mark wan thesis the mechanism of discoidin …...the mechanism of discoidin domain receptor 1...
TRANSCRIPT
The Mechanism of Discoidin Domain Receptor 1 Mediated Vascular Calcification
By
Mark H. Wan
A thesis submitted in conformity with the requirements for the degree of Master of Science
Graduate Department of Laboratory Medicine and Pathobiology University of Toronto
© Copyrighted by Mark H. Wan 2013
ii
The Mechanism of Discoidin Domain Receptor 1 Mediated Vascular Calcification
Mark H. Wan
Master of Science
Graduate Department of Laboratory Medicine and Pathobiology University of Toronto
2013
ABSTRACT
Introduction: Activation of Runt Related Transcription Factor 2 (RUNX2) is required for
transdifferentiation of Vascular Smooth Muscle Cells (VSMCs) into a calcifying osteoblast‐like
phenotype. Our lab showed that deletion of Discoidin Domain Receptor 1 (Ddr1), decreased
atherosclerotic vascular calcification in the Ldlr‐/‐ mouse.
Hypothesis: DDR1 regulates RUNX2 activity by affecting microtubule organization during VSMC
mediated calcification.
Results: Ddr1‐/‐ VSMCs show reduced RUNX2 activity when compared to Ddr1+/+ VSMCs.
Addition of the microtubule‐destabilizing agent nocodazole inhibited both RUNX2 activity and
nuclear localization in Ddr1+/+ VSMCs. Addition of the microtubule‐stabilizing agent taxol
rescued RUNX2 nuclear localization in Ddr1‐/‐ VSMCs. Despite this, Taxol was unable to rescue
RUNX2 activity as it eliminated activity in both genotypes.
Conclusion: These findings indicate that under osteogenic conditions, Ddr1 deletion impedes
the dynamic instability required for the maintenance of microtubule architecture. This prevents
RUNX2 nuclear localization and transcriptional activation in VSMCs.
iii
DEDICATIONS
“Affection without sentiment, authority without cruelty, discipline without aggression, humor without ridicule, sacrifice without obligation, companionship without possessiveness."
‐ William E. Blatz
For my mother Lily and my father Jeremy, who truly embody everything it means to be
great parents. Since the very beginning they poured their hearts and souls into me, and their only wish is to see me happy. I could not have asked for better parents. They have always been my biggest fans and most fervent supporters. It is to both of them that I dedicate my thesis.
iv
ACKNOWLEDGEMENTS The accomplishment of this thesis was not my work alone, but only possible through the mentorship, support and help from many good people. In this section I will try to adequately express gratitude to each of them. Firstly, my work and this thesis would not be possible without my supervisor, Dr. Michelle Bendeck. Not only has she provided me with great ideas to drive my work, but has always driven me to accomplish my best with much enthusiasm and patience. Michelle mentored me with utmost dedication and often going above and beyond what was expected of a supervisor to ensure I received the most out of my training. I could not thank her enough for all the opportunities, countless documents she has edited and the protocols she has read through. Many thanks go out to Dr. James Eubanks who decided to take on the incredibly difficult task of mentoring an enthusiastic second year undergrad with no research experience. His tutelage has set the foundation for every subsequent research project I have taken on. I am also truly thankful to Dr. Scott Heximer for first providing me the opportunity to work in his lab as a summer student and a project student. He has always welcomed my many visits and chats and later served on my committee and provided much excellent advice crucial to this project. Another person who deserves thanks is Dr. Steffen‐Sebastian Bolz. While I’ve never been his student in any official capacity, he has often taken an altruistic interest both in my career and development as a person. His constant advice and encouragement are much to be thankful for. I am grateful for the advice I have received from my committee members Drs. Rita Kandel and Craig Simmons to complete this project. Dr. Simmons also helped to provide me with the many opportunities in the MATCH program which has certainly helped my graduate training. This work also would not have been possible without my friends and colleagues both past and present from the Bendeck lab. I cannot thank all of you enough for making this journey fun and for all the hours you all put in to help shape my training. Particular thanks go out to Dr. Guangpei Hou, whose experimental advice and technical prowess has been crucial to the completion of this project. Thanks also go out to Dr. Antonio Rocca, whose job was not to give any experimental help, answers to questions or mentorship, but gave them often and selflessly. Finally, this work would not have been possible without Drs. Cecilia Giachelli and Mei Speer who helped to troubleshoot this project and provide us with our platform of investigation. Thanks also go out to Dr. Renny Franceschi who provided us with the RUNX2 activity vector for my project.
v
TABLE OF CONTENTS
ABSTRACT DEDICATIONS
ACKNOWLEDGEMENTS iv
TABLE OF CONTENTS v
LIST OF ABBREVIATIONS viii
LIST OF FIGURES x
INTRODUCTION 1
ATHEROSCLEROSIS 1
i Atherosclerosis
ii The Role of Collagens in Atherosclerosis 2
TYPE I COLLAGEN RECEPTORS 3
i Integrins 4
ii Discoidin Domain Receptors 4
iii DDR Signalling 5 iv DDR Function In Smooth Muscle Cells 7 v DDR1 in Atherosclerosis and Calcification iv DDR2 in Bone Mineralization and Arthritis 10 VASCULAR CALCIFICATION 11
i Types of Vascular Calcification 11
ii Models of Vascular Calcification 13
iii Mechanism of Vascular Calcification 15
iv The Role of Phosphate in Vascular Calcification 18
RUNX2 20
i RUNX2 Structure and Function 20
ii RUNX2 Transcriptional Targets 21
iii Regulation of RUNX2 22
iv RUNX2 in Vascular Calcification 23
8
iii
ii
vi
MICROTUBULES 24
i Structure and Organization of Microtubules 24
ii Regulation of Microtubule Dynamics 25
iii Microtubules and Nuclear Transport 26
RATIONALE AND HYPOTHESIS 28
MATERIALS AND METHODS 29
Chemical and Reagents 29
Smooth Muscle Cell Isolation and Culturing 29
In Vitro Calcification Assay 30
Calcium Quantification 31
Reverse Transcription and Quantitative Real Time PCR 33
Immunoblotting 34
Immunocytochemistry 36
Immunofluorescent Quantification 37
Plasmid Preparation and Transformation 37
Transfection 38
Luciferase Assay 39
Statistical Analyses 39
RESULTS 40
i DDR1 Deficiency Attenuates Calcification Potential of Vascular Smooth Muscle Cells 40 ii DDR1 Deficiency Alters Expression of Phosphate Handling Genes in Vascular Smooth Muscle Cells at an Early, but not Late Stage of Calcification 40 iii DDR1 Deficiency Alters the Expression of TenC at Late Stage Calcification, but not αSMA and Atf4 42 iv DDR1 Deficiency Results in Reduced RUNX2 Activity and Osteocalcin Expression in Vascular Smooth Muscle Cells 43 v P38 Inhibition Reduces RUNX2 Activity and Osteocalcin mRNA Expression, but not RUNX2 Nuclear Localization in DDR1 Expressing Vascular Smooth Muscle Cells 45
vi DDR1 Deficiency Impairs phosphoP38 Nuclear Localization 46 DDR1 Regulates RUNX2 Nuclear Localization and Activity by Modulating a Dynamically Unstable Microtubule Network in Vascular Smooth Muscle Cells 46
vii
DISCUSSION 49
i DDR1 in Matrix Calcification 49
ii Phosphate Handling in DDR1 Mediated Calcification 50
iii Smooth Muscle Cell Phenotype in DDR1 Mediated Calcification 51
iv DDR1 Mediated RUNX2 Activation 54
v Microtubules in DDR1 Mediated RUNX2 Activity 55
vi Future Directions 58
vii Conclusions 61
FIGURES REFERENCES
86
62
viii
LIST OF ABBREVIATIONS αSMA – Alpha Smooth Muscle Actin ADP – Adenosine Diphosphosphate ALP – Alkaline Phosphatase AMP – Adenosine Monophosphate APOE – Apolipoprotein E ARP – Acidic Ribsomal Protein ATP – Adenosine Triphosphate ATF4 – Activated Transcription Factor 4 BMP2 – Bone Morphogenic Protein 2 BSA – Bovine Serum Albumin CD40 – Clusters of Differentiation 40 cDNA – Complementary DNA CKD – Chronic Kidney Disease COL1A1 – Type 1 Collagen Alpha 1 COL1A2 – Type 1 Collagen Alpha 2 COX ‐ Cyclooxygenase Ct – Cycle Threshold DDR – Discoidin Domain Receptor DMSO – Dimethyl Sulfoxide DNA – Deoxyribo Nucleic Acid EC – Endothelial Cells ESRD – End Stage Renal Disease ER – Endoplasmic Reticulum ERK‐ Extracellular Related Kinase FAK – Focal Adhesion Kinase FGFR1 – Fibroblast Growth Factor Receptor 1 GDP – Guanosine Diphosphate GTP – Guanosine Triphosphate HDAC – Histone Deacetylase HIF1α – Hypoxia Inducible Factor 1 Alpha IBSP – Integrin Binding Bone Sialoprotein IGF‐1 – Insulin‐like Growth Factor 1 IL – Interleukin LDL – Low Density Lipoprotein LDLR – Low Density Lipoprotein Receptor Mθ – Macrophage MAP – Non Motor Microtubule Associated Protein MAPK – Mitogen Activated Protein Kinase MEKi – MEK Inhibition MGP – Matrix GLA Protein MMP – Metallomatrix Proteases MTOC – Microtubule Organizing Center
ix
NFκB – Nuclear Factor Kappa B NO – Nitric Oxide iNOS – Inflammatory Nitric Oxide Synthase NPP1 ‐ Neuroecto Pyrophosphatase 1 OCN – Osteocalcin OCT4 – Octamer Binding Transcription Factor 4 ON ‐ Osteonectin OPN – Osteopontin OSE – Osteoblast Sensitive Element OxLDL – Oxidized Low Density Lipoprotein P38i – P38 Inhibition PBS – Phosphate Buffered Saline PBST – Phosphate Buffered Saline Tween 20 PCR – Polymerase Chain Reaction PFAv‐ Paraformaldehyde Pi – Inogranic Phosphate PiT – Sodium Dependant Phosphate Co‐Transporter PPi – Pyrophosphate PTH – Parathyroid Hormone qRT‐PCR – Quantitative Real Time PCR RBP – Retinoblastoma Protein RNA – Ribo Nucleic Acid RLU – Relative Luminescence Unit ROS – Reactive Oxidative Species RTK – Receptor Tyrosine Kinase RUNX – Runt Related Transcription Factor SHCA – Src Homology 2 Domain Containing A SMED ‐ Spondylo‐Meta‐Epiphyseal Dysplasia SOX9 – Sry Box 9 SSEA1 – Stage Specific Embryonic Antigen 1 STAT1‐ Signal Transducer and Activator of Transcription 1 TBST – Tris Buffered Saline Tween 20 TENC – Tenascin C TGFβ – Transforming Growth Factor Beta TNFα – Tumour Necrosis Factor Alpha VSMC – Vascular Smooth Muscle Cell VCAM – Vascular Cell Adhesion Molecule VEGF – Vascular Endothelial Growth Factor
x
LIST OF FIGURES Figure 1: Diagram of Normal Artery and Artery Affected by Atherosclerosis
Figure 2: Diagram of DDR1 Isoforms in Humans and Mice
Figure 3: Diagram of Molecules Which Regulate Pi and PPi Levels
Figure 4: Diagram of RUNX2 Isoforms in Humans and Mice
Figure 5: Diagram of RUNX2 Stimulatory and Inhibitory Phosphoserine Residues Figure 6: Diagram of Microtubule Dynamics Figure 7: Ddr1‐/‐ VSMCs Show Decreased Calcifying Potential Figure 8: Ddr1‐/‐ VSMCs Show Increased Pit‐1 mRNA Expression and Decreased Alp and Npp1 mRNA Expression at Early but not Late Stage Calcification Figure 9: Ddr1‐/‐ VSMC Show no Change in αSMA Protein and Atf4 mRNA Expression, but Show Increased TenC mRNA Expression at Late but Not Early Stage Calcification Figure 10: Ddr1‐/‐ VSMCs Show No Changes in RUNX2 Protein and mRNA Expression Figure 11: Ddr1‐/‐ VSMCs Show Reduced RUNX2 Activity and Ocn mRNA Expression Figure 12: Ddr1‐/‐ VSMCs Show Reduced RUNX2 Nuclear Localization Figure 13: RUNX2 Activity and mRNA Levels of Osteocalcin are Sensitive to P38i in Ddr1+/+
VSMCs Figure 14: Nuclear Localization of RUNX2 in Ddr1+/+ and Ddr1‐/‐ VSMCs Was Not Affected by MEKi or P38i Figure 15: Phospho ERK1/2 Signalling is Decreased in Ddr1‐/‐ VSMCs at Late but Not Early Stage Calcification Figure 16: PhosphoP38 Signalling is Decreased in Ddr1‐/‐ VSMCs at Late but Not Early Stage Calcification Figure 17: Ddr1‐/‐ VSMCs Show Reduced PhosphoP38 Nuclear Localization Figure 18: Ddr1‐/‐ VSMCs Show Increased Cell Spreading and Reduced α Tubulin Polymerization
xi
Figure 19: Ddr1‐/‐ VSMCs Show Reduced RUNX2 Co‐Localization with Microtubules Figure 20: Taxol Increased RUNX2 Nuclear Localization in Ddr1+/+ and Ddr1‐/‐ VSMCs While Nocodazole, Prevented RUNX2 Nuclear Localization in Ddr1+/+ VSMCs Figure 21: Nocodazole or Taxol Attenuated RUNX2 Activity in Ddr1+/+ and Ddr1‐/‐ VSMCs but Increased Ocn mRNA Expression Figure 22: Proposed model of DDR1 regulation of RUNX2 activity in VSMC
1
INTRODUCTION
ATHEROSCLEROSIS
Atherosclerosis begins with dysfunction of the endothelial layer, which leads to the
increased permeability of the endothelial cells (EC)1. This leads to the accumulation of low
density lipoprotein (LDL) underneath the EC monolayer of the arterial wall, forming a fatty
streak2. This results in activation of the endothelial layer, which recruits monocytes3. Upon
egression, monocytes, differentiate into macrophages (Mθs)4. After the Mθs have entered the
subendothelial space, they phagocytose low density lipoprotein (LDL) and oxidize LDL to
produce Oxidized LDLs (oxLDLs) 5. These cells thus become Mθ derived foam cells. In addition,
they also promote Vascular Smooth Muscle Cells (VSMCs) to migrate and proliferate from the
medial wall into the subendothelial space6(Figure 1). During this process, Mθs and VSMCs will
actively synthesize matrix proteins such as fibrillar collagens7. These extracellular proteins serve
many roles in atherosclerosis such as migratory lattices for the cells and also forming a fibrous
cap which prevents plaque from rupturing7.
As the lesion grows in size, the vessel lumen attempts to accommodate the bulging
plaque through eccentric grow, if the lesion growth exceeds that of vessel growth, the vessel
becomes increasingly stenosed or occluded at which point the blood can no longer perfuse
tissues distal to the blockage8. In a vulnerable plaque, matrix degrading enzymes along with the
hemodynamic stresses from systole and diastole increasing the probability of rupture9.
The probability of rupture is increased with the development of atherosclerotic or
intimal calcification10, which develops in advanced plaques. This process begins with the
development of a necrotic core as result of dead cell accumulation which is susceptible to the
2
initiation of calcification11. As calcification occurs, the plaque is destabilized from the increased
focal stress points10. Once rupture occurs, the clotting cascade initiates12. The result is a
thrombus overlaying the lesion and occluding the vessel12. If the clot travels downstream it
becomes an embolus which can block the blood supply to other tissues12. Depending on the
location of the block, this can either lead to an infarction, stroke or a myocardial infarction12.
The Role of Collagens in Atherosclerosis:
Collagens are expressed by both VSMCs and Mθs during atherosclerosis and are the
main protein constituent of atherosclerotic plaques, comprising ~60% of the total protein13.
These molecules are composed of three alpha helical chains containing Gly‐X‐Y repeats, where
X and Y are usually proline or hydroxyproline residues14. Type IV collagens line the basement
membrane of ECs and VSMCs in normal vessels15. Upon plaque development, the expression of
Type IV collagen increases and it serves as an attachment and migration substrate for both Mθs
and VSMCs15. Other collagens such as types I and III fibrillar collagens, provide tensile strength
to the plaque15. They are expressed in low amounts in the early plaque, but dramatically
increase in expression driving the advanced stages of plaque development15. Likewise Types V
and VI collagens also increase in expression with plaque age, however their role within the
plaque is largely unknown15. Type VIII collagen is also highly upregulated during atherosclerotic
development16. This was later found to responsible for the growth and migration of VSMCS and
thus is a critical component of the injury process16.
The importance of collagens to atherosclerosis was first hinted in the diverse roles it
played in VSMCs and Mθs. Studies showed that inhibitors of collagen synthesis attenuate VSMC
3
migration, which indicated that VSMC required collagen for migration17. Indeed migration
assays which were later performed on VSMCs and Mθs revealed that Types I18 and VIII19
collagens are able to induce migration. While adherence to collagens by VSMCs is required for
migration, collagenase activity is also required for the process. Collagenases are required to
release focal adhesion matrix contacts and allow cell translocation17. Degraded type I collagen
serves as a chemoattractant for migrating VSMCs and Mθs20. In addition, type I collagen also
induces monocyte maturation. Experiments which plated monocytes on type I collagen
increased cell spreading and increased expression of cell adhesion molecules21. Together these
studies underscore the importance of collagens, and in particular type I collagen in the biology
of VSMCs and Mθs, which are major cell types in atherosclerosis.
Type I collagen is also critical to the maintenance of the structural integrity of an
atherosclerotic plaque. Type I collagen, along with type III collagen, form a fibrous cap which
provides plaque rupture20. As a result, plaque instability is induced by degrading type I collagen
fibers at the shoulder regions of the atherosclerotic plaque22‐24. While type I collagen plays an
essential role in stabilizing the plaque, studies on the type I collagenase MMP1 showed that its
expression reduced atherosclerotic burden25. Furthermore, type I collagen may also contribute
to plaque rupture, as it likely plays a major role in atherosclerotic calcification.
TYPE I COLLAGEN RECEPTORS:
Cell signalling by type I collagen is mediated by cell surface receptors. There are two
main families of type I collagen receptors, the integrins: α1, α2 and β1 and the Discoidin
Domain Receptors (DDRs).
4
Integrins:
Integrins are a family of glycoproteins which form heterodimers consisting of one α and
one β subunit. In addition to binding type I collagen, integrins are also involved in assembling
secreted type I collagen monomers into fibrils26. Upon binding to type I collagen molecules,
integrins are able to signal using intracellular signalling pathways. Since integrins do not contain
any kinase activity, they will either interact with intracellular kinases such as SRC, Integrin
Linked Kinases (ILK) or Focal Adhesion Kinase (FAK)27. Crosstalk with Receptor Tyrosine Kinases
(RTKs) is also another method integrins use to elicit signalling27. Type I collagen signalling
through integrins has been linked to many atherosclerotic processes such as mediating
leukocyte adhesions, foam cell formation and smooth muscle cell proliferation28.
Discoidin Domain Receptors:
In atherosclerosis, VSMCs and Mθs express DDRs, a class of collagen receptors29. DDRs
are a group of RTKs characterized by an extracellular discoidin domain, which mediates binding
to triple helical collagens30. The discoidin domain was first sequenced on the protein discoidin
1, which is a lectin found on the slime mold Dictostelium Discoidium31. Since its first discovery,
the discoidin domain has later been identified in blood coagulation factors V and VIII and also
neuropilins 1 and 2 which are involved in axonal guidance and angiogenesis32.
DDRs were first discovered during a search for RTKs that were highly expressed in
malignant cancer cell lines30. To date two DDR genes have been identified in humans and mice,
DDR1 and DDR230. DDR1 contains five splice variants which are widely expressed named a‐e,
5
and an additional splice variant found in rat testes30 (Figure 2). Conversely, DDR2 has only one
form30.
Although there is considerable overlap between the ligands of the DDRs, each receptor
is able to bind its own repertoire of triple helical collagens. DDR1 binds types I‐V and VIII
collagens, while DDR2 binds types I‐III and X collagen30.
DDR Signalling:
DDR1 phosphorylation leads to downstream signalling and activation of cellular
pathways. The DDR proteins facilitate binding through an N terminal discoidin domain which is
attached to the extracellular stalk domain, a transmembrane domain and a cytoplasmic
tyrosine kinase domain at the C Terminus30. Upon binding triple helical collagen and
dimerization, the intracellular kinase domains autotransphosphorylate each other and initiate
signalling30. In comparison to most RTKs, DDR1 have delayed phosphorylation kinetics.
Depending on the cell type, maximum phosphorylation can occur at an hour to several hours
after initial stimulation with collagen33. This is a process that is dependent on cell adherence, as
cells grown in suspension typically show peak phosphorylation at much earlier time point than
cells grown on a monolayer34. Similarly, DDR2 receptor phosphorylation kinetics are also slow
and occur in a process which is not well understood35.
Studies also show that DDR1 crosstalks with integrin, G‐protein coupled receptors and
WNT signalling pathways in order to fine tune specific cell responses to ligand activation36, 37. A
recent proteomic study determined that DDR2 may cross talk with Insulin Receptor38. Ligand
6
activation of DDRs are also cell type dependent and is likely the result of differences in
intracellular adaptors.
DDR1 has been shown to interact with several adaptor proteins such as Src homology 2
containing protein A (SHCA)39, NCK2, and CAS which are recruited to DDR1 upon
phosphorylation to facilitate signalling40. However, not all adaptors are recruited upon DDR1
activation. Adaptors such as WW‐domain containing protein 141 and Protein phosphatase 1
regulatory subunit42 dissociate from DDR1 upon activation, followed by the activation of
Mitogen Activated Protein Kinase (MAPK) ERK signalling pathways in HEK 293 cells upon
stimulation with type IV collagen. DDR1 regulation of the MAPK Extracellular Related Kinase
1/2 (ERK1/2) has been identified with many DDR1 processes. DDR1 signalling is important in
regulation of mouse embryonic stem cells in an undifferentiated state by type I collagen43. In
this study, the presence of DDR1 was responsible for maintaining the expression of NANOG,
Octamer Binding Transciption Factor 4(OCT4) and Stage Specific Embryonic Antigen 1 (SSEA‐1).
This process is dependent on DDR1 signalling through ERK1/2. DDR1 –ERK1/2 signalling is also
required for the upregulation of MMP10 expression induced by type I collagen in human lung
fibroblasts44. Furthermore, our own laboratory discovered that DDR1 phosphorylates ERK1/2
upon stimulation with type I collagen in VSMCs18. ERK1/2 has also been linked as a downstream
component of DDR2 signalling in RUNX2 activation45. In a study on mouse osteoblasts, the
authors showed that DDR2 null cells showed decreased calcification, Ocn expression and
RUNX2 activity46. P38 another MAPK has also been linked to DDR1 signalling in various cells
types. In mouse microglia, type I collagen induced DDR1 signalling through P38 is required for
inflammatory induction of Clusters of Differentiaation 40 (CD40), Cyclooxygenase 2 (COX2) and
7
MMP947. DDR1 is also important in the P38 induction of Nitric Oxide (NO) production in murine
macrophages through the upregulation of Inflammatory Nitric Oxide Synthase (iNOS)48. The
inflammatory role of DDR1‐P38 signalling is further explored in a study demonstrating that it is
crucial for the activation of Nuclear Factor Kappa B (NFκB) in mouse monocytes49. As ERK and
P38 signalling are fundamentally important to functioning of many different pathways in many
different cell types, DDR1 signalling may contribute to a number of key processes in normal
function and in pathology.
DDR Function in Smooth Muscle Cells
Our laboratory discovered that, denudation of DDR deficient mice showed reduced
intimal thickening compared to WT mice50. This suggests an important role of DDR1 in vascular
injury. Later it was found that both DDR1 and DDR2 are present in the atherosclerotic plaques
of humans and primates51. These studies implied an important role of DDR1 in VSMCs. Indeed,
VSMCs were later found to express a high level of DDR1 in vivo and in vitro. In DDR1 KO VSMCs,
migration, proliferation and adhesion were significantly decreased52. In addition, KO VSMCs
also showed decreased expression of proatherogenic MMPs 2 and 952. These deficiencies in the
DDR1 KO cells were later found to be rescued with DDR1 transfection. These studies all
suggested that DDR1 played an important role in VSMCs during atherosclerosis.
Studies on DDR2 in VSMCs by others have demonstrated that hypoxia induces the P38
signalling pathway to promote MYC and Myc Associated factor X binding to DDR2 promoter53.
The process increases the expression of DDR2 and increases the migration of VSMCs. Another
study furthering the understanding of this mechanism showed that TNFα also contributes to
8
DDR2 upregulation during VSMC migration54. The process by which DDR2 facilitates migration
in VSMCs is revealed to be dependent on MMP2 activity. In support of DDR2s migratory role,
another study also showed that DDR2 functions to reduce Focal Adhesion Kinase activity in
VSMCs55. Despite these identified migratory roles of DDR2, similar studies carried out to
investigate the role of DDR2 in VSMCs by our laboratory showed that deletion of Ddr2 did not
affect cell migration, proliferation or migration in VSMCs56. These studies may reflect the
differences of the model system used to study DDR2 function. Our lab studied DDR2 function
through the use of carotid VSMCs from DDR2 KO mice, while the mentioned studied used
VSMCs from rat aorta. Furthermore, the aforementioned studies performed migratory
experiments in hypoxia whilst our study was carried out under normoxia.
DDR1 in Atherosclerosis and Calcification:
Studies by our laboratory have demonstrated that DDR1 has important and different
functions in VSMCs and monocytes in atherosclerosis. DDR1 expression and function was
investigated in the Ldlr‐/‐ model of atherosclerosis 57. Ddr1‐/‐; Ldlr‐/‐ mice exhibited smaller
plaques that were rich in fibrous proteins such as fibrillar collagen and elastin compared with
Ddr1+/+; Ldlr‐/‐ mice 29. These studies also revealed that the increase in fibrous proteins was due
to increased synthesis of matrix molecules such as type I collagen by VSMCs. It was revealed
that Ddr1‐/‐; Ldlr‐/‐ mice had decreased decrease in the expression of Monocyte Chemotactic
Protein 1 (MCP‐1) and Vascular Cell Adhesion Molecule 1 (VCAM‐1). Furthermore, this study
also indicated that Ddr1‐/‐; Ldlr‐/‐ Mθs showed decreased mRNA expression of MMP2, 8, 13 and
14. These results suggested that DDR1 deficiency in Mθs prevented plaque invasion.
9
A later study by our laboratory confirmed that Mθs deficient in DDR1 have reduced
invasion into the plaque as a result of decreased attachment to type IV collagen and response
to Monocyte Chemoattractant Protein 1 (MCP‐1)58. Overall these experiments showed that
Mθs deficient in DDR1 reduced atherogenesis. While the function of DDR1 indicates that it
functions to increase attachment and plaque invasion in Mθs, studies carried out by our lab
demonstrate other functions in VSMCs. A later study carried out by our lab demonstrated that
DDR1 deficiency in the vessel increases both matrix production and plaque size29. This is
presumed to be carried out by VSMCs which show increased synthesis of procollagen I mRNA
and also had a net increase of MMP and net decrease of TIMP mRNA production in the DDR1
deficient VSMCs, which allowed for increased matrix production and remodelling in the plaque.
Taken together these studies indicate that DDR1 plays important roles in modulating the
infiltration and inflammation of Mθs and suppressing matrix remodelling and synthesis in
VSMCs.
Another hallmark of atherosclerosis which was discovered in the course of studies on
DDR1 in atherosclerosis showed that Ddr1+/+; Ldlr‐/‐ mice had indications of increased markers
of chondrogenic differentiation, Sry Box 9 (SOX9) and type X collagen, along with increased
propensity to calcification of the atherosclerotic plaques. By contrast, Ddr1‐/‐; Ldlr‐/‐ mice
showed decreased plaque calcification59. Together, these studies indicated that DDR1 plays an
important functional role in atherogenesis and the development of atherosclerotic calcification.
10
DDR2 in Bone Mineralization and Arthritis:
DDR2 has been implicated in several studies related to bone development and
osteoarthritis. The role of DDR2 in bone biology was first hinted when a screen of RTKs of
human osteoarthritis chondrocytes which showed an upregulation of the DDR2 mRNA
transcript when compared to un‐diseased chondrocytes60. Indeed later studies showed that
DDR2 expression is increased in the cartilage joints of the collagen type IX and XI deficient
mouse models of osteoarthritis61. It was also found in these studies that DDR2 when bound to
type II collagen acted to induce the expression of MMP13 in these chondrocytes, which is
hypothesized to contribute to the mechanism of osteoarthritis62. Further evidence
demonstrating the importance of DDR2 in chondrocyte biology is revealed that upon
stimulation with type II collagen, DDR2 was able to upregulate the mRNA expression of the
proinflammatory cytokine IL‐660. This process was dependant on DDR2 signalling through ERK,
P38 and Nuclear Factor κ B (NFκB).
The consequence of DDR2 mutation was also identified with a clinical manifestation in
calcification abnormalities. Spondylo‐meta‐epiphyseal dysplasia (SMED) is a autosomal
recessive disease characterized by short limbs and a host of skeletal abnormalities. This disease
was recently mapped to mutations within the Ddr2 gene which affected the tyrosine kinase
domain of the receptor63. Although the mechanisms have not been identified, it is likely that
these mutations affect DDR2 signalling. The role of DDR2 in calcification was solidified in a study
demonstrating that DDR2 signals to activate RUNX2 through ERK signalling in chondrocytes45.
Together, these studies indicate an important role for DDR2 in calcification. Although DDR1 has
11
not been studied as extensively in the context of calcification, the similarities between
receptors suggest a probable role of DDR1 in calcification.
VASCULAR CALCIFICATION
Types of Vascular Calcification:
There are three main types of vascular calcification: 1) intimal calcification which is seen
in patients with atherosclerosis, 2) medial calcification, also known as Monckeberg’s
arteriosclerosis and 3) calciphylaxis, a systemic calcification of the vascular system. Although
these are separate and distinct forms of vascular calcification, there is frequent overlap of these
diseases in pathology where several forms of calcification can be found together64. Such is the
case when medial calcification occurs with atherosclerotic calcification.
Atherosclerotic calcification is a feature of advanced atherosclerotic plaques, and
manifests as calcification in the intimal layer65. Although there is limited evidence showing that
calcification may occur as a result of circulating progenitors with a calcifying phenotype, most
evidence suggests that mineralization in atherosclerosis is affected by two major cell types,
VSMCs66 and Mθs67. Although both cell types are found adjacent to the calcified region, there is
no evidence that Mθs transdifferentiate into an osteoblast phenotype under these conditions.
VSMCs are the main candidate for driving mineralization as they are phenotypically plastic and
derive from the same mesenchymal precursor cells as osteoblasts68, 69. Evidence suggests that
the expression of RUNX2 promotes VSMC conversion to an osteochondrogenic phenotype70. In
mature atherosclerotic calcifications, the tissue can appear as bone71. In extreme cases, this
tissue is complete with vascularisation and a hematopoietically active marrow71.
12
Medial calcification of the arteries can happen as an entity by itself, or as a result of
intimal calcification in advanced atherosclerosis72. When medial calcification occurs in
association with atherosclerosis, it usually lies in the medial layer adjacent to the
atherosclerotic calcification. Histologically, hydroxyl apatite crystals depositions occur along the
internal elastic lamina. While the mechanism of medial calcification is not as well defined as in
atherosclerotic calcification, current work suggests that RUNX2 plays a role. The consequences
of the medial calcification result in the hardening of the arterial wall known as
arteriosclerosis10. As a result, pulse pressure is increased and therefore chronic stress can
further exacerbate kidney failure in addition to heart failure10.
Calciphylaxis is another form of vascular calcification that occurs in the medial layer.
Similar to medial calcification, calciphylaxis may also arise as result of End Stage Renal Disease
(ESRD)10. In addition to ESRD, calciphylaxis may also arise from non‐uremic disease conditions
such as hyperthyroidism, liver cirrhosis, inflammatory diseases such as rheumatoid arthritis and
Crohn’s disease or as a secondary complication of chemotherapy10. Despite their similarities,
calciphylaxis and medial calcification represent vastly different pathologies. Calciphylaxis refers
to the systemic calcification of all blood vessels at the medial layer and non‐bone extravascular
tissues, whereas medial calcification occurs primarily in the major arteries10. As the calcification
is systemic, this leads to poor perfusion of the tissues resulting in tissue necrosis10. Also, the
hardened blood vessels which can also lead to thrombosis and increased blood pressure, which
ultimately results as increased pressure on cardiac pulmonary system and ultimately heart
failure.
13
Models of Vascular Calcification:
Calcification is studied using models which emulate the pathophysiologic perturbations
which occur that ultimately lead to calcification. The factors which result in calcification stem
from a hormonal imbalance such as high levels of Parathyroid Hormone (PTH),
hyperphosphatemia, loss of calcification inhibitors, drug usage and as a secondary complication
of a disease; because of this most models of calcification perturb either one or more of these
factors. In vivo studies have largely been carried out in rabbit, rat and mice, while in vitro
studies have been carried out on VSMCs, and organ cultures of varying species origin73. As it is
relatively easy to genetically manipulate a mouse, various genes which inhibit calcification such
as Fetuin A74, Matrix GLA Protein (MGP)75 or Osteopontin (OPN)76 have been knocked out in
order to produce models of calcification. However, because some of these genes are expressed
in most tissues, extraskeletal mineralization is not limited to vascular tissue, but is extensive
throughout the soft tissues of the animal. This can be addressed by using a smooth muscle cell
specific gene such as SM22α to drive the cre lox recombinase system77. Genes with high
penetrance such as Feutin A KO animals can produce vascular calcification without further
manipulation78. However animal models making use of genes with lower penetrance such as
OPN knockout mice, need to be combined with other perturbations in order to produce a
vascular calcification model78.
A commonly used disease model to study calcification is atherosclerosis. Most
atherosclerotic calcification studies are performed by feeding animals a high fat and cholesterol
diet to develop atherosclerosis. As such there have been many studies on atherosclerosis using
this model with rabbits, rats, and mice. Mice however, are inherently resistant to developing
14
atherosclerosis even with an atherogenic diet 79. As a result, an atherogenic diet is
accompanied with genetic manipulation deleting either the LDL receptor or Apolipoprotein E
(APOE), by deleting the LDL receptor or APOE you are able to increase the circulating levels of
cholesterol to promote atherosclerosis79. As the animals develop atherosclerosis and the
complications that come with it, calcific deposition will occur in the intima of the plaque. In
addition, various factors can also be studied while the animals are undergoing calcification. In
mice, genes can be manipulated to study their affects on atherosclerotic calcification.
Medial calcification commonly occurs in association with ESRD, because this is the case,
animals models of Chronic Kidney Disease (CKD) combined with a high phosphate diet are
commonly used to study medial calcification. Renal failure is most commonly induced by the
5/6 nephorectomy, which involves the surgical removal of one kidney along with the partial
ablation of the remaining kidney either by partial nephorectomy, posterior or anterior renal
artery ligation or by electrocauterization on the cortex of the remaining kidney80. As a result of
the reduced nephron number combined with the high phosphate diet, the animals will develop
severe hyperphosphatemia80. Thus, the resulting process promotes calcification of the arteries.
Due to the complexity of this surgery it is usually done in larger animals such as rabbits and rats,
but there has been some success using this technique with mice81.
The effects of high phosphate can also be studied in VSMCs82 and aortic ring sections83
in tissue culture. Prior to studies on calcification of vascular tissues these cultures were used to
stimulate mineralization in long bones, calvaria, and cells of the osteoblast lineage84. The major
benefit of this technique is the relatively fast endpoint in which you can acquire data. These
cultures typically involve supplementation of phosphate either in the inorganic form as Pi85 or
15
as β Glycerol Phosphate86. In addition the process can be accelerated with supplementation of
ascorbic acid to induce matrix synthesis to provide the calcifying cells with an osteoid87. Also
dexamethsone can be added which helps to inhibit the VSMC proliferation and promote the
ossifying phenotype88. As a result, in vitro assays provide the potential to study the effects of
cytokines, various ligands, extracellular matrix and other extracellular forces such as shear and
tension which can provide mechanistic cues for vascular calcification.
Mechanisms of Vascular Calcification:
The mechanism for vascular calcification was long thought to be the passive deposition
of hydroxy apatite crystals in the blood vessel walls; while this may hold true for calciphylaxis,
more recent evidence shows that both atherosclerotic and medial calcification is an active
cellular osteogenic process10. This process is regulated by the osteochondrogenic transcription
factor RUNX2 and regulation of its activity89.
During atherosclerotic calcification, the VSMCs respond to many different stimuli which
can promote osteochondrogenic differentiation. Oxidized LDL particles (OxLDL) which are in
abundance within the atherosclerotic plaque are known to activate Toll Like Receptor 2 to
activate the osteogenic process90. In addition, because atherosclerosis is an inflammatory
vascular disease, many inflammatory cells such as macrophages and other leukocytes which
secrete cytokines such as IL691 and TGFβ92 are a part of the inflammatory process. These
molecules have shown to elevate the activity of RUNX2 and ultimately calcification. Other
factors such as cell death from apoptosis and necrosis also contribute to calcification. Upon
death these cells leave behind apoptotic bodies, which can provide the seeding points for
16
amorphous calcium phosphate deposition11. Not only do amorphous crystals provide the
primary seed sites for apatite crystals, but the crystals themselves are able to promote
osteochondrogenic transdifferentiation through an unknown process93. Furthermore, physical
factors such as shear stress are able to regulate calcification as both regions of low shear and
low shear in vitro environments promote osteogenic conversion in cells94. Ultimately these
factors result in VSMCs transdifferentiation into an osteochondrogenic phenotype. These cells
which are the precursors to chondrocyte and osteoblast‐like cells synthesize a matrix which is
permissive to calcification called the osteoid95. This matrix is rich with type I collagen and
calcium binding proteins such as Osteocalcin (OCN), Osteonectin (ON) and OPN96. At later
stages it is hypothesized to mirror endochondral ossification, in which the chondrogenic‐like
cells die while the osteoblast‐like cells predominate97. The osteoblast‐like cells cause
mineralization of the osteoid by secretion of calcium phosphate containing matrix vesicles98.
During this period, osteoblast‐like cells produce proteins such as ALP99 and PiT‐1100 in order to
increase intracellular Pi levels and mineralize the matrix. Initially this matrix is largely
amorphous calcium phosphate, over time the amorphous calcium phosphate is replaced with
hydroxyl apatite and later develops into bone‐like tissue101.
In contrast to atherosclerotic plaque calcification, medial calcification largely stems as a
secondary complication from a pathological state. These states include atherosclerosis,
parathyroidism, diabetes mellitus and hyperphosphatemia from ESRD caused by CKD 10. The
most common cause of medial calcification is CKD102. In CKD, the function of the kidney is
compromised and it is unable to secrete phosphate leading to elevation of phosphate
concentration in the serum, the resulting hyperphosphatemia leads to medial calcification103.
17
While it was previously thought that medial calcification occured through an osteogenic
independent process, recent studies by Giachelli and colleagues have demonstrated that
activation of the RUNX2 osteogenic gene program occurs in medial calcification81.
Medial calcification can also arise as a result of deficiency of tissue inhibitors of
calcification. Genetic deletion of tissue inhibitors of calcification such as Neuroecto
Pyrophosphatase 1 (NPP1)104, and MGP105 results in ectopic calcification in both humans and
mice. Furthermore, studies show that deficiency of circulating inhibitors such as Feutin A74 and
Klotho106 also promote medial calcification. The use of drugs which prevent inhibiting function
may also be the cause of medial calcification. Warfarin is a drug that is commonly used to treat
anticoagulation in patients with cardiovascular problems. It works by preventing the γ
carboxylation of several clotting factors. As MGP also requires γ carboxylation, this process is
hindered as well107.
Although atherosclerotic and medial calcifications are initiated through separate
pathways, both ultimately converge upon the expression of RUNX2. With respect to medial
calcification, whether RUNX2 contributes to the calcification process or is the result of
activation by calcium phosphate depositions is unclear. As in medial calcification, calcium
deposition occurs in association with degraded elastin fibers in the internal elastic lamina, and
not reminiscent of the osteoid synthesis in atherosclerotic calcification108. However, in
atherosclerotic calcification, studies have demonstrated that the activation of RUNX2 is
indispensible for osteochondrogenic differentiation109.
18
The Role of Phosphate in Vascular Calcification:
Hyperphosphatemia leads to medial calcification and calciphylaxis and can exacerbate
atherosclerotic calcification10. Previously, it was thought that hyperphosphatemia causes
calcification primarily due to the increased inorganic phosphate (Pi) available for calcium
phosphate formation. While this is the case for calciphylaxis, later studies, revealed a crucial
role for Pi110. Furthermore, culturing VSMCs in high Pi resulted in a VSMC phenotypic change to
an osteoblastic/chondrogenic phenotype which was accompanied by the expression of RUNX2
and markers of calcification such as OCN, OPN and Alkaline Phosphatase (ALP)69. Despite this,
the precise mechanism by which Pi is able to activate the osteogenic gene program is unknown.
Current evidence demonstrates that the propensity of vascular calcification is dependent upon
the careful balance of Pi and pyrophosphate (PPi)111 (Figure 3).
PPi is an inhibitor of calcification that acts by inhibiting the nucleation and propagation
of hydroxy apatite crystals111. The intracellular levels of PPi in VSMCs are determined largely by
the expression of NPP1112. The primary function of NPPs is to catabolise ATP to produce PPi104.
In healthy vascular tissue NPP1 is highly expressed and its levels are reduced during vascular
calcification113. In VSMCs, NPP1 is the major protein that carries out the reaction converting
ATP to AMP and PPi114. NPP1 deficiencies in mice have manifested themselves in severe soft
tissue calcification, arterial calcification and spontaneous aortic calcification104. In humans,
NPP1 deficiency has been linked to infantile medial arterial calcification115. As NPP1 competes
with Alkaline Phosphatase (ALP) for the ATP substrate, it was thought that NPP1 prevented
calcification by competing for ATP. However, studies showed that PPi alone was able to prevent
nucleation of calcium phosphate crystals in cell culture116. These studies revealed that balance
19
of PPi and Pi levels are important in ultimately regulating calcification and not substrate
competition between NPP1 and ALP. Furthermore, the PPi which is produced by NPP1 can be
hydrolyzed into two free Pi molecules by ALP, negating the competition hypothesis117. In
addition to ATP and PPi, ALP is also able to generate free Pi by the hydrolysis of ADP, AMP118
and phosphate groups on phosphoproteins119. Unsurprisingly, ALP has been observed in many
studies to be indispensible in VSMC mediated calcification120, 121. ALP in normal physiology is
required for osteoblasts to create bone tissue122. High serum concentration of ALP is often
indicative of rapid bone development, such as in childhood123. ALP deficiency in human
osteoblasts causes hypophosphatasia often with fatal results124. If a patient with ALP deficiency
survives to early childhood, they will have a severe form of rickets and early loss of deciduous
teeth.
Another major determinant of intracellular Pi levels is the transport of Pi into the VSMC.
Intracellular phosphate transport is mediated primarily by the family of Sodium Dependant
Phosphate Co‐Transporters (PiTs)125. Type I and type II PiTs are expressed primarily in the
kidney and intestinal epithelium126, while type III PiTs are less tissue specific and play a major
role in Pi transport within VSMCs127. There are two transporters within the type III class; they
are PiT‐1 and PiT‐2128. PiT‐1 is highly expressed in VSMCs, but PiT‐2 is expressed to a far lesser
degree128. The importance of PiT‐1 to vascular calcification was first revealed when it was
discovered that PiT‐1 is necessary for VSMC to osteochondrogenic conversion in high Pi
treatment129.
Together these studies on NPP1, ALP and PiT‐1 underscore the importance of Pi in VSMC
mediated calcification (Figure 3).
20
RUNX2
RUNX2 Structure and Function:
The Runx2 gene contains two promoters which give rise to two isoforms, named type I
and type II130. These isoforms contain an identical DNA binding domain130 ,an α helical
structure for microtubule binding and translocation and a nuclear targeting matrix (NTM) 131.
Both RUNX2 isoforms also contain a 200 amino acid sequence called a PEST domain (Figure 4)
131. The PEST domain is destabilizing and is hypothesized to facilitate degradation through
either the proteosome132 or calpain pathways133.
These isoforms differ in the N‐terminus134. The type I N‐terminus contains the MRIPV
amino acid sequence131, 135. In type II RUNX2, the N‐terminus contains the MASNS amino acid
sequence instead, 133. How the N terminus affects isoform differences is unknown. 136 During
early calcification the type I isoform is favoured, but as calcification progresses, type II
expression predominates134. It is not known how or why this process occurs, but both are
important for bone development.
RUNX2 functions in the ossification process by committing mesenchymal stem cells to
become osteochondrogenic cells137. Although RUNX2 can bind DNA as a monomer, RUNX2
must bind DNA with Osterix in a heterodimeric complex for osteoblast differentiation138. It can
also function in the chondrocyte lineage pathway, bound with SOX 9 instead139. The importance
of RUNX2 in bone development can be seen as Runx2 null mice are embryonic lethal due to a
failure in skeletal tissue ossification and Runx2+/‐ mice develop cranial and collarbone
malformations termed cledocranial dysplasia140. Although RUNX2 is indispensible for
calcification, overexpression of RUNX2 in mice did not show the opposite phenotype of RUNX2
21
deficiency. While the number of osteoblasts increased in RUNX2 transgenic animals, osteoid
development and proper mineralization was impaired resulting in osteopenia141. From these
studies it can be concluded that precise regulation RUNX2 activity is required for osteoblasts
and chondrocytes during bone development.
RUNX2 Transcriptional Targets:
RUNX2 binds to a DNA consensus sequence 5’CCACA3’ called Osteoblast Sensitive
Elements (OSE)142, in the promoter regions of target genes, such as Type I Collagen alpha 1
(Col1a1)143, Type I Collagen alpha 2 (Col1a2)143, Opn144, Integrin Binding Bone Sialoprotein
(Ibsp)145, Ocn143 and Vascular Endothelial Growth Factor (Vegf)146. As the binding of OSE by
RUNX2 represents a clear indication of RUNX2 activity, OSEs have been used in promoter
reporter constructs to demonstrate RUNX2 activity147. Most of the identified genes which are
activated by RUNX2 are major constituents of the bone osteoid or play functional roles in
calcification. For example, type I collagen is an important structural component of the bone
osteoid148. Additionally, type I collagen is also ligand for pro‐osteogenic signalling149. Other gene
products such as OPN, IBSP and OCN contain calcium phosphate binding sites, are hypothesized
for facilitating calcium phosphate nucleation and provide the physical means for osteoclasts to
attach and remodel bone150. Another RUNX2 target, VEGF is promotes angiogenesis, which is
important for the vascularisation of bone151. As such, many of the mRNAs of these RUNX2
target genes have been used as a measure of its activity in studies on osteogenic development
and in vascular calcification. However it is also important to note that the expressions of these
genes are likely influenced as well by other transcriptional activators. Vegf mRNA expression for
22
example can be induced by Hypoxia Inducible Factor 1 alpha (HIF1α), which is RUNX2
independent152.
Regulation of RUNX2:
RUNX2 is regulated by transcription of the gene to produce mRNA. Bone Morphogenic
Protein 2 (BMP2), which is an important pathway to osteogenic development, regulates the
transcription of Runx2 mRNA153. Other molecules which regulate the expression of Runx2
mRNA include Retinoblastoma Protein (RBP)154, and Insulin‐like Growth Factor 1 (IGF1)155.
In addition to regulating the expression levels of RUNX2, the activity of RUNX2 is also
regulated by many signalling pathways. Signalling pathways which have been identified to
positively regulate RUNX2 activity include BMP2, Fibroblast Growth Factor Receptor 1 (FGFR1),
Interleukin 1 β (IL‐1β) and Transforming Growth Factor β (TGFβ)156. These pathways ultimately
lead to the phosphorylation and activation of RUNX2. Regulation of RUNX2 activity occurs
through phosphorylation of serine sites, the identified sites all follow the sequence of proline‐
serine‐theronine (Figure 5). Recent work showed that RUNX2 contains three clearly identified
phosphoserine sites (S14, S104 and S451), which have inhibitory and stimulatory effects, both
S104 and S451 are inhibitory while S14 is stimulatory70 (Figure 5). While many kinases have
been identified as crucial to RUNX2 phosphorylation, only the MAPKs, ERK1/2 and P38 have
been identified as direct kinases for S14147. ERK1/2 and P38 have also been identified as direct
RUNX2 binding partners147. Furthermore, ERK1/2 and P38 have also been identified as crucial
MAPKs for the maturation of osteoblasts further solidifying their roles in calcification82. While
the exact subcellular location where phosphorylation of RUNX2 occurs has not been studied,
23
P38 and ERK1/2 typically phosphorylate transcription factors after they have entered the
nucleus.
To elicit its effects on calcification, RUNX2 is directed to the nucleus from the cytoplasm
70. From there RUNX2 is able to bind to DNA elements and induce gene expression. To enter the
nucleus RUNX2 interacts with the microtubule network via its N terminus for transport131. Upon
entry to the nucleus, RUNX2 activate genes in the osteogenic process. These processes
demonstrate the complexity and redundancy of mechanisms by which RUNX2 activity can be
regulated.
RUNX2 in Vascular Calcification:
RUNX2 is required for osteogenic modulated vascular calcification109, and
osteochondrogenic transdifferentiation in VSMCs157. Many of the factors which activate RUNX2
in osteochondrogenic development in bone tissue also activate RUNX2 in VSMCs. In most
vascular calcification studies, RUNX2 activity is detected by phosphorylated serine residues,
promoter‐report construct or transcriptional activation of RUNX2 targets. Such targets include
Col1a1, Col1a2, Opn, and Ocn. As the vascularisation of advanced vascular calcification, Vegf
has become the marker for late stage vascular calcification10. Although RUNX2 activation in
VSMCs promotes the expression of osteochondrogenic genes and ossification,
transdifferentiated VSMCs still maintain many smooth muscle markers such as αSMA, SM22α,
myocardin and smooth muscle myosin heavy chain69, 158.
24
MICROTUBULES
As previously discussed, microtubules are key components which regulate the
translocation of RUNX2 into the nucleus. Because the association between RUNX2 and
microtubules is a relatively novel finding and sparsely defined, this latter part of the review will
be dedicated to fully exploring microtubules.
Structure and Organization of Microtubules:
Microtubules along with actin and intermediate filaments are a part of the eukaryotic
cellular cytoskeleton. Microtubules are typically organized at the Microtubule Organizing
Center (MTOC), which is composed of γ‐tubulin, and situated at the centriole159. Typically,
nucleation of microtubules occurs at the MTOC. However in cells which lack MTOC,
microtubules nucleate at discrete points in the cytoplasm through a poorly understood
process159.
The microtubule network is made up of repeating subunits of α and β‐tubulin dimers,
and forms an imperfect helical orientation of 13 tubulin dimers per turn (Figure 6). Overall
these form the microtubule structure which is a hollow cylinder. This cylinder has a diameter of
25 nm and can grow up to 25 µm in length160. Microtubule filaments have polarity with + or –
ends which have β‐tubulin or α‐tubulin exposed respectively160. This orientation is essential for
the biological function of microtubules. The orientation of microtubules is such that subunits
are added from the + end160. Typically, the + end refers to the end at the MTOC and the – end
refers to the end extending to the cell periphery. This orientation is particularly important for
the motor proteins dyenin and kinesin which shuttle cargo from + to – and – to + respectively.
25
Microtubules polymerize when α and β‐tubulin dimers are bound to GTP to form a stabilized
GTP cap at the – end (Figure 6)161. Hydrolysis of GTP to GDP within β‐tubulin, promotes a bent
conformation of the proteins hence promoting depolymerisation161. This active assembly and
disassembly of microtubules is termed dynamic instability.
Regulation of Microtubule Dynamics:
The stability of microtubules can be regulated by interactions with other proteins such
as motor proteins and non motor microtubule associated proteins. These proteins may or may
not influence the GTPase activity of β‐tubulin. Motor proteins which affect the stability of
microtubule are the kinesins. Kinesins from the kinesin 13 family play major roles in regulating
microtubules. Their primary function is to catalyze the intrinsic property of microtubules to
depolymerize162. They are able to accomplish this by binding to the curved ends of microtubule
filaments to promote instability in an ATP dependant process163.
Microtubules dynamics are also regulated by post transcriptional modifications which
are catalyzed by various microtubule associated proteins. The major ones include
detyrosination, Δ2‐tubulin generation and acetylation. Detyrosination of α‐tubulin refers to the
removal of the terminal Tyr residue at the C‐Terminus of α‐tubulin, thus rendering Glu the new
terminal residue164. This process is mediated by an as of yet unidentified detyrosinase.
Detyrosinated α‐tubulin favours the polymerized state in microtubules. However this process is
reversible, as Tubulin Tyr ligase can re‐add Tyr to the Glu residue164. In order to make α‐tubulin
permanently favour polymerization, detyrosinated α‐tubulin is processed by cytosolic
carboxypeptidase. This removes the Glu residue in α‐tubulin giving rise to Δ2‐tubulin164.
26
Detyrosinated tubulin and Δ2‐tubulin are commonly found in terminally differentiated neuronal
cells where the dynamic cycling of microtubules is not required164‐166.
Another microtubule modification that is associated with stability is the acetylation of α‐
tubulin. Acetylation is a process where an acetyl group is conjugated to K40 on α‐tubulin by a
currently unknown acetyl transferase. HDAC6 is the only deacetylase known to regulate α‐
tubulin acetylation167. Evidence to support this comes from HDAC6 overexpression which
resulted in a global decrease in tubulin acetylation and increased susceptibility to destabilizing
agents such as nocodazole168. It is hypothesized that acetylation of K40 site induces a
conformational change, which affects the α‐tubulin‐β‐tubulin dimer binding to prevent
polymerization of microtubules164. The regulation of microtubule polymerization is crucial for
biological function.
Microtubules and Nuclear Transport:
Though it was previously thought that nuclear localization occurred via Brownian
motion, recent work shows that nuclear translocation for a variety of different molecules
requires the use of microtubules. For examples signal transduction kinases such as P38 use the
microtubule network in order to translocate into the nucleus169. Recent work on steroid
hormone receptors has demonstrated that they also require microtubules for nuclear
transport170.
Microtubules are also required for the translocation for transcription factors including
RUNX2131 and HIF1α171. RUNX2 is only able to bind to polymerized microtubules, but not
monomeric tubulin131. The study also showed that RUNX2 associates with the microtubule
27
network via the N terminal amino acid residues 1 – 361. However, it is not known whether
RUNX2 associates directly with the microtubule network or through a protein complex. A study
on HIF1α and steroid hormone receptor translocation to the nucleus by microtubules showed
that both molecules bind to Dynamitin, which is a part of the dynein activator complex170, 171.
This complex is able to bind both dynein and kinesin and thus is used for bidirectional transport
of organelles and vesicles on microtubules within the cell172. RUNX2 may also bind to Dynamitin
for transport.
Interestingly, although transcription factors require microtubules, purely polymerizing
microtubules is not sufficient for nuclear translocation for most transcription factors131, 171. For
most nuclear transcription factors, a state of dynamic instability is required for nuclear
translocation. Why dynamic instability is required for the translocation of proteins is not
known.
Mitochondrial translocation also requires dynamic instability for translocation within
the cell. Mitochondria interact with positive end proteins which latch on and carry them into
the microtubule network160. A similar latching process may be required for transcription factors
as well. Upon arrival at the nuclear pore complex, the protein is recognized and translocated
within by a poorly understood process.
In meschenchymal stem cells, DDR1 is crucial for forming dendritic extensions by
microtubules173. As transcription factors such as RUNX2, HIF1α and signalling molecules such as
ERK, P38 and STATs all require microtubule translocation; this might represent a master
regulatory network for controlling nuclear activity of transcription factors by DDR1.
28
RATIONALE AND HYPOTHESIS
Type I collagen has been associated with both osteochondrogenic differentiation and
vascular calcification149, 174. The alpha 1 integrin receptor has been implicated in regulating
collagen induced osteochondrogenic differentiation in bone, however, the receptor by which
type I collagen elicits its osteochondrogenic effects in VSMCs has not been identified. In a
published study, our group observed that Ddr1‐/‐; Ldlr‐/‐ mice had reduced atherosclerotic
calcification59. As type I collagen is a ligand for DDR1, this suggested a novel mechanism by
which type I collagen can promote osteochondrogenic transdifferentiation. Our laboratory also
reported that Ddr1‐/‐ VSMCs showed attenuated calcifying potential, with some indication of
decreased osteochondrogenic transdifferentition, such as decreased ALP activity59. This led me
to perform a screen on Ddr1+/+ and Ddr1‐/‐ VSMCs to identify genes which might be involved in
osteochondrogenic transdifferentiation. My screens revealed that Ocn, showed a significant
reduction in Ddr1‐/‐ cells. This result led me to investigate the mechanism by which DDR1
influenced Ocn expression. Ocn is a target gene of the osteochondrogenic transcription factor
RUNX2, and RUNX2 activity is regulated by MAPKs such as ERK1/2 and P38147, 175‐177. DDR1 has
been shown to signal through those MAPKs18, 43, 44, 48. As this represented a possible link
between DDR1 and RUNX2 I hypothesized that: DDR1 activates RUNX2 by signalling through
MAPKs in VSMCs during calcification. However later in the course of my studies I discovered
that the absence of DDR1 diminished RUNX2 activity at early stage calcification in spite of no
DDR1 dependant differences in ERK1/2 and P38 signalling at that time point. In addition, P38
inhibition reduced Ocn expression and RUNX2 activity in DDR1 expressing cells. My studies also
revealed that Ddr1‐/‐ VSMCs show reduced RUNX2 and phosphoP38 nuclear localization.
29
Furthermore, close observations of my Ddr1‐/‐ VSMCs showed increased cell spreading which
was absent in Ddr1+/+ VSMCs. As nuclear translocation of RUNX2 and phosphoP38 and
maintenance of cell morphology require a microtubule network131, 171, I postulated that there
were differences in microtubule structure between the Ddr1+/+ and Ddr1‐/‐ VSMCs. Supporting
this, an article published by Lund et. al. showed that mescenchymal stem cells sense type I
collagen using a DDR1 stabilized microtubule network173. A prior study also demonstrates that
RUNX2 requires a microtubule network for nuclear translocation131. Together these studies
suggest that DDR1 may able to regulate the translocation and thus the activity of RUNX2. Thus I
later hypothesized that: DDR1 regulates RUNX2 activity by microtubule organization in VSMCs
during calcification.
MATERIALS AND METHODS
Chemicals and Reagents
All chemicals and reagents were purchased from Sigma‐Aldrich (St. Louis, MO) unless
otherwise stated.
Smooth Muscle Cell Isolation and Culturing
Mouse VSMCs were isolated using elastase type III digest from Ddr1+/+ and Ddr1‐/‐
C57BL6 mice carotid arteries as described in Hou et. al.50 and propagated in a medium
composed of high glucose Dulbecco’s modified Eagle’s Medium (Gibco) and supplemented with
10% heat inactivated fetal bovine serum, 1% antibiotic‐antimycotic (culturing media) at 37°C
30
with 5% CO2 in 75 cm2 culture flasks. Cells were passaged upon growth to approximately 90%
confluency. To passage the cells, culture media was aspirated and the cells were washed 3
times using Dulbeco’s phosphate buffered saline with MgCl2 and CaCl2 (PBS). After washing, the
cells were passaged through enzymatic digest by incubating the cells with 2mL of 0.02% trypsin
for 1 min at 37°C to detach cells from the culture flask. The cell suspension was removed and
placed in 50 mL falcon tubes (BD Falcon) containing culture media that has been warmed to
37°C. This cell suspension was centrifuged at 1,000 rpm for 5 mins. The supernatant was
aspirated and the cell pellet was redistributed into 3 new culture flasks, each containing 10 mL
of culture media. The periodicity by which the cells were passaged was approximately 4 days
after each passage. Experiments were performed from these isolated VSMCs from passage 6
through to passage 10.
In Vitro Calcification Assay
VSMCs from passages 6 to 10 were seeded at 1563 cells/cm2 and cultured in culturing
media. After 72 hours the culturing media was washed out using PBS and replaced with media
consisting of high glucose Dulbecco’s modified Eagle’s Medium (Gibco) supplemented with 3%
heat inactivated fetal bovine serum (Hyclone), 1% antibiotic‐antimycotic (Penicillin‐
Streptomycin/Amphotericin B) (Gibco) and 2.4 mM Pi (osteogenic media). In parallel
experiments involving MAPK pathway inhibition, 10 mM stocks of PD098059 (MEK inhibitor)
and SB203580 (P38 inhibitor) were made in Dimethyl Sulfoxide (DMSO) and diluted at 1:1,000
to create a final concentration of 10 µM within the osteogenic media. Controls to inhibitor
experiments were prepared by adding DMSO at a 1:1,000 dilution in the media. In experiments
31
involving microtubule destabilization, 10 mM stock of nocodazole was made in DMSO and
diluted at 1:1,000 to create a final concentration of 10 µM. To stabilize microtubules, a 10 mM
stock of taxol was made in DMSO and diluted at 1:1,000 to create a final concentration of 10
µM. These cells were incubated for 0, 2, 4 or 12 days after the addition of the osteogenic media
and then used for calcium quantification, RT‐PCR, immunoblotting, immunohistochemistry and
luciferase activity assays.
Calcium Quantification
The calcium content of calcified VSMCs was determined using the o‐cresophthalein
complexone colorimetric assay (Clinotech Diagnostics). The assay was performed according to
manufacturer protocols. Initially, the assayed cells were incubated with 500 µL of 0.6 N HCl per
well in a 6 well culture plate (BD Falcon) at 4°C overnight. The HCl was added to release the
calcium phosphate crystals. The resultant HCl solution was diluted 50 fold with 0.6N HCl.
Following this, the diluted sample was added in triplicate, 10 µL each in a clear 96 well plate (BD
Falcon). To quantify the amount of calcium within each sample, a standard curve generated
using the calcium standard solution with O‐cresopthalein contained within the assay kit ranging
from 0 to 7.5 mg/ml. 100 µL of the prepared calcium quantification reagent was pipetted into
each well containing experimental and standard curve samples and incubated for 5 mins at
room temperature. O‐cresopthalein functions by binding calcium within the solution and forms
a purple coloured complex. As a result, the amount of calcium was determined by measuring
intensity of purple colour at 540 nm. The measured values were calculated by interpolating the
absorbance values to the calculated standard curve formula.
32
Measured calcium content was normalized to total cellular protein. Protein content was
measured using a microBCA assay according to manufacturer protocols (Pierce). To obtain total
cellular protein, decalcified cells were washed 3 times with PBS and incubated with 0.1 N NaOH
at room temperature for 1 hour to solubilise the protein. To prepare the protein samples
collected in NaOH for quantification, the samples were diluted 1:200 with 0.1 N NaOH. To
correlate absorbance values to protein concentration a BSA standard ranging from 0 to 40
µg/mL was prepared according to manufacturer protocol by diluting the supplied BSA stock
with 0.1 NaOH. The BSA reaction was performed by pipetting 100 µL samples and the prepared
BSA standard in triplicate on a 96 well plate. To visualize the protein content with in each
sample 100 µL of prepared BCA reagent was pipetted into each of the sample and standard
containing wells and incubated for 1.5 hours at 37°C. In a BCA reaction Cu2+ ions are reduced by
peptide bonds to Cu+ in a concentration dependant manner. The Cu+ is bound by bicinchronic
acid to a form a purple complex, which absorbs light at 562 nm. The measured values were
calculated by interpolating the absorbance values to the calculated standard curve formula.
In addition, calcium content was also normalized to total cellular DNA using the
Fluoreporter dsDNA quantification kit (Molecular Probes). Following calcium extraction with
0.6N HCl and washed with PBS, cells are incubated with 500 µL of ddH2O per well and incubated
at ‐80°C for 1 hour. Following this the samples were thawed and incubated at 37°C for 1 hour.
To quantify the amount of DNA a standard curve ranging from 0 to 10 ng/µL was prepared
according to manufacture protocol using the supplied salmon sperm DNA stock. Each of the
samples and DNA standards were pipetted in triplicate on a black walled 96 well plate (BD
falcon). To visualize the DNA content 100 µL of Hoechst TNE buffer was pipette into each well
33
containing a sample or standard and incubated at room temperature for 5 mins. The Hoechst
dye within the solution is able to bind dsDNA and thus providing a means to visualize the DNA.
After incubation, the plate was read on a HTS fluorescent microplate reader (PherSTAR) by
setting excitation and emission filters at 360 nm and 460 nm respectively. The measured values
were calculated by interpolating the absorbance values to the calculated standard curve
formula.
Reverse Transcription and Quantitative Real Time‐PCR
RNA was isolated from VSMCs using the RNeasy RNA isolation kit (Qiagen). RNA was
later quantified using the NanoDrop 2000c (Thermo Scientific). To prepare the RNA for reverse
transcription, 1 µg of RNA was treated with DNase I (Fermentas) for 30 mins at 37°C in supplied
buffer (10 mM Tris‐HCl pH 7.5, 2.5 mM MgCl2 and 0.1 mM CaCl2) to remove genomic DNA.
DNase I was inactivated by the addition of 1 µg of EDTA and heat inactivation for 10 mins at
65°C. This solution was reverse transcribed into cDNA using the Superscript First Strand
Synthesis Kit (Invitrogen) with 100 ng of random hexamers, 200 nM dNTP, 10 mM DTT, 40 u of
RNaseOUT and 50 u of Superscript II in a buffer containing 50 mM Tris‐HCl, pH 8.3, 75 mM KCl
and 3 mM MgCl2. qRT‐PCR reactions were performed using SYBR advantage qRT‐PCR premix
kit (Clonetech) on the Quantitative Real Time (qRT)‐PCR using the ABI 7900 HT Fast Real‐Time
PCR System (Applied Biosystems). Individual reactions were performed in 10 µL with 4 ng of
cDNA and 450 nM of the following forward and reverse primers DDR1 (FWD 5’‐
CTGCTCTTTACTGAAGGCTC‐3’, REV 5’‐CAGGCCATAGCGGCACTTGG‐3’), PIT‐1 (FWD 5’‐
GCGCCCTTCCGCGGGCTTT‐3’, REV 5’ GTCTCCGGCCCGGCCGCTTGCCCGC‐3’), NPP1 (FWD 5’‐
34
CTCGGTTGAGACCCACTGATG‐3’, REV 5’‐GCTCCCGGCAAGAAAGATTT‐3’), ALP (FWD 5’‐
AGACACAAGCATTCCCACTAT‐3’, REV 5’‐CACCATCTCGGAGAGCG‐3’), RUNX2 (FWD
5’‐GATGCTCTGTTTCTTTCTTTCAGG‐3’, REV 5’‐CTCCAGCATTTCATGCTAGT‐3’), ATF4 (FWD
5’‐TTGACCACGTTGGATGACAC‐3’, REV 5’‐CAGAGATATCAACTTCACTGCCTA‐3’), OCN (FWD
5’‐GTAGTGAACAGACTCCGGC‐3’, REV 5’‐AGTGATACCGTAGATGCGT‐3’) TENC (FWD
5’‐ACCGCAGAGAAGAATTTTGG‐3’, REV 5’‐TCCCCATGGTCTTGTAGGTC‐3’) and Acidic Ribsomal
Protein (ARP) (FWD 5’‐AGACCTCCTTCTTCCAGGCTTT‐3’), (REV 5’‐
CCCACCTTGTCTCCAGTCTTTATC‐3’). The conditions for the PCR reaction were 95°C for 10 mins
followed by 40 cycles of 95°C for 15 seconds and 60°C for 1 min. Following PCR, a dissociation
step was performed by heating the samples to 95°C for 15 secs and then cooling them to 60°C
for 15 secs and then heating again to 95°C for 15 secs. The data collected was used to generate
the dissociation curve from which the cycle threshold (Ct) was determined using SDS 2.3
software (Applied Biosystems). To determine relative expression levels the comparative Ct
method (2‐ΔΔCt) was used as previously described59.
Immunoblotting
Protein was isolated from smooth muscle cells incubated in osteogenic conditions for
either 2 or 12 days using the cell lysis buffer (20 mM Tris‐HCl, (pH 7.5), 150 mM NaCl, 1 mM
Na2EDTA, 1 mM EGTA, 1% Triton, 2.5 mM sodium pyrophosphate, 1 mM betaglycerophosphate,
1 mM Na3VO4, 1 µg/ml leupeptin) (Cell Signalling Technology). Using the microBCA assay as
described above, appropriate sample volumes of sample were added to 2X sample buffer (0.1
M Tris pH 6.8, 2% SDS, 20% glycerol, 0.4% bromophenol blue, 5% β‐mercaptoethanol) and
35
boiled for 5 mins. Samples were loaded on and separated with a 10% TGX separating gel (Bio‐
Rad) and were subsequently transferred to a nitrocellulose membrane (Bio‐Rad) at 100 V for 1
hour at 4°C. Membranes were incubated in blocking buffer (5% non‐fat dry skim milk, in Tris
Buffered Saline‐Tween 20 (TBST)) for 1 hour at room temperature, followed by washing in TBST
and incubated in the appropriate primary and secondary antibodies in 2.5% non‐fat dry skim
milk in TBST. The antibodies were diluted at the following concentrations: rabbit polyclonal
anti‐ERK1/2 (Cell Signalling Technology) (1:1000), rabbit polyclonal anti‐pERK1/2(Cell Signalling
Technology) (1:250), rabbit polyclonal anti‐P38(Cell Signalling Technology) (1:250), rabbit
polyclonal anti‐pP38 (Cell Signalling Technology) (1:250), mouse monoclonal anti‐RUNX2
(Abcam) (1:250), rabbit polyclonal anti‐β actin(Abcam) (1:1000), HRP conjugated goat anti‐
rabbit (Cell Signalling Technology) (1:1000) and HRP conjugated horse anti‐mouse (Cell
Signalling Technology) (1:1,000). Membranes probed with antibodies was incubated with
western blot chemiluminescence reagent Plus (Perkin Elmer), the membrane was stripped by
incubating Restore Plus western blot stripping buffer (Pierce) for 30 mins at 37°C, Washed with
TBST and re‐probed for next protein of interest. Immunoblots were imaged using an Image
Station 4000MM Pro (Kodak) and quantified using Image J software (NIH). To normalize,
densitometry values obtained were divided by the β‐actin loading control densitometry values.
pERK1/2 and pP38 expression values were obtained by dividing the β‐actin normalized values of
the phospho forms to the β‐actin normalized values of total forms.
36
Immunocytochemistry
VSMCs were grown using the calcification assay conditions on 8 chambered culture
slides (BD Falcon) to visualize cells after immunocytochemistry staining. The cells were
subsequently washed with PBS and fixed with 5% PFA (Paraformaldehyde) for 10 mins at room
temperature followed by a PBS wash and permeablized with PBS containing 0.25% Triton X‐100
for 10 mins. After permeabilization the cells were washed again and blocked with 1% BSA with
PBS and 0.1% Tween (PBST). The cells were later incubated with mouse monoclonal anti‐RUNX2
(Abcam) (1:100), rabbit polyclonal anti‐pP38 (Cell Signalling Technology) (1:100) and rabbit
polyclonal anti‐α tubulin (Abcam) (1:100) in 1% BSA PBST overnight at 4°C. The primary
antibody solution was decanted and the cells were washed with PBS after overnight incubation.
Secondary anti‐mouse Alexa 488 conjugated goat antibody (Invitrogen) (1:100), secondary anti‐
rabbit Alexa 488 conjugated donkey antibody (Invitrogen) (1:100), secondary anti‐mouse Alexa
568 conjugated goat antibody (Invitrogen) (1:100), and secondary anti‐rabbit Alexa 568
conjugated donkey antibody (Invitrogen) (1:100) in 1% BSA PBST were added and incubated for
1 hour at room temperature in the dark. Alternatively, VSMC membrane was visualized by
incubating Alexa 647 Wheat Germ Agglutinin (WGA) (Invitrogen) (1:1,000) in 1% BSA PBST was
added and incubated for 1 hour at room temperature in the dark.
Following secondary antibody or WGA incubation the cells were washed again with PBS
and then incubated for 5 mins with Hoechst in PBS (1:1,000) to stain the nucleus. After Hoechst
incubation the cells are washed with PBS and fixed with CC mount (Invitrogen) and sealed with
a #1 thickness glass cover slide (VWR) and clear nail polish. The cells were visualized by confocal
fluorescence microscopy using the Nikon AR1 laser scanning confocal microscope (Nikon).
37
Immunofluorescence Quantification
To quantify the extent of nuclear localization of RUNX2 in the immunofluorescence
images, Image J was used to measure the intensity of the fluorescent signals. First using the
free form tools, the cytoplasm of the cell was traced out, and the average intensity of the signal
is measured and tabulated. Next, using free form tools the nucleus was also selected and the
average intensity of the signal is measured and tabulated. The average intensity of the nucleus
is then divided by the average intensity of the cytoplasm to obtain the nuclear:cytoplasmic
ratio. For each experiment, one field of view was captured at 400X magnification from each of 3
separate wells. For each field of view 3 cells were chosen at random for quantification. The
images were quantified in a fashion blinded to both genotype and treatment. The values
obtained for each individual experiment were averaged and treated as one replicate. To obtain
data for statistical analysis a total of 3 separate experiments were analyzed.
Plasmid Preparation and Transformation
The 6OSE2 Luc plasmid to assess RUNX2 activity was a kind gift from Dr. Renny
Franceschi from University of Michigan Ann Arbor 147. To prepare 6OSE2 plasmid for
experiments, the plasmid was transformed into E‐coli. To perform the transformation, a 100 µL
aliquot of XL‐1 Blue Competent E‐Coli Cells (Strategene) was thawed on ice. After thawing, 1.7
µL of β‐mercaptoethanol was added to thawed cells and swirled gently. The cells were allowed
to incubate on ice for an additional 10 mins. To propagate the plasmid 20 ng of 6OSE2 Luc was
added to the thawed cell mixture and incubated on ice for 30 mins. A heat pulse of 42°C was
applied to the cell mixture for 45 secs. Following the heat pulse, the cells were incubated on ice
38
for 2 mins. 37°C Pre‐warmed 0.9 mL of SOC media (Invitrogen) was added to the cell mixture.
The cell containing SOC (Invitrogen) mixture was incubated at 37°C for 1 hour with shaking at
250 rpm. The transformed cells were plated on LB‐ampicillin agar plates prepared from a LB
stock containing 100 mg/mL of ampicillin to select for cells containing the plasmid. Cells were
grown overnight at 37 °C to allow growth. After growth period a single colony was selected and
subsequently cultured in a starter culture of 40 mL of LB‐ampicillin at 37 °C with constant
shaking at 100 rpm for 4 hours. After which, the starter culture was placed in a larger LB‐
ampicillin culture containing 1 L of media at 37 °C with constant shaking at 100 rpm overnight.
Plasmid DNA was extracted from the overnight culture using a Maxi Prep DNA isolation kit
(Invitrogen) according to manufacturer instructions. Isolated Plasmid DNA was quantified using
the NanoDrop 2000c spectrophotometer (Thermo Scientific) for future preparations.
Transfection
VSMCs were grown in culturing media in 24 well plates (BD Falcon) for 2 days. Following
the incubation, calcifying media was removed and cells were washed with PBS. For each well of
cells the following mixture was prepared: 5 µg of 6OSE2 Luc plasmid was diluted in 100 µL of
Opti‐MEM (Gibo), then 0.5 µL of Plus Reagent (Invitrogen) was added to the mixture and
allowed to incubate at room temperature for 15 mins. For each volume of the above prepared
mixture, 2.5 µL of Lipofectamine LTX (Invitrogen) was added and incubated for 30 mins at room
temperature. Following incubation, the subsequent mixture was added to 500 µL of antibiotic
free DMEM with 10% FBS and cultured for an additional 24 hours to complete the transfection.
After transfection, cells were cultured according to the calcification assay conditions.
39
Luciferase Activity Assay
Following calcification assay culturing, transfected cells were assayed for RUNX2 activity
with reagents provided with the Dual Luciferase Assay Kit (Promega). Each well of cells was
lysed by incubating 100 µL of passive lysis buffer (Promega) for 30 mins with gentle shaking. To
measure the activity of the lysate samples, 20 µL of the obtained lysate was pre‐dispensed in
triplicate on to a black walled 96 well plate. The Luminat LB9507 luminometer (Berthold
Technologies) added the following reagents automatically: First, 100 µL of Luciferase Assay
Reagent II (Promega) to measure the activity of firefly luciferase. After a 2 second delay, Stop
and Glo Reagent (Promega) was added to quench the firefly luciferase signal and provide the
substrate for the renilla luciferase signal. Relative Luminescence Units (RLUs) were obtained by
normalizing the firefly luciferase signal driven by the activity of RUNX2 to the renilla luciferase
signal generated by the CMV promoter on each plasmid.
Statistical Analyses
Statistics were performed using Sigma Stat (SysStat Software). Data was presented as ±
SEM. To test for statistical significance, in vitro calcification assays and densitometry on
immunoblots were analyzed using a paired Student’s t‐test. qRT‐PCR, nuclear localization
quantification and luciferase assays were analyzed using Two way ANOVA. Pairwise
comparisons between groups were made following the Two way ANOVA using Tukey’s posthoc
test. Statistical significance was set at P<0.05 for all tests.
40
RESULTS
DDR1 Deficiency Attenuates Vascular Smooth Muscle Cell Calcification of Matrix
To study the role of DDR1 in VSMC mediated calcification, I utilized a high phosphate
osteogenic model of in vitro calcification adapted from Speer et. al.69. After 12 days of culture,
total calcium content in the extracellular matrix in Ddr1‐/‐ VSMCs was reduced by 43% when
compared with Ddr1+/+ VSMCs after normalizing to total protein level (Figure 7A). Ddr1‐/‐ VSMCs
showed a 38% reduction in calcium content compared to Ddr1+/+ VSMCs after normalizing to
total DNA level (Figure 7B). It is clear that VSMC mediated calcification is diminished in the
absence of DDR1.
DDR1 Deficiency Alters Expression of Phosphate Handling Genes in Vascular Smooth Muscle
Cells at an Early, but not at a Late Stage of Calcification
Previous research has suggested that vascular calcification results from an imbalance of
the Pi to PPi levels in VSMCs10, which is in turn due to an imbalance of phosphate regulators. I
therefore hypothesized that the blunted calcification observed in Ddr1‐/‐ VSMCs was due to
altered expression of genes involved in phosphate handling. The mRNA expression of various
phosphate regulatory molecules was measured using qRT‐PCR. I first looked at Pit‐1 and Alp,
which are involved in increasing intracellular phosphate levels. Pit‐1 mRNA was increased in
Ddr1‐/‐ VSMCs after 2 days in osteogenic culture, but by 12 days Pit‐1 mRNA expression was
decreased and there were comparable levels in both genotypes (Figure 8A). I saw a significant
decrease in the mRNA expression of Alp at 2 days in Ddr1‐/‐ VSMCs but by 12 days, both Ddr1‐/‐
and Ddr1+/+ VSMCs increased their expression and there were no differences when comparing
41
the two cell types (Figure 8B). This data shows that deficiency of DDR1 affects expression of Pit‐
1 and Alp at an early stage, but by late stage calcification the gene expression of Pit‐1 and Alp
mRNA in Ddr1‐/‐ VSMCs catches up to Ddr1+/+ VSMC levels.
Next I looked at mRNA levels of the pyrophosphatase, Npp1, which inhibits calcification
by promoting the accumulation of PPi. I saw a significant decrease in Npp1 mRNA expression in
Ddr1‐/‐ VSMCs at the 2 day time point, but expression levels in Ddr1‐/‐ and Ddr1+/+ VSMCs
increased and there were no significant differences between the cell types at 12 days (Figure
8C). This indicated to me that loss of DDR1 affects the mRNA expression of npp1 in VSMCs
during early stage calcification, but not at late stage calcification.
Together my results indicate that loss of DDR1 impacts the mRNA expression of Pit‐1,
Alp, Npp1 at an early point of calcification, but not at a late stage. This indicates that DDR1
delays the gene expression of phosphate handling genes in response to high phosphate. If these
changes in mRNA expression carry forward to protein expression changes, they will likely
impact phosphate handling, nucleation of calcium apatite and thus vascular calcification.
In my experiments I saw a decrease in Alp expression in Ddr1‐/‐ VSMCs. This could lead
to decreased Pi production by these VSMCs and thus less calcium phosphate crystal deposition.
I also observed a decrease of Npp1 expression in Ddr1‐/‐ VSMCs, which seems counterintuitive
as these cells clearly show decreased calcification. However PPi produced by NPP1 can also be a
substrate for ALP, producing two Pi molecules178. Therefore decreased NPP1 levels in Ddr1‐/‐
VSMCs may also lead indirectly to less vascular calcification. With respect to Pit‐1 expression
levels, Pit‐1 mRNA expression is likely increased in the Ddr1‐/‐ VSMCs in response to the
decrease in available Pi. Over time, these gene expression differences as a result of DDR1
42
absence are overcome. Nevertheless, the early expression of phosphate handling proteins may
provide a quick start to calcium phosphate nucleation in Ddr1+/+ VSMCs. Therefore while the
Ddr1‐/‐ VSMCs may eventually match the Ddr1+/+ VSMCs in phosphate handling, they are
unlikely to reach the same extent of matrix calcification.
DDR1 Deficiency Increases the Expression of TenC at Late Stage Calcification, but not αSMA
and Atf4
During vascular calcification, there is VSMC transdifferentiation into an osteoblast‐like
phenotype with the gradual loss of the VSMC phenotype69. I wanted to determine whether loss
of DDR1 attenuated VSMC phenotypic transition whereas DDR1 is permissive to
transdifferentiation. To address this I first examined the protein levels of the smooth muscle
cell marker Alpha Smooth Muscle Actin (αSMA). Immunoblotting to detect αSMA revealed no
significant differences comparing Ddr1‐/‐ and Ddr1+/+ VSMCs at the 12 day timepoint (Figure
9A). This indicated to me that deficiency of DDR1 did not alter the expression of αSMA under
osteogenic conditions.
During vascular calcification, VSMC proliferation is inhibited to promote
osteochondrogenic transdifferentiation179. I therefore hypothesized that Ddr1 deletion may
increase VSMC proliferation, thus preventing transdifferentiation. To determine this I examined
two markers of smooth muscle proliferation, Activated Transcription Factor 4 (Atf4)180 and
Tenascin C (TenC)181. qRT‐PCR of Ddr1‐/‐ and Ddr1+/+ VSMCs showed no differences in the
expression of Atf4 between the genotypes at either time point (Figure 9B). However there was
43
a decrease in the mRNA expression of Atf4 from 2 days to 12 days in both Ddr1‐/‐ and Ddr1+/+
VSMCs.
Following this, I examined the mRNA expression of TenC under the same osteogenic
conditions and saw no differences between Ddr1‐/‐ and Ddr1+/+ VSMCs at 2 days in osteogenic
culture. However, by 12 days the Ddr1‐/‐ VSMCs showed a significant increase in the expression
of TenC (Figure 9C). Taken together these results imply that loss of DDR1 does not alter the
expression of αSMA or Atf4. However, the loss of DDR1 does increase the expression of TenC at
late stage calcification. In previous literature, it has been shown that TENC promotes VSMC
proliferation through an autocrine and paracrine process181. As the calcification of VSMCs
requires proliferation to be halted in order for transdifferentiation into an osteochondrogenic
phenotype, the high expression of TenC mRNA in Ddr1‐/‐ VSMCs may result in increased TENC
protein expression, thus restricting transdifferentiation by promoting proliferation.
DDR1 Deficiency Results in Reduced RUNX2 Activity and Osteocalcin Expression in Vascular
Smooth Muscle Cells
The transcription factor RUNX2 mediates the transdifferentiation of VSMCs to an
osteoblast‐like phenotype69, 173. Therefore, I was interested in determining whether DDR1
regulates either RUNX2 expression or its activity. The mRNA expression of Runx2 increased
after culturing in osteogenic media and peaked at 2 days in both Ddr1‐/‐ and Ddr1+/+
VSMCs(Figure 10A). By 12 days however the Runx2 mRNA diminished back to pre osteogenic
culturing levels. Despite the upregulation in Runx2 mRNA expression with osteogenic culture,
the levels of Runx2 mRNA in Ddr1‐/‐ and Ddr1+/+ VSMCs showed no differences at any time
44
point. Immunoblotting revealed that RUNX2 protein expression increased from 2 days to 12
days in VSMCs. However, similar to mRNA levels there were no significant differences in RUNX2
protein expression between Ddr1‐/‐ or Ddr1+/+ VSMCs at either the 2 day (Figure 10B), or the 12
day (Figure 10C) time points. Overall, this indicated to me that DDR1 did not change the
expression of RUNX2.
To determine whether RUNX2 activity was altered between Ddr1‐/‐ and Ddr1+/+ VSMCs, a
luciferase activity assay was performed on cells that were transfected with 6OSE2‐LUC
construct. The 6OSE2‐LUC construct measures RUNX2 activity using a firefly luciferase reporter.
Ddr1‐/‐ VSMCs showed a decrease in RUNX2 activity compared to Ddr1+/+ VSMCs even prior to
the addition of osteogenic media (Figure 11A). The RUNX2 activity remained unchanged in
Ddr1‐/‐ VSMCs after 2 days of osteogenic treatment, whereas RUNX2 activity in Ddr1+/+ VSMCs
drastically increased (Figure 11A). These experiments indicated that Ddr1 deletion prevented
RUNX2 activation under osteogenic conditions.
Next, I wanted to determine whether the expression of Ocn, a gene regulated by
RUNX2, would reflect the results observed with the activity assay. Ocn mRNA expression
showed a decrease in expression in Ddr1‐/‐ VMSCs at 0, 2 and 12 days in osteogenic culture
when compared to Ddr1+/+ VSMCs (Figure 11B). In addition I saw an increase in Ocn expression
in Ddr1+/+ VSMCs after osteogenic treatment. This increase was absent in Ddr1‐/‐ VMSCs. Taken
together these results suggest that VSMCs were undergoing osteogenic transdifferentiation and
the absence of DDR1 attenuated the process.
45
P38 inhibition Reduces RUNX2 Activity and Osteocalcin mRNA Expression, but not RUNX2
Nuclear Localization in DDR1 Expressing Vascular Smooth Muscle Cells
The effective nuclear transport of RUNX2 is important in regulating its activity. To test
whether DDR1 affects the nuclear transport of RUNX2 I immunostained for RUNX2 in Ddr1‐/‐
and Ddr1+/+ VSMCs in osteogenic culture. There was a significant decrease in RUNX2 nuclear
localization in Ddr1‐/‐ VSMCs compared with Ddr1+/+ VSMCs at both 2 and 12 days (Figures 12A,
B and C). Next, since ERK1/2 and P38 are activated downstream of DDR1 signalling and are
activators of RUNX2 5, 9, 17, 18, I determined whether inhibiting MEK or P38 would alter DDR1
dependant RUNX2 activity (Figure 13A) and Ocn mRNA expression. Treatment with MEKi did
not alter RUNX2 activity or Ocn mRNA levels (Figure 13B) in Ddr1‐/‐ or Ddr1+/+ VSMCs compared
to their respective DMSO controls. However, treatment with P38i significantly decreased
RUNX2 activity in the Ddr1‐/‐ and Ddr1+/+ VSMCs (Figure 13A). P38i significantly reduced Ocn
expression levels in Ddr1+/+ VSMCs, but did not alter levels in Ddr1‐/‐ VSMCs (Figure 13B). The
differences in RUNX2 activity and Ocn mRNA expression were not due to change in Runx2
mRNA expression, as P38 inhibition treatment failed to alter RUNX2 mRNA (Figure 13C) This
shows that P38 inhibition decreased RUNX2 activity and not expression of mRNA.
Next I determined whether the decrease in RUNX2 activity after P38 inhibition was due
to altered RUNX2 nuclear localization. I found that neither MEKi (Figure 14B) nor P38i (Figure
14C) affected RUNX2 nuclear localization in Ddr1‐/‐ and Ddr1+/+ VSMCs. Thus, although P38
signalling may play a role in the activation of RUNX2, it does not affect the nuclear localization
of RUNX2 under osteogenic conditions in VSMCs. This suggested to me that DDR1 regulates the
nuclear localization of RUNX2 by another mechanism.
46
DDR1 Deficiency Impairs phosphoP38 Nuclear Localization
I wanted to understand ERK and P38 signalling in DDR1 modulation of osteogenically
cultured VSMCs, because DDR1 was previously shown to activate ERK1/2 and P389, 17, 18. To
address this, I performed immunoblotting of ERK1/2 and P38 in Ddr1‐/‐ and Ddr1+/+ VSMC.
There were no differences in ERK1/2 or P38 phosphorylation comparing Ddr1‐/‐ and Ddr1+/+
VSMCs after 2 days of culture (Figures 15A and 16A). However, by 12 days Ddr1‐/‐ VSMCs had
significant decreases in ERK1/2 and P38 phosphorylation compared to Ddr1+/+ VSMCs (Figures
15B and 16B).
Although I did not see an alteration in P38 phoshorylation at 2 days between Ddr1‐/‐ and
Ddr1+/+ VSMCs (Figure 15A), I saw a decrease in RUNX2 activity in Ddr1‐/‐ VSMCs compared to
Ddr1+/+ VSMCs (Figure 11A). In addition, I saw that RUNX2 activity in Ddr1+/+ VSMCs was
sensitive to P38 inhibition (Figure 13A). Therefore I hypothesized that in the absence of DDR1
the delivery of phosphoP38 to RUNX2 in the nucleus might be hindered. Indeed, phospho P38
failed to localize to the nucleus in Ddr1‐/‐ VSMCs, but was present in the nucleus of Ddr1+/+
VSMCs (Figure 17 A and B).
DDR1 Regulates RUNX2 Nuclear Localization and Activity by Modulating a Dynamically
Unstable Microtubule Network in Vascular Smooth Muscle Cells
The loss of DDR1 disrupts both phosphoP38 and RUNX2 nuclear localization, Therefore,
DDR1 likely regulates a mechanism common to both proteins. Previously, the expression of
DDR1 was shown to regulate the microtubule network in mesenchymal stem cells15. As
microtubules regulate the nuclear localization of phosphoP38 and RUNX2131, 182, I wanted to
47
determine whether the microtubule network was disrupted in DDR1‐deficient VSMCs in
osteogenic culture. I first analyzed cell morphology by examining Ddr1‐/‐ and Ddr1+/+ VSMCs in
osteogenic media using differential interference contrast microscopy. I saw that Ddr1+/+ VSMCs
cultured in osteogenic media for 2 days had a more slender cell shape, compared to Ddr1‐/‐
VSMCs of the same cohort, which exhibited a more spread cell shape (Figure 18A). Also, α‐
tubulin staining of Ddr1‐/‐ VSMCs after 2 (Figure 18B) and 12 days (Figure 18C) showed diffuse
cytoplasmic staining, while Ddr1+/+ VSMCs showed strong staining of well organized
microtubules. Since I observed differences in microtubule organization between Ddr1+/+ and
Ddr1‐/‐ VSMCs, I next wanted to determine whether RUNX2 binding to the microtubule network
was affected. Immunostaining revealed that RUNX2 shows relatively little co‐localization with
the diffuse cytoplasmic tubulin in Ddr1‐/‐ VSMCs, whereas RUNX2 strongly co‐localizes with the
well defined microtubules in the Ddr1+/+ VSMCs (Figure 19). This indicates that RUNX2 does not
bind to diffuse cytoplasmic tubulin.
I next hypothesized that experimental disruption of microtubules to prevent
microtubule polymerization would prevent the nuclear localization and activity of RUNX2 in
Ddr1+/+ VSMCs. I chose to use nocodazole, as it binds to α and β‐tublin heterodimers to
accelerate GTPase activity in β‐tubulin to prevent polymerization into microtubules183. After
depolymerisation of microtubules with nocodazole there was a significant reduction in nuclear
localization of RUNX2 in the Ddr1+/+ VSMCs (Figure 20B ) compared to DMSO treated controls
(Figure 20A). I then analyzed RUNX2 activity and Ocn mRNA expression in Ddr1‐/‐ and Ddr1+/+
VSMCs treated with nocodazole. Treatment with nocodazole did not affect RUNX2 activity in
Ddr1‐/‐ VSMCs, but significantly decreased RUNX2 activity in Ddr1+/+ VSMCs (Figure 21A). This
48
indicated that DDR1 mediated RUNX2 activity is sensitive to microtubule depolymerization.
However Ocn mRNA expression under nocodazole treatment did not follow that of RUNX2
activity. Instead nocodazole significantly increased Ocn mRNA expression in both Ddr1‐/‐ and
Ddr1+/+ VSMCs (Figure 21B). Thus Ocn mRNA expression may be influenced by factors other
than RUNX2 after microtubule destabilization.
I next hypothesized that stabilizing microtubules with the microtubule stabilizing agent
would affect RUNX2 localization and activity in a DDR1 dependant manner in VSMCs. Taxol, the
microtubule stabilizing agent binds a hydrophobic pocket in β‐tubulin to promote
polymerization in absence of GTP184. The addition of taxol was able to rescue the nuclear
localization of RUNX2 in Ddr1‐/‐ VSMCs, (Figure 20C). I therefore hypothesized that RUNX2
activity and Ocn mRNA in Ddr1‐/‐ VSMCs would be similarly rescued by the addition of taxol.
However, treatment with taxol did not affect RUNX2 activity in Ddr1‐/‐ VSMCs, but significantly
decreased RUNX2 activity in Ddr1+/+ VSMCs (Figure 21A). By contrast, taxol significantly
increased Ocn mRNA expression in both Ddr1‐/‐ and Ddr1+/+ VSMCs (Figure 21B). These results
suggest that Ocn mRNA expression is subject to other factors related to microtubule kinetics,
which are independent of RUNX2 activity in VSMCs. I then wanted to demonstrate that
treatments with nocodazole and taxol decreased RUNX2 activity by altering microtubule
stability but not by decreasing RUNX2 expression. Runx2 mRNA levels showed that nocodazole
or taxol treatment did not diminish levels, but rather increased expression (Figure 21C). Taken
together the results suggest that DDR1 mediated RUNX2 activity by stabilizing microtubule
translocation. Hindrance of RUNX2 activity by microtubule depolymerizing or stabilizing agents
49
indicates the DDR1 mediated activation of RUNX2 requires the dynamic instability of
microtubules.
DISCUSSION
DDR1 in matrix calcification
In the current study, I examined the role of DDR1 in VSMC mediated calcification. To
begin, I utilized an in vitro osteogenic model and examined the functional consequences of
Ddr1 deletion in VSMCs on matrix calcification. I found that Ddr1 deletion blunts calcifying
potential in VSMCs cultured in osteogenic media for 12 days. Previous work has linked type I
collagen to VSMC transdifferentiation into an osteoblast‐like calcifying phenotype 20, 149, 185.
However, work on another type I collagen receptor, integrin β1 demonstrated that it
functioned to prevent calcification instead149. While the exact mechanism by which integrin β1
is able to inhibit calcification is not entirely clear, the pro‐ or anti‐calcifying properties of type I
collagen are likely to be receptor‐specific. DDR1 represents a receptor through which type I
collagen may be able to promote a calcifying phenotype. Although DDR1 interacts with type I
collagen, it also interacts with types II‐VI and VIII collagens, all of which are present in the
atherosclerotic plaque30. I therefore could not clearly identify which collagen type(s) interacted
with DDR1 to promote osteochondrogenic differentiation in my study. Although it has not been
implicated in VSMCs transdifferentiation or calcification, type II collagen is able to induce
chondrogenesis in mesenchymal stem cells186 and adipocytes187. Other studies, including a
previous study from our laboratory, have demonstrated that Type II collagen is deposited
50
within calcified regions the atheroscleroslerotic plaque59, 188. This represents another collagen
which may regulate the osteochondrogenic transdifferentiation process in VSMCs. My study is
the first to demonstrate that DDR1 is able to accelerate the calcifying potential of VSMCs and
provides a receptor through which collagens may be able to regulate vascular calcification.
Phosphate Handling in DDR1 Mediated Calcification
I observed alterations in the expression of Pit‐1, Alp and Npp1 in Ddr1‐/‐ VSMCs during
early but not late stage calcification. My results showed that during early stage calcification
both Alp and Npp1 expression were decreased in the Ddr1‐/‐ VSMCs while the expression of Pit‐
1 was increased. While the decrease and increase in Npp1 and Pit‐1 expression in the Ddr1‐/‐
VSMCs respectively did not seem to be in line with the decreased calcification observed, it is
likely the decrease in Alp expression was able to overcome the Npp1 and Pit‐1 expression levels
to decrease calcification in the Ddr1‐/‐ VSMCs. Furthermore, the increase in Pit‐1 expression in
the Ddr1‐/‐ VSMCs maybe a compensatory mechanism in response to the decrease in Alp
expression in order to increase intracellular Pi levels. Although at late stage calcification both
Ddr1+/+ and Ddr1‐/‐ VSMCs had no significant changes in expression of these genes, the early
changes in phosphate handling genes were likely sufficient to create the differences in
calcification between Ddr1+/+ and Ddr1‐/‐ VSMCs.
I also observed in the course of this study, changes in the mRNA levels of Pit‐1, Alp and
Npp1 from early to late stage calcification. As Alp is a marker for osteogenic transdifferentiation
it was not surprising to see its levels increase as calcification progresses as is reported in
51
literature121. Conversely, Pit‐1 mRNA expression was reduced as calcification progressed. While
the change of expression in Pit‐1 through the progression of calcification has not been
described in literature, it is possible that there are sufficient intracellular levels of Pi at late
stage calcification and hence the reduction in Pit‐1 mRNA expression at that time point.
Another possibility for the reduction of Pit‐1 expression is to compensate for the increased Alp
expression at that time.
Finally, the decrease in Npp1 mRNA levels over the course of vascular calcification has
been described in literature. In most studies the decrease of NPP1 expression is the cause of
vascular calcification104, 189. As a result NPP1 expression decreases with the progression of
calcification. While it has been determined this is the cause of vascular calcification, it is
unknown whether this also occurs as a result of calcification. While my results may seem
contradictory to current literature, it may also be possible that Npp1 increases in expression as
a negative feedback mechanism to prevent calcification. Opn is an example of a gene which is
associated with the development of vascular calcification and is commonly used as a marker of
the osteochondrogenic phenotype93. However it was discovered that OPN expression inhibited
vascular calcification82, 190. Together it appears that DDR1 alters the expression of Pit‐1, Alp and
NPP1 mRNA levels during calcifying conditions, which likely affects vascular calcification.
Smooth Muscle Phenotype in DDR1 Mediated Calcification
VSMCs are the primary cell type which transdifferentiates into an osteoblast like cell
type to promote calcification of the vascular tissue69. As the VSMCs transdifferentiates into a
52
osteoblast like cell, its smooth muscle characteristics are lost69. I determined whether or not
DDR1 affects the smooth muscle phenotype by examining the expression of smooth muscle cell
markers. I determined in my studies that absence of DDR1 does not increase the expression of
the smooth muscle marker αSMA. While it is generally accepted that VSMC markers decrease in
expression as vascular calcification progresses, most documented studies show that they will
continue to express VSMC markers69, 158. Another explanation would be that DDR1 has no affect
on the VSMC phenotype during calcification. Despite this, I have not performed an exhaustive
screen of all VSMC markers. Validated markers of the VSMC phenotype are SM22α, Myocardin
and SM‐MHC69, 158. Although useful as a gauge to determine VSMC phenotype, many studies
have misidentified cells as being VSMC on the basis of an inadequate screen of markers191‐193.
Furthermore, VSMCs have also been misidentified as other cell types due to a temporary loss of
markers during injury or migration194, 195. Despite this, experimental evidence from Giachelli
and colleagues showed that expression of VSMC markers does not impact osteochondrogenic
pathways or hinder calcification158. Thus, the expression level of VSMC markers likely has no
impact the calcifying potential in my study.
I have also observed a significant increase in the expression of TenC mRNA, but no
change in the expression of Atf4 in Ddr1‐/‐ VSMCs. As TENC acts to promote VSMC
proliferation181, Ddr1‐/‐ VSMCs might have higher rates of proliferation under my experimental
conditions, however I did not measure proliferation in my studies ATF4 is also a marker of
proliferation180, but unlike TenC, I did not see any significant differences between Ddr1+/+ and
Ddr1‐/‐ VSMCs. Despite this, the increased TenC levels in Ddr1‐/‐ VSMCs may be sufficient to
promote a higher rate of proliferation in Ddr1‐/‐ VSMCs. In line with my study, Curat et. al.
53
previously reported that Ddr1‐/‐ mesangial cells show higher rates of proliferation than Ddr1+/+
cells196. As osteochondrogenic transdifferentiation occurs, VSMC proliferation needs to be
suppressed179. This continued signal for proliferation in the Ddr1‐/‐ VSMCs may serve to inhibit
calcification in these VSMCs. Together; these results indicate that DDR1 does not alter VSMC
phenotype under calcifying conditions, but may suppress VSMC proliferation to allow for
osteochondrogenic conversion. However I have not yet measured proliferation.
During endochondral ossification, the osteoid is synthesized by chondrocytes,and is later
calcified by osteoblasts97. While there is limited evidence that chondrocytes transdifferentiate
into osteoblasts197, it is believed that osteoblasts stem largely from the common
osteochondrogenic progenitor198. As VSMCs in atherosclerotic calcification frequently express
both osteoblast and chondrocyte markers, the predominately held view is that atherosclerotic
calcification mirrors endochondral ossification. Indeed, a previous study by our laboratory
showed that chondrocyte markers such as SOX9, Type II and Type X collagens were expressed in
atherosclerotic plaques59. In this present study I also observed that DDR1 deletion decreases
osteoblast markers such as Ocn and Alp expression. However I did not screen for chondrocyte
markers. During arterial calcification, it is not clear whether the expression of osteoblast and
chondrocyte markers represents one homogenous population of cells performing both roles, or
two separate populations of cells each with separate roles. Futhermore, prior work on
ossification suggests that the chondrogenic phenotype antagonizes the osteoblast phenotype
through the expression of SOX9199. Therefore, if DDR1 is able to alter the expression of SOX9,
this represents a mechanism by which DDR1 can regulate the osteoblast phenotype and thus
calcification. Despite the incomplete information regarding VSMC transdifferentiation during
54
atherosclerotic calcification, it is clear that during atherosclerotic calcification VSMCs lose
smooth muscle characteristics, and have transdifferentiated into cells which have
characteristics of chondrocytes and or osteoblasts.
DDR1 Mediated RUNX2 activation
In this study I did not observe any DDR1 dependant differences of RUNX2 mRNA or
protein expression. However I did see an increase in Runx2 mRNA expression which peaked at
day 2 and declined back to baseline levels by day 12. RUNX2 protein levels increased steadily
from 2 to 12 days. The increase in RUNX2 protein over the course of vascular calcification is well
documented and is consistent with my findings109, 157. It is possible that Runx2 mRNA expression
is decreased as a result of negative feedback from the RUNX2 protein that accumulates. It is
noted that RUNX2 changes isoforms from the long lived type I to the rapidly degraded type II at
late stage calcification in bone development134. It would therefore be expected that Runx2
mRNA levels would be increased in order to replenish the rapidly degraded type II RUNX2
molecules at late stage calcification. However, this was not observed in my study. While the
conversion of type I to type II RUNX2 has been established in bone development134, this process
may not occur in VSMCs undergoing osteochondrogenic trandifferentiation. Furthermore, the
RUNX2 isoform switch may occur at a much later time point in VSMCs which this study has not
analyzed.
There were no genotype‐dependent differences in RUNX2 mRNA or protein levels, but I
did see a significant decrease in RUNX2 activity in Ddr1‐/‐ VSMCs. Paradoxically I saw that
55
RUNX2 activity was sensitive to P38i treatment despite not showing any differences in activated
phosphoP38 levels at that stage of calcification. I attributed this to the regulation of RUNX2 by
microtubules, which will be discussed in depth later. Even though P38 activation may not play a
role in activating RUNX2 in a DDR1 dependant manner at this stage, it cannot be discounted as
having an effect in DDR1 regulation of RUNX2. I observed a significant decrease in phosphoP38
levels at the 12 day time point. It is conceivable that while VSMCs are in culture they
synthesized much more matrix by day 12 and hence there is an increase in DDR1 receptor
activation by day 12 due to increased number of DDR1 ligands in the cell culture. As P38
signalling has been demonstrated as an important activator of RUNX2147, 177, it would likely alter
RUNX2 activity at this stage. Unfortunately the effect of P38 inhibition on RUNX2 activity at late
stage calcification could not be assessed due the sensitivity of Ddr1‐/‐ VSMCs to cell death under
P38 inhibition.
I also found that inhibition of ERK1/2 signalling did not affect RUNX2 activity in VSMCs
Previous literature reported that ERK1/2 is an activator of RUNX2, and a even more potent
activator of RUNX2 than P38147 in mesenchymal stem cells and osteoblasts.. Taken together, my
results demonstrate for the first time that DDR1 is able to regulate the activity RUNX2; however
this is not dependent upon ERK1/2 or P38 activity at early stages of calcification.
Microtubles in DDR1 Mediated RUNX2 activity
Previously there have been no works linking DDR1 to the modulation of microtubules or
the nuclear translocation of proteins. My study provides evidence that DDR1 is able to regulate
56
the microtubule network which controls RUNX2 nuclear localization and activity. As the
regulation of vascular calcification by any component of the cytoskeleton has never been
explored, my work on the regulation of RUNX2 by DDR1 via microtubules may represent a
paradigm shift for further investigations on vascular calcification.
My study also showed that phosphoP38, another protein requiring microtubules for
nuclear transport, also showed reduced nuclear localization in Ddr1‐/‐ VSMCs. While it was
previously thought that most proteins translocate into the nucleus through Brownian motion,
there is a growing body of literature that shows many proteins such as steroid hormone
receptors200, transcription factors131, 131, 171 and signalling kinases182 which require the use of
microtubules.
I have shown that RUNX2 co‐localizes with the microtubule network in the Ddr1+/+
VSMCs, but not in Ddr1‐/‐ VSMCs. This is consistent with studies by Pockwinse et. al. who
showed that RUNX2 binds to microtubules131. Thus, it was not surprising to see that
depolymerizing microtubules by treatment with nocodazole prevented both RUNX2 nuclear
localization and activity in Ddr1+/+ VSMCs. Many previous studies have shown that either
stabilizing or preventing the polymerization of microtubules alters transport of protein cargoes.
In contrast, I found that stabilizing microtubules by treatment with taxol rescues RUNX2
nuclear localization in Ddr1‐/‐ VSMCs. This finding was consistent with the results of a study by
Pockwinse et. al who showed that taxol treatment increased nuclear population of RUNX2 and
decreased its cytoplasmic population 131. As microtubule facilitated nuclear transport is a
relatively new area of investigation, there may be other proteins whose transport into the
57
nucleus depends solely on stabilization. Despite rescuing nuclear localization in Ddr1‐/‐ VSMCs,
the stabilization of microtubules by taxol was unable to restore RUNX2 activity. Since taxol acts
to lock microtubules in a stable but static state, it prevents dynamic instability. Therefore my
results suggest that activation of RUNX2 may require the dynamic instability of microtubules.
Although we have not tested this, a possible explanation for this is that phosphoP38 requires
dynamic instability as previously demonstrated182. As RUNX2 phosphorylation by phosphoP38
likely occurs in the nucleus, restriction of phosphoP38 in the cytosol will effectively prevent
RUNX2 activation. Another possibility is that the incubation time with taxol used in this assay
was not sufficient to allow the newly translocated RUNX2 to act upon the reporter construct.
Unfortunately, longer term treatments with taxol result in cell death and detachment so I am
not able to assess the long term effects on RUNX2 activity and in vitro calcification.
In my studies I found that Ocn mRNA expression was increased with either taxol or
nocodazole treatment in both genotypes. Thus Ocn expression was induced even in the
absence of RUNX2 activation. A possible explanation lies with the differences between the
reporter promoter construct I used and the Ocn promoter. Although both 6OSE2‐LUC and the
Ocn promoter contain six OSEs, the Ocn promoter may contain binding sites for other
transcriptional elements. Although numerous pathways have been shown to activate Ocn
expression, they all converge on RUNX2 activation201‐204. It is entirely possible that when
manipulating microtubules, factors which have not yet been identified act on Ocn expression.
Vegf is another gene which contains OSEs, it however also contains HIF1α Responsive Elements
which are activated by HIF1α and entirely independent of RUNX2205. Since the 6OSE2‐Luc
construct is only activated by RUNX2; its measured activity it is likely more accurate.
58
DDR1 regulation of microtubules, although poorly explored, it is not unprecedented131.
There have also been many documented cases where DDR1 is shown to regulate processes
related to microtubule function. Cell spreading for example is a process that is regulated by a
dynamic microtubule network206. In this study, I observed Ddr1‐/‐ VSMCs also showed more cell
spreading than Ddr1+/+ VSMCs. This finding is consistent with a previous study which showed
that DDR1 suppressed the spreading of HEK293 cells 207. Studies on VSMCs have also hinted at
DDR1 regulation of microtubules. Cell migration is a process that depends on microtubule
action208. As previously discussed in the introduction, Ddr1‐/‐ VSMCs show significantly reduced
cell migration52, 131. Taken together, previous literature and my results suggest that DDR1
regulates dynamic instability in microtubules which is required for RUNX2 nuclear translocation
and activity during VSMC calcification.
Future Directions
In the current study I identified a mechanism by which DDR1 is able to control the
osteogenic process, determining that DDR1 mediate the nuclear translocation of RUNX2 via the
microtubule network. Also of interest is whether kinase activity of DDR1 is required to
modulate the microtubule network and RUNX2 nuclear translocation. Although there are many
functions of DDR1 which require the phosphorylation, other functions such VSMC attachment
to matrix and controlling the collective migration of cells does not require its kinase activity209.
Therefore it may be possible that DDR1 regulates microtubules through a kinase independent
pathway. This can easily be tested by adding the RTK phosphorylation inhibitor, Imatinib in
59
VSMCs during osteogenic culture. Although Imatinib inhibits all RTKs, it most strongly inhibits
the phosphorylation of DDR1210. This method can be used to determine whether or not the
translocation of RUNX2 requires the phosphorylation of DDR1. To further validate this
mechanism, a kinase dead construct of DDR1 can be transfected into Ddr1‐/‐ VSMCs cultured in
osteogenic conditions in order to determine if it affects RUNX2 translocation.
Work on osteogenic development shows that type I collagen is a pro‐osteogenic
molecule10, 93, 209. While my work implies that DDR1 represents a functional link between type I
collagen and RUNX2 activity, further work is need to determine this. Although DDR1 is a Type I
collagen receptor, it is also able to bind other collagen ligands185. Although studies have shown
that VSMCs synthesize type I collagen in vitro, VSMCs also synthesize other collagens which are
DDR1 ligands185, 211. In order to fully demonstrate which collagen is able to elicit the pro‐
osteogenic effects through DDR1, VSMCs can be cultured in osteogenic conditions with the
addition of separate collagen ligands for DDR1. This will help determine which DDR1 ligand(s) if
any affect the microtubule network which impacts RUNX2 translocation.
Furthermore, despite numerous lines of evidence suggesting that DDR1 impacts
microtubule related functions such as cell migration, invasion and proliferation 18, 29, 57, 58, 131, 212,
the mechanism by which DDR1 regulates microtubules, during osteogenic conditions or
otherwise remains unknown. My study suggests that there is a difference in the dynamic
instability between microtubules in Ddr1‐/‐ and Ddr1+/+ VSMCs. However, in my study I was only
able to examine microtubule structure at the endpoints of calcification due to available
materials and reagents. Further studies can be carried out to examine image the microtubule
60
network throughout the calcification process with Tubulin Tracker (Life Technologies) which is
flurophore conjugated to a microtubule binding compound which does not hinder microtubule
formation. This compound will allow me to determine microtubule growth rates and cycling as
modulated by DDR1 through live cell microscopy. In addition, the dynamic instability of
microtubules is regulated by two main factors, acetylation of tubulin and the binding of
microtubule associated proteins to tubulin213. Hypothetically DDR1 could regulate either or
both of these factors to regulate dynamic instability of microtubules. This can further be
investigated by using an acetyled tubulin antibody for immunofluoresce and immunoblotting in
the cultured cells. Candidate microtubule associated proteins can also be identified through
mass spectrometry.
It is interesting to note that under circumstances where there are large concentrations
of the transcription factor in question, the requirement for microtubules in nuclear localization
is uncoupled171, 214. Since RUNX2 expression in both DDR1 expressing and Ddr1‐/‐ VSMCs
increases over time in osteogenic culture, it would be interesting to investigate whether
microtubule requirement for nuclear localization is lost over time. This can also be determined
by overexpressing RUNX2 in DDR1 expressing and Ddr1‐/‐ VSMCs in osteogenic culture and
determining if RUNX2 overexpression eliminates the need for DDR1 regulation of microtubules
for nuclear translocation.
61
Conclusions
My study is the first to identify the mechanistic ability of DDR1 to regulate VSMC
calcification. DDR1 is crucial in the regulation of osteochondrogenic transcriptional activity
through its regulation of RUNX2 activity in VSMCs. I have demonstrated that DDR1 regulates
the dynamic instability of microtubules in VSMCs. In absence of DDR1, microtubule
polymerization is hindered and thus RUNX2 is unable to bind to unpolymerized tubulin
molecules, thus unable to enter the nucleus. I have also demonstrated that inducing
microtubule polymerization in DDR1 KO VSMCs, rescues nuclear localization. Although the
finding that polymerized microtubules results in RUNX2 nuclear translocation is not
unprecedented131, this study is the first to show that rescuing RUNX2 nuclear localization does
not rescue activity. Although the mechanism has not been thoroughly investigated the
cytoplasmic abundance of phosphoP38 during the absence of DDR1, hints that DDR1 regulation
of microtubule has an effect on phosphoP38 localization as well. The effect of DDR1 regulation
of microtubules in VSMCs does not seem to be limited to nuclear translocation alone. Cell
spreading a common phenomenon in cells with hindered microtubule regulation215, is also
prevalent in the absence of DDR1 in VSMCs in my study. In conclusion my study demonstrates
that DDR1 a novel collagen receptor regulates VSMC calcification through microtubule
moderation of RUNX2 activity. The proposed model of DDR1 regulation of RUNX2 activity is
illustrated (Figure 22).
62
FIGURES
Figure 1: Diagram of Normal Artery and Artery Affected by Atherosclerosis
This diagram depicts a normal artery (left) and an artery affected by Atherosclerosis. Diagram
was taken from http://www.pharmaceutical‐networking.com.
63
Figure 2: Diagram of DDR1 Isoforms in Humans and Mice
This diagram depicts the five different DDR1 isoforms in humans and mice. Disocoidin Domain
(DD), Transmembrane (TM), Juxatamembrane (JM), Tyrosine Kinase (TK). Green represents 37
AA insert in the juxtamembrane region and red represents the six AA insert in the tyrosine
kinase region.
64
Figure 3: Diagram of Molecules Which Regulate Pi and PPi Levels
This diagram depicts molecules which regulate Pi and PPi levels within a cell and how they
ultimately give rise or inhibit calcification.
65
Figure 4: Diagram of RUNX2 Isoforms in Humans and Mice
This diagram depicts the two different RUNX2 isoforms in humans and mice. Amino Terminus
(NT), RUNT DNA binding domain, Nuclear Translocation Matrix (NTM) and Destabilizing PEST
domain. Red represents the region which contains the microtubule binding site.
66
Figure 5: Diagram of RUNX2 Stimulatory and Inhibitory Phosphoserine Residues
This diagram depicts the known stimulatory and inhibitory phosphoserine residues on RUNX2
and how they are activated by P38 and ERK1/2.
67
Figure 6: Diagram of Microtubule Dynamics
This diagram depicts the dynamic instability of microtubules. GTP bound tubulin promotes
polymerization, while GDP bound tubulin promotes disassembly. Diagram taken from Nature
Reviews Cancer 4, 253‐265 (April 2004)
68
Figure 7: Ddr1‐/‐ VSMCs Show Decreased Calcifying Potential: (A) Calcium content of Ddr1+/+
(black bars) and Ddr1‐/‐ (gray bars) VSMC cultures measured by O‐cresophathelein based
colourimetric assay after 12 days and normalized to total cellular protein using a BCA
colourimetic assay. (B) Calcium content of Ddr1+/+ (black bars) and Ddr1‐/‐ (gray bars) VSMC
cultures measured by O‐cresophathelein based colourimetric assay after 12 days
andnormalized to total cellular DNA using a Hoechst nuclear dye based assay. Data shown are
means ± SEM; n=3
A B
69
Figure 8: Ddr1‐/‐ VSMCs Show Increased Pit‐1 mRNA Expression and Decreased Alp and Npp1
mRNA Expression at Early but not Late Stage Calcification: (A) Ddr1+/+ (black bars) and Ddr1‐/‐
(gray bars) VSMC Pit‐1 mRNA levels at 2 and 12 days measured using qRT‐PCR. (B) Ddr1+/+ (black
bars) and Ddr1‐/‐ (gray bars) VSMC Alp mRNA levels at 2 and 12 days measured using qRT‐PCR.
(C) Ddr1+/+ (black bars) and Ddr1‐/‐ (gray bars) VSMC Npp1 mRNA levels at 2 and 12 days
measured using qRT‐PCR. * indicates a significant difference between groups (P<0.05). Data
shown are means ± SEM; n=3.
A B
C
Pit‐1 Alp
Npp1
70
Figure 9: Ddr1‐/‐ VSMC Show no Change in αSMA Protein and Atf4 mRNA Expression, but
Show Increased TenC mRNA Expression at Late but Not Early Stage Calcification: (A) Ddr1+/+
(black bars) and Ddr1‐/‐ (gray bars) VSMC αSMA protein levels at 12 days measured using
immunoblotting. (B) Ddr1+/+ (black bars) and Ddr1‐/‐ (gray bars) VSMC Atf4 mRNA levels at 2 and
12 days measured using qRT‐PCR. (C) Ddr1+/+ (black bars) and Ddr1‐/‐ (gray bars) VSMC TenC
mRNA levels at 2 and 12 days measured using qRT‐PCR. * indicates a significant difference
between groups (P<0.05). Data shown are means ± SEM; n=3.
A
B
C
αSMA
Atf4
TenC
71
Figure 10: Ddr1‐/‐ VSMCs Show No Changes in RUNX2 Protein and mRNA Expression: (A)
Ddr1+/+ (black bars) and Ddr1‐/‐ (gray bars) VSMC cultures from 2, 0 and 12 days was measured
for Runx2 mRNA using qRT‐PCR. (B) RUNX2 protein levels from Ddr1+/+ (black bars) and Ddr1‐/‐
(gray bars) VSMC measured with immunoblotting at 2 days. (C) RUNX2 protein levels from
Ddr1+/+ (black bars) and Ddr1‐/‐ (gray bars) VSMC were measured with immunoblotting at 12
days. * indicates a significant difference between groups (P<0.05). Data shown are means ±
SEM; n=3.
A
B C
Runx2 mRNA
2 Days RUNX2 Protein 12 Days RUNX2 Protein
72
Figure 11: Ddr1‐/‐ VSMCs Show Reduced RUNX2 Activity and Ocn mRNA Expression: (A) RUNX2
activity of Ddr1+/+ (black bars) and Ddr1‐/‐ (gray bars) VSMCs was measured using a luciferase
assay of 0 and 12 day cultures (B) mRNA levels of Ocn in Ddr1+/+ (black bars) and Ddr1‐/‐ (gray
bars) VSMCs was measured using qRT‐PCR 0, 2 and 12 day cultures. * indicates a significant
difference between groups (P<0.05). Data shown are means ± SEM; n=3.
BA Osteocalcin mRNA RUNX2 Activity
73
Figure 12: Ddr1‐/‐ VSMCs Show Reduced RUNX2 Nuclear Localization: (A) Ddr1+/+ and Ddr1‐/‐
cells from 2 days were 4% PFA fixed and stained for RUNX2 (Green) Nuclei were stained with
Hoechst (blue). Scale bars = 20 µm. (B) Ddr1+/+ and Ddr1‐/‐ cells from 12 days were 4% PFA fixed
and stained for RUNX2 (Green) Nuclei were stained with Hoechst (blue). Scale bars = 20 µm.
(C)Quantification of RUNX2 nuclear and cytoplasmic amounts from VSMC in 2 and 12 days
under osteogenic conditions Ddr1+/+ (black bars) and Ddr1‐/‐ (gray bars). * indicates a significant
difference between groups (P<0.05). Data shown are means ± SEM; n=3.
A B
C
2 Days 12 Days
74
Figure 13: RUNX2 Activity and mRNA Levels of Osteocalcin are Sensitive to P38i in Ddr1+/+
VSMCs: (A) RUNX2 activity of Ddr1+/+ (black bars) and Ddr1‐/‐ (gray bars) VSMC cultures treated
with DMSO, 10 µM PD098059 (MEK inhibitor) or SB203580 (P38 inhibitor) measured using
luciferase activity assay. (B) Ocn mRNA of Ddr1+/+ (black bars) and Ddr1‐/‐ (gray bars) VSMC
cultures treated with DMSO, 10 µM PD098059 (MEK inhibitor) or SB203580 (P38 inhibitor)
measured using qRT‐PCR. (C) Runx2 mRNA of Ddr1+/+ (black bars) and Ddr1‐/‐ (gray bars) VSMC
cultures treated with DMSO, 10 µM PD098059 (MEK inhibitor) or SB203580 (P38 inhibitor)
measured using qRT‐PCR. * indicates a significant difference between groups (P<0.05). Data
shown are means ± SEM; n=3.
A B
C
RUNX2 Activity Ocn mRNA
Runx2 mRNA
75
Figure 14: Nuclear Localization of RUNX2 in Ddr1+/+ and Ddr1‐/‐ VSMCs Was Not Affected by
MEKi or P38i: (A) Ddr1+/+ and Ddr1‐/‐ cells treated with DMSO were 4% PFA fixed and stained for
RUNX2 (Green) Nuclei were stained with Hoechst (blue). Scale bars = 20 µm. (B) Ddr1+/+ and
Ddr1‐/‐ cells treated with 10 µM PD098059 (MEK inhibitor) were 4% PFA fixed and stained for
RUNX2 (Green) Nuclei were stained with Hoechst (blue). Scale bars = 20 µm. (C) Ddr1+/+ and
Ddr1‐/‐ cells treated with 10 µM SB203580 (P38 inhibitor) were 4% PFA fixed and stained for
RUNX2 (Green) Nuclei were stained with Hoechst (blue). Scale bars = 20 µm. (D) Quantification
of RUNX2 nuclear and cytoplasmic amounts of Ddr1+/+ (black bars) and Ddr1‐/‐ (gray bars) VSMC
A B
C D
DMSO MEKi
P38i
76
cultures in 2 and 12 days under osteogenic conditions. * indicates a significant difference
between groups (P<0.05). Data shown are means ± SEM; n=3.
77
Figure 15: Phospho ERK1/2 Signalling is Decreased in Ddr1‐/‐ VSMCs at Late but Not Early
Stage Calcification: (A) PhosphoERK1/2 protein levels from Ddr1+/+ (black bars) and Ddr1‐/‐ (gray
bars) VSMC measured with immunoblotting at 2 days. (B) PhosphoERK1/2 protein levels from
Ddr1+/+ (black bars) and Ddr1‐/‐ (gray bars) VSMC were measured with immunoblotting at 12
days. * indicates a significant difference between groups (P<0.05). Data shown are means ±
SEM; n=3.
A
B
2 Days
12 Days
78
Figure 16: PhosphoP38 Signalling is Decreased in Ddr1‐/‐ VSMCs at Late but Not Early Stage
Calcification: (A) PhosphoP38 protein levels from Ddr1+/+ (black bars) and Ddr1‐/‐ (gray bars)
VSMC measured with immunoblotting at 2 days. (B) PhosphoP38 protein levels from Ddr1+/+
(black bars) and Ddr1‐/‐ (gray bars) VSMC were measured with immunoblotting at 12 days. *
indicates a significant difference between groups (P<0.05). Data shown are means ± SEM; n=3.
A
B 12 Days
2 Days
79
Figure 17: Ddr1‐/‐ VSMCs Show Reduced PhosphoP38 Nuclear Localization: (A) Ddr1+/+ and
Ddr1‐/‐ cells from 2 days were 4% PFA fixed and stained for phosphoP38 (Green) Nuclei were
stained with Hoechst (blue). Scale bars = 20 µm. (B) Ddr1+/+ and Ddr1‐/‐ cells from 12 days were
4% PFA fixed and stained for phosphoP38 (Green) Nuclei were stained with Hoechst (blue).
Scale bars = 20 µm. (C)Quantification of phosphoP38 nuclear and cytoplasmic amounts from
VSMC in 2 and 12 days under osteogenic conditions Ddr1+/+ (black bars) and Ddr1‐/‐ (gray bars).
* indicates a significant difference between groups (P<0.05). Data shown are means ± SEM;
n=3.
A B
C
2 Days 12 Days
80
Figure 18: Ddr1‐/‐ VSMCs Show Increased Cell Spreading and Reduced α Tubulin
Polymerization: (A) Ddr1+/+ and Ddr1‐/‐ cells from 2 days were 4% PFA fixed and imaged using
DIC microscopy. Scale bars = 20 µm. (B) Ddr1+/+ and Ddr1‐/‐ cells from 2 days were 4% PFA fixed
and stained for α Tubulin (Green) Nuclei were stained with Hoechst (blue). Scale bars = 20 µm.
(C) Ddr1+/+ and Ddr1‐/‐ cells from 12 days were 4% PFA fixed and stained for α Tubulin (Green)
Nuclei were stained with Hoechst (blue). Scale bars = 20 µm.
A B
C 12 Days
2 Days
81
Figure 19: Ddr1‐/‐ VSMCs Show Reduced RUNX2 Co‐Localization With Microtubules: Ddr1+/+
and Ddr1‐/‐ VSMC cultures in 2 days under osteogenic conditions (2.4 mM Pi) were 4% PFA fixed
and stained for α Tubulin (Green) and RUNX2 (Red). Scale bars = 40 µm.
82
Figure 20: Taxol Increased RUNX2 Nuclear Localization in Ddr1+/+ and Ddr1‐/‐ VSMCs While
Nocodazole, Prevented RUNX2 Nuclear Localization in Ddr1+/+ VSMCs: (A) Ddr1+/+ and Ddr1‐/‐
cells treated with DMSO were 4% PFA fixed and stained for RUNX2 (Green) Nuclei were stained
with Hoechst (blue). Scale bars = 20 µm. (B) Ddr1+/+ and Ddr1‐/‐ cells treated with 10 µM
nocodazole (microtubule destabilizing agent) were 4% PFA fixed and stained for RUNX2
(Green) Nuclei were stained with Hoechst (blue). Scale bars = 20 µm. (C) Ddr1+/+ and Ddr1‐/‐ cells
treated with 10 µM taxol (microtubule stabilizing agent) were 4% PFA fixed and stained for
RUNX2 (Green) Nuclei were stained with Hoechst (blue). Scale bars = 20 µm. (D) Quantification
A B
C D
DMSO Nocodazole
Taxol
83
of RUNX2 nuclear and cytoplasmic amounts of Ddr1+/+ (black bars) and Ddr1‐/‐ (gray bars) VSMC
cultures in 2 and 12 days under osteogenic conditions. * indicates a significant difference
between groups (P<0.05). Data shown are means ± SEM; n=3.
84
Figure 21: Nocodazole or Taxol Attenuated RUNX2 Activity in Ddr1+/+ and Ddr1‐/‐ VSMCs but
Increased Ocn mRNA Expression: (A) RUNX2 activity of Ddr1+/+ (black bars) and Ddr1‐/‐ (gray
bars) VSMC cultures treated with DMSO, 10 µM nocodazole or taxol measured using luciferase
activity assay. (B) Ocn mRNA of Ddr1+/+ (black bars) and Ddr1‐/‐ (gray bars) VSMC cultures
treated with DMSO, 10 µM nocodazole or taxol measured using qRT‐PCR. (C) Runx2 mRNA of
Ddr1+/+ (black bars) and Ddr1‐/‐ (gray bars) VSMC cultures treated with DMSO, 10 µM
nocodazole or taxol measured using qRT‐PCR. * indicates a significant difference between
groups (P<0.05). Data shown are means ± SEM; n=3.
A B
C
RUNX2 Activity Ocn mRNA
Runx2 mRNA
85
Figure 22: Proposed model of DDR1 regulation of RUNX2 activity in VSMCs: The
schematic highlights the possible pathway by which DDR1 acts to regulate RUNX2 in VSMCs.
DDR1 is essential for organizing a microtubule network for the nuclear localization of RUNX2.
86
REFERENCES
(1) Davignon J, Ganz P. Role of endothelial dysfunction in atherosclerosis. Circulation 2004 June 15;109(23 Suppl 1):III27‐III32.
(2) Stary HC, Chandler AB, Dinsmore RE, Fuster V, Glagov S, Insull W, Jr., Rosenfeld ME, Schwartz CJ, Wagner WD, Wissler RW. A definition of advanced types of atherosclerotic lesions and a histological classification of atherosclerosis. A report from the Committee on Vascular Lesions of the Council on Arteriosclerosis, American Heart Association. Arterioscler Thromb Vasc Biol 1995 September;15(9):1512‐31.
(3) Nakashima Y, Raines EW, Plump AS, Breslow JL, Ross R. Upregulation of VCAM‐1 and ICAM‐1 at atherosclerosis‐prone sites on the endothelium in the ApoE‐deficient mouse. Arterioscler Thromb Vasc Biol 1998 May;18(5):842‐51.
(4) Boyle JJ. Macrophage activation in atherosclerosis: pathogenesis and pharmacology of plaque rupture. Curr Vasc Pharmacol 2005 January;3(1):63‐8.
(5) Boullier A, Bird DA, Chang MK, Dennis EA, Friedman P, Gillotre‐Taylor K, Horkko S, Palinski W, Quehenberger O, Shaw P, Steinberg D, Terpstra V, Witztum JL. Scavenger receptors, oxidized LDL, and atherosclerosis. Ann N Y Acad Sci 2001 December;947:214‐22.
(6) Galkina E, Ley K. Immune and inflammatory mechanisms of atherosclerosis (*). Annu Rev Immunol 2009;27:165‐97.
(7) Newby AC, Zaltsman AB. Fibrous cap formation or destruction‐‐the critical importance of vascular smooth muscle cell proliferation, migration and matrix formation. Cardiovasc Res 1999 February;41(2):345‐60.
(8) Varnava AM, Mills PG, Davies MJ. Relationship between coronary artery remodeling and plaque vulnerability. Circulation 2002 February 26;105(8):939‐43.
(9) Katritsis DG, Pantos J, Efstathopoulos E. Hemodynamic factors and atheromatic plaque rupture in the coronary arteries: from vulnerable plaque to vulnerable coronary segment. Coron Artery Dis 2007 May;18(3):229‐37.
(10) Sage AP, Tintut Y, Demer LL. Regulatory mechanisms in vascular calcification. Nat Rev Cardiol 2010 September;7(9):528‐36.
(11) 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 November 24;87(11):1055‐62.
87
(12) Fernandez‐Ortiz A, Badimon JJ, Falk E, Fuster V, Meyer B, Mailhac A, Weng D, Shah PK, Badimon L. Characterization of the relative thrombogenicity of atherosclerotic plaque components: implications for consequences of plaque rupture. J Am Coll Cardiol 1994 June;23(7):1562‐9.
(13) Smith EB. THE INFLUENCE OF AGE AND ATHEROSCLEROSIS ON THE CHEMISTRY OF AORTIC INTIMA.2. COLLAGEN AND MUCOPOLYSACCHARIDES. J Atheroscler Res 1965 March;5:241‐8.
(14) Shoulders MD, Raines RT. Collagen structure and stability. Annu Rev Biochem 2009;78:929‐58.
(15) Voss B, Rauterberg J. Localization of collagen types I, III, IV and V, fibronectin and laminin in human arteries by the indirect immunofluorescence method. Pathol Res Pract 1986 October;181(5):568‐75.
(16) Adiguzel E, Hou G, Mulholland D, Hopfer U, Fukai N, Olsen B, Bendeck M. Migration and growth are attenuated in vascular smooth muscle cells with type VIII collagen‐null alleles. Arterioscler Thromb Vasc Biol 2006 January;26(1):56‐61.
(17) Rocnik EF, Chan BM, Pickering JG. Evidence for a role of collagen synthesis in arterial smooth muscle cell migration. J Clin Invest 1998 May 1;101(9):1889‐98.
(18) Lu KK, Trcka D, Bendeck MP. Collagen stimulates discoidin domain receptor 1‐mediated migration of smooth muscle cells through Src. Cardiovasc Pathol 2011 March;20(2):71‐6.
(19) Hou G, Mulholland D, Gronska MA, Bendeck MP. Type VIII collagen stimulates smooth muscle cell migration and matrix metalloproteinase synthesis after arterial injury. Am J Pathol 2000 February;156(2):467‐76.
(20) Adiguzel E, Ahmad PJ, Franco C, Bendeck MP. Collagens in the progression and complications of atherosclerosis. Vasc Med 2009 February;14(1):73‐89.
(21) Wesley RB, Meng X, Godin D, Galis ZS. Extracellular matrix modulates macrophage functions characteristic to atheroma: collagen type I enhances acquisition of resident macrophage traits by human peripheral blood monocytes in vitro. Arterioscler Thromb Vasc Biol 1998 March;18(3):432‐40.
(22) Choi ET, Collins ET, Marine LA, Uberti MG, Uchida H, Leidenfrost JE, Khan MF, Boc KP, Abendschein DR, Parks WC. Matrix metalloproteinase‐9 modulation by resident arterial cells is responsible for injury‐induced accelerated atherosclerotic plaque development in apolipoprotein E‐deficient mice. Arterioscler Thromb Vasc Biol 2005 May;25(5):1020‐5.
88
(23) Kuzuya M, Nakamura K, Sasaki T, Cheng XW, Itohara S, Iguchi A. Effect of MMP‐2 deficiency on atherosclerotic lesion formation in apoE‐deficient mice. Arterioscler Thromb Vasc Biol 2006 May;26(5):1120‐5.
(24) Luttun A, Lutgens E, Manderveld A, Maris K, Collen D, Carmeliet P, Moons L. Loss of matrix metalloproteinase‐9 or matrix metalloproteinase‐12 protects apolipoprotein E‐deficient mice against atherosclerotic media destruction but differentially affects plaque growth. Circulation 2004 March 23;109(11):1408‐14.
(25) Lemaitre V, O'Byrne TK, Borczuk AC, Okada Y, Tall AR, D'Armiento J. ApoE knockout mice expressing human matrix metalloproteinase‐1 in macrophages have less advanced atherosclerosis. J Clin Invest 2001 May;107(10):1227‐34.
(26) Li S, Van Den Diepstraten C, D'Souza SJ, Chan BM, Pickering JG. Vascular smooth muscle cells orchestrate the assembly of type I collagen via alpha2beta1 integrin, RhoA, and fibronectin polymerization. Am J Pathol 2003 September;163(3):1045‐56.
(27) Hynes RO. Integrins: versatility, modulation, and signaling in cell adhesion. Cell 1992 April 3;69(1):11‐25.
(28) Hultgardh‐Nilsson A, Durbeej M. Role of the extracellular matrix and its receptors in smooth muscle cell function: implications in vascular development and disease. Curr Opin Lipidol 2007 October;18(5):540‐5.
(29) Franco C, Ahmad PJ, Hou G, Wong E, Bendeck MP. Increased cell and matrix accumulation during atherogenesis in mice with vessel wall‐specific deletion of discoidin domain receptor 1. Circ Res 2010 June 11;106(11):1775‐83.
(30) Vogel W, Gish GD, Alves F, Pawson T. The discoidin domain receptor tyrosine kinases are activated by collagen. Mol Cell 1997 December;1(1):13‐23.
(31) Springer WR, Cooper DN, Barondes SH. Discoidin I is implicated in cell‐substratum attachment and ordered cell migration of Dictyostelium discoideum and resembles fibronectin. Cell 1984 December;39(3 Pt 2):557‐64.
(32) Kiedzierska A, Smietana K, Czepczynska H, Otlewski J. Structural similarities and functional diversity of eukaryotic discoidin‐like domains. Biochim Biophys Acta 2007 September;1774(9):1069‐78.
(33) L'hote CG, Thomas PH, Ganesan TS. Functional analysis of discoidin domain receptor 1: effect of adhesion on DDR1 phosphorylation. FASEB J 2002 February;16(2):234‐6.
(34) L'Hote CG, Thomas PH, Ganesan TS. Functional analysis of discoidin domain receptor 1: effect of adhesion on DDR1 phosphorylation. FASEB J 2002 February;16(2):234‐6.
89
(35) Ikeda K, Wang LH, Torres R, Zhao H, Olaso E, Eng FJ, Labrador P, Klein R, Lovett D, Yancopoulos GD, Friedman SL, Lin HC. Discoidin domain receptor 2 interacts with Src and Shc following its activation by type I collagen. J Biol Chem 2002 May 24;277(21):19206‐12.
(36) Dejmek J, Dib K, Jonsson M, Andersson T. Wnt‐5a and G‐protein signaling are required for collagen‐induced DDR1 receptor activation and normal mammary cell adhesion. Int J Cancer 2003 January 20;103(3):344‐51.
(37) Wang CZ, Su HW, Hsu YC, Shen MR, Tang MJ. A discoidin domain receptor 1/SHP‐2 signaling complex inhibits alpha2beta1‐integrin‐mediated signal transducers and activators of transcription 1/3 activation and cell migration. Mol Biol Cell 2006 June;17(6):2839‐52.
(38) Iwai LK, Chang F, Huang PH. Phosphoproteomic analysis identifies insulin enhancement of discoidin domain receptor 2 phosphorylation. Cell Adh Migr 2012 November 15;7(2).
(39) Vogel W, Gish GD, Alves F, Pawson T. The discoidin domain receptor tyrosine kinases are activated by collagen. Mol Cell 1997 December;1(1):13‐23.
(40) Koo DH, McFadden C, Huang Y, Abdulhussein R, Friese‐Hamim M, Vogel WF. Pinpointing phosphotyrosine‐dependent interactions downstream of the collagen receptor DDR1. FEBS Lett 2006 January 9;580(1):15‐22.
(41) Hansen C, Greengard P, Nairn AC, Andersson T, Vogel WF. Phosphorylation of DARPP‐32 regulates breast cancer cell migration downstream of the receptor tyrosine kinase DDR1. Exp Cell Res 2006 December 10;312(20):4011‐8.
(42) Hilton HN, Stanford PM, Harris J, Oakes SR, Kaplan W, Daly RJ, Ormandy CJ. KIBRA interacts with discoidin domain receptor 1 to modulate collagen‐induced signalling. Biochim Biophys Acta 2008 March;1783(3):383‐93.
(43) Suh HN, Han HJ. Collagen I regulates the self‐renewal of mouse embryonic stem cells through alpha2beta1 integrin‐ and DDR1‐dependent Bmi‐1. J Cell Physiol 2011 December;226(12):3422‐32.
(44) Ruiz PA, Jarai G. Collagen I induces discoidin domain receptor (DDR) 1 expression through DDR2 and a JAK2‐ERK1/2‐mediated mechanism in primary human lung fibroblasts. J Biol Chem 2011 April 15;286(15):12912‐23.
(45) Zhang Y, Su J, Yu J, Bu X, Ren T, Liu X, Yao L. An essential role of discoidin domain receptor 2 (DDR2) in osteoblast differentiation and chondrocyte maturation via modulation of Runx2 activation. J Bone Miner Res 2011 March;26(3):604‐17.
90
(46) Lin KL, Chou CH, Hsieh SC, Hwa SY, Lee MT, Wang FF. Transcriptional upregulation of DDR2 by ATF4 facilitates osteoblastic differentiation through p38 MAPK‐mediated Runx2 activation. J Bone Miner Res 2010 November;25(11):2489‐503.
(47) Seo MC, Kim S, Kim SH, Zheng LT, Park EK, Lee WH, Suk K. Discoidin domain receptor 1 mediates collagen‐induced inflammatory activation of microglia in culture. J Neurosci Res 2008 April;86(5):1087‐95.
(48) Kim SH, Lee S, Suk K, Bark H, Jun CD, Kim DK, Choi CH, Yoshimura T. Discoidin domain receptor 1 mediates collagen‐induced nitric oxide production in J774A.1 murine macrophages. Free Radic Biol Med 2007 February 1;42(3):343‐52.
(49) Matsuyama W, Watanabe M, Shirahama Y, Oonakahara K, Higashimoto I, Yoshimura T, Osame M, Arimura K. Activation of discoidin domain receptor 1 on CD14‐positive bronchoalveolar lavage fluid cells induces chemokine production in idiopathic pulmonary fibrosis. J Immunol 2005 May 15;174(10):6490‐8.
(50) Hou G, Vogel W, Bendeck MP. The discoidin domain receptor tyrosine kinase DDR1 in arterial wound repair. J Clin Invest 2001 March;107(6):727‐35.
(51) Ferri N, Carragher NO, Raines EW. Role of discoidin domain receptors 1 and 2 in human smooth muscle cell‐mediated collagen remodeling: potential implications in atherosclerosis and lymphangioleiomyomatosis. Am J Pathol 2004 May;164(5):1575‐85.
(52) Hou G, Vogel WF, Bendeck MP. Tyrosine kinase activity of discoidin domain receptor 1 is necessary for smooth muscle cell migration and matrix metalloproteinase expression. Circ Res 2002 June 14;90(11):1147‐9.
(53) Chen SC, Wang BW, Wang DL, Shyu KG. Hypoxia induces discoidin domain receptor‐2 expression via the p38 pathway in vascular smooth muscle cells to increase their migration. Biochem Biophys Res Commun 2008 October 3;374(4):662‐7.
(54) Shyu KG, Wang BW, Chang H. Hyperbaric oxygen activates discoidin domain receptor 2 via tumour necrosis factor‐alpha and the p38 MAPK pathway to increase vascular smooth muscle cell migration through matrix metalloproteinase 2. Clin Sci (Lond) 2009 April;116(7):575‐83.
(55) Bhadriraju K, Chung KH, Spurlin TA, Haynes RJ, Elliott JT, Plant AL. The relative roles of collagen adhesive receptor DDR2 activation and matrix stiffness on the downregulation of focal adhesion kinase in vascular smooth muscle cells. Biomaterials 2009 December;30(35):6687‐94.
91
(56) Hou G, Wang D, Bendeck MP. Deletion of discoidin domain receptor 2 does not affect smooth muscle cell adhesion, migration, or proliferation in response to type I collagen. Cardiovasc Pathol 2012 May;21(3):214‐8.
(57) Franco C, Hou G, Ahmad PJ, Fu EY, Koh L, Vogel WF, Bendeck MP. Discoidin domain receptor 1 (ddr1) deletion decreases atherosclerosis by accelerating matrix accumulation and reducing inflammation in low‐density lipoprotein receptor‐deficient mice. Circ Res 2008 May 23;102(10):1202‐11.
(58) Franco C, Britto K, Wong E, Hou G, Zhu SN, Chen M, Cybulsky MI, Bendeck MP. Discoidin domain receptor 1 on bone marrow‐derived cells promotes macrophage accumulation during atherogenesis. Circ Res 2009 November 20;105(11):1141‐8.
(59) Ahmad PJ, Trcka D, Xue S, Franco C, Speer MY, Giachelli CM, Bendeck MP. Discoidin domain receptor‐1 deficiency attenuates atherosclerotic calcification and smooth muscle cell‐mediated mineralization. Am J Pathol 2009 December;175(6):2686‐96.
(60) Klatt AR, Zech D, Kuhn G, Paul‐Klausch B, Klinger G, Renno JH, Schmidt J, Malchau G, Wielckens K. Discoidin domain receptor 2 mediates the collagen II‐dependent release of interleukin‐6 in primary human chondrocytes. J Pathol 2009 June;218(2):241‐7.
(61) Lam NP, Li Y, Waldman AB, Brussiau J, Lee PL, Olsen BR, Xu L. Age‐dependent increase of discoidin domain receptor 2 and matrix metalloproteinase 13 expression in temporomandibular joint cartilage of type IX and type XI collagen‐deficient mice. Arch Oral Biol 2007 June;52(6):579‐84.
(62) Xu L, Peng H, Wu D, Hu K, Goldring MB, Olsen BR, Li Y. Activation of the discoidin domain receptor 2 induces expression of matrix metalloproteinase 13 associated with osteoarthritis in mice. J Biol Chem 2005 January 7;280(1):548‐55.
(63) Bargal R, Cormier‐Daire V, Ben‐Neriah Z, Le MM, Sosna J, Melki J, Zangen DH, Smithson SF, Borochowitz Z, Belostotsky R, Raas‐Rothschild A. Mutations in DDR2 gene cause SMED with short limbs and abnormal calcifications. Am J Hum Genet 2009 January;84(1):80‐4.
(64) McCullough PA, Chinnaiyan KM, Agrawal V, Danielewicz E, Abela GS. Amplification of atherosclerotic calcification and Monckeberg's sclerosis: a spectrum of the same disease process. Adv Chronic Kidney Dis 2008 October;15(4):396‐412.
(65) BEADENKOPF WG, DAOUD AS, LOVE BM. CALCIFICATION IN THE CORONARY ARTERIES AND ITS RELATIONSHIP TO ARTERIOSCLEROSIS AND MYOCARDIAL INFARCTION. Am J Roentgenol Radium Ther Nucl Med 1964 October;92:865‐71.
(66) Wada T, McKee MD, Steitz S, Giachelli CM. Calcification of vascular smooth muscle cell cultures: inhibition by osteopontin. Circ Res 1999 February 5;84(2):166‐78.
92
(67) Tintut Y, Patel J, Territo M, Saini T, Parhami F, Demer LL. Monocyte/macrophage regulation of vascular calcification in vitro. Circulation 2002 February 5;105(5):650‐5.
(68) Iyemere VP, Proudfoot D, Weissberg PL, Shanahan CM. Vascular smooth muscle cell phenotypic plasticity and the regulation of vascular calcification. J Intern Med 2006 September;260(3):192‐210.
(69) Speer MY, Yang HY, Brabb T, Leaf E, Look A, Lin WL, Frutkin A, Dichek D, Giachelli CM. Smooth muscle cells give rise to osteochondrogenic precursors and chondrocytes in calcifying arteries. Circ Res 2009 March 27;104(6):733‐41.
(70) Tanaka T, Sato H, Doi H, Yoshida CA, Shimizu T, Matsui H, Yamazaki M, Akiyama H, Kawai‐Kowase K, Iso T, Komori T, Arai M, Kurabayashi M. Runx2 represses myocardin‐mediated differentiation and facilitates osteogenic conversion of vascular smooth muscle cells. Mol Cell Biol 2008 February;28(3):1147‐60.
(71) Schinke T, Karsenty G. Vascular calcification‐‐a passive process in need of inhibitors. Nephrol Dial Transplant 2000 September;15(9):1272‐4.
(72) McCullough PA, Chinnaiyan KM, Agrawal V, Danielewicz E, Abela GS. Amplification of atherosclerotic calcification and Monckeberg's sclerosis: a spectrum of the same disease process. Adv Chronic Kidney Dis 2008 October;15(4):396‐412.
(73) Smith JD, Breslow JL. The emergence of mouse models of atherosclerosis and their relevance to clinical research. J Intern Med 1997 August;242(2):99‐109.
(74) Merx MW, Schafer C, Westenfeld R, Brandenburg V, Hidajat S, Weber C, Ketteler M, Jahnen‐Dechent W. Myocardial stiffness, cardiac remodeling, and diastolic dysfunction in calcification‐prone fetuin‐A‐deficient mice. J Am Soc Nephrol 2005 November;16(11):3357‐64.
(75) Luo G, Ducy P, McKee MD, Pinero GJ, Loyer E, Behringer RR, Karsenty G. Spontaneous calcification of arteries and cartilage in mice lacking matrix GLA protein. Nature 1997 March 6;386(6620):78‐81.
(76) Mazzali M, Hughes J, Dantas M, Liaw L, Steitz S, Alpers CE, Pichler RH, Lan HY, Giachelli CM, Shankland SJ, Couser WG, Johnson RJ. Effects of cyclosporine in osteopontin null mice. Kidney Int 2002 July;62(1):78‐85.
(77) Sun Y, Byon CH, Yuan K, Chen J, Mao X, Heath JM, Javed A, Zhang K, Anderson PG, Chen Y. Smooth muscle cell‐specific runx2 deficiency inhibits vascular calcification. Circ Res 2012 August 17;111(5):543‐52.
93
(78) Wallin R, Wajih N, Greenwood GT, Sane DC. Arterial calcification: a review of mechanisms, animal models, and the prospects for therapy. Med Res Rev 2001 July;21(4):274‐301.
(79) Smith JD, Breslow JL. The emergence of mouse models of atherosclerosis and their relevance to clinical research. J Intern Med 1997 August;242(2):99‐109.
(80) Shobeiri N, Adams MA, Holden RM. Vascular calcification in animal models of CKD: A review. Am J Nephrol 2010;31(6):471‐81.
(81) El‐Abbadi MM, Pai AS, Leaf EM, Yang HY, Bartley BA, Quan KK, Ingalls CM, Liao HW, Giachelli CM. Phosphate feeding induces arterial medial calcification in uremic mice: role of serum phosphorus, fibroblast growth factor‐23, and osteopontin. Kidney Int 2009 June;75(12):1297‐307.
(82) Wada T, McKee MD, Steitz S, Giachelli CM. Calcification of vascular smooth muscle cell cultures: inhibition by osteopontin. Circ Res 1999 February 5;84(2):166‐78.
(83) Lomashvili KA, Cobbs S, Hennigar RA, Hardcastle KI, O'Neill WC. Phosphate‐induced vascular calcification: role of pyrophosphate and osteopontin. J Am Soc Nephrol 2004 June;15(6):1392‐401.
(84) Rawlinson SC, Mosley JR, Suswillo RF, Pitsillides AA, Lanyon LE. Calvarial and limb bone cells in organ and monolayer culture do not show the same early responses to dynamic mechanical strain. J Bone Miner Res 1995 August;10(8):1225‐32.
(85) Jono S, McKee MD, Murry CE, Shioi A, Nishizawa Y, Mori K, Morii H, Giachelli CM. Phosphate regulation of vascular smooth muscle cell calcification. Circ Res 2000 September 29;87(7):E10‐E17.
(86) Kircelli F, Peter ME, Sevinc OE, Celenk FG, Yilmaz M, Steppan S, Asci G, Ok E, Passlick‐Deetjen J. Magnesium reduces calcification in bovine vascular smooth muscle cells in a dose‐dependent manner. Nephrol Dial Transplant 2012 February;27(2):514‐21.
(87) Ganta DR, McCarthy MB, Gronowicz GA. Ascorbic acid alters collagen integrins in bone culture. Endocrinology 1997 September;138(9):3606‐12.
(88) Reil TD, Sarkar R, Kashyap VS, Sarkar M, Gelabert HA. Dexamethasone suppresses vascular smooth muscle cell proliferation. J Surg Res 1999 July;85(1):109‐14.
(89) Enomoto H, Enomoto‐Iwamoto M, Iwamoto M, Nomura S, Himeno M, Kitamura Y, Kishimoto T, Komori T. Cbfa1 is a positive regulatory factor in chondrocyte maturation. J Biol Chem 2000 March 24;275(12):8695‐702.
94
(90) Su X, Ao L, Shi Y, Johnson TR, Fullerton DA, Meng X. Oxidized low density lipoprotein induces bone morphogenetic protein‐2 in coronary artery endothelial cells via Toll‐like receptors 2 and 4. J Biol Chem 2011 April 8;286(14):12213‐20.
(91) Iwasaki K, Komaki M, Mimori K, Leon E, Izumi Y, Ishikawa I. IL‐6 induces osteoblastic differentiation of periodontal ligament cells. J Dent Res 2008 October;87(10):937‐42.
(92) 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 May;93(5):2106‐13.
(93) Sage AP, Lu J, Tintut Y, Demer LL. Hyperphosphatemia‐induced nanocrystals upregulate the expression of bone morphogenetic protein‐2 and osteopontin genes in mouse smooth muscle cells in vitro. Kidney Int 2011 February;79(4):414‐22.
(94) Balachandran K, Sucosky P, Yoganathan AP. Hemodynamics and mechanobiology of aortic valve inflammation and calcification. Int J Inflam 2011;2011:263870.
(95) Mori‐Akiyama Y, Akiyama H, Rowitch DH, de CB. Sox9 is required for determination of the chondrogenic cell lineage in the cranial neural crest. Proc Natl Acad Sci U S A 2003 August 5;100(16):9360‐5.
(96) Yoshiki S. A simple histological method for identification of osteoid matrix in decalcified bone. Stain Technol 1973 September;48(5):233‐8.
(97) Fujita T, Azuma Y, Fukuyama R, Hattori Y, Yoshida C, Koida M, Ogita K, Komori T. Runx2 induces osteoblast and chondrocyte differentiation and enhances their migration by coupling with PI3K‐Akt signaling. J Cell Biol 2004 July 5;166(1):85‐95.
(98) Dean DD, Schwartz Z, Bonewald L, Muniz OE, Morales S, Gomez R, Brooks BP, Qiao M, Howell DS, Boyan BD. Matrix vesicles produced by osteoblast‐like cells in culture become significantly enriched in proteoglycan‐degrading metalloproteinases after addition of beta‐glycerophosphate and ascorbic acid. Calcif Tissue Int 1994 May;54(5):399‐408.
(99) Shioi A, Katagi M, Okuno Y, Mori K, Jono S, Koyama H, Nishizawa Y. Induction of bone‐type alkaline phosphatase in human vascular smooth muscle cells: roles of tumor necrosis factor‐alpha and oncostatin M derived from macrophages. Circ Res 2002 July 12;91(1):9‐16.
(100) Li X, Yang HY, Giachelli CM. Role of the sodium‐dependent phosphate cotransporter, Pit‐1, in vascular smooth muscle cell calcification. Circ Res 2006 April 14;98(7):905‐12.
95
(101) Villa‐Bellosta R, Millan A, Sorribas V. Role of calcium‐phosphate deposition in vascular smooth muscle cell calcification. Am J Physiol Cell Physiol 2011 January;300(1):C210‐C220.
(102) Gipstein RM, Coburn JW, Adams DA, Lee DB, Parsa KP, Sellers A, Suki WN, Massry SG. Calciphylaxis in man. A syndrome of tissue necrosis and vascular calcification in 11 patients with chronic renal failure. Arch Intern Med 1976 November;136(11):1273‐80.
(103) Fitzpatrick LA, Severson A, Edwards WD, Ingram RT. Diffuse calcification in human coronary arteries. Association of osteopontin with atherosclerosis. J Clin Invest 1994 October;94(4):1597‐604.
(104) Johnson K, Polewski M, van ED, Terkeltaub R. Chondrogenesis mediated by PPi depletion promotes spontaneous aortic calcification in NPP1‐/‐ mice. Arterioscler Thromb Vasc Biol 2005 April;25(4):686‐91.
(105) El‐Maadawy S, Kaartinen MT, Schinke T, Murshed M, Karsenty G, McKee MD. Cartilage formation and calcification in arteries of mice lacking matrix Gla protein. Connect Tissue Res 2003;44 Suppl 1:272‐8.
(106) Ohnishi M, Nakatani T, Lanske B, Razzaque MS. Reversal of mineral ion homeostasis and soft‐tissue calcification of klotho knockout mice by deletion of vitamin D 1alpha‐hydroxylase. Kidney Int 2009 June;75(11):1166‐72.
(107) Price PA, Faus SA, Williamson MK. Warfarin causes rapid calcification of the elastic lamellae in rat arteries and heart valves. Arterioscler Thromb Vasc Biol 1998 September;18(9):1400‐7.
(108) Micheletti RG, Fishbein GA, Currier JS, Singer EJ, Fishbein MC. Calcification of the internal elastic lamina of coronary arteries. Mod Pathol 2008 August;21(8):1019‐28.
(109) Sun Y, Byon CH, Yuan K, Chen J, Mao X, Heath JM, Javed A, Zhang K, Anderson PG, Chen Y. Smooth muscle cell‐specific runx2 deficiency inhibits vascular calcification. Circ Res 2012 August 17;111(5):543‐52.
(110) Mathew S, Tustison KS, Sugatani T, Chaudhary LR, Rifas L, Hruska KA. The mechanism of phosphorus as a cardiovascular risk factor in CKD. J Am Soc Nephrol 2008 June;19(6):1092‐105.
(111) Huang MS, Sage AP, Lu J, Demer LL, Tintut Y. Phosphate and pyrophosphate mediate PKA‐induced vascular cell calcification. Biochem Biophys Res Commun 2008 September 26;374(3):553‐8.
96
(112) Villa‐Bellosta R, Wang X, Millan JL, Dubyak GR, O'Neill WC. Extracellular pyrophosphate metabolism and calcification in vascular smooth muscle. Am J Physiol Heart Circ Physiol 2011 July;301(1):H61‐H68.
(113) Nitschke Y, Weissen‐Plenz G, Terkeltaub R, Rutsch F. Npp1 promotes atherosclerosis in ApoE knockout mice. J Cell Mol Med 2011 November;15(11):2273‐83.
(114) Terkeltaub R. Physiologic and pathologic functions of the NPP nucleotide pyrophosphatase/phosphodiesterase family focusing on NPP1 in calcification. Purinergic Signal 2006 June;2(2):371‐7.
(115) Rutsch F, Ruf N, Vaingankar S, Toliat MR, Suk A, Hohne W, Schauer G, Lehmann M, Roscioli T, Schnabel D, Epplen JT, Knisely A, Superti‐Furga A, McGill J, Filippone M, Sinaiko AR, Vallance H, Hinrichs B, Smith W, Ferre M, Terkeltaub R, Nurnberg P. Mutations in ENPP1 are associated with 'idiopathic' infantile arterial calcification. Nat Genet 2003 August;34(4):379‐81.
(116) Johnson K, Goding J, van ED, Sali A, Hu SI, Farley D, Krug H, Hessle L, Millan JL, Terkeltaub R. Linked deficiencies in extracellular PP(i) and osteopontin mediate pathologic calcification associated with defective PC‐1 and ANK expression. J Bone Miner Res 2003 June;18(6):994‐1004.
(117) Shinozaki T, Pritzker KP. Regulation of alkaline phosphatase: implications for calcium pyrophosphate dihydrate crystal dissolution and other alkaline phosphatase functions. J Rheumatol 1996 April;23(4):677‐83.
(118) Wass M, Butterworth PJ. Nucleoside pyrophosphatase activity associated with pig kidney alkaline phosphatase. Biochem J 1971 October;124(5):891‐6.
(119) Chan JR, Stinson RA. Dephosphorylation of phosphoproteins of human liver plasma membranes by endogenous and purified liver alkaline phosphatases. J Biol Chem 1986 June 15;261(17):7635‐9.
(120) Balcerzak M, Hamade E, Zhang L, Pikula S, Azzar G, Radisson J, Bandorowicz‐Pikula J, Buchet R. The roles of annexins and alkaline phosphatase in mineralization process. Acta Biochim Pol 2003;50(4):1019‐38.
(121) Narisawa S, Harmey D, Yadav MC, O'Neill WC, Hoylaerts MF, Millan JL. Novel inhibitors of alkaline phosphatase suppress vascular smooth muscle cell calcification. J Bone Miner Res 2007 November;22(11):1700‐10.
(122) Yohay DA, Zhang J, Thrailkill KM, Arthur JM, Quarles LD. Role of serum in the developmental expression of alkaline phosphatase in MC3T3‐E1 osteoblasts. J Cell Physiol 1994 March;158(3):467‐75.
97
(123) Turan S, Topcu B, Gokce I, Guran T, Atay Z, Omar A, Akcay T, Bereket A. Serum alkaline phosphatase levels in healthy children and evaluation of alkaline phosphatase z‐scores in different types of rickets. J Clin Res Pediatr Endocrinol 2011;3(1):7‐11.
(124) Fallon MD, Teitelbaum SL, Weinstein RS, Goldfischer S, Brown DM, Whyte MP. Hypophosphatasia: clinicopathologic comparison of the infantile, childhood, and adult forms. Medicine (Baltimore) 1984 January;63(1):12‐24.
(125) Li X, Giachelli CM. Sodium‐dependent phosphate cotransporters and vascular calcification. Curr Opin Nephrol Hypertens 2007 July;16(4):325‐8.
(126) Villa‐Bellosta R, Sorribas V. Compensatory regulation of the sodium/phosphate cotransporters NaPi‐IIc (SCL34A3) and Pit‐2 (SLC20A2) during Pi deprivation and acidosis. Pflugers Arch 2010 February;459(3):499‐508.
(127) Villa‐Bellosta R, Bogaert YE, Levi M, Sorribas V. Characterization of phosphate transport in rat vascular smooth muscle cells: implications for vascular calcification. Arterioscler Thromb Vasc Biol 2007 May;27(5):1030‐6.
(128) Villa‐Bellosta R, Bogaert YE, Levi M, Sorribas V. Characterization of phosphate transport in rat vascular smooth muscle cells: implications for vascular calcification. Arterioscler Thromb Vasc Biol 2007 May;27(5):1030‐6.
(129) Li X, Yang HY, Giachelli CM. Role of the sodium‐dependent phosphate cotransporter, Pit‐1, in vascular smooth muscle cell calcification. Circ Res 2006 April 14;98(7):905‐12.
(130) Sudhakar S, Katz MS, Elango N. Analysis of type‐I and type‐II RUNX2 protein expression in osteoblasts. Biochem Biophys Res Commun 2001 August 10;286(1):74‐9.
(131) Pockwinse SM, Rajgopal A, Young DW, Mujeeb KA, Nickerson J, Javed A, Redick S, Lian JB, Van Wijnen AJ, Stein JL, Stein GS, Doxsey SJ. Microtubule‐dependent nuclear‐cytoplasmic shuttling of Runx2. J Cell Physiol 2006 February;206(2):354‐62.
(132) Belizario JE, Alves J, Garay‐Malpartida M, Occhiucci JM. Coupling caspase cleavage and proteasomal degradation of proteins carrying PEST motif. Curr Protein Pept Sci 2008 June;9(3):210‐20.
(133) Wang N, Chen W, Linsel‐Nitschke P, Martinez LO, Agerholm‐Larsen B, Silver DL, Tall AR. A PEST sequence in ABCA1 regulates degradation by calpain protease and stabilization of ABCA1 by apoA‐I. J Clin Invest 2003 January;111(1):99‐107.
(134) Sudhakar S, Katz MS, Elango N. Analysis of type‐I and type‐II RUNX2 protein expression in osteoblasts. Biochem Biophys Res Commun 2001 August 10;286(1):74‐9.
98
(135) Cohen MM, Jr. Perspectives on RUNX genes: an update. Am J Med Genet A 2009 December;149A(12):2629‐46.
(136) Terry A, Kilbey A, Vaillant F, Stewart M, Jenkins A, Cameron E, Neil JC. Conservation and expression of an alternative 3' exon of Runx2 encoding a novel proline‐rich C‐terminal domain. Gene 2004 July 7;336(1):115‐25.
(137) Lee KS, Kim HJ, Li QL, Chi XZ, Ueta C, Komori T, Wozney JM, Kim EG, Choi JY, Ryoo HM, Bae SC. Runx2 is a common target of transforming growth factor beta1 and bone morphogenetic protein 2, and cooperation between Runx2 and Smad5 induces osteoblast‐specific gene expression in the pluripotent mesenchymal precursor cell line C2C12. Mol Cell Biol 2000 December;20(23):8783‐92.
(138) Igarashi M, Kamiya N, Hasegawa M, Kasuya T, Takahashi T, Takagi M. Inductive effects of dexamethasone on the gene expression of Cbfa1, Osterix and bone matrix proteins during differentiation of cultured primary rat osteoblasts. J Mol Histol 2004 January;35(1):3‐10.
(139) Zhou G, Zheng Q, Engin F, Munivez E, Chen Y, Sebald E, Krakow D, Lee B. Dominance of SOX9 function over RUNX2 during skeletogenesis. Proc Natl Acad Sci U S A 2006 December 12;103(50):19004‐9.
(140) Otto F, Thornell AP, Crompton T, Denzel A, Gilmour KC, Rosewell IR, Stamp GW, Beddington RS, Mundlos S, Olsen BR, Selby PB, Owen MJ. Cbfa1, a candidate gene for cleidocranial dysplasia syndrome, is essential for osteoblast differentiation and bone development. Cell 1997 May 30;89(5):765‐71.
(141) Choi JY, Pratap J, Javed A, Zaidi SK, Xing L, Balint E, Dalamangas S, Boyce B, van Wijnen AJ, Lian JB, Stein JL, Jones SN, Stein GS. Subnuclear targeting of Runx/Cbfa/AML factors is essential for tissue‐specific differentiation during embryonic development. Proc Natl Acad Sci U S A 2001 July 17;98(15):8650‐5.
(142) Porte D, Tuckermann J, Becker M, Baumann B, Teurich S, Higgins T, Owen MJ, Schorpp‐Kistner M, Angel P. Both AP‐1 and Cbfa1‐like factors are required for the induction of interstitial collagenase by parathyroid hormone. Oncogene 1999 January 21;18(3):667‐78.
(143) Tsuji K, Ito Y, Noda M. Expression of the PEBP2alphaA/AML3/CBFA1 gene is regulated by BMP4/7 heterodimer and its overexpression suppresses type I collagen and osteocalcin gene expression in osteoblastic and nonosteoblastic mesenchymal cells. Bone 1998 February;22(2):87‐92.
(144) Inman CK, Shore P. The osteoblast transcription factor Runx2 is expressed in mammary epithelial cells and mediates osteopontin expression. J Biol Chem 2003 December 5;278(49):48684‐9.
99
(145) Javed A, Barnes GL, Jasanya BO, Stein JL, Gerstenfeld L, Lian JB, Stein GS. runt homology domain transcription factors (Runx, Cbfa, and AML) mediate repression of the bone sialoprotein promoter: evidence for promoter context‐dependent activity of Cbfa proteins. Mol Cell Biol 2001 April;21(8):2891‐905.
(146) Kwon TG, Zhao X, Yang Q, Li Y, Ge C, Zhao G, Franceschi RT. Physical and functional interactions between Runx2 and HIF‐1alpha induce vascular endothelial growth factor gene expression. J Cell Biochem 2011 December;112(12):3582‐93.
(147) Ge C, Yang Q, Zhao G, Yu H, Kirkwood KL, Franceschi RT. Interactions between extracellular signal‐regulated kinase 1/2 and P38 map kinase pathways in the control of RUNX2 phosphorylation and transcriptional activity. J Bone Miner Res 2011 November 9.
(148) Yoshiki S. A simple histological method for identification of osteoid matrix in decalcified bone. Stain Technol 1973 September;48(5):233‐8.
(149) 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 December;18(12):1964‐71.
(150) Chen Y, Bal BS, Gorski JP. Calcium and collagen binding properties of osteopontin, bone sialoprotein, and bone acidic glycoprotein‐75 from bone. J Biol Chem 1992 December 5;267(34):24871‐8.
(151) Street J, Bao M, deGuzman L, Bunting S, Peale FV, Jr., Ferrara N, Steinmetz H, Hoeffel J, Cleland JL, Daugherty A, van BN, Redmond HP, Carano RA, Filvaroff EH. Vascular endothelial growth factor stimulates bone repair by promoting angiogenesis and bone turnover. Proc Natl Acad Sci U S A 2002 July 23;99(15):9656‐61.
(152) Ahluwalia A, Tarnawski AS. Critical role of hypoxia sensor‐‐HIF‐1alpha in VEGF gene activation. Implications for angiogenesis and tissue injury healing. Curr Med Chem 2012;19(1):90‐7.
(153) Knippenberg M, Helder MN, Zandieh DB, Wuisman PI, Klein‐Nulend J. Osteogenesis versus chondrogenesis by BMP‐2 and BMP‐7 in adipose stem cells. Biochem Biophys Res Commun 2006 April 14;342(3):902‐8.
(154) Monroe DG, Hawse JR, Subramaniam M, Spelsberg TC. Retinoblastoma binding protein‐1 (RBP1) is a Runx2 coactivator and promotes osteoblastic differentiation. BMC Musculoskelet Disord 2010;11:104.
(155) Yang S, Xu H, Yu S, Cao H, Fan J, Ge C, Fransceschi RT, Dong HH, Xiao G. Foxo1 mediates insulin‐like growth factor 1 (IGF1)/insulin regulation of osteocalcin
100
expression by antagonizing Runx2 in osteoblasts. J Biol Chem 2011 May 27;286(21):19149‐58.
(156) Long F. Building strong bones: molecular regulation of the osteoblast lineage. Nat Rev Mol Cell Biol 2012 January;13(1):27‐38.
(157) Tanaka T, Sato H, Doi H, Yoshida CA, Shimizu T, Matsui H, Yamazaki M, Akiyama H, Kawai‐Kowase K, Iso T, Komori T, Arai M, Kurabayashi M. Runx2 represses myocardin‐mediated differentiation and facilitates osteogenic conversion of vascular smooth muscle cells. Mol Cell Biol 2008 February;28(3):1147‐60.
(158) Speer MY, Li X, Hiremath PG, Giachelli CM. Runx2/Cbfa1, but not loss of myocardin, is required for smooth muscle cell lineage reprogramming toward osteochondrogenesis. J Cell Biochem 2010 July 1;110(4):935‐47.
(159) de FH, Bouissou A, Perez F. Interplay between microtubule dynamics and intracellular organization. Int J Biochem Cell Biol 2012 February;44(2):266‐74.
(160) de FH, Bouissou A, Perez F. Interplay between microtubule dynamics and intracellular organization. Int J Biochem Cell Biol 2012 February;44(2):266‐74.
(161) Gebremichael Y, Chu JW, Voth GA. Intrinsic bending and structural rearrangement of tubulin dimer: molecular dynamics simulations and coarse‐grained analysis. Biophys J 2008 September;95(5):2487‐99.
(162) Ghosh‐Roy A, Goncharov A, Jin Y, Chisholm AD. Kinesin‐13 and tubulin posttranslational modifications regulate microtubule growth in axon regeneration. Dev Cell 2012 October 16;23(4):716‐28.
(163) Mulder AM, Glavis‐Bloom A, Moores CA, Wagenbach M, Carragher B, Wordeman L, Milligan RA. A new model for binding of kinesin 13 to curved microtubule protofilaments. J Cell Biol 2009 April 6;185(1):51‐7.
(164) Janke C, Kneussel M. Tubulin post‐translational modifications: encoding functions on the neuronal microtubule cytoskeleton. Trends Neurosci 2010 August;33(8):362‐72.
(165) Hammond JW, Huang CF, Kaech S, Jacobson C, Banker G, Verhey KJ. Posttranslational modifications of tubulin and the polarized transport of kinesin‐1 in neurons. Mol Biol Cell 2010 February 15;21(4):572‐83.
(166) Paturle‐Lafanechere L, Manier M, Trigault N, Pirollet F, Mazarguil H, Job D. Accumulation of delta 2‐tubulin, a major tubulin variant that cannot be tyrosinated, in neuronal tissues and in stable microtubule assemblies. J Cell Sci 1994 June;107 ( Pt 6):1529‐43.
101
(167) Hubbert C, Guardiola A, Shao R, Kawaguchi Y, Ito A, Nixon A, Yoshida M, Wang XF, Yao TP. HDAC6 is a microtubule‐associated deacetylase. Nature 2002 May 23;417(6887):455‐8.
(168) Matsuyama A, Shimazu T, Sumida Y, Saito A, Yoshimatsu Y, Seigneurin‐Berny D, Osada H, Komatsu Y, Nishino N, Khochbin S, Horinouchi S, Yoshida M. In vivo destabilization of dynamic microtubules by HDAC6‐mediated deacetylation. EMBO J 2002 December 16;21(24):6820‐31.
(169) Gong X, Ming X, Deng P, Jiang Y. Mechanisms regulating the nuclear translocation of p38 MAP kinase. J Cell Biochem 2010 August 15;110(6):1420‐9.
(170) Vafopoulou X. Ecdysteroid receptor (EcR) is associated with microtubules and with mitochondria in the cytoplasm of prothoracic gland cells of Rhodnius prolixus (Hemiptera). Arch Insect Biochem Physiol 2009 December;72(4):249‐62.
(171) Carbonaro M, Escuin D, O'Brate A, Thadani‐Mulero M, Giannakakou P. Microtubules regulate hypoxia‐inducible factor‐1alpha protein trafficking and activity: implications for taxane therapy. J Biol Chem 2012 April 6;287(15):11859‐69.
(172) Burkhardt JK, Echeverri CJ, Nilsson T, Vallee RB. Overexpression of the dynamitin (p50) subunit of the dynactin complex disrupts dynein‐dependent maintenance of membrane organelle distribution. J Cell Biol 1997 October 20;139(2):469‐84.
(173) Lund AW, Stegemann JP, Plopper GE. Mesenchymal Stem Cells Sense Three Dimensional Type I Collagen through Discoidin Domain Receptor 1. Open Stem Cell J 2009;1:40‐53.
(174) Alonso M, Claros S, Becerra J, Andrades JA. The effect of type I collagen on osteochondrogenic differentiation in adipose‐derived stromal cells in vivo. Cytotherapy 2008;10(6):597‐610.
(175) Bokui N, Otani T, Igarashi K, Kaku J, Oda M, Nagaoka T, Seno M, Tatematsu K, Okajima T, Matsuzaki T, Ting K, Tanizawa K, Kuroda S. Involvement of MAPK signaling molecules and Runx2 in the NELL1‐induced osteoblastic differentiation. FEBS Lett 2008 January 23;582(2):365‐71.
(176) Ge C, Xiao G, Jiang D, Franceschi RT. Critical role of the extracellular signal‐regulated kinase‐MAPK pathway in osteoblast differentiation and skeletal development. J Cell Biol 2007 February 26;176(5):709‐18.
(177) Lee HW, Suh JH, Kim HN, Kim AY, Park SY, Shin CS, Choi JY, Kim JB. Berberine promotes osteoblast differentiation by Runx2 activation with p38 MAPK. J Bone Miner Res 2008 August;23(8):1227‐37.
102
(178) Shinozaki T, Pritzker KP. Regulation of alkaline phosphatase: implications for calcium pyrophosphate dihydrate crystal dissolution and other alkaline phosphatase functions. J Rheumatol 1996 April;23(4):677‐83.
(179) Nakano‐Kurimoto R, Ikeda K, Uraoka M, Nakagawa Y, Yutaka K, Koide M, Takahashi T, Matoba S, Yamada H, Okigaki M, Matsubara H. Replicative senescence of vascular smooth muscle cells enhances the calcification through initiating the osteoblastic transition. Am J Physiol Heart Circ Physiol 2009 November;297(5):H1673‐H1684.
(180) Malabanan KP, Kanellakis P, Bobik A, Khachigian LM. Activation transcription factor‐4 induced by fibroblast growth factor‐2 regulates vascular endothelial growth factor‐A transcription in vascular smooth muscle cells and mediates intimal thickening in rat arteries following balloon injury. Circ Res 2008 August 15;103(4):378‐87.
(181) Jones PL, Crack J, Rabinovitch M. Regulation of tenascin‐C, a vascular smooth muscle cell survival factor that interacts with the alpha v beta 3 integrin to promote epidermal growth factor receptor phosphorylation and growth. J Cell Biol 1997 October 6;139(1):279‐93.
(182) Gong X, Ming X, Deng P, Jiang Y. Mechanisms regulating the nuclear translocation of p38 MAP kinase. J Cell Biochem 2010 August 15;110(6):1420‐9.
(183) Mejillano MR, Shivanna BD, Himes RH. Studies on the nocodazole‐induced GTPase activity of tubulin. Arch Biochem Biophys 1996 December 1;336(1):130‐8.
(184) Xu K, Luduena RF. Characterization of nuclear betaII‐tubulin in tumor cells: a possible novel target for taxol. Cell Motil Cytoskeleton 2002 September;53(1):39‐52.
(185) Vogel WF, Abdulhussein R, Ford CE. Sensing extracellular matrix: an update on discoidin domain receptor function. Cell Signal 2006 August;18(8):1108‐16.
(186) Chen CW, Tsai YH, Deng WP, Shih SN, Fang CL, Burch JG, Chen WH, Lai WF. Type I and II collagen regulation of chondrogenic differentiation by mesenchymal progenitor cells. J Orthop Res 2005 March;23(2):446‐53.
(187) Lu Z, Doulabi BZ, Huang C, Bank RA, Helder MN. Collagen type II enhances chondrogenesis in adipose tissue‐derived stem cells by affecting cell shape. Tissue Eng Part A 2010 January;16(1):81‐90.
(188) Naik V, Leaf EM, Hu JH, Yang HY, Nguyen NB, Giachelli CM, Speer MY. Sources of cells that contribute to atherosclerotic intimal calcification: an in vivo genetic fate mapping study. Cardiovasc Res 2012 June 1;94(3):545‐54.
(189) Rutsch F, Ruf N, Vaingankar S, Toliat MR, Suk A, Hohne W, Schauer G, Lehmann M, Roscioli T, Schnabel D, Epplen JT, Knisely A, Superti‐Furga A, McGill J, Filippone M,
103
Sinaiko AR, Vallance H, Hinrichs B, Smith W, Ferre M, Terkeltaub R, Nurnberg P. Mutations in ENPP1 are associated with 'idiopathic' infantile arterial calcification. Nat Genet 2003 August;34(4):379‐81.
(190) Speer MY, Chien YC, Quan M, Yang HY, Vali H, McKee MD, Giachelli CM. Smooth muscle cells deficient in osteopontin have enhanced susceptibility to calcification in vitro. Cardiovasc Res 2005 May 1;66(2):324‐33.
(191) Woodcock‐Mitchell J, Mitchell JJ, Low RB, Kieny M, Sengel P, Rubbia L, Skalli O, Jackson B, Gabbiani G. Alpha‐smooth muscle actin is transiently expressed in embryonic rat cardiac and skeletal muscles. Differentiation 1988 December;39(3):161‐6.
(192) Cintorino M, Bellizzi de ME, Leoncini P, Tripodi SA, Xu LJ, Sappino AP, Schmitt‐Graff A, Gabbiani G. Expression of alpha‐smooth‐muscle actin in stromal cells of the uterine cervix during epithelial neoplastic changes. Int J Cancer 1991 April 1;47(6):843‐6.
(193) Cintorino M, Vindigni C, Del Vecchio MT, Tosi P, Frezzotti R, Hadjistilianou T, Leoncini P, Silvestri S, Skalli O, Gabbiani G. Expression of actin isoforms and intermediate filament proteins in childhood orbital rhabdomyosarcomas. J Submicrosc Cytol Pathol 1989 July;21(3):409‐19.
(194) Thyberg J, Blomgren K, Roy J, Tran PK, Hedin U. Phenotypic modulation of smooth muscle cells after arterial injury is associated with changes in the distribution of laminin and fibronectin. J Histochem Cytochem 1997 June;45(6):837‐46.
(195) Thyberg J, Blomgren K, Hedin U, Dryjski M. Phenotypic modulation of smooth muscle cells during the formation of neointimal thickenings in the rat carotid artery after balloon injury: an electron‐microscopic and stereological study. Cell Tissue Res 1995 September;281(3):421‐33.
(196) Curat CA, Vogel WF. Discoidin domain receptor 1 controls growth and adhesion of mesangial cells. J Am Soc Nephrol 2002 November;13(11):2648‐56.
(197) Thesingh CW, Groot CG, Wassenaar AM. Transdifferentiation of hypertrophic chondrocytes into osteoblasts in murine fetal metatarsal bones, induced by co‐cultured cerebrum. Bone Miner 1991 January;12(1):25‐40.
(198) Aaron RK, Ciombor DM. Acceleration of experimental endochondral ossification by biophysical stimulation of the progenitor cell pool. J Orthop Res 1996 July;14(4):582‐9.
(199) Dy P, Wang W, Bhattaram P, Wang Q, Wang L, Ballock RT, Lefebvre V. Sox9 directs hypertrophic maturation and blocks osteoblast differentiation of growth plate chondrocytes. Dev Cell 2012 March 13;22(3):597‐609.
104
(200) Perrot‐Applanat M, Lescop P, Milgrom E. The cytoskeleton and the cellular traffic of the progesterone receptor. J Cell Biol 1992 October;119(2):337‐48.
(201) Jang WG, Kim EJ, Kim DK, Ryoo HM, Lee KB, Kim SH, Choi HS, Koh JT. BMP2 protein regulates osteocalcin expression via Runx2‐mediated Atf6 gene transcription. J Biol Chem 2012 January 6;287(2):905‐15.
(202) Yang S, Xu H, Yu S, Cao H, Fan J, Ge C, Fransceschi RT, Dong HH, Xiao G. Foxo1 mediates insulin‐like growth factor 1 (IGF1)/insulin regulation of osteocalcin expression by antagonizing Runx2 in osteoblasts. J Biol Chem 2011 May 27;286(21):19149‐58.
(203) Shimoyama A, Wada M, Ikeda F, Hata K, Matsubara T, Nifuji A, Noda M, Amano K, Yamaguchi A, Nishimura R, Yoneda T. Ihh/Gli2 signaling promotes osteoblast differentiation by regulating Runx2 expression and function. Mol Biol Cell 2007 July;18(7):2411‐8.
(204) Matsuguchi T, Chiba N, Bandow K, Kakimoto K, Masuda A, Ohnishi T. JNK activity is essential for Atf4 expression and late‐stage osteoblast differentiation. J Bone Miner Res 2009 March;24(3):398‐410.
(205) Ahluwalia A, Tarnawski AS. Critical role of hypoxia sensor‐‐HIF‐1alpha in VEGF gene activation. Implications for angiogenesis and tissue injury healing. Curr Med Chem 2012;19(1):90‐7.
(206) Rhee S, Jiang H, Ho CH, Grinnell F. Microtubule function in fibroblast spreading is modulated according to the tension state of cell‐matrix interactions. Proc Natl Acad Sci U S A 2007 March 27;104(13):5425‐30.
(207) Huang Y, Arora P, McCulloch CA, Vogel WF. The collagen receptor DDR1 regulates cell spreading and motility by associating with myosin IIA. J Cell Sci 2009 May 15;122(Pt 10):1637‐46.
(208) Watanabe T, Noritake J, Kaibuchi K. Regulation of microtubules in cell migration. Trends Cell Biol 2005 February;15(2):76‐83.
(209) Hidalgo‐Carcedo C, Hooper S, Chaudhry SI, Williamson P, Harrington K, Leitinger B, Sahai E. Collective cell migration requires suppression of actomyosin at cell‐cell contacts mediated by DDR1 and the cell polarity regulators Par3 and Par6. Nat Cell Biol 2011 January;13(1):49‐58.
(210) Day E, Waters B, Spiegel K, Alnadaf T, Manley PW, Buchdunger E, Walker C, Jarai G. Inhibition of collagen‐induced discoidin domain receptor 1 and 2 activation by imatinib, nilotinib and dasatinib. Eur J Pharmacol 2008 December 3;599(1‐3):44‐53.
105
(211) Schlumberger W, Thie M, Rauterberg J, Robenek H. Collagen synthesis in cultured aortic smooth muscle cells. Modulation by collagen lattice culture, transforming growth factor‐beta 1, and epidermal growth factor. Arterioscler Thromb 1991 November;11(6):1660‐6.
(212) Janke C, Bulinski JC. Post‐translational regulation of the microtubule cytoskeleton: mechanisms and functions. Nat Rev Mol Cell Biol 2011 December;12(12):773‐86.
(213) Desai A, Mitchison TJ. Microtubule polymerization dynamics. Annu Rev Cell Dev Biol 1997;13:83‐117.
(214) Sablina AA, Chumakov PM, Levine AJ, Kopnin BP. p53 activation in response to microtubule disruption is mediated by integrin‐Erk signaling. Oncogene 2001 February 22;20(8):899‐909.
(215) Watanabe T, Noritake J, Kaibuchi K. Regulation of microtubules in cell migration. Trends Cell Biol 2005 February;15(2):76‐83.