP against calcification. These fundamental findings are essential for future mechanistic

studies of CNP at the molecular level and may perhaps eventually lead to the development of a

new treatment option.

8.2. Future Work

Given the current results, there are a number of suggested directions for future investigations, as

detailed in the following sections.

8.2.1. Determination of Changes in Valve Matrix Mechanics in vivo

The significance of matrix mechanics in modulating phenotype and transcriptional regulation of

VICs is evident in this thesis. To bridge the gap between our in vitro findings and the in vivo

relevancy, it will be important to identify the association between changes in valve matrix

117

mechanics and the progression of CAVD. Changes in mechanical properties of valve tissue

during CAVD can be determined using a micropipette aspiration technique similar to that

described by Matsumoto et al.16. AVs at different stages of disease development can be obtained

from porcine animal model fed with an atherogenic diet for various durations. The local elastic

moduli of normal AVs (i.e., animal fed a normal diet) and those of early-, intermediate- and late-

disease stage AVs can be measured by the micropipette aspiration method.

Immunohistochemical staining can be performed to characterize the pathological differentiation

of VICs temporally with respect to disease progression. Changes in the mechanical properties of

normal and diseased valve tissues can then be correlated with the extent of pathological

differentiation of VICs.

8.2.2. Improvement of the Cell Culture System

Although the collagen-based cell culture system was functional for all tests conducted, there

exist a number of limitations. Some of the limitations are: 1) the duration of cell culture was

limited to prevent substantial collagen degradation; 2) there exists a difference in the total

amount of collagen available on the two matrices, which may affect the ability of cells to spread

at later time points; and 3) fine tuning of stiffness is not possible with the existing system.

Presumably, changes in valve matrix mechanics involve a wide range of stiffness. To study the

effect of a range of physiologically relevant stiffness identified from Section 8.2.1, a culture

system that can be fine-tuned to provide a wide range of stiffness while maintaining surface

chemistry and to provide substrates with the same total collagen available over the given culture

conditions is needed. Polyacrylamide (PA) substrates are promising candidate materials as they

can provide a wide range of stiffnesses, while maintaining similar surface chemistry (reviewed in 10). However, initial efforts in our lab with standard surface modification methods failed to

provide appropriate surface adhesiveness to VICs. Recently, we successfully modified the

surface modification procedure and improved the adhesiveness to primary VICs. Our

preliminary study with primary VICs cultured on collagen-coated PA substrates with stiffnesses

of 11 kPa, 22 kPa, 50 kPa and 144 kPa showed that calcification by VICs was more prominent

on substrates with stiffnesses of 22 kPa and 50 kPa in comparison to those cultured on substrate

with stiffnesses of 11 kPa and 144 kPa (Figure B.3. in Appendix B). This preliminary result

suggests the possibility of culturing VICs on PA substrates, which can be tuned to the stiffness of

valve tissues measured at various disease stages. Such an in vitro study will provide a means to

118

identify the molecular determinants for mechanically regulated phenotypic drift of and

calcification by VICs on materials that closely resemble the mechanical properties of native

normal and diseased valves.

8.2.3. Effect of CNP treatment at Different Stages of Disease Progression

This thesis identified a cellular basis responsible for the protective effects of CNP against

CAVD. The next logical step is to identify the molecular mechanism responsible for the

inhibitory effects of CNP in the pathological differentiation of VICs. We hypothesized that CNP

mediated its cellular response via the NPR-B/cGMP signaling pathway. To address this

hypothesis, we have begun siRNA transfection experiments to manipulate the expression of

NPR-B receptor in VICs. Once those are completed, it will be important to evaluate the effect of

CNP: 1) in vitro by culturing cells on matrices with stiffnesses that represent various disease

stages; and 2) in vivo to evaluate the effectiveness of CNP treatment given at different stages of

CAVD. Because of the heterogeneity of VICs and the ability of these cells to differentiate into

various phenotypes over the disease progression, their response to CNP may vary depending on

the time at which treatment is administered. Studies have reported that the responsiveness of

cells to CNP depends on their commitment to certain lineages249. Presumably, the ability of VICs

to differentiate into various phenotypes is regulated in part by matrix stiffness and is altered as

the disease progresses. To test this, CNP can be applied to VICs cultured on matrices with

stiffnesses that represent early-, immediate- and late-stages of the disease. Such a culture system

can provide information regarding the ability of CNP to suppress pathological changes of VICs

when subjected to matrix stiffnesses that are physiologically relevant. Subsequently, these data

may help determine the preferred CNP treatment time point over the course of disease

development. Such data can serve as an initial guideline for the selection of treatment regimen

for in vivo tests. Mice models may be suitable for in vivo CNP studies, because genetic mutation

with mice can be done with ease. Initial work has verified the feasibility of dissecting mouse

aortic valves and isolating VICs from the valve leaflets (Figure B.4 and B.5 in Appendix B).

8.2.4. Identification of Transcriptional Pathways that Regulate Pathological

Differentiation of VICs

The current microarray study identified differential regulation of transcripts by matrix stiffness

and CNP. The study further confirmed the pro-osteogenic nature of our compliant matrices.

119

However, to fully benefit from the hypothesis generating power of microarray experiments and

to understand the transcriptional regulatory pathways that are involved in mechanical and

biochemical modulations, gene expression network analysis can be done to reveal important

phenomenological link between the expression of different genes. To do so, re-annotation of all

entries of the porcine microarray chips based on BLAST and cross-referencing of porcine

sequences to the human genome is necessary. Based on the re-annotation, canonical pathway

analysis can be performed to identify transcript networks with putative significance in CAVD.

120

References

1. WHO World Health Report 2003: World Health Organization. Geneva, Switzerland; 2003.

2. Stewart BF, Siscovick D, Lind BK, Gardin JM, Gottdiener JS, Smith VE, Kitzman DW, Otto CM. Clinical factors associated with calcific aortic valve disease. Cardiovascular Health Study. J Am Coll Cardiol. 1997;29:630-634.

3. Otto CM, Lind BK, Kitzman DW, Gersh BJ, Siscovick DS. Association of aortic-valve sclerosis with cardiovascular mortality and morbidity in the elderly. N Engl J Med. 1999;341:142-147.

4. Rajamannan NM, Sangiorgi G, Springett M, Arnold K, Mohacsi T, Spagnoli LG, Edwards WD, Tajik AJ, Schwartz RS. Experimental hypercholesterolemia induces apoptosis in the aortic valve. J Heart Valve Dis. 2001;10:371-374.

5. Tanaka K, Sata M, Fukuda D, Suematsu Y, Motomura N, Takamoto S, Hirata Y, Nagai R. Age-associated aortic stenosis in apolipoprotein E-deficient mice. J Am Coll Cardiol. 2005;46:134-141.

6. Simmons CA, Grant GR, Manduchi E, Davies PF. Spatial heterogeneity of endothelial phenotypes correlates with side-specific vulnerability to calcification in normal porcine aortic valves. Circ Res. 2005;96:792-799.

7. Georges PC, Hui JJ, Gombos Z, McCormick ME, Wang AY, Uemura M, Mick R, Janmey PA, Furth EE, Wells RG. Increased stiffness of the rat liver precedes matrix deposition: implications for fibrosis. Am J Physiol Gastrointest Liver Physiol. 2007;293:G1147-1154.

8. Shroff RC, Shah V, Hiorns MP, Schoppet M, Hofbauer LC, Hawa G, Schurgers LJ, Singhal A, Merryweather I, Brogan P, Shanahan C, Deanfield J, Rees L. The circulating calcification inhibitors, fetuin-A and osteoprotegerin, but not matrix Gla protein, are associated with vascular stiffness and calcification in children on dialysis. Nephrol Dial Transplant. 2008;23:3263-3271.

9. Suki B, Bates JH. Extracellular matrix mechanics in lung parenchymal diseases. Respir Physiol Neurobiol. 2008;163:33-43.

10. Yip CYY, Chen J-H, Simmons CA. Engineering substrate mechanics to regulate cell response. In: Khademhosseini A, Borenstein J, Toner M, Takayama S, eds. . Norwood, Mass: Artech Hourse; 2008.

11. Arora PD, Narani N, McCulloch CA. The compliance of collagen gels regulates transforming growth factor-beta induction of alpha-smooth muscle actin in fibroblasts. Am J Pathol. 1999;154:871-882.

121

12. Wells RG, Discher DE. Matrix elasticity, cytoskeletal tension, and TGF-beta: the insoluble and soluble meet. Sci Signal. 2008;1:pe13.

13. Hinton RB, Jr., Lincoln J, Deutsch GH, Osinska H, Manning PB, Benson DW, Yutzey KE. Extracellular matrix remodeling and organization in developing and diseased aortic valves. Circ Res. 2006;98:1431-1438.

14. Otto CM, Kuusisto J, Reichenbach DD, Gown AM, O'Brien KD. Characterization of the early lesion of 'degenerative' valvular aortic stenosis. Histological and immunohistochemical studies. Circulation. 1994;90:844-853.

15. Lee TC, Midura RJ, Hascall VC, Vesely I. The effect of elastin damage on the mechanics of the aortic valve. J Biomech. 2001;34:203-210.

16. Matsumoto T, Abe H, Ohashi T, Kato Y, Sato M. Local elastic modulus of atherosclerotic lesions of rabbit thoracic aortas measured by pipette aspiration method. Physiol Meas. 2002;23:635-648.

17. Loomba RS, Arora R. Statin Therapy and Aortic Stenosis: A Systematic Review of the Effects of Statin Therapy on Aortic Stenosis. Am J Ther. 2009.

18. Olsson AG. Lipid lowering and aortic valve disease. Curr Atheroscler Rep. 2009;11:377-383.

19. Peltonen TO, Taskinen P, Soini Y, Rysa J, Ronkainen J, Ohtonen P, Satta J, Juvonen T, Ruskoaho H, Leskinen H. Distinct downregulation of C-type natriuretic peptide system in human aortic valve stenosis. Circulation. 2007;116:1283-1289.

20. Thubrikar MJ. The aortic valve. Boca Raton, Florida: CRC Press, Inc.; 1990.

21. Stella JA, Sacks MS. On the biaxial mechanical properties of the layers of the aortic valve leaflet. J Biomech Eng. 2007;129:757-766.

22. Sacks MS, Smith DB, Hiester ED. The aortic valve microstructure: effects of transvalvular pressure. J Biomed Mater Res. 1998;41:131-141.

23. Vesely I, Noseworthy R. Micromechanics of the fibrosa and the ventricularis in aortic valve leaflets. J Biomech. 1992;25:101-113.

24. Schoen FJ. Aortic valve structure-function correlations: role of elastic fibers no longer a stretch of the imagination. J Heart Valve Dis. 1997;6:1-6.

25. Vesely I. The role of elastin in aortic valve mechanics. J Biomech. 1998;31:115-123.

26. Rodriguez KJ, Masters KS. Regulation of valvular interstitial cell calcification by components of the extracellular matrix. J Biomed Mater Res A. 2009;90:1043-1053.

27. Statistics Canada. Causes of Death 2003. http://www.statcan.ca/bsolc/english/bsolc?catno=84-208-X.

122

28. Economic burden of illness in Canada,1998.: Health Canada; 2002.

29. International Cardiovascular Disease Statistics.: American Heart Association; 2004.

30. Heart and Stroke Foundation of Canada Web site. www.heartandstroke.ca.

31. Freeman RV, Otto CM. Spectrum of calcific aortic valve disease: pathogenesis, disease progression, and treatment strategies. Circulation. 2005;111:3316-3326.

32. Aronow WS, Ahn C, Shirani J, Kronzon I. Comparison of frequency of new coronary events in older subjects with and without valvular aortic sclerosis. Am J Cardiol. 1999;83:599-600, A598.

33. World Population Ageing 1950-2050. : Population Division, DESA, United Nations.; 2007.

34. Barasch E, Gottdiener JS, Marino Larsen EK, Chaves PH, Newman AB. Cardiovascular morbidity and mortality in community-dwelling elderly individuals with calcification of the fibrous skeleton of the base of the heart and aortosclerosis (The Cardiovascular Health Study). Am J Cardiol. 2006;97:1281-1286.

35. Rossi A, Bertagnolli G, Cicoira M, Golia G, Zanolla L, Santini F, Cemin C, Ferrario G, Zardini P. Association of aortic valve sclerosis and coronary artery disease in patients with severe nonischemic mitral regurgitation. Clin Cardiol. 2003;26:579-582.

36. Otto CM, O'Brien KD. Why is there discordance between calcific aortic stenosis and coronary artery disease? Heart. 2001;85:601-602.

37. Demer LL. Cholesterol in vascular and valvular calcification. Circulation. 2001;104:1881-1883.

38. Cury RC, Ferencik M, Hoffmann U, Ferullo A, Moselewski F, Abbara S, Booth SL, O'Donnell CJ, Brady TJ, Achenbach S. Epidemiology and association of vascular and valvular calcium quantified by multidetector computed tomography in elderly asymptomatic subjects. Am J Cardiol. 2004;94:348-351.

39. Massy ZA, Maziere C, Kamel S, Brazier M, Choukroun G, Tribouilloy C, Slama M, Andrejak M, Maziere JC. Impact of inflammation and oxidative stress on vascular calcifications in chronic kidney disease. Pediatr Nephrol. 2005;20:380-382.

40. Pohle K, Maffert R, Ropers D, Moshage W, Stilianakis N, Daniel WG, Achenbach S. Progression of aortic valve calcification: association with coronary atherosclerosis and cardiovascular risk factors. Circulation. 2001;104:1927-1932.

41. Ladislas R. Matrix remodeling and atherosclerosis effect of age Fibrogenesis: Cellular and Molecular Basis: Springer US; 2005.

42. Fondard O, Detaint D, Iung B, Choqueux C, Adle-Biassette H, Jarraya M, Hvass U, Couetil JP, Henin D, Michel JB, Vahanian A, Jacob MP. Extracellular matrix

123

remodelling in human aortic valve disease: the role of matrix metalloproteinases and their tissue inhibitors. Eur Heart J. 2005;26:1333-1341.

43. O'Brien KD, Reichenbach DD, Marcovina SM, Kuusisto J, Alpers CE, Otto CM. Apolipoproteins B, (a), and E accumulate in the morphologically early lesion of 'degenerative' valvular aortic stenosis. Arterioscler Thromb Vasc Biol. 1996;16:523-532.

44. Mehrabian M, Demer LL, Lusis AJ. Differential accumulation of intimal monocyte-macrophages relative to lipoproteins and lipofuscin corresponds to hemodynamic forces on cardiac valves in mice. Arterioscler Thromb. 1991;11:947-957.

45. Thubrikar MJ, Aouad J, Nolan SP. Patterns of calcific deposits in operatively excised stenotic or purely regurgitant aortic valves and their relation to mechanical stress. Am J Cardiol. 1986;58:304-308.

46. Thubrikar MJ, Nolan SP, Aouad J, Deck JD. Stress sharing between the sinus and leaflets of canine aortic valve. Ann Thorac Surg. 1986;42:434-440.

47. Rabkin-Aikawa E, Farber M, Aikawa M, Schoen FJ. Dynamic and reversible changes of interstitial cell phenotype during remodeling of cardiac valves. J Heart Valve Dis. 2004;13:841-847.

48. Mohler ER, 3rd, Gannon F, Reynolds C, Zimmerman R, Keane MG, Kaplan FS. Bone formation and inflammation in cardiac valves. Circulation. 2001;103:1522-1528.

49. Rajamannan NM, Subramaniam M, Rickard D, Stock SR, Donovan J, Springett M, Orszulak T, Fullerton DA, Tajik AJ, Bonow RO, Spelsberg T. Human aortic valve calcification is associated with an osteoblast phenotype. Circulation. 2003;107:2181-2184.

50. Poggianti E, Venneri L, Chubuchny V, Jambrik Z, Baroncini LA, Picano E. Aortic valve sclerosis is associated with systemic endothelial dysfunction. J Am Coll Cardiol. 2003;41:136-141.

51. Young EW, Wheeler AR, Simmons CA. Matrix-dependent adhesion of vascular and valvular endothelial cells in microfluidic channels. Lab Chip. 2007;7:1759-1766.

52. Bairati A, DeBiasi S. Presence of a smooth muscle system in aortic valve leaflets. Anat Embryol (Berl). 1981;161:329-340.

53. Merryman WD, Huang HY, Schoen FJ, Sacks MS. The effects of cellular contraction on aortic valve leaflet flexural stiffness. J Biomech. 2006;39:88-96.

54. Mohler ER, 3rd, Chawla MK, Chang AW, Vyavahare N, Levy RJ, Graham L, Gannon FH. Identification and characterization of calcifying valve cells from human and canine aortic valves. J Heart Valve Dis. 1999;8:254-260.

124

55. Pho M, Lee W, Watt DR, Laschinger C, Simmons CA, McCulloch CA. Cofilin is a marker of myofibroblast differentiation in cells from porcine aortic cardiac valves. Am J Physiol Heart Circ Physiol. 2008;294:H1767-1778.

56. Cimini M, Rogers KA, Boughner DR. Aortic valve interstitial cells: an evaluation of cell viability and cell phenotype over time. J Heart Valve Dis. 2002;11:881-887.

57. Messier RH, Jr., Bass BL, Aly HM, Jones JL, Domkowski PW, Wallace RB, Hopkins RA. Dual structural and functional phenotypes of the porcine aortic valve interstitial population: characteristics of the leaflet myofibroblast. J Surg Res. 1994;57:1-21.

58. Chen JH, Yip CY, Sone ED, Simmons CA. Identification and characterization of aortic valve mesenchymal progenitor cells with robust osteogenic calcification potential. Am J Pathol. 2009;174:1109-1119.

59. Cushing MC, Liao JT, Anseth KS. Activation of valvular interstitial cells is mediated by transforming growth factor-beta1 interactions with matrix molecules. Matrix Biol. 2005;24:428-437.

60. Komuro T. Re-evaluation of fibroblasts and fibroblast-like cells. Anat Embryol (Berl). 1990;182:103-112.

61. Taylor PM, Allen SP, Yacoub MH. Phenotypic and functional characterization of interstitial cells from human heart valves, pericardium and skin. J Heart Valve Dis. 2000;9:150-158.

62. Lester W, Rosenthal A, Granton B, Gotlieb AI. Porcine mitral valve interstitial cells in culture. Lab Invest. 1988;59:710-719.

63. Mulholland DL, Gotlieb AI. Cell biology of valvular interstitial cells. Can J Cardiol. 1996;12:231-236.

64. Cushing MC, Mariner PD, Liao JT, Sims EA, Anseth KS. Fibroblast growth factor represses Smad-mediated myofibroblast activation in aortic valvular interstitial cells. Faseb J. 2008;22:1769-1777.

65. Walker GA, Masters KS, Shah DN, Anseth KS, Leinwand LA. Valvular myofibroblast activation by transforming growth factor-beta: implications for pathological extracellular matrix remodeling in heart valve disease. Circ Res. 2004;95:253-260.

66. Visconti RP, Ebihara Y, LaRue AC, Fleming PA, McQuinn TC, Masuya M, Minamiguchi H, Markwald RR, Ogawa M, Drake CJ. An in vivo analysis of hematopoietic stem cell potential: hematopoietic origin of cardiac valve interstitial cells. Circ Res. 2006;98:690-696.

67. Latif N, Sarathchandra P, Taylor PM, Antoniw J, Yacoub MH. Molecules mediating cell-ECM and cell-cell communication in human heart valves. Cell Biochem Biophys. 2005;43:275-287.

125

68. Yang X, Fullerton DA, Su X, Ao L, Cleveland JC, Jr., Meng X. Pro-osteogenic phenotype of human aortic valve interstitial cells is associated with higher levels of Toll-like receptors 2 and 4 and enhanced expression of bone morphogenetic protein 2. J Am Coll Cardiol. 2009;53:491-500.

69. Jian B, Narula N, Li QY, Mohler ER, 3rd, Levy RJ. Progression of aortic valve stenosis: TGF-beta1 is present in calcified aortic valve cusps and promotes aortic valve interstitial cell calcification via apoptosis. Ann Thorac Surg. 2003;75:457-465; discussion 465-456.

70. Soini Y, Satta J, Maatta M, Autio-Harmainen H. Expression of MMP2, MMP9, MT1-MMP, TIMP1, and TIMP2 mRNA in valvular lesions of the heart. J Pathol. 2001;194:225-231.

71. Kaden JJ, Dempfle CE, Grobholz R, Tran HT, Kilic R, Sarikoc A, Brueckmann M, Vahl C, Hagl S, Haase KK, Borggrefe M. Interleukin-1 beta promotes matrix metalloproteinase expression and cell proliferation in calcific aortic valve stenosis. Atherosclerosis. 2003;170:205-211.

72. Kaden JJ, Dempfle CE, Grobholz R, Fischer CS, Vocke DC, Kilic R, Sarikoc A, Pinol R, Hagl S, Lang S, Brueckmann M, Borggrefe M. Inflammatory regulation of extracellular matrix remodeling in calcific aortic valve stenosis. Cardiovasc Pathol. 2005;14:80-87.

73. Osman L, Chester AH, Amrani M, Yacoub MH, Smolenski RT. A novel role of extracellular nucleotides in valve calcification: a potential target for atorvastatin. Circulation. 2006;114:I566-572.

74. Clark-Greuel JN, Connolly JM, Sorichillo E, Narula NR, Rapoport HS, Mohler ER, 3rd, Gorman JH, 3rd, Gorman RC, Levy RJ. Transforming growth factor-beta1 mechanisms in aortic valve calcification: increased alkaline phosphatase and related events. Ann Thorac Surg. 2007;83:946-953.

75. Canfield AE, Doherty MJ, Wood AC, Farrington C, Ashton B, Begum N, Harvey B, Poole A, Grant ME, Boot-Handford RP. Role of pericytes in vascular calcification: a review. Z Kardiol. 2000;89 Suppl 2:20-27.

76. Watson KE, Bostrom K, Ravindranath R, Lam T, Norton B, Demer LL. TGF-beta 1 and 25-hydroxycholesterol stimulate osteoblast-like vascular cells to calcify. J Clin Invest. 1994;93:2106-2113.

77. Proudfoot D, Skepper JN, Hegyi L, Bennett MR, Shanahan CM, Weissberg PL. Apoptosis regulates human vascular calcification in vitro: evidence for initiation of vascular calcification by apoptotic bodies. Circ Res. 2000;87:1055-1062.

78. Kim KM. Apoptosis and calcification. Scanning Microsc. 1995;9:1137-1175; discussion 1175-1138.

79. Yip CY, Chen JH, Zhao R, Simmons CA. Calcification by valve interstitial cells is regulated by the stiffness of the extracellular matrix. Arterioscler Thromb Vasc Biol. 2009;29:936-942.

126

80. Hunter GK, Hauschka PV, Poole AR, Rosenberg LC, Goldberg HA. Nucleation and inhibition of hydroxyapatite formation by mineralized tissue proteins. Biochem J. 1996;317 ( Pt 1):59-64.

81. Hauschka PV, Wians FH, Jr. Osteocalcin-hydroxyapatite interaction in the extracellular organic matrix of bone. Anat Rec. 1989;224:180-188.

82. Edep ME, Shirani J, Wolf P, Brown DL. Matrix metalloproteinase expression in nonrheumatic aortic stenosis. Cardiovasc Pathol. 2000;9:281-286.

83. Lee YS, Chou YY. Endothelial alterations and senile calcific aortic stenosis: an electron microscopic observation. Proc Natl Sci Counc Repub China B. 1997;21:137-143.

84. Kaden JJ, Bickelhaupt S, Grobholz R, Haase KK, Sarikoc A, Kilic R, Brueckmann M, Lang S, Zahn I, Vahl C, Hagl S, Dempfle CE, Borggrefe M. Receptor activator of nuclear factor kappaB ligand and osteoprotegerin regulate aortic valve calcification. J Mol Cell Cardiol. 2004;36:57-66.

85. Taipale J, Saharinen J, Hedman K, Keski-Oja J. Latent transforming growth factor-beta 1 and its binding protein are components of extracellular matrix microfibrils. J Histochem Cytochem. 1996;44:875-889.

86. Corcoran ML, Stetler-Stevenson WG. Tissue inhibitor of metalloproteinase-2 stimulates fibroblast proliferation via a cAMP-dependent mechanism. J Biol Chem. 1995;270:13453-13459.

87. Hayakawa T, Yamashita K, Ohuchi E, Shinagawa A. Cell growth-promoting activity of tissue inhibitor of metalloproteinases-2 (TIMP-2). J Cell Sci. 1994;107 ( Pt 9):2373-2379.

88. Jian B, Jones PL, Li Q, Mohler ER, 3rd, Schoen FJ, Levy RJ. Matrix metalloproteinase-2 is associated with tenascin-C in calcific aortic stenosis. Am J Pathol. 2001;159:321-327.

89. Jones PL, Jones FS. Tenascin-C in development and disease: gene regulation and cell function. Matrix Biol. 2000;19:581-596.

90. Satta J, Melkko J, Pollanen R, Tuukkanen J, Paakko P, Ohtonen P, Mennander A, Soini Y. Progression of human aortic valve stenosis is associated with tenascin-C expression. J Am Coll Cardiol. 2002;39:96-101.

91. Kaden JJ, Bickelhaupt S, Grobholz R, Vahl CF, Hagl S, Brueckmann M, Haase KK, Dempfle CE, Borggrefe M. Expression of bone sialoprotein and bone morphogenetic protein-2 in calcific aortic stenosis. J Heart Valve Dis. 2004;13:560-566.

92. Osman L, Yacoub MH, Latif N, Amrani M, Chester AH. Role of human valve interstitial cells in valve calcification and their response to atorvastatin. Circulation. 2006;114:I547-552.

93. Mallat Z, Tedgui A. Unbalanced RANKL/RANK pathway in aortic valve sclerosis. J Mol Cell Cardiol. 2004;36:17-19.

127

94. Alexopoulos A, Bravou V, Peroukides S, Kaklamanis L, Varakis J, Alexopoulos D, Papadaki H. Bone regulatory factors NFATc1 and Osterix in human calcific aortic valves. Int J Cardiol. 2008.

95. Ghazvini-Boroujerdi M, Clark J, Narula N, Palmatory E, Connolly JM, DeFelice S, Xu J, Jian B, Hazelwood S, Levy RJ. Transcription factor Egr-1 in calcific aortic valve disease. J Heart Valve Dis. 2004;13:894-903.

96. O'Brien KD, Shavelle DM, Caulfield MT, McDonald TO, Olin-Lewis K, Otto CM, Probstfield JL. Association of angiotensin-converting enzyme with low-density lipoprotein in aortic valvular lesions and in human plasma. Circulation. 2002;106:2224-2230.

97. Gerber IL, Legget ME, West TM, Richards AM, Stewart RA. Usefulness of serial measurement of N-terminal pro-brain natriuretic peptide plasma levels in asymptomatic patients with aortic stenosis to predict symptomatic deterioration. Am J Cardiol. 2005;95:898-901.

98. Hofbauer LC, Schoppet M. Clinical implications of the osteoprotegerin/RANKL/RANK system for bone and vascular diseases. Jama. 2004;292:490-495.

99. Tang Y, Urs S, Liaw L. Hairy-related transcription factors inhibit Notch-induced smooth muscle alpha-actin expression by interfering with Notch intracellular domain/CBF-1 complex interaction with the CBF-1-binding site. Circ Res. 2008;102:661-668.

100. Nigam V, Srivastava D. Notch1 Represses Osteogenic Pathways in Aortic Valve Cells. J Mol Cell Cardiol. 2009.

101. Nakao K, Ogawa Y, Suga S, Imura H. Molecular biology and biochemistry of the natriuretic peptide system. II: Natriuretic peptide receptors. J Hypertens. 1992;10:1111-1114.

102. Koller KJ, Lowe DG, Bennett GL, Minamino N, Kangawa K, Matsuo H, Goeddel DV. Selective activation of the B natriuretic peptide receptor by C-type natriuretic peptide (CNP). Science. 1991;252:120-123.

103. Potter LR, Abbey-Hosch S, Dickey DM. Natriuretic peptides, their receptors, and cyclic guanosine monophosphate-dependent signaling functions. Endocr Rev. 2006;27:47-72.

104. Moura LM, Rocha-Goncalves F, Zamorano JL, Barros I, Bettencourt P, Rajamannan N. New cardiovascular biomarkers: clinical implications in patients with valvular heart disease. Expert Rev Cardiovasc Ther. 2008;6:945-954.

105. Wu C, Wu F, Pan J, Morser J, Wu Q. Furin-mediated processing of Pro-C-type natriuretic peptide. J Biol Chem. 2003;278:25847-25852.

106. Wollert KC, Yurukova S, Kilic A, Begrow F, Fiedler B, Gambaryan S, Walter U, Lohmann SM, Kuhn M. Increased effects of C-type natriuretic peptide on contractility

128

and calcium regulation in murine hearts overexpressing cyclic GMP-dependent protein kinase I. Br J Pharmacol. 2003;140:1227-1236.

107. Tsuji T, Kunieda T. A loss-of-function mutation in natriuretic peptide receptor 2 (Npr2) gene is responsible for disproportionate dwarfism in cn/cn mouse. J Biol Chem. 2005;280:14288-14292.

108. Yasoda A, Komatsu Y, Chusho H, Miyazawa T, Ozasa A, Miura M, Kurihara T, Rogi T, Tanaka S, Suda M, Tamura N, Ogawa Y, Nakao K. Overexpression of CNP in chondrocytes rescues achondroplasia through a MAPK-dependent pathway. Nat Med. 2004;10:80-86.

109. Kennedy JA, Hua X, Mishra K, Murphy GA, Rosenkranz AC, Horowitz JD. Inhibition of calcifying nodule formation in cultured porcine aortic valve cells by nitric oxide donors. Eur J Pharmacol. 2009;602:28-35.

110. Matsuo H, Furuya M. C-type natriuretic protein inhibits intimal thickening after vascular injury. Ann N Y Acad Sci. 1997;811:45-47.

111. Murakami S, Nagaya N, Itoh T, Fujii T, Iwase T, Hamada K, Kimura H, Kangawa K. C-type natriuretic peptide attenuates bleomycin-induced pulmonary fibrosis in mice. Am J Physiol Lung Cell Mol Physiol. 2004;287:L1172-1177.

112. Soeki T, Kishimoto I, Okumura H, Tokudome T, Horio T, Mori K, Kangawa K. C-type natriuretic peptide, a novel antifibrotic and antihypertrophic agent, prevents cardiac remodeling after myocardial infarction. J Am Coll Cardiol. 2005;45:608-616.

113. Astor BC, Kaczmarek RG, Hefflin B, Daley WR. Mortality after aortic valve replacement: results from a nationally representative database. Ann Thorac Surg. 2000;70:1939-1945.

114. Sharony R, Grossi EA, Saunders PC, Schwartz CF, Ciuffo GB, Baumann FG, Delianides J, Applebaum RM, Ribakove GH, Culliford AT, Galloway AC, Colvin SB. Aortic valve replacement in patients with impaired ventricular function. Ann Thorac Surg. 2003;75:1808-1814.

115. Grube E, Schuler G, Buellesfeld L, Gerckens U, Linke A, Wenaweser P, Sauren B, Mohr FW, Walther T, Zickmann B, Iversen S, Felderhoff T, Cartier R, Bonan R. Percutaneous aortic valve replacement for severe aortic stenosis in high-risk patients using the second- and current third-generation self-expanding CoreValve prosthesis: device success and 30-day clinical outcome. J Am Coll Cardiol. 2007;50:69-76.

116. Kvidal P, Bergstrom R, Malm T, Stahle E. Long-term follow-up of morbidity and mortality after aortic valve replacement with a mechanical valve prosthesis. Eur Heart J. 2000;21:1099-1111.

117. Rosenhek R, Rader F, Loho N, Gabriel H, Heger M, Klaar U, Schemper M, Binder T, Maurer G, Baumgartner H. Statins but not angiotensin-converting enzyme inhibitors delay progression of aortic stenosis. Circulation. 2004;110:1291-1295.

129

118. O'Brien KD, Probstfield JL, Caulfield MT, Nasir K, Takasu J, Shavelle DM, Wu AH, Zhao XQ, Budoff MJ. Angiotensin-converting enzyme inhibitors and change in aortic valve calcium. Arch Intern Med. 2005;165:858-862.

119. Wu B, Elmariah S, Kaplan FS, Cheng G, Mohler ER, 3rd. Paradoxical effects of statins on aortic valve myofibroblasts and osteoblasts: implications for end-stage valvular heart disease. Arterioscler Thromb Vasc Biol. 2005;25:592-597.

120. Monzack EL, Gu X, Masters KS. Efficacy of simvastatin treatment of valvular interstitial cells varies with the extracellular environment. Arterioscler Thromb Vasc Biol. 2009;29:246-253.

121. Cowell SJ, Newby DE, Prescott RJ, Bloomfield P, Reid J, Northridge DB, Boon NA. A randomized trial of intensive lipid-lowering therapy in calcific aortic stenosis. N Engl J Med. 2005;352:2389-2397.

122. Tammen H, Zucht HD, Budde P. Oncopeptidomics--a commentary on opportunities and limitations. Cancer Lett. 2007;249:80-86.

123. Farmer JA. Intensive lipid lowering with simvastatin and ezetimibe in aortic stenosis (the SEAS trial). Curr Atheroscler Rep. 2009;11:82-83.

124. Helske S, Laine M, Kupari M, Lommi J, Turto H, Nurmi L, Tikkanen I, Werkkala K, Lindstedt KA, Kovanen PT. Increased expression of profibrotic neutral endopeptidase and bradykinin type 1 receptors in stenotic aortic valves. Eur Heart J. 2007;28:1894-1903.

125. Arishiro K, Hoshiga M, Negoro N, Jin D, Takai S, Miyazaki M, Ishihara T, Hanafusa T. Angiotensin receptor-1 blocker inhibits atherosclerotic changes and endothelial disruption of the aortic valve in hypercholesterolemic rabbits. J Am Coll Cardiol. 2007;49:1482-1489.

126. Peterson JT. The importance of estimating the therapeutic index in the development of matrix metalloproteinase inhibitors. Cardiovasc Res. 2006;69:677-687.

127. Butcher JT, Simmons CA, Warnock JN. Mechanobiology of the aortic heart valve. J Heart Valve Dis. 2008;17:62-73.

128. Smith KE, Metzler SA, Warnock JN. Cyclic strain inhibits acute pro-inflammatory gene expression in aortic valve interstitial cells. Biomech Model Mechanobiol. 2009.

129. Yeung T, Georges PC, Flanagan LA, Marg B, Ortiz M, Funaki M, Zahir N, Ming W, Weaver V, Janmey PA. Effects of substrate stiffness on cell morphology, cytoskeletal structure, and adhesion. Cell Motil Cytoskeleton. 2005;60:24-34.

130. Engler AJ, Sen S, Sweeney HL, Discher DE. Matrix elasticity directs stem cell lineage specification. Cell. 2006;126:677-689.

130

131. Khatiwala CB, Peyton SR, Metzke M, Putnam AJ. The regulation of osteogenesis by ECM rigidity in MC3T3-E1 cells requires MAPK activation. J Cell Physiol. 2007;211:661-672.

132. Khatiwala CB, Peyton SR, Putnam AJ. Intrinsic mechanical properties of the extracellular matrix affect the behavior of pre-osteoblastic MC3T3-E1 cells. Am J Physiol Cell Physiol. 2006;290:C1640-1650.

133. Ingber DE. Mechanobiology and diseases of mechanotransduction. Ann Med. 2003;35:564-577.

134. Rai US, Singh RK. Synthesis and mechanical characterization of polymer-matrix composites containing calcium carbonate/white cement filler. Materials Letters. 2004;58:235-240.

135. Bauccio M. ASM Engineered Materials Reference Book. ASM International; 1994.

136. Engler AJ, Griffin MA, Sen S, Bonnemann CG, Sweeney HL, Discher DE. Myotubes differentiate optimally on substrates with tissue-like stiffness: pathological implications for soft or stiff microenvironments. J Cell Biol. 2004;166:877-887.

137. Wells RG. The role of matrix stiffness in regulating cell behavior. Hepatology. 2008;47:1394-1400.

138. Paszek MJ, Zahir N, Johnson KR, Lakins JN, Rozenberg GI, Gefen A, Reinhart-King CA, Margulies SS, Dembo M, Boettiger D, Hammer DA, Weaver VM. Tensional homeostasis and the malignant phenotype. Cancer Cell. 2005;8:241-254.

139. Choquet D, Felsenfeld DP, Sheetz MP. Extracellular matrix rigidity causes strengthening of integrin-cytoskeleton linkages. Cell. 1997;88:39-48.

140. Nayal A, Webb DJ, Horwitz AF. Talin: an emerging focal point of adhesion dynamics. Curr Opin Cell Biol. 2004;16:94-98.

141. Legate KR, Montanez E, Kudlacek O, Fassler R. ILK, PINCH and parvin: the tIPP of integrin signalling. Nat Rev Mol Cell Biol. 2006;7:20-31.

142. Pankov R, Cukierman E, Katz BZ, Matsumoto K, Lin DC, Lin S, Hahn C, Yamada KM. Integrin dynamics and matrix assembly: tensin-dependent translocation of alpha(5)beta(1) integrins promotes early fibronectin fibrillogenesis. J Cell Biol. 2000;148:1075-1090.

143. Schlunck G, Han H, Wecker T, Kampik D, Meyer-ter-Vehn T, Grehn F. Substrate rigidity modulates cell matrix interactions and protein expression in human trabecular meshwork cells. Invest Ophthalmol Vis Sci. 2008;49:262-269.

144. Hinz B, Dugina V, Ballestrem C, Wehrle-Haller B, Chaponnier C. Alpha-smooth muscle actin is crucial for focal adhesion maturation in myofibroblasts. Mol Biol Cell. 2003;14:2508-2519.

131

145. Hinz B. Masters and servants of the force: the role of matrix adhesions in myofibroblast force perception and transmission. Eur J Cell Biol. 2006;85:175-181.

146. Chiquet M, Gelman L, Lutz R, Maier S. From mechanotransduction to extracellular matrix gene expression in fibroblasts. Biochim Biophys Acta. 2009;1793:911-920.

147. Chen CS. Mechanotransduction - a field pulling together? J Cell Sci. 2008;121:3285-3292.

148. Geiger B, Spatz JP, Bershadsky AD. Environmental sensing through focal adhesions. Nat Rev Mol Cell Biol. 2009;10:21-33.

149. Keselowsky BG, Collard DM, Garcia AJ. Surface chemistry modulates focal adhesion composition and signaling through changes in integrin binding. Biomaterials. 2004;25:5947-5954.

150. Mercier I, Lechaire JP, Desmouliere A, Gaill F, Aumailley M. Interactions of human skin fibroblasts with monomeric or fibrillar collagens induce different organization of the cytoskeleton. Exp Cell Res. 1996;225:245-256.

151. Arnold M, Cavalcanti-Adam EA, Glass R, Blummel J, Eck W, Kantlehner M, Kessler H, Spatz JP. Activation of integrin function by nanopatterned adhesive interfaces. Chemphyschem. 2004;5:383-388.

152. Chen CS, Mrksich M, Huang S, Whitesides GM, Ingber DE. Geometric control of cell life and death. Science. 1997;276:1425-1428.

153. Dalby MJ, Riehle MO, Johnstone H, Affrossman S, Curtis AS. In vitro reaction of endothelial cells to polymer demixed nanotopography. Biomaterials. 2002;23:2945-2954.

154. Vogel V, Sheetz M. Local force and geometry sensing regulate cell functions. Nat Rev Mol Cell Biol. 2006;7:265-275.

155. Engler A, Bacakova L, Newman C, Hategan A, Griffin M, Discher D. Substrate compliance versus ligand density in cell on gel responses. Biophys J. 2004;86:617-628.

156. Galbraith CG, Sheetz MP. Forces on adhesive contacts affect cell function. Curr Opin Cell Biol. 1998;10:566-571.

157. Pelham RJ, Jr., Wang Y. Cell locomotion and focal adhesions are regulated by substrate flexibility. Proc Natl Acad Sci U S A. 1997;94:13661-13665.

158. Zhong C, Chrzanowska-Wodnicka M, Brown J, Shaub A, Belkin AM, Burridge K. Rho-mediated contractility exposes a cryptic site in fibronectin and induces fibronectin matrix assembly. J Cell Biol. 1998;141:539-551.

159. Wong AH, Amabile PG, Yuksel E, Waugh JM, Dake MD. Perivascular release of insulin-like growth factor-1 limits neointima formation in the balloon-injured artery by redirecting smooth muscle cell migration. J Vasc Interv Radiol. 2001;12:347-350.

132

160. Galbraith CG, Yamada KM, Sheetz MP. The relationship between force and focal complex development. J Cell Biol. 2002;159:695-705.

161. Peyton SR, Putnam AJ. Extracellular matrix rigidity governs smooth muscle cell motility in a biphasic fashion. J Cell Physiol. 2005;204:198-209.

162. Wang N, Ingber DE. Control of cytoskeletal mechanics by extracellular matrix, cell shape, and mechanical tension. Biophys J. 1994;66:2181-2189.

163. Guo WH, Wang YL. Retrograde fluxes of focal adhesion proteins in response to cell migration and mechanical signals. Mol Biol Cell. 2007;18:4519-4527.

164. Khatiwala CB, Kim PD, Peyton SR, Putnam AJ. ECM compliance regulates osteogenesis by influencing MAPK signaling downstream of RhoA and ROCK. J Bone Miner Res. 2009;24:886-898.

165. Fringer J, Grinnell F. Fibroblast quiescence in floating or released collagen matrices: contribution of the ERK signaling pathway and actin cytoskeletal organization. J Biol Chem. 2001;276:31047-31052.

166. Paszek MJ, Weaver VM. The tension mounts: mechanics meets morphogenesis and malignancy. J Mammary Gland Biol Neoplasia. 2004;9:325-342.

167. Georges PC, Janmey PA. Cell type-specific response to growth on soft materials. J Appl Physiol. 2005;98:1547-1553.

168. Georges PC, Miller WJ, Meaney DF, Sawyer ES, Janmey PA. Matrices with compliance comparable to that of brain tissue select neuronal over glial growth in mixed cortical cultures. Biophys J. 2006;90:3012-3018.

169. Zamir E, Geiger B. Molecular complexity and dynamics of cell-matrix adhesions. J Cell Sci. 2001;114:3583-3590.

170. Zaidel-Bar R, Itzkovitz S, Ma'ayan A, Iyengar R, Geiger B. Functional atlas of the integrin adhesome. Nat Cell Biol. 2007;9:858-867.

171. Assoian RK, Klein EA. Growth control by intracellular tension and extracellular stiffness. Trends Cell Biol. 2008;18:347-352.

172. Frisch SM, Screaton RA. Anoikis mechanisms. Curr Opin Cell Biol. 2001;13:555-562.

173. Comoglio PM, Boccaccio C, Trusolino L. Interactions between growth factor receptors and adhesion molecules: breaking the rules. Curr Opin Cell Biol. 2003;15:565-571.

174. Buxton PG, Bitar M, Gellynck K, Parkar M, Brown RA, Young AM, Knowles JC, Nazhat SN. Dense collagen matrix accelerates osteogenic differentiation and rescues the apoptotic response to MMP inhibition. Bone. 2008;43:377-385.

175. Rodolfo C, Piacentini M. Does cytoskeleton 'Akt' in apoptosis? Cell Death Differ. 2002;9:477-478.

133

176. Wells RG. The role of matrix stiffness in hepatic stellate cell activation and liver fibrosis. J Clin Gastroenterol. 2005;39:S158-161.

177. Kong HJ, Polte TR, Alsberg E, Mooney DJ. FRET measurements of cell-traction forces and nano-scale clustering of adhesion ligands varied by substrate stiffness. Proc Natl Acad Sci U S A. 2005;102:4300-4305.

178. Benton JA, Kern HB, Anseth KS. Substrate properties influence calcification in valvular interstitial cell culture. J Heart Valve Dis. 2008;17:689-699.

179. Hansen LK, Wilhelm J, Fassett JT. Regulation of hepatocyte cell cycle progression and differentiation by type I collagen structure. Curr Top Dev Biol. 2006;72:205-236.

180. Semler EJ, Ranucci CS, Moghe PV. Mechanochemical manipulation of hepatocyte aggregation can selectively induce or repress liver-specific function. Biotechnol Bioeng. 2000;69:359-369.

181. Tomasek JJ, Gabbiani G, Hinz B, Chaponnier C, Brown RA. Myofibroblasts and mechano-regulation of connective tissue remodelling. Nat Rev Mol Cell Biol. 2002;3:349-363.

182. Zaman MH, Trapani LM, Sieminski AL, Mackellar D, Gong H, Kamm RD, Wells A, Lauffenburger DA, Matsudaira P. Migration of tumor cells in 3D matrices is governed by matrix stiffness along with cell-matrix adhesion and proteolysis. Proc Natl Acad Sci U S A. 2006;103:10889-10894.

183. Wolf K, Mazo I, Leung H, Engelke K, von Andrian UH, Deryugina EI, Strongin AY, Brocker EB, Friedl P. Compensation mechanism in tumor cell migration: mesenchymal-amoeboid transition after blocking of pericellular proteolysis. J Cell Biol. 2003;160:267-277.

184. Peyton SR, Kim PD, Ghajar CM, Seliktar D, Putnam AJ. The effects of matrix stiffness and RhoA on the phenotypic plasticity of smooth muscle cells in a 3-D biosynthetic hydrogel system. Biomaterials. 2008;29:2597-2607.

185. Li Z, Dranoff JA, Chan EP, Uemura M, Sevigny J, Wells RG. Transforming growth factor-beta and substrate stiffness regulate portal fibroblast activation in culture. Hepatology. 2007;46:1246-1256.

186. Imanaka K, Takamoto S, Ohtsuka T, Oka T, Furuse A, Omata S. The stiffness of normal and abnormal mitral valves. Ann Thorac Cardiovasc Surg. 2007;13:178-184.

187. Huang Z, Li J, Jiang Z, Qi Y, Tang C, Du J. Effects of adrenomedullin, C-type natriuretic peptide, and parathyroid hormone-related peptide on calcification in cultured rat vascular smooth muscle cells. J Cardiovasc Pharmacol. 2003;42:89-97.

188. Inoue A, Hiruma Y, Hirose S, Yamaguchi A, Furuya M, Tanaka S, Hagiwara H. Stimulation by C-type natriuretic peptide of the differentiation of clonal osteoblastic MC3T3-E1 cells. Biochem Biophys Res Commun. 1996;221:703-707.

134

189. Quirk RA, Kellam B, Bhandari RN, Davies MC, Tendler SJ, Shakesheff KM. Cell-type-specific adhesion onto polymer surfaces from mixed cell populations. Biotechnol Bioeng. 2003;81:625-628.

190. Hubbell JA, Massia SP, Desai NP, Drumheller PD. Endothelial cell-selective materials for tissue engineering in the vascular graft via a new receptor. Biotechnology (N Y). 1991;9:568-572.

191. Iwasaki Y, Tabata E, Kurita K, Akiyoshi K. Selective cell attachment to a biomimetic polymer surface through the recognition of cell-surface tags. Bioconjug Chem. 2005;16:567-575.

192. Gu X, Masters KS. Role of the MAPK/ERK pathway in valvular interstitial cell calcification. Am J Physiol Heart Circ Physiol. 2009;296:H1748-1757.

193. Bellows CG, Melcher AH, Aubin JE. Contraction and organization of collagen gels by cells cultured from periodontal ligament, gingiva and bone suggest functional differences between cell types. J Cell Sci. 1981;50:299-314.

194. Christopherson GT, Song H, Mao HQ. The influence of fiber diameter of electrospun substrates on neural stem cell differentiation and proliferation. Biomaterials. 2009;30:556-564.

195. Koyama H, Raines EW, Bornfeldt KE, Roberts JM, Ross R. Fibrillar collagen inhibits arterial smooth muscle proliferation through regulation of Cdk2 inhibitors. Cell. 1996;87:1069-1078.

196. Nishiyama T, Horii I, Nakayama Y, Ozawa T, Hayashi T. A distinct characteristic of the quiescent state of human dermal fibroblasts in contracted collagen gel as revealed by no response to epidermal growth factor alone, but a positive growth response to a combination of the growth factor and saikosaponin b1. Matrix. 1990;10:412-419.

197. Ichii T, Koyama H, Tanaka S, Kim S, Shioi A, Okuno Y, Raines EW, Iwao H, Otani S, Nishizawa Y. Fibrillar collagen specifically regulates human vascular smooth muscle cell genes involved in cellular responses and the pericellular matrix environment. Circ Res. 2001;88:460-467.

198. Mauch C, Adelmann-Grill B, Hatamochi A, Krieg T. Collagenase gene expression in fibroblasts is regulated by a three-dimensional contact with collagen. FEBS Lett. 1989;250:301-305.

199. Mauch C, Hatamochi A, Scharffetter K, Krieg T. Regulation of collagen synthesis in fibroblasts within a three-dimensional collagen gel. Exp Cell Res. 1988;178:493-503.

200. Ravanti L, Heino J, Lopez-Otin C, Kahari VM. Induction of collagenase-3 (MMP-13) expression in human skin fibroblasts by three-dimensional collagen is mediated by p38 mitogen-activated protein kinase. J Biol Chem. 1999;274:2446-2455.

135

201. Romanic AM, Adachi E, Kadler KE, Hojima Y, Prockop DJ. Copolymerization of pNcollagen III and collagen I. pNcollagen III decreases the rate of incorporation of collagen I into fibrils, the amount of collagen I incorporated, and the diameter of the fibrils formed. J Biol Chem. 1991;266:12703-12709.

202. Prockop DJ, Kivirikko KI. Collagens: molecular biology, diseases, and potentials for therapy. Annu Rev Biochem. 1995;64:403-434.

203. Liu X, Wu H, Byrne M, Krane S, Jaenisch R. Type III collagen is crucial for collagen I fibrillogenesis and for normal cardiovascular development. Proc Natl Acad Sci U S A. 1997;94:1852-1856.

204. Gullberg DE, Lundgren-Akerlund E. Collagen-binding I domain integrins--what do they do? Prog Histochem Cytochem. 2002;37:3-54.

205. Heino J, Huhtala M, Kapyla J, Johnson MS. Evolution of collagen-based adhesion systems. Int J Biochem Cell Biol. 2009;41:341-348.

206. Kramer RH, Marks N. Identification of integrin collagen receptors on human melanoma cells. J Biol Chem. 1989;264:4684-4688.

207. Camper L, Hellman U, Lundgren-Akerlund E. Isolation, cloning, and sequence analysis of the integrin subunit alpha10, a beta1-associated collagen binding integrin expressed on chondrocytes. J Biol Chem. 1998;273:20383-20389.

208. Velling T, Kusche-Gullberg M, Sejersen T, Gullberg D. cDNA cloning and chromosomal localization of human alpha(11) integrin. A collagen-binding, I domain-containing, beta(1)-associated integrin alpha-chain present in muscle tissues. J Biol Chem. 1999;274:25735-25742.

209. Kim JK, Xu Y, Xu X, Keene DR, Gurusiddappa S, Liang X, Wary KK, Hook M. A novel binding site in collagen type III for integrins alpha1beta1 and alpha2beta1. J Biol Chem. 2005;280:32512-32520.

210. Vuoriluoto K, Jokinen J, Kallio K, Salmivirta M, Heino J, Ivaska J. Syndecan-1 supports integrin alpha2beta1-mediated adhesion to collagen. Exp Cell Res. 2008;314:3369-3381.

211. Shrivastava A, Radziejewski C, Campbell E, Kovac L, McGlynn M, Ryan TE, Davis S, Goldfarb MP, Glass DJ, Lemke G, Yancopoulos GD. An orphan receptor tyrosine kinase family whose members serve as nonintegrin collagen receptors. Mol Cell. 1997;1:25-34.

212. Vogel W, Gish GD, Alves F, Pawson T. The discoidin domain receptor tyrosine kinases are activated by collagen. Mol Cell. 1997;1:13-23.

213. Shah DN, Recktenwall-Work SM, Anseth KS. The effect of bioactive hydrogels on the secretion of extracellular matrix molecules by valvular interstitial cells. Biomaterials. 2008;29:2060-2072.

136

214. Chester AH, Taylor PM. Molecular and functional characteristics of heart-valve interstitial cells. Philos Trans R Soc Lond B Biol Sci. 2007;362:1437-1443.

215. Liu AC, Joag VR, Gotlieb AI. The emerging role of valve interstitial cell phenotypes in regulating heart valve pathobiology. Am J Pathol. 2007;171:1407-1418.

216. O'Brien KD. Pathogenesis of calcific aortic valve disease: a disease process comes of age (and a good deal more). Arterioscler Thromb Vasc Biol. 2006;26:1721-1728.

217. Speer MY, Giachelli CM. Regulation of cardiovascular calcification. Cardiovasc Pathol. 2004;13:63-70.

218. Mohler ER, 3rd. Mechanisms of aortic valve calcification. Am J Cardiol. 2004;94:1396-1402, A1396.

219. Filip DA, Nistor A, Bulla A, Radu A, Lupu F, Simionescu M. Cellular events in the development of valvular atherosclerotic lesions induced by experimental hypercholesterolemia. Atherosclerosis. 1987;67:199-214.

220. Mathieu P, Voisine P, Pepin A, Shetty R, Savard N, Dagenais F. Calcification of human valve interstitial cells is dependent on alkaline phosphatase activity. J Heart Valve Dis. 2005;14:353-357.

221. Discher DE, Janmey P, Wang YL. Tissue cells feel and respond to the stiffness of their substrate. Science. 2005;310:1139-1143.

222. Ringe J, Kaps C, Schmitt B, Buscher K, Bartel J, Smolian H, Schultz O, Burmester GR, Haupl T, Sittinger M. Porcine mesenchymal stem cells. Induction of distinct mesenchymal cell lineages. Cell Tissue Res. 2002;307:321-327.

223. Chai H, Zhou W, Lin P, Lumsden A, Yao Q, Chen C. Ginsenosides block HIV protease inhibitor ritonavir-induced vascular dysfunction of porcine coronary arteries. Am J Physiol Heart Circ Physiol. 2005;288:H2965-2971.

224. Grinnell F, Zhu M, Carlson MA, Abrams JM. Release of mechanical tension triggers apoptosis of human fibroblasts in a model of regressing granulation tissue. Exp Cell Res. 1999;248:608-619.

225. Xia H, Nho RS, Kahm J, Kleidon J, Henke CA. Focal adhesion kinase is upstream of phosphatidylinositol 3-kinase/Akt in regulating fibroblast survival in response to contraction of type I collagen matrices via a beta 1 integrin viability signaling pathway. J Biol Chem. 2004;279:33024-33034.

226. Schoen FJ. Evolving concepts of cardiac valve dynamics: the continuum of development, functional structure, pathobiology, and tissue engineering. Circulation. 2008;118:1864-1880.

227. Theriot JA, Mitchison TJ. Actin microfilament dynamics in locomoting cells. Nature. 1991;352:126-131.

137

228. Frame M, Norman J. A tal(in) of cell spreading. Nat Cell Biol. 2008;10:1017-1019.

229. Arora PD, McCulloch CA. Dependence of collagen remodelling on alpha-smooth muscle actin expression by fibroblasts. J Cell Physiol. 1994;159:161-175.

230. Hinz B, Celetta G, Tomasek JJ, Gabbiani G, Chaponnier C. Alpha-smooth muscle actin expression upregulates fibroblast contractile activity. Mol Biol Cell. 2001;12:2730-2741.

231. Tian B, Lessan K, Kahm J, Kleidon J, Henke C. beta 1 integrin regulates fibroblast viability during collagen matrix contraction through a phosphatidylinositol 3-kinase/Akt/protein kinase B signaling pathway. J Biol Chem. 2002;277:24667-24675.

232. Clarke MC, Littlewood TD, Figg N, Maguire JJ, Davenport AP, Goddard M, Bennett MR. Chronic apoptosis of vascular smooth muscle cells accelerates atherosclerosis and promotes calcification and medial degeneration. Circ Res. 2008;102:1529-1538.

233. Benton JA, Kern HB, Leinwand LA, Mariner PD, Anseth KS. Statins Block Calcific Nodule Formation of Valvular Interstitial Cells by Inhibiting {alpha}-Smooth Muscle Actin Expression. Arterioscler Thromb Vasc Biol. 2009.

234. Aikawa E, Nahrendorf M, Sosnovik D, Lok VM, Jaffer FA, Aikawa M, Weissleder R. Multimodality molecular imaging identifies proteolytic and osteogenic activities in early aortic valve disease. Circulation. 2007;115:377-386.

235. Kunz-Schughart LA, Wenninger S, Neumeier T, Seidl P, Knuechel R. Three-dimensional tissue structure affects sensitivity of fibroblasts to TGF-beta 1. Am J Physiol Cell Physiol. 2003;284:C209-219.

236. McDonald TO, Gerrity RG, Jen C, Chen HJ, Wark K, Wight TN, Chait A, O'Brien KD. Diabetes and arterial extracellular matrix changes in a porcine model of atherosclerosis. J Histochem Cytochem. 2007;55:1149-1157.

237. Simionescu A, Simionescu DT, Vyavahare NR. Osteogenic responses in fibroblasts activated by elastin degradation products and transforming growth factor-beta1: role of myofibroblasts in vascular calcification. Am J Pathol. 2007;171:116-123.

238. Watson KE, Parhami F, Shin V, Demer LL. Fibronectin and collagen I matrixes promote calcification of vascular cells in vitro, whereas collagen IV matrix is inhibitory. Arterioscler Thromb Vasc Biol. 1998;18:1964-1971.

239. Mohler ER, 3rd, Adam LP, McClelland P, Graham L, Hathaway DR. Detection of osteopontin in calcified human aortic valves. Arterioscler Thromb Vasc Biol. 1997;17:547-552.

240. Fang Y, Mohler ER, 3rd, Hsieh E, Osman H, Hashemi SM, Davies PF, Rothblat GH, Wilensky RL, Levitan I. Hypercholesterolemia suppresses inwardly rectifying K+ channels in aortic endothelium in vitro and in vivo. Circ Res. 2006;98:1064-1071.

138

241. Walsh BJ, Thornton SC, Penny R, Breit SN. Microplate reader-based quantitation of collagens. Anal Biochem. 1992;203:187-190.

242. Pandey KN. Biology of natriuretic peptides and their receptors. Peptides. 2005;26:901-932.

243. Naruko T, Ueda M, van der Wal AC, van der Loos CM, Itoh H, Nakao K, Becker AE. C-type natriuretic peptide in human coronary atherosclerotic lesions. Circulation. 1996;94:3103-3108.

244. Chusho H, Tamura N, Ogawa Y, Yasoda A, Suda M, Miyazawa T, Nakamura K, Nakao K, Kurihara T, Komatsu Y, Itoh H, Tanaka K, Saito Y, Katsuki M, Nakao K. Dwarfism and early death in mice lacking C-type natriuretic peptide. Proc Natl Acad Sci U S A. 2001;98:4016-4021.

245. Tamura N, Doolittle LK, Hammer RE, Shelton JM, Richardson JA, Garbers DL. Critical roles of the guanylyl cyclase B receptor in endochondral ossification and development of female reproductive organs. Proc Natl Acad Sci U S A. 2004;101:17300-17305.

246. Inoue A, Hiruma Y, Hirose S, Yamaguchi A, Hagiwara H. Reciprocal regulation by cyclic nucleotides of the differentiation of rat osteoblast-like cells and mineralization of nodules. Biochem Biophys Res Commun. 1995;215:1104-1110.

247. Suda M, Tanaka K, Fukushima M, Natsui K, Yasoda A, Komatsu Y, Ogawa Y, Itoh H, Nakao K. C-type natriuretic peptide as an autocrine/paracrine regulator of osteoblast. Evidence for possible presence of bone natriuretic peptide system. Biochem Biophys Res Commun. 1996;223:1-6.

248. Yeh LC, Zavala MC, Lee JC. C-type natriuretic peptide enhances osteogenic protein-1-induced osteoblastic cell differentiation via Smad5 phosphorylation. J Cell Biochem. 2006;97:494-500.

249. Suda M, Komatsu Y, Tanaka K, Yasoda A, Sakuma Y, Tamura N, Ogawa Y, Nakao K. C-Type natriuretic peptide/guanylate cyclase B system in rat osteogenic ROB-C26 cells and its down-regulation by dexamethazone. Calcif Tissue Int. 1999;65:472-478.

250. Han B, Fixler R, Beeri R, Wang Y, Bachrach U, Hasin Y. The opposing effects of endothelin-1 and C-type natriuretic peptide on apoptosis of neonatal rat cardiac myocytes. Eur J Pharmacol. 2003;474:15-20.

251. Suenobu N, Shichiri M, Iwashina M, Marumo F, Hirata Y. Natriuretic peptides and nitric oxide induce endothelial apoptosis via a cGMP-dependent mechanism. Arterioscler Thromb Vasc Biol. 1999;19:140-146.

252. Itoh T, Nagaya N, Murakami S, Fujii T, Iwase T, Ishibashi-Ueda H, Yutani C, Yamagishi M, Kimura H, Kangawa K. C-type natriuretic peptide ameliorates monocrotaline-induced pulmonary hypertension in rats. Am J Respir Crit Care Med. 2004;170:1204-1211.

139

253. Miyoshi H, Umeshita K, Sakon M, Imajoh-Ohmi S, Fujitani K, Gotoh M, Oiki E, Kambayashi J, Monden M. Calpain activation in plasma membrane bleb formation during tert-butyl hydroperoxide-induced rat hepatocyte injury. Gastroenterology. 1996;110:1897-1904.

254. Martin SJ, O'Brien GA, Nishioka WK, McGahon AJ, Mahboubi A, Saido TC, Green DR. Proteolysis of fodrin (non-erythroid spectrin) during apoptosis. J Biol Chem. 1995;270:6425-6428.

255. Oliveros JC. VENNY. An interactive tool for comparing lists with Venn Diagrams.; 2007.

256. Shindo K, Kawashima N, Sakamoto K, Yamaguchi A, Umezawa A, Takagi M, Katsube K, Suda H. Osteogenic differentiation of the mesenchymal progenitor cells, Kusa is suppressed by Notch signaling. Exp Cell Res. 2003;290:370-380.

257. Garg V, Muth AN, Ransom JF, Schluterman MK, Barnes R, King IN, Grossfeld PD, Srivastava D. Mutations in NOTCH1 cause aortic valve disease. Nature. 2005;437:270-274.

258. Steitz SA, Speer MY, Curinga G, Yang HY, Haynes P, Aebersold R, Schinke T, Karsenty G, Giachelli CM. Smooth muscle cell phenotypic transition associated with calcification: upregulation of Cbfa1 and downregulation of smooth muscle lineage markers. Circ Res. 2001;89:1147-1154.

259. Rajamannan NM, Subramaniam M, Springett M, Sebo TC, Niekrasz M, McConnell JP, Singh RJ, Stone NJ, Bonow RO, Spelsberg TC. Atorvastatin inhibits hypercholesterolemia-induced cellular proliferation and bone matrix production in the rabbit aortic valve. Circulation. 2002;105:2660-2665.

260. Javed A, Bae JS, Afzal F, Gutierrez S, Pratap J, Zaidi SK, Lou Y, van Wijnen AJ, Stein JL, Stein GS, Lian JB. Structural coupling of Smad and Runx2 for execution of the BMP2 osteogenic signal. J Biol Chem. 2008;283:8412-8422.

261. Phillips JE, Gersbach CA, Wojtowicz AM, Garcia AJ. Glucocorticoid-induced osteogenesis is negatively regulated by Runx2/Cbfa1 serine phosphorylation. J Cell Sci. 2006;119:581-591.

262. Moioli EK, Hong L, Mao JJ. Inhibition of osteogenic differentiation of human mesenchymal stem cells. Wound Repair Regen. 2007;15:413-421.

263. Ghayor C, Ehrbar M, San Miguel B, Gratz KW, Weber FE. cAMP enhances BMP2-signaling through PKA and MKP1-dependent mechanisms. Biochem Biophys Res Commun. 2009;381:247-252.

264. Chambers RC, Leoni P, Kaminski N, Laurent GJ, Heller RA. Global expression profiling of fibroblast responses to transforming growth factor-beta1 reveals the induction of inhibitor of differentiation-1 and provides evidence of smooth muscle cell phenotypic switching. Am J Pathol. 2003;162:533-546.

140

265. Serini G, Gabbiani G. Mechanisms of myofibroblast activity and phenotypic modulation. Exp Cell Res. 1999;250:273-283.

266. Cagnin S, Biscuola M, Patuzzo C, Trabetti E, Pasquali A, Laveder P, Faggian G, Iafrancesco M, Mazzucco A, Pignatti PF, Lanfranchi G. Reconstruction and functional analysis of altered molecular pathways in human atherosclerotic arteries. BMC Genomics. 2009;10:13.

267. Schneider JG, von Eynatten M, Parhofer KG, Volkmer JE, Schiekofer S, Hamann A, Nawroth PP, Dugi KA. Atorvastatin improves diabetic dyslipidemia and increases lipoprotein lipase activity in vivo. Atherosclerosis. 2004;175:325-331.

268. Ueno N, Ling N, Ying SY, Esch F, Shimasaki S, Guillemin R. Isolation and partial characterization of follistatin: a single-chain Mr 35,000 monomeric protein that inhibits the release of follicle-stimulating hormone. Proc Natl Acad Sci U S A. 1987;84:8282-8286.

269. de Winter JP, ten Dijke P, de Vries CJ, van Achterberg TA, Sugino H, de Waele P, Huylebroeck D, Verschueren K, van den Eijnden-van Raaij AJ. Follistatins neutralize activin bioactivity by inhibition of activin binding to its type II receptors. Mol Cell Endocrinol. 1996;116:105-114.

270. Itoh S, Itoh F, Goumans MJ, Ten Dijke P. Signaling of transforming growth factor-beta family members through Smad proteins. Eur J Biochem. 2000;267:6954-6967.

271. You L, Kruse FE. Differential effect of activin A and BMP-7 on myofibroblast differentiation and the role of the Smad signaling pathway. Invest Ophthalmol Vis Sci. 2002;43:72-81.

272. Iemura S, Yamamoto TS, Takagi C, Uchiyama H, Natsume T, Shimasaki S, Sugino H, Ueno N. Direct binding of follistatin to a complex of bone-morphogenetic protein and its receptor inhibits ventral and epidermal cell fates in early Xenopus embryo. Proc Natl Acad Sci U S A. 1998;95:9337-9342.

273. Sadeghi MM, Collinge M, Pardi R, Bender JR. Simvastatin modulates cytokine-mediated endothelial cell adhesion molecule induction: involvement of an inhibitory G protein. J Immunol. 2000;165:2712-2718.

274. Sugiyama S, Kugiyama K, Matsumura T, Suga S, Itoh H, Nakao K, Yasue H. Lipoproteins regulate C-type natriuretic peptide secretion from cultured vascular endothelial cells. Arterioscler Thromb Vasc Biol. 1995;15:1968-1974.

275. Kohno M, Yokokawa K, Yasunari K, Kano H, Minami M, Ueda M, Yoshikawa J. Effect of natriuretic peptide family on the oxidized LDL-induced migration of human coronary artery smooth muscle cells. Circ Res. 1997;81:585-590.

276. Kohno M, Yasunari K, Maeda K, Kano H, Minami M, Hanehira T, Yoshikawa J. Effects of cardiac natriuretic peptides on oxidized low-density lipoprotein- and

141

lysophosphatidylcholine-induced human mesangial cell migration. Hypertension. 2000;35:971-977.

277. Chrisman TD, Perkins DT, Garbers DL. Identification of a potent serum factor that causes desensitization of the receptor for C-Type natriuretic peptide. Cell Commun Signal. 2003;1:4.

278. Yoder AR, Kruse AC, Earhart CA, Ohlendorf DH, Potter LR. Reduced ability of C-type natriuretic peptide (CNP) to activate natriuretic peptide receptor B (NPR-B) causes dwarfism in lbab -/- mice. Peptides. 2008;29:1575-1581.

142

Appendix A A. Protocols

A.1. Fabrication of Type I Collagen Matrices

(Modified from Bellows CG, Melcher AH, Aubin JE. Contraction and organization of collagen

gels by cells cultured from periodontal ligament, gingiva and bone suggest functional differences

between cell types. J Cell Sci. 1981;50:299-314.)

Purpose: To synthesize collagen matrices with different mechanical properties, but similar

biochemical properties.

Reagents:

10x concentrated sterile DMEM (Sigma D7777) made with distilled water

0.25 M sterile NaHCO3 (Sigma 223530) buffer made with distilled water

0.01 M sterile NaOH (Sigma S2770) made with distilled water

FBS (HyClone FSSP9749370, lot # KRA25425)

1% Penicillin/Streptomycin (P/S) mixture (Sigma P4333)

Bovine dermal Type I collagen (Advanced Biomatrix, Part No. 5005)

Equipment:

24 well plate

12mm coverslips

Small tweezers

Procedure:

1. Sterilize coverslips with ethanol burner and place in 24 well plate

2. For 24-well plate, combine and vortex to mix:

a. 0.6mL of 10x concentrated DMEM

b. 0.6mL of 0.25 M NaHCO3

c. 0.6mL of FBS

d. 0.6mL of P/S mixture

e. 0.24mL of 0.01 M NaOH

3. Add 5mL of Type I collagen; pipette up and down gently to mix.

4. Pipette gel mixture onto plate:

a. Thick/compliant gel – 500 L in each well of a 24-well plate

143

b. Thin/stiff gel – 500 L in each well of a 24-well plate, let sit for 1 minute and

then remove excess gel mixture by aspiration

5. Incubate at 37oC, 5% CO2 overnight for polymerization

A.2. Scanning Electron Microscopy

Purpose: To study the microstructure of specimens

Reagents:

Sterile PBS with calcium chloride and magnesium chloride (Sigma P5655)

10% Neutral Buffered Formalin (NBF)

- 100 mL formaldehyde (Sigma F1635)

- 900 mL distilled water

- 4g sodium phosphate monobasic (Sigma S8282)

- 6.5g sodium phosphate dibasic anhydrous (Sigma S0876)

Ethanol (30%, 50%, 70%, 90%, 95% and 100%)

Liquid carbon dioxide

Conductive paint

Equipment:

Polaron CPD7501 critical point drying system

Polaron SC515 SEM coating system

Scanning Electron Microscope (Hitachi Instrument Model S-2500)

Aluminum stubs

Procedures:

1. Wash samples (2-3 times) with sterile PBS

2. Fix samples with sterile 10% NBF for 30 minutes (note: sterilize NBF with 0.2m filter)

3. Dehydrate sample at room temperature in a series of ethanol washes at 30%, 50%, 70%*,

95%* and 100% ethanol for 30 minutes each

4. Critical point dry the sample with liquid carbon dioxide in a Polaron CPD7501

5. Mount the samples on aluminum stubs and place conductive paint on the end of the

samples

6. Sputter coat with gold using a Polaron SC515 SEM coating system

7. Examine the sample at 1,000X to 5,000X magnification using a scanning electron

microscope

144

* Note: samples may be store in 70% or 95% ethanol at 4oC degree for a few weeks before the

last step of dehydration.

A.3. Hydroxyproline Assay

Purpose: To determine the approximate collagen content within a sample by quantifying the

amount of hydroxyproline (OH-Pro).

Overview:

This colourimetric assay is used to quantitate the amount of hydroxyproline (OH-Pro) in papain

digested samples subjected to acid hydrolysis. Since collagen contains approximately 8-10%

hydroxyproline, this assay can be used to determine the approximate amount of collagen within a

sample. Free hydroxyproline is released from protein and peptides by acid hydrolysis. The

hydroxlate is then neutralized. The hydroxyproline is oxidized into a pyrrole with chloramine T.

This intermediate turns pink in colour with the addition of Ehrlich’s Reagent (4-

dimethylaminobenaldehyde).

Reagents:

Papain digestion buffer

- 0.272 g ammonium acetate

- 0.038 g Na2 EDTA 2H2O

- 0.031 g DL-dithiothreitol (DTT)

- Add 100 mL distilled water

- Adjust pH to 6.2 with acetic acid or sodium hydroxide

- Store at 4oC

Papain (25 mg/mL Sigma P3125)

0.001 N and 6 N hydrochloride acid (HCl)

5.7 N NaOH

L-hydroxyproline standard (either cis or trans variants)

Hydroxyproline assay buffer

- 5 g citric acid

- 1.2 mL glacial acetic acid

- 7.23 g sodium acetate

- 3.4 g sodium hydroxide

145

- Bring up to 100 mL with deionized H2O and adjust pH to 6 with acetic acid or sodium

hydroxide

- Store at 4oC for up to 2-3 weeks

0.05 N chloramine-T

- Dissolve 0.282 g chloramine-T in 4 mL deionized H2O

- Add 6 mL methyl-cellosolve (2-methoxyethanol) and 10 mL of assay buffer

- Chloramine-T solution must be prepared fresh each time

3.15 N perchloric acid

- In a fume hood, dilute 4.6 mL of 70% perchloric acid in 14.6 mL deionized H2O

- Store in a glass container at room temperature

Ehrlich’s Reagent

- Dissolve 4 g of Ehrlich’s reagent ((p-Dimethylaminobenzaldehyde) in 20 mL of methyl-

cellosolve

- Dilution is facilitated by heating the mixture to no greater than 60C

- Ehrlich’s reagent must be prepared fresh each time

Equipment:

96 well microplate

Plate reader (for absorbance measurement)

Procedures:

I. Papain digestion

1. Pre-heat the waterbath to 65C

2. Dilute papain stock solution to desire concentration with papain digestion buffer. For in

vitro cell cultures, a concentration of 40 g/mL is recommended. For in vivo samples, a

concentration of 80 g/mL is recommended.

3. Add 600 L papain digestion solution to each sample and transfer to eppendorf tubes

4. Wrap each tube with paraffin and a plastic microtube cover

5. Digest the samples at 65oC for 48 hours

6. Store the samples at –20oC until ready for hydroxyproline assay

146

II. Acid hydrolysis of samples

7. Pre-heat the test tube heater to 100oC

8. Transfer 100 L of papain digested sample into a Pyrex glass tube with a Teflon lined

screw cap

9. Add an equal amount of 6N HCl (e.g. add 100 L of sample with 100 L 6N HCl)

10. Incubate the samples in the test tube holder at 110oC for 18 hours

11. Neutralize the hydroxylate by adding the same amount of NaOH (e.g. 100 L 5.7N

NaOH required for 100 L sample)

12. Make a 1/60 dilution of the hydroxylate by bringing up the volume to 6 mL with

deionized H2O. (e.g. add 5700 L for 100 L of sample)

13. Aliquot 600 L of hydroxylate into three separate eppendorf tubes

III. Preparation of L-hydroxyproline standards

14. Prepare hydroxyproline stock solution by dissolving L-hydroxyproline standard to a final

concentration of 100 g/mL in 0.001N HCl. Store in aliquot of 200 L each at –70oC.

15. Prepare a fresh set of OH-Pro standards by diluting the 100 g/mL hydroxyproline stock

solution in distilled water. Keep the standards on ice.

Serial Dilution for Hydroxyproline Standard Initial Volume 2400

Concentrations Volume Aliquot Less Aliquot Mass Add Volume

[g/mL] [L] [L] [L] [g] [L]

5 2400 600 1800 9 450 4 2250 600 1650 6.6 550 3 2200 600 1600 4.8 800 2 2400 600 1800 3.6 1800 1 3600 600 3000 3 3000 0.5 6000 600 5400 2.7 Concentration of Standard 100 [g/mL] Volume of Standard 120 [L] Volume of Distilled Water 2280 [L]

147

IV. Colour development for microplate protocol

16. For each tube (samples and standards), in the same order, add the following:

- 300 L 0.05 N chloramine-T (mix and let stand for 20 minutes)

- 300 L 3.15 N perchloric acid (mix and let stand for 5 minutes)

- 300 L Ehrlich’s Reagent (mix and heat at 60C for 20 minutes)

17. Cool tubes in cold tap water for 5 minutes

18. Load 200 L of standards ad samples into a 96 well microplate

19. Measure the absorbance at 560 nm using a plate reader

V. Determination of hydroxyproline content

20. Determine concentration of sample from standard curve and multiply by the dilution

factor (i.e. 6000 L/ volume of aliquot in L)

A.4. Valvular Interstitial Cell Isolation

Purpose: To isolation pure population of valve interstitial cells from pig aortic valves.

Reagents:

1% Penicillin/Streptomycin (P/S) mixture (Sigma P4333) 0.1mg/ml Amphotericin B (Sigma A9528) Sterile and non-sterile phosphate buffered saline (PBS) with calcium chloride and

magnesium chloride (Sigma P5655) 0.125% trypsin with EDTA (Sigma T4049) TESCA buffer (50mM TES (Sigma T5691), 0.36mM Calcium chloride (Sigma C5670), pH

7.4 at 37oC) 150 units/ml of collagenase (Sigma TC0130) Dulbecco’s Modified Eagle’s Medium (DMEM, Sigma D7777) Fetal Bovine Serum (FBS, HyClone FSSP9749370, lot # KRA25425) Trypan blue solution (Sigma T8154) Equipment:

Large and small dissection scissors Large and small tweezers Scalpel Biohazard waste bags

148

Dissection tray

40 m and 70 m cell strainers

Cell scrapers Sterile petri dishes 15mL centrifuge tubes 50mL centrifuge tubes Hemocytometer or ViCell cell viability analyzer Procedure:

Step 1-3 can be done outside the hood

1. Store pig hearts in non-sterile PBS with Ca2+/Mg2+ until ready for dissection

2. Cut heart in half

3. Use large dissection scissors to cut open aorta until all three leaflets are visible

4. Use small scissors and tweezers to remove individual leaflets and then rinse the leaflets

with PBS with Ca2+/Mg2+ and 1% P/S and 0.1mg/ml Amphotericin B

Step 4 and up must be done in the cell culture hood

5. Rinse leaflets (2-3 times) with sterile PBS with Ca2+/Mg2+ and 1% P/S and 0.1mg/ml

Amphotericin B, hold leaflets in last wash

6. To remove endothelial cells (ECs):

a. Place 3 leaflets per 5mL of 150 units/mL collagenase solution reconstituted in

TESCA buffer

b. Incubate for 20 minutes at 37oC, 5% CO2

c. Transfer leaflets to a new tube containing 0.125% trypsin and incubate for 7

minutes at 37oC, 5% CO2

7. Vortex at maximum speed for 1 minute

8. Place leaflets in a petri dish and scrape leaflet surfaces using a cell scraper

9. In another clean petri dish, rinse (2-3 times) away loose ECs with sterile PBS with

Ca2+/Mg2+ and 1% P/S and 0.1mg/ml Amphotericin B

10. In another clean petri dish, soak leaflets in 0.125% trypsin for 1-2 minutes

11. Transfer three leaflets at a time into a new petri dish and mince into small pieces with

scissors

12. Transfer leaflets pieces to 5ml of 150 units/mL collagenase solution reconstituted in PBS

with Ca2+/Mg2+

149

13. Incubate for 2 hours at 37oC, 5% CO2

14. Vortex at maximum speed for 1 minute

15. Strain tissue using 70 m cell strainers (1 strainer/3 leaflets)

16. Rinse filter once with equal volume (i.e. 5mL) of DMEM

17. Centrifuge cell to pellet (Speed: 1150 rpm, 7 minutes)

18. Resuspend in supplemented DMEM (DMEM + 10% FBS + 1% P/S)

19. Count viable and dead cells with hemocytometer or with Vi-Cell cell viability analyzer

20. Seed cells with the desire cell density

21. Two hours after plating, check cells to see if adherent

22. Remove medium and dead cells by replacing the media with fresh supplemented DMEM

the following day

A.5. Cryopresevation of VICs Purpose: To store viable animal cells long-term.

Regents:

Dulbecco’s Modified Eagle’s Medium (DMEM, Sigma D7777)

Sterile PBS without calcium chloride and magnesium chloride (Sigma P5655)

0.125% trypsin with EDTA (Sigma T4049)

Cryoprotective medium (90% supplemented DMEM and 10% of dimethyl sulfoxide

(DMSO, Sigma D2650))

Trypan blue solution (Sigma T8154)

Equipment:

15mL centrifuge tubes

2mL cryogenic vials

Hemocytometer or ViCell cell viability analyzer

Procedure:

Prior to freezing, the cells should be maintained in an actively growing state to ensure maximum

health and a good recovery. Using a microscope, quickly check the general appearance of the

culture. Also check culture with unaided eye to look for contaminants (bacteria, small fungal

colonies that may be floating at the medium-air interface).

1. Pre-label the estimated number of cryogenic vials required (assuming 1-2 x 106 cells/vial)

with the following information:

150

a. Name of the cell type

b. Passage number

c. Date

d. Cell density

e. Your name

2. Harvest cells by trypsinization and count total cell number using a hemocytometer or Vi-

Cell cell viability analyzer

3. Remove the supernatant from the centrifuged cells and resuspend the cell pellet with

sufficient cryoprotective medium to give a final cell concentration of 1-2 x 106 cells/vial.

4. Place the vials in a controlled rate of freezing container overnight in the –80oC freezer.

5. The next day, transfer the vials to a liquid nitrogen freezer for permanent storage.

A.6. Releasing Cells from Collagen Matrices

Purpose: To release the cells cultured on the surface of collagen matrices for nucleic acid

extraction, proliferation assay or protein extraction

Reagents:

Collagenase (Sigma C0130)

TESCA buffer (TESCA buffer (50mM TES (Sigma T5691), 0.36mM Calcium chloride

(Sigma C5670), pH 7.4 at 37oC)


Equipment:

4 oC Centrifuge

Procedures:

1. Prepare 500 units/mL of collagenase with TESCA buffer, protect away from light

a. For 24 well plate, each well will require 1 mL of collagenase mixture

2. Remove cell culture media

3. Rinse samples (2-3 times) with sterile PBS with Ca2+/Mg2+

4. Incubate samples with 500 units/mL of collagenase at 37oC:

a. Thick/compliant matrices – incubate for 1.5 hours, pipette the matrices up and

down every 15-20 minutes to facilitate the digestion

b. Thin/stiff matrices – incubate for 20 minutes

151

5. Once incubation is completed (i.e. collagen matrices should be digested and become a

liquid), transfer the digested mixtures (i.e. degraded collagen, collagenase solution and

cell suspension) to eppendorf tubes

6. Pellet cells by centrifugation at 4oC, 900 x g for 15 minutes

7. Remove supernatant

8. Resuspend cell pellets with sterile PBS with Ca2+ /Mg2+ or RNase/ DNase PBS if

samples are used for RNA extraction

9. Centrifuge at 4 oC, 900 x g for 10 minutes

10. Remove supernatant

11. Store cell pellets at –80oC or proceed to assay

A.7. Cellular Proliferation Assay (Modified from manufacturer protocol “CyQuant® NF Cell Proliferation Assay kit”, Invitrogen.)

Purpose: To determine the extent of proliferation based on measurement of cellular DNA

content via fluorescent dye binding.

Reagents:

CyQuant® NF cell proliferation assay kit (Invitrogen C35006)


RNAse/DNase free deionized water

Equipment:

4 oC Centrifuge

FLUOstar OPTIMA fluorescence plate reader

Procedures:

1. After releasing cells from collagen matrices using collagenase (See protocol A.6).

Sediment cells by centrifugation (300 x g for 10 minutes) and wash cell pellets with

sterile PBS

2. Primary VICs of known cell numbers (100-20000 cells/tube) can be used to generate

standard curve

3. Prepare 11 mL of 1x HBSS buffer by diluting 2.2 mL of 5x HBSS buffer (Component C)

with 8.8 mL of RNAse/DNase free deionized water

4. Prepare 2x dye binding solution by adding 22 L of CyQuant® NF dye reagent

(Component A)

152

5. Resuspend cell pellets in 1x HBSS buffer and dispense 50 L aliquots of suspension

containing 100-10000 cells into microplate wells

6. Dispense 50 L of 2X dye binding solution into each microplate well

7. Cover the microplate with tin foil and incubate at 37 oC for 60 minutes

8. Measure the fluorescence intensity of each sample using a fluorescence microplate reader

with excitation at ~485 nm and emission detection at ~530 nm with in 2 hours after

incubation

A.8. Cell Viability Assay (Modified from manufacturer protocol “LIVE/ DEAD ® Viability/Cytotoxicity Kit for

mammalian cells”, Invitrogen.)

Purpose: To determine live and dead cells with two fluorescence probes that recognize

intracellular esterase activity and plasma membrane integrity.

Reagents:

LIVE/ DEAD ® Viability/Cytotoxicity Kit for mammalian cells (Invitrogen MP03224)


Supplemented DMEM

Equipment:

Fluorescence microscope with excitation/emission wavelengths at 494/517nm and

528/617nm

Procedures:

1. Warm the dye reagent to room temperature

2. Wash adherent cells with sterile PBS to remove esterase accumulated in the serum

3. For controls, kill the cells using 4% formalin for 5-10 minutes

4. Dilute calcein AM and Ethidium Homodimer (EthD-1) to 4 M and 2 M,

respectively, with freshly made supplemented DMEM

5. Add 500 L of the diluted calcein AM and EthD-1 mixture to each well

6. Cover the culture plate with tin foil and incubate at 37 oC for 30 minutes

7. Immediate examine the staining under a fluorescence microscope. Live and dead cells

can be visualized with excitation/emission wavelengths at 494/517nm (i.e. Green) and

528/617nm (i.e. Red), respectively.

153

8. Samples can be mounted on clean microscope slides and sealed under a coverglass to

prevent evaporation for storage.

A.9. Alkaline Phosphatase Staining

Purpose: To determine alkaline phosphatase activity in culture, which is a marker for osteoblast

differentiation.

Reagents:

ALP harvest buffer:

- 10 mM Tris-HCl, pH 7.4 (Sigma T-3253 Trizma HCl, MW = 157.6)

- 0.2% NP40 or IgePal

- Mix the two chemicals in one tube/glass bottle. Store at 4 degree fridge.

ALP assay buffer:

- 2 mM PMSF (Sigma P7626 Phenylmethylsulfonyl fluoride)

Make a stock of 200mM PMSF with 100% EtOH. Store at 4 degree fridge. Add

PMSF to mixture (a+b) right before use to give a final concentration of 2mM PMSF.

(i.e. 1:100 dilution).

ALP assay buffer:

- 100 mM glycine (MW = 75.07). For 500 mL, add 3.754 g

- 1 mM MgCl2 (hexahydrate form, FW = 203.3). For 500 mL, add 0.1017g

- deionized water to 500 mL

Adjust pH to 10.5, store at 4Â°C

pNPP substrate:

- 50 mM pNPP (Sigma 4744 p-Nitrophenylphosphate disodium hexaH2O, FW=371.1)can be

found in -20 degree freezer with the ALP staining reagents. Add 0.4639g to 24.5mL

deionized water.

pNP standard (reaction product)

- stock solution is 10 mM pNP (Sigma 7660 p-Nitrophenol solution), can be found in 4

degree fridge

0.1N NaOH (reaction stop solution)

- If the stock solution is more concentrated than 0.1N, dilute with deionized water

154

Procedure:

1. Wash cells with cold PBS

2. Add:

- 250uL lysis buffer to cells for 6-well plate



3. Rocking wells to cover all cells

4. Incubate 15 minutes, then scrape and collect lysate in 1.5mL centrifuge tube

5. Vortex lysate for 30 seconds and keep ON ICE or store at -20 oC

6. Centrifuge for 10 minutes at 13,000 RPM

7. Transfer sample volume of supernatant to 2 new 1.5mL centrifuge tubes.

For VICs cultured in OS media,

- 45ul for the ALP activity assay

- 5uL for micro BCA assay

8. Prepare a control (blank) sample with the same volume of lysis buffer only

9. To samples and the blank, add ALP assay buffer to make a total volume of:

a. 400uL for 6 well plate samples

b. 200uL for 12 well plate samples

c. 100uL for 24 well plate samples

10. The next steps are TIMED, so leave 20 seconds between tubes to allow enough

time for the procedure

a. Add 50mM pNPP to a tube, vortex well to mix, and incubate in 37 oC

water bath for 10-60 minutes (this is a guideline only; monitor colour for

change to yellow)

- 100uL of pNPP for 6 well plate samples



b. Repeat step (a) for each tube, leaving 20 seconds between tubes

11.While waiting for reaction to complete, prepare pNP standard solutions

diluted in assay buffer

a. Standards of 0-500uM pNP

b Make 1 mL of 500uM pNP by mixing 50uL of 10 mM pNP, 450uL water, and

155

500 uL assay buffer

c. Make 1 mL of 250 mM pNP by 1:2 dilution of 500 uM pNP solution with assay buffer

(i.e., 500 uL of 500 uM pNP + 500 uL assay buffer)

d. Continue to make serial dilutions of 125, 62.5, 31.25, 15.625, and 0 uM pNP.

12. To stop the reaction, add 0.1N NaOH (500ul for 6 well plate samples, 250ul

for 12 well plate samples, 125ul for 24 well plate) to the first tube

incubated and vortex. Repeat for each tube, leaving 20 seconds between tubes

so that all incubation times are equal. Also add 500uL 0.1N NaOH to each of

the standards.

14. Read the absorbance at 405nm with a spectrophotometer or microplate reader.

15. Activity is expressed in units of mol pNP per minute. Remember to take into

account the dilution factor.

A.10. Indirect Immunostaining Protocol

Purpose: To visualize the expression of protein being investigated

Reagents:


10% Neutral Buffered Formalin (NBF)

- 100 mL formaldehyde (Sigma F1635)

- 900 mL distilled water

- 4g sodium phosphate monobasic (Sigma S8282)

- 6.5g sodium phosphate dibasic anhydrous (Sigma S0876)

0.1% Triton X-100 (Sigma T8532, diluted with deionized H2O)

3% Bovine serum albumin (BSA, Sigma A9647, reconstituted in PBS)

Antibody for the protein of interest (Primary antibody)

Fluorescent conjugated secondary antibody (e.g. if host of primary antibody is mouse, use

anti-mouse secondary antibody)

10% Serum from host of secondary antibody (diluted with PBS)

Permafluor anti-fade mounting medium

10 ug/mL Hoechst nuclear stain (Invitrogen H3570, diluted with 10% serum)

156

Equipment:

Coverglass

Humidification chamber (i.e. Gladware container lined with wet paper towels)

Fluorescence microscope

Procedures:

I. Fixation and permeabilization

1. Remove media

2. Rinse samples (2-3 times) with PBS

3. Fix samples with 10% NBF for 15-30 minutes (depending on the porosity of the samples)

at room temperature

4. Remove fixative and rinse twice with PBS for 5 minutes each

5. Permeabilize with 0.1% Triton X-100 for 5 minutes

6. Rinse twce with PBS for 5 minutes

7. Fixed and permeabilized cells can be stored in PBS at 4oC if necessary

II. Primary antibody staining

8. Block with 3% BSA for 20 minutes at 37oC

9. Dilute primary antibody with 3% BSA to working concentration

10. Remove blocking solution and apply primary antibody. Do the following if:

a. Cytospin samples - Use wax pen to mark circle around cell splatter and then apply

sufficient reagent to cover the cells

b. Culture on coverglass – Pipet a droplet of antibody on a clean paraffin surface (50

L for a 12mm coverglass). Place the coverglass with cells facing down, onto the

antibody droplet.

c. Collagen matrice samples – Stain directly in the wells. Do not remove the

constrained matrices from the wells.

11. Incubate 60 minutes at 37 oC or overnight at 4 oC.

III. Secondary antibody staining

12. Wash slides twice with PBS for 10 minutes

13. Make 10% serum/PBS using serum from host of secondary antibody

14. Dilute secondary antibody in 10% serum to working concentration

157

15. Block slides with 10% serum for 30 minutes at room temperature

16. Apply secondary antibody to slides for one hour at room temperature in humidification

chamber

17. Rinse with PBS for 5 minutes

18. Incubate with Hoechst diluted 1:1000 in PBS for 5 minutes to stain nuclei

19. Rinse with PBS for 5 minutes

20. Briefly rinse with distilled water to remove salts from PBS

21. Apply Permaflour mounting medium and mount coverslips with nail polish*

* Note: for collagen matrices, carefully remove the matrices from the wells. Pipet a droplet

of Permaflour mounting medium onto a microscope slide and then place the surface with

cells onto the mounting medium. Do not mount with coverslips.

Primary antibody Secondary antibody Antibody Dilution

factor Incubation

time Antibody Dilution

factor Incubation time

-SMA 1:100 1 hour @ 37oC or overnight @ 4oC

Anti-mouse

1:100 1 hour @ RT

Fluorescent-conjugated Phalloidin

1:100 1 hour @ 37oC or overnight @ 4oC

na na na

CNP 1:10 3 hour @ room temperature

Anti-goat 1:100 1 hour @ room temperature

Cbfa1/Runx2 1:100 3 hour @ room temperature

Anti-rabbit

1:100 1 hour @ room temperature

A.11. Western Blot Purpose: To quantify the expression of a protein of interest

Reagents:

Cold Sterile PBS with calcium chloride and magnesium chloride (Sigma P5655)

10x lysis buffer

100 mM Phenylmethanesulfonyl fluoride (Sigma P7626, reconstituted in 100% ethanol)

Micro BCA Protein Assay Reagent (Pierce 23235)

30% acrylamide and bis-crylamide(acry/bis)

Resolving gel buffer (1.5 M Tris-base)

158

- 18.15 g Tris-base (Sigma 77-86-1)

- 60 mL Deionized H2O

- Adjust pH with 6 N HCl to pH 8.8

- Bring volume to 100mL and store at 4oC

Stacking gel buffer

- 6 g Tris-base (Sigma 77-86-1)

- 60 mL Deionized H2O


- Bring volume to 100mL and store at 4oC

0.5% (wt/vol) and 10% (wt/vol) sodium dodecyl sulfate (SDS, Sigma L3771, reconstituted in

deionized H2O)

10% APS

TEMED (Sigma

10x running buffer

- 30.3 g Tris-base (Sigma 77-86-1)

- 144 g Glycine (Sigma G8898)

- Bring to 1 litre with deionized H2O, store at 4oC

1M Tris-base

- 12 g Tris-base (Sigma 77-86-1)


- Bring volume to 100 mL with deionized H2O, store at 4oC

5x Laemmli loading dye

- 1 M Tris-base

- 10 g SDS (Sigma L3771)

- 50 mL Glycerol

- 250 mg Bromophenol blue

- 20 mL -mercaptoethanol (Sigma M3148)

- Store in glass bottle at room temperature

Protein ladder (Fermentas SM0671)

Methanol (Sigma 179957)

Protein transfer buffer (** must be made one day in advance)

- 11.64 g Tris-base

159

- 5.86 g Glycine

- 20% Methanol

- 7.5 mL 10% SDS

- Bring the volume to 1 litre with deionized H2O and store at 4 oC

10x TBS

- Tris-base

- NaCl

- Bring the volume to 1 litre with deionized H2O and store at 4 oC

1% Bovine serum albumin (BSA Sigma A9647, reconstituted in 1x TBST buffer)

Amersham ECL Plus™ chemilumiscece western blotting detection reagent

Equipment:

Pre-chilled cell scrapers

Pre-chilled eppendorf tubes

96 well plate

Filter papers

Polyvinylidene fluoride (PVDF) transfer membranes (BioRad 1620177)

SNAP i.d.TM protein detection system

Western blotting film

Procedures:

I. Protein extraction

1. Prepare fresh 1x lysis buffer for each protein extraction. For 1 mL of 1x lysis buffer,

combine:

a. 100 L 10x lysis buffer

b. 10 L 100 mM PMSF

c. 890 L deionized H2O

2. Place culture on ice

3. Remove media

4. Rinse (2-3 times) with cold PBS

5. Add 1x lysis buffer

a. For 24 well plate, add 30 L per well

b. For 12 well plate, add 100 L per well

c. For 6 well plate, add 200 L per well

160

6. Remove cells from culture by using a pre-chilled cell scrape. Rinse cell scrape with cold

PBS between samples.

7. Transfer lysate to a pre-chilled eppendorf tubes

8. Incubate on ice and place on a shaker set at maximum speed for 30 minutes

9. Vortex the samples every 5 minutes

II. Quantification of protein concentration

10. Prepare diluted BSA albumin standards

Vial Volume of diluent Volume and source of BSA Final BSA concentration

A 450 L 50 L of stock (2 mg/mL) 200 g/mL

B 400 L 100 L of vial A dilution 40 g/mL

C 250 L 250 L of vial B dilution 20 g/mL

D 250 L 250 L of vial C dilution 10 g/mL

E 250 L 250 L of vial D dilution 5 g/mL

F 250 L 250 L of vial E dilution 2.5 g/mL

G 300 L 200 L of vial F dilution 1 g/mL

H 250 L 250 L of vial G dilution 0.5 g/mL

I 500 L -- 0 g/mL

11. Prepare BCA working reagent by mixing 25 parts of micro BCA reagent MA, 24 parts

Reagent MB with 1 part of Reagent MC (25:24:1, Reagent MA:MB:MC)

12. Pipette 50 L of each standard or unknown sample into a microplate well

13. Add 50 L of working reagent to each well and mix plate thoroughly on a plate shaker

for 30 seconds

14. Incubate at 37oC for 2 hours

15. Cool plate to room temperature

16. Measure the absorbance at or near 562 nm on a plate reader

17. Subtract the average 562 nm absorbance reading of the Blank standard replicates from

the 562 nm reading of all other individual standard and unknown sample replicates

18. Prepare a standard curve by plotting the average Blank-corrected 562 nm reading for

each BSA standard versus its concentration in g/mL. Use the standard curve to

161

determine the protein concentration of each unknown sample. The linear working range

is 2-40 g/mL.

III. Protein electrophoresis

19. Cast 10% resolving gel by combining the reagents in the following order:

a. 7.9 mL deionized H2O

b. 6.7 mL 30% acryl/bis

c. 5 mL resolving gel buffer

d. 200 L 10% SDS

e. 200 L of 10% APS

f. 8 L TEMED

g. Mix gently on ice and immediately pour into the gel cassette. Fill the gel cassette

up to 1 cm under the comb.

h. Overlay the gel with 0.5% SDS

i. Allow the gel to polymerize for 45-60 minutes

j. Rinse the gel completely with deionized H2O to remove SDS

k. Remove excess water with filter paper

20. Cast 5% stacking gel by combining:

a. 76.8 mL deionized H2O

b. 1.7 mL 30% acryl/bis

c. 1.25 mL stacking gel buffer

d. 100 L 10% SDS

e. 100 L of 10% APS

f. 10 L TEMED

g. Mix gently on ice and immediately pour on top of the polymerized resolving gel

h. Place gel comb into the cassette

i. Overlay the gel with 0.5% SDS

l. Allow the gel to polymerize for 45-60 minutes

m. Rinse the gel completely with deionized H2O to remove SDS

n. Wrap the gel with plastic wrap and then store in humidifying chamber at 4oC or

proceed to protein electrophoresis

21. Prepare 1x running buffer by combining:

162

a. 100 mL 10x running buffer

b. 10 mL 10% SDS

c. 890 mL Deionized H2O

d. Warm buffer to room temperature

22. Boil a beaker of water to 95%

23. Dilute samples to the desire concentration with 1x lysis buffer and 5x loading dye. Keep

samples on ice at all time. (e.g. For 60 g/ 40 L in a lane, combine X L of sample, 8

L of 5x loading dye and 40 L - 8 L - X L of 1x lysis buffer)

24. Pipette up and down to mix

25. Boil diluted samples at 95 oC for 5 minutes

26. Cool to room temperature and spin down the samples with a bench top centrifuge

27. Load the samples to the gel (i.e. 40 l per lane for a 1.5 mm gel)

28. Load 10 l of ladder to the gel

29. Run gel with running buffer for 1.5 hour at 170 V and 0.04 A

IV. Protein transfer to membrane

30. Prepare filter pads, filter papers and membrane for protein transfer

a. Filter pads:

i. Rinse with warm tap H2O

ii. Soak in distilled H2O for 5 minutes

iii. Soak in deionized H2O for 5 minutes

iv. Soak in cold transfer buffer for at least 30 minutes

b. Filter papers

i. Soak in distilled H2O for 5 minutes

ii. Soak in deionized H2O for 5 minutes

iii. Soak in cold transfer buffer for at least 30 minutes

c. PDVF membranes (*do not touch the membrane with gloves, use tweezers)

i. Soak in distilled H2O for 5 minutes

ii. Soak in deionized H2O for 5 minutes

iii. Soak in cold transfer buffer for at least 30 minutes

31. Assemble the transfer tank on ice with stir bar according the following diagram

32. Run transfer tank at 100 V, 0.35 A for 1hour and 5 minutes

163

33. In the mean while, prepare 1x TBST by combing:

a. 1 mL Tween 20

b. 100 mL 10x TBS buffer

c. Bring the volume to 1 litre with deionized H2O, store at 4oC

34. Upon completion of protein transfer, rinse membrane twice with 1x TBST

35. Membranes can be wrapped with plastic wrap and store at –20oC or proceed to

immunoblot

Figure A.1. Assembly of the protein transfer tank

V. Immunoblot

36. Rinse membrane twice with 1x TBST

37. Prepare the SNAP i.d.TM protein detection system accordingly to user’s manual

38. Prepare 1% (wt/vol) BSA with 1x TBST (i.e. the blocking agent, ~ 50 mL per membrane)

39. Dilute primary antibody with 1% BSA and store on ice (see dilution chart for details, ~3

mL per membrane)

40. Dilute secondary antibody with 1% BSA and store on ice (see dilution chart for details,

~3mL per membrane)

41. Place membranes onto SNAP i.d.TM protein detection system

42. Block membranes with 1% BSA for 20-30s

43. Add 3 mL of diluted primary antibody to each membrane and incubate for 10 minutes at

room temperature. At this point, prepare the developing machine (i.e. ensure there is

sufficient developer, fixer, water and turn on the machine) and warm the

chemiluminescence reagents to room temperature.

164

44. Wash membranes (3-5 times) with approximately 30 mL of 1x TBST. Turn on the

vacuum pump for 20-30 seconds in-between each wash to remove the reagents.

45. Add 3 mL of diluted secondary antibody to each membrane and incubate for 10 minutes

at room temperature

46. Wash membranes (3-5 times) with approximately 30 mL of 1x TBST. Turn on the

vacuum pump for 20-30 seconds in-between each wash to remove the reagents.

47. Remove the membranes to plastic tray and keep membranes in 1x TBST

VI. Chemiluminescence detection (* must be done in the dark)

48. For each membrane, mix 4 mL of Amersham ECL Plus™ Reagent A with 100 L of

Reagent B in the dark. Store the detection mixture in a tube wrapped with tin foil.

49. In the dark room, prepare three pieces of clean plastic wrap

50. Place membrane onto a clean plastic wrap

51. Pour 4 mL of detection mixture onto each membrane

52. Incubate for 1-5 minutes

53. Remove membrane with tweezers and remove excess detection reagents by tapping the

edge of the membrane on a paper towel

54. Place the membranes onto another clean plastic wrap

55. Wrap the membrane with the plastic wrap and fold the edges to prevent the membrane

from drying up

56. Place the wrapped membrane onto a cassette

57. Carefully place a piece of film on top. Expose the film what an appropriate length of

time.

58. Develop the film immediate

59. Repeat the procedure for optimal exposure time

60. Rinse (2-3 times) membranes with 1x TBST, wrap with plastic wrap and store at –80oC

(* membranes can be used again in the future with proper handling and storage)

VII. Image analysis

61. Scan the developed film

62. Load the image into Image J

63. Select “analyze” function

165

64. Select “gel” function

65. Set the location of each lane

66. Select “plot lanes”

67. Measure area of each lane by using the “wand” tool

68. Do the same with the housekeeping protein

69. Normalize protein expression with the housekeeping protein expression level

VIII. Antibody dilution chart

Primary antibody Secondary antibody

Name and supplier Dilution factor

Expected product size

(kDa)

Name and supplier

Dilution factor

GAPDH (Stressgen CSA-335E)

1:3000

36 HRP anti-mouse 1:3000

Akt (Cell signal #9272)

1:1000 60 HRP anti-rabbit 1:3000

Phosphorylated Akt (Cell signal #9271)


P38 (Cell signal #9212)


Phosphorylated P38 (Cell signal #9211)


-SMA (Sigma A2547)

1:10000 42 HRP anti-mouse 1:3000

166

A.12. Primer Sequences for PCR and qRT-PCR

Gene name and accession

number

Primer sequence Annealing temperature

( oC)

Product size (bp)

NPR-B DQ487044.1

Left primer: 5’-agcattaccgtaccctggtg-3’ Right primer: 5’-tagtgaggccggtcatcatgt-3’

60 142

CNP, M64758.1

Left primer: 5’-accgactccagca-3’ Right primer: 5’-ataaagtggccag-3’

60 103

Osteonectin, AW436132

Left primer: 5’-tccggatctttcctttgctttcta-3’ Right primer: 5’-ccttcacatcgtggcaagagtttg-3’

60 187

Osteocalcin, AW346755

Left primer: 5’-tcaaccccgactgcgacgag-3’ Right primer 5’-ttggagcagctgggatgatgg-3’

60 106

GAPDH, AF017079

Left primer: 5’-tgtaccaccaactgcttggc-3’ Right primer 5’-ggcatggactgtggtcatgag-3’

60 86

TGF-1 receptor I, AB182260.1

Left primer: 5’-gacggcattccagtgtttct-3’ Right primer: 5’-tgcacatacaaatggcctgt-3’

60 169

TGF-1 receptor II, EF396957.1

Left primer: 5’-cagggaagaacgttcatggt-3’ Right primer 5’-ccaaccaaagctgagtccat-3’

60 128

167

Appendix B B. Preliminary Data

B.1. The Effect of Statins on the Expression of CNP by VICs Objective: Statins are lipoprotein-lowering agents and are potential therapeutics for CAVD.

Intriguingly, statins display similar effects as CNP on the differentiation of VICs, inhibiting

myofibroblast120, 233 and osteoblast differentiation of VICs in vitro92. This prompts the question

of whether there exists a possible molecular association between CNP and statins, leading to

similar biological effects on VICs. We therefore evaluated the expression of CNP in VICs

treated with or without simvastatin as a means to provide a mechanistic explanation of statin-

mediated protective effects on VICs.

Methods: The effect of statin treatment on CNP expression by VICs was evaluated. Simvastatin

was activated prior to use by alkaline hydrolysis with NaOH and ethanol273. Cells were cultured

in complete media or calcifying media with (1 M) or without activated simvastatin for up to 14

days. Morphological changes were evaluated with bright field microscopy. RNA was extracted

after three days of treatment and qRT-PCR was performed with primers for CNP (Accession

number: M64768, forward primer: 5’-accgactccagca-3’ and reverse primer: 5’-ataaagtggccag-3’).

Transcriptional expression was quantified by the comparative Ct method as previously described.

Results: A three-day treatment with simvastatin significantly upregulated CNP transcript

expression in cells cultured in complete media (Figure B.1). Statin treatment had no detectable

effect on CNP transcript expression in cells cultured on calcifying media (Figure B.1), despite its

ability to inhibit aggregate formation after 14 days in culture (Figure B.2: A, B and D).

168

Figure B.1. Expression of CNP transcript after three days of simvastatin treatment

A relative gene expression level of less than one indicates lower exprssion with simvastatin

treatment relative to that without treatment. * P < 0.05.

Discussion: Similar to CNP, statins have recently been shown to suppress aggregate formation

that is associated with myofibroblast or osteoblast differentiation of VICs in vitro. We found that

expression of CNP by VICs cultured in calcifying media was not augmented by simvastatin,

despite the observed inhibition of calcific aggregate formation with simvastatin treatment. These

data suggest that statins might regulate osteogenic differentiation of VICs via a signaling

pathway independent of CNP. Statins could interfere with other signaling pathways that regulate

osteogenic differentiation of VICs. For instance, statins promote the breakdown of extracellular

ATP to adenosine and trigger signaling via the P1 purinergic receptor, which leads to the

inhibition of osteogenic differentiation of VICs in vitro73. In contrast, we identified an

upregulation of CNP expression by simvastatin, when VICs were cultured in conditions that

favoured myofibroblast differentiation. Although this result is intriguing, additional investigation

is necessary to further identify if CNP signaling mediates, in part, the anti-myofibrogenic effect

of statins. Whether CNP signaling pathway interacts with the HMG-CoA reductase pathway has

yet to be investigated, but statins have been shown to decrease aggregate formation associated

with myofibroblasts independent of the HMG-CoA reductase pathway120. While little is known

regarding the regulation of CNP production by VICs, CNP expression by endothelial cells has

been studied extensively274. It has been shown that oxidized LDL and its extracted lipids reduce

secretion of CNP by cultured vascular endothelial cells. Various in vitro studies suggest that

CNP could act to suppress the atherogenic activity of both oxidized LDL and the bioactive

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

Complete media Osteogenic media

Re

lativ

e fo

ld e

xpre

ssio

n(T

rea

ted

/un

tre

ate

d)

Re

lati

ve C

NP

exp

ress

ion

(A

U)

*

Complete media Calcifying media

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

Complete media Osteogenic media

Re

lativ

e fo

ld e

xpre

ssio

n(T

rea

ted

/un

tre

ate

d)

Re

lati

ve C

NP

exp

ress

ion

(A

U)

*


Re

lati

ve C

NP

exp

ress

ion

(A

U)

*


169

lysophospholipids275, 276. Lipophilic signaling molecules known to be associated with high-

density lipoprotein (HDL) such as sphingosine-1-phosphate have also been showed to suppress

CNP/NPR-B signaling277. Data from these studies suggest a close relation between lipoproteins

and CNP signaling, which should be investigated in the future in order to improve our

fundamental understanding of CNP and its involvement in valve cell biology.

B.2. Culturing Primary VICs on Polyacrylamide Substrates

Objective: An alternative cell culture system that can be fine-tuned to a wider range of stiffness

was implemented for culturing of VICs. Previously, we found poor adhesion of primary VICs to

polyacrylamide (PA) substrates coated with monomeric collagen. It is evident that VICs adhere

well on fibrillar type I collagen matrices, and therefore we tested if coating PA substrates with

thin fibrillar type I collagen matrices could improve the adhesivness of primary VICs to the PA

substrates.

Method: The process to fabricate the PA gels was described in the study by

Khatiwala et al132. Briefly, stock solutions of 40% acrylamide and 2% bis-acrylamide (Bio-Rad

Laboratories) were used. Different volumes of acrylamide and bis-acrylamide were mixed with

sterile de-ionized water and 10 M HEPES (pH 7.5). The ratio of acrylamide and bis-acrylamide

determines the stiffness of PA gels. Next, photoinitiator (10% ammonium persulfate; Bio-Rad

Laboratories) and radical stabilizer TEMED (Bio-Rad Laboratories) were added at 1/200 volume

and 1/2000 volume respectively. Immediately, the solution was syringe filtered using a 0.22 μm

filter. The mixture was pipetted onto an adhesive film and then covered with a surfasil-treated

top coverslip. The gels were polymerized in sterile biosafety cabinet for 10-15 minuties. Once

polymerized, the top cover slips were removed and the gels were surface functionalized with N-

sulfosucciniidyl-6-(4’-azido-2’-nitrophenylamino) hexanoate (sulfo-SANPAH; Pierce

Biotechnology), a UV-light sensitive heterobifunctional crosslinker. The photoinitiator was

activated through exposure to UV light (Blak-Ray; UVP) at 365 nm for 12 min. Once activated,

the gels were rinsed with PBS and then thin fibrillar type I collagen was placed on the surface of

the gels.

170

Figure B.2. Inhibition of aggregate formation by simavastatin treatment

(A and C) Few aggregates were formed in culture with statin treatment by day 9 and 14. (B and

D) In the absence of simvastatin, aggregates formed readily by day 9 and seemed to increase in

size by day 14. (E) Statin treatment significantly reduced aggregate formation by VICs cultured

in calcifying media in comparison to those of untreated samples. * P < 0.05.

020406080

100120140

Day 9 Day 14Treatment duration

Nu

mb

er

of a

gg

reg

ate

s Without simavastatin With simvastatin

A. B.

C. D.

Da

y 9

Da

y 1

4

E.

**

020406080

100120140


Nu

mb

er

of a

gg

reg

ate


A. B.

C. D.

Da

y 9

Da

y 1

4

E.

020406080

100120140


Nu

mb

er

of a

gg

reg

ate


A. B.

C. D.

Da

y 9

Da

y 1

4

E.

**

171

Results: The adhesion of VICs to PA substrates coated with fibrillar type I collage was

significantly improved compared to those of coated with monomeric collagen. We found that the

morphology varied when cells were cultured on PA substrates with different stiffnesses. When

primary VICs were cultured in calcifying media on fibrillar type I collagen-coated PA substrates

of various stiffnesses (11 kPa, 22 kPa, 50 kPa and 144 kPa), calcification by VICs was most

prominent on substrates with stiffness of 22 kPa and 50 kPa (Figure B.3). Cells also formed

aggregates on PA substrate with stiffnesses of 11 kPa and 144 kPa, but those aggregates stained

weakly for ARS (Figure B.3).

Discussion: By lining PA gels of different stiffnesses with thin type I collagen matrices, primary

VICs were able to adhere to and proliferate on these culture surfaces. Importantly, cells were

able to response to the stiffness of the substrate underlying the thin collagen matrices. With this

approach, the surface chemistry is decoupled from substrate mechanics. The total amount of

collagen is the same on all PA gels. We further observed, quantitatively, calcification by VICs

was most prominent on substrates with stiffness of 22 kPa and 50 kPa. Cell aggregates found on

the more compliant gels (11 kPa) stained weakly for ARS. Engler et al has shown that bone

marrow-derived MSCs underwent osteogenic differentiation perferentially on substrates with

stiffness range of 25 – 40 kPa130. Further experiments should be performed to determine if VICs

respond to substrate stiffness in a similar manner as MSCs, in which calcification occurs via

osteogenic differentiation on substrates with stiffness of 22 kPa and 50 kPa. As described in

Chapter Five, compliant collagen matrices (E ~ 30 kPa) promote osteogenic-dependent

calcification by VICs, it is expected that VICs cultured on substrates with stiffness of 22 kPa and

50 kPa would also likely calcify via osteogenic differentiation. The preliminary data support the

use of PA gels to test the matrix stiffness effects on mediating the response of primary VICs.

This experimental technique will be directly applicable to the suggested future work desapcribed

in Chapter Eight.

172

Figure B.3. Calcification by primary VICs on PA substrates with different stiffnesses

ARS staining for calcium was more intense in aggregates found on substrates with stiffness of 22

kPa and 50 kPa.

B.3. Isolation of VICs from Mouse Aortic Valve

Objective: In vitro data showed that CNP protected against pathological differentiation of VICs

into myofibroblasts and osteoblasts. In vivo testing is necessary to determine the effectiveness of

CNP in protecting against CAVD and also to identify possible molecular mechanisms. One

possible in vivo model is to use genetically modified mice to evaluate the role of CNP signaling

components in regulating the progression of valve calcification. For example, long bone

abnormality (lbab-/-) mice can be used to study if the anti-calcific effect of CNP depends on the

activation of NPR-B. These mice have a single point mutation that converts an arginine to a

glycine in a conserved coding region of the CNP gene, reducing the ability of CNP to bind to

NPR-B278. In order to use mice as models of CAVD, it is important to first establish methods to

dissect aortic valves from mice. Further, it may be useful to isolate VICs from genetically

modified mice for furture molecular studies.

Methods: Aortic valves from 8-11 weeks old mice were dissected under a stereomicroscope.

Briefly, the heart was removed from the mouse and placed in sterile PBS (Figure B.4). The aorta

was cut open carefully and the aortic valve leaflets were removed and placed in sterile PBS. The

valve leaflets were paraffin embedded, sectioned and stained with Masson trichrome for the

evaluation of valve matrix organization and composition. Individual valve leaflets were placed

on collagen-coated TCPS, which preferentially leads to the explant of VICs, and cultured in

complete media.

173

Results: Mouse aortic valve leaflets were successfully dissected. Histological analysis showed

VICs permeate the valve matrix. Masson trichrome staining showed abundant collagen in the

ventricular side of the mouse AV (Green, Figure B.5: A). The explant method using collagen-

coated TCPS successfully isolated VICs from the mouse leaflets (Figure B.5: B). The explanted

cells stained positive for - SMA (Green, Figure B.5: C), consistent with the phenotype of VICs

from pigs and humans when cultured on a stiff substrate.

Figure B.4. Isolation of mouse aortic valve

The size of a mouse heart is approximately 10 mm in length. In order to dissect the aortic valve,

the aortic arch was first identified (labeled AA) and the aorta was cut open. The aortic valve

(labeled AV) can be found between the aorta and the left ventricle.

Discussion: It is possible to dissect individual aortic valve leaflets from mice and to isolate VICs

by using an explant method. It is, therefore, feasible to use mouse cells in vitro to study the

molecular mechanisms responsible for CNP signaling in CAVD.

174

Figure B.5. Mouse aortic valve and VICs

(A) Masson trichrome histochemical staining of a paraffin-embedded mouse aortic valve,

showing the transverse section of the leaflet. (B) Mouse aortic VICs were isolated by an explant

method on collagen-coated TCPS to enrich for VICs over VECs, which were previously shown

not to adhere well to type I collagen51. (C) On stiff TCPS, the mouse VICs express -SMA stress

fibres (green), consistent with the phenotype of VICs from pigs and humans.

A.

B. C.

A.

B. C.

A.

B. C.

Aortic side

Ventricular side

pathology of calcific aortic valve disease: the role of … · 2013. 12. 18. · calcific aortic...

Documents