FUNCTIONAL INTERACTIONS BETWEEN THE

DISCOIDIN DOMAIN RECEPTOR 1 AND β1 INTEGRINS

by

Lisa Alexandra Staudinger

A thesis submitted in conformity with the

requirements for the degree of

Master of Science

Graduate Department of Dentistry

University of Toronto

© Copyright by Lisa Alexandra Staudinger 2013

ii

Functional Interactions Between the Discoidin Domain Receptor 1 and β1 Integrins

Lisa Alexandra Staudinger

Master of Science

Graduate Department of Dentistry

University of Toronto

2013

Abstract

The rate limiting step of phagocytosis is the binding of collagen to specific receptors, which

include β1 integrins and the discoidin domain receptor 1 (DDR1). While these two receptors may

interact, the functional nature of these interactions is not defined. We examined the effects of

DDR1 over-expression on β1 integrin function and determined that DDR1 over-expression

enhanced cell attachment through β1 integrins. These data are consistent with data showing that

DDR1 over-expression enhanced cell-surface, but not total, β1 integrin expression and

activation. As shown by experiments with endoglycosidase H, DDR1 over-expression increased

glycosylation of the β1 integrin subunit. Collectively these data indicate that DDR1 enhances β1

integrin interactions with fibrillar collagen, possibly by affecting the processing and trafficking

of β1 integrins to the cell surface. Our data provide insight into the mechanisms by which

fibrotic conditions such as cyclosporine A-induced gingival overgrowth are regulated.

iii

Acknowledgments

My Master of Science has been the most challenging, most tiring, but by far, the most rewarding

two years of my life. However, I cannot take sole credit for the completion of my degree, as this

work would not have been possible if not for the support of so many wonderful individuals. First

and foremost, I would like to thank Dr. McCulloch for taking a chance and allowing me to work

in his lab. His immense intelligence and passion for science have been an inspiration, and his

guidance and support, invaluable. I have felt it a privilege from day one to be working under his

supervision. I am more than grateful to Dr. Moriarty for being endlessly generous with her time

and for always providing me with valuable advice and direction. I am thankful to my committee

member Dr. Bendeck for her more than helpful suggestions. Thank you to Wilson Lee for always

making me feel like the funniest person alive by laughing at my jokes, no matter how bad they

were, and of course, for the endless hours he has put in to help make my thesis complete. For

much of the contents of this work I must also thank Stephen Spano, working with him was

always a blast. I would like to say thanks to Carol Laschinger for teaching me how to make the

perfect western blot and to my lab-mates Hamid Mohammadi, Dhaarmini Rajshankar, Qin

Wang, Vanessa Pinto, Dominik Fritz, Ilana Talior and Pam Arora for their advice, friendship and

fun times in the lab. I would like to thank Andrew Peters for inspiring me to not be afraid to

pursue science, for keeping me motivated, and for his constant support and encouragement. Last

but not least, I would like to thank my family. My grandparents Grete and Peter for their

financial, and of course emotional investment in my future, and to my brother PJ and parents

Gail and Herb for their unconditional love and support.

iv

Table of Contents

Abstract ………………………………………………………………………….... ii

Acknowledgements ……………………………………………………….……… iii

Table of Contents …………………………………………………………………. iv

List of Diagrams …………………………………………………………………… vii

List of Figures …………………………………………………………………….. viii

List of Appendices ………………………………………………………………... xii

Abbreviations …………………………………………………………………….. xiii

Chapter 1 – Literature Review ………………………………………………….. 1

I. Soft Connective Tissue ……………………………………………….. 1

A. Structure and Function of the Extracellular Matrix ...…………….... 1

i. Collagens …………………………………………………… 2

B. Cells of the Soft Connective Tissue ………………………………… 3

II. Homeostasis …………………………………………………………… 3

A. Function ……………...…………………………………………….. 4

B. Regulatory Mechanisms…………………………………………….. 4

i. Matrix Metalloproteinases…………………………………... 5

III. Disorders of Homeostasis …………………………………………….. 6

A. Mechanisms of Fibrosis…………………………………………….. 6

B. Drug-Induced Gingival Overgrowth………………………………… 7

v

i. Cyclosporin A-Induced Gingival Overgrowth……………… 8

IV. Phagocytosis …………………………………………………………… 9

A. Mechanism of action………………………………………………… 9

V. Receptors ……………………………………………………………… 12

A. Collagen binding: The Rate Limiting Step in Matrix Remodeling…. 12

B. β1 Integrins…………………………………………………………. 13

i. Structure and Function……………………………………… 13

ii. Integrin Signaling and Regulation of Activity……………… 15

iii. Glycosylation……………………………………………….. 16

C. DDR1……………………………………………………………….. 18

i. Structure and Function……………………………………… 18

ii. Isoforms…………………………………………………….. 20

iii. Signaling Pathways…………………………………………. 21

iv. DDR1 and Fibrosis…………………………………………. 22

VI. DDR1 Interactions ……………………………………………………. 23

A. DDR1 and Metalloproteinases……………………………………… 23

B. DDR1 and β1 Integrins……………………………………………… 24

Statement of the Problem ……………………………………………………….... 26

Chapter 2 – Introduction ………………………………………………………… 27

Materials and Methods……………………………………………………………. 30

Results ……………………………………………………………………………… 37

vi

Discussion ………………………………………………………………………….. 45

Figure Legends ……………………………………………………………………. 51

Figures ……………………………………………………………………………... 57

Future Directions and Conclusions…………………………………………….… 68

Appendix …………………………………………………………………………... 71

Appendix Legends ………………………………………………………………… 72

Appendix Figures…………………………………………………………………... 74

References …………………………………………………………………….…… 77

vii

List of Diagrams

Diagram 1. A severe case of Cyclosporin A-induced gingival overgrowth…….. 9

Diagram 2. Intracellular pathway of collagen degradation: Phagocytosis…….... 10

Diagram 3. The β1 integrin family of extracellular matrix receptors …………… 14

Diagram 4. β1 integrin activation in response to collagen binding……………… 16

Diagram 5. Structure of DDR1 transcript variants and isoforms …………..…… 21

viii

List of Figures

Figure 1. ……………………………………………………………………… 57

A. Immunoblot of DDR1 expression

B. Immunoblot of β1 integrin expression

Figure 2. ……………………………………………………………………… 58

A. Histogram of collagen-coated bead binding after 1 hour

B. Histogram of collagen-coated bead binding after 2, 4 and 24 hours

C. Histogram of fibronectin-coated bead binding after 1 hour

D. Histogram of collagen-coated bead binding after incubation with CD29 blocking

antibody

E. Histogram of cell surface α2 integrin expression

Figure 3. ……………………………………………………………………… 59

A. Histogram of floating gel collagen contraction assay

B. Histogram of attached gel contraction assay

Figure 4. ……………………………………………………………………… 60

A. Images of in vitro cell migration assay

B. Line plot of reduction in scratch width

ix

C. Histogram of mean slope of scratch width

Figure 5. ……………………………………………………………………… 61

A. TIRF images of activated β1 integrin staining in focal adhesions

B. Histogram of stained β1 integrin total cell area

C. Histogram of number of activated β1 integrin containing focal adhesions

D. Histogram of stained activated β1 integrin focal adhesion length

E. Histogram of stained area of activated β1 integrin containing focal adhesions

Figure 6. ……………………………………………………………………… 62

A. TIRF images of talin stained focal adhesions

B. Histogram of stained talin total cell area

C. Histogram of number of stained talin containing focal adhesions

D. Histogram of stained talin focal adhesion length

E. Histogram of stained area of talin containing focal adhesions

Figure 7. ……………………………………………………………………… 63

A. TIRF images of paxillin stained focal adhesions

B. Histogram of paxillin stained total cell area

x

C. Histogram of number of paxillin containing focal adhesions

D. Histogram of stained paxillin focal adhesion length

E. Histogram of stained area of paxillin containing focal adhesions

Figure 8. ……………………………………………………………………… 64

A. TIRF images of vinculin stained focal adhesions

B. Histogram of vinculin stained total cell area

C. Histogram of number of vinculin containing focal adhesions

D. Histogram of stained vinculin focal adhesion length

E. Histogram of stained area of vinculin containing focal adhesions

Figure 9. ……………………………………………………………………… 64

A. Histogram of total β1 integrin surface expression

B. Histogram of activated β1 integrin expression of cells plated on collagen

C. Histogram of activated β1 integrin expression of cells plated plastic

D. Histogram of activated β1 integrin expression of cells in suspension

Figure 10. Immunoblot of β1 integrin expression after treatment with the disintegrin

jararhagin…………………………………………………………… 66

xi

Figure 11. Immunoblot of β1 integrin expression after treatment with endoglycosidase H

…………………………………………………………………….... 67

xii

List of Appendices

Appendix 1 …………………………………………………………….... 74

A. Histogram of collagen-coated bead binding in cyclosporin A treated

human gingival fibroblasts after 1 hour

B. Histogram of collagen-coated bead binding in cyclosporin A treated cells

after 1 hour

Appendix 2 …………………………………………………………….... 75

A. Histogram of collagen-coated bead binding in cyclosporin A treated cells

after overnight plating inside collagen gels

B. Histogram of floating collagen gel contraction assay of cyclosporin A

treated cells

C. Histogram of attached collagen gel contraction assay of cyclosporin A

treated cells

Appendix 3. ………………………………………………………………. 76

A. Line plot of reduction in scratch width of cyclosporin A treated cells

B. Histogram of mean slope of scratch width of cyclosporin A treated cells

xiii

Abbreviations

ACER2 Alkaline ceramidase 2

ATP Adenosine triphosphate

BCA Bicinchoninic acid

BSA Bovine serum albumin

Cdc42 Cell division control protein 42 homolog

CsA Cyclosporin A

DDR1 & -2 Discoidin Domain receptor 1 and 2

ECM Extracellular Matrix

ER Endoplasmic reticulum

ERK Extracellular signal-regulated kinase

FAK Focal adhesion kinase

FITC Fluorescein isothiocyanate

GnT N-acetylglucosamine-glycosyltransferase

GTPase Guanosine triphosphatase

HEPES 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid

HGF Human gingival fibroblast

IG Immunoglobulin

JAK2 Janus kinase 2

JNK c-Jun NH2 -terminal kinase

LAMP-2 Lysosomal associated membrane protein 2

MAPK Mitogen activated kinase

MDCK Madin-Darby canine kidney epithelial cell line

MMP Matrix metalloproteinase

mRNA Messenger ribonucleic acid

NF-κB Nuclear factor kappa-light-chain-enhancer of activated B cell

PBS Phosphate buffered saline

PI3K Phosphatidylinositide 3-kinase

PMSF Phenylmethylsulfonyl fluoride

RNA Ribonucleic acid

RTK Receptor tyrosine kinase

RT-PCR Reverse transcription polymerase chain reaction

SDS-PAGE Sodium dodecyl sulfate polyacrylamide gel electrophoresis

STAT5 Signal transducer and activator of transcription 5

TGF-β Transforming growth factor beta

TIMP Tissue inhibitors of metalloproteinase

TIRF Total internal reflection fluorescence

TNF-α Tumor necrosis factor alpha

TNT Tris-HCL, NaCl, Tween

TRAF6 TNF receptor associated factor 6

V-ATPase Vacuolar-type H+-adenosine triphosphatase

SEM Standard error of the mean

1

Chapter 1

Literature Review

I. Soft Connective Tissue

I. A. Structure and Function of the Extracellular Matrix

The extracellular matrix (ECM) is the major structural component of connective tissues (Alberts,

2002) providing support and attachment for organs, ligaments, fascia and other specialized

components of organs and tissues (Hynes, 2009). The ECM acts to support and orient layers of

cells at the basement membrane and in the interstitium provides a substrate for cell migration.

ECM molecules are ligands for cell adhesion and signaling through cell surface receptors. The

mechanical properties of the ECM regulate many aspects of cell behavior (Hynes, 2009).

Further, the ECM is essential for the differentiation, proliferation, survival, polarity, and

migration of cells with which it is in contact (Alberts, 2002; Hynes, 2009; Kim et al., 2011;

Woessner, 1993).

Mammalian genomes encode hundreds of ECM proteins, including collagens, laminin,

fibronectin, proteoglycans and glycosaminoglycans (Alberts, 2002; Hynes, 2009; Kinane, 2000;

Woessner, 1993). Variations in the number, organization, composition and distribution of ECM

proteins contribute to the distinctive structure and function of connective tissues (Hynes, 2009).

The most abundant component of the ECM is the family of collagens (Eckes et al., 1999; Kim et

al., 2011; Kinane, 2000; Perez-Tamayo, 1978), which together endow the ECM with impressive

structural and information-processing properties.

2

I. A. i. Collagens

Collagens are a family of proteins with at least 29 known members (Kim et al., 2011). Type I

collagen makes up approximately 80% of the global extracellular matrix in mammals (Eckes et

al., 1999) although minor collagens such as type II, III, V and XI are also present in defined

tissues and in special circumstances such as in development and wound healing (Eckes et al.,

1999; Eyre, 2004; Kim et al., 2011; Narayanan and Page, 1983). All collagens share a common

structural motif of helical fibrils that are formed by three protein subunits. The primary function

of collagens is to act as the structural support for connective tissues; collagens also act as binding

partners for other ECM proteins (Kim et al., 2011).

Interstitial collagens consist of three α chains composed of approximately 1000 residues with

repeating Gly-X-Y triplets, where X and Y are most commonly proline and hydroxyproline,

respectively (Chung et al., 2004). Due to the high imino-acid content and the tri-peptide unit

repeats, each α chain assumes a left-handed helical conformation, producing a 95 kDa pro-

collagen α chain subunit. The three left-handed α chains intertwine with one another to form a

right-handed collagen fibril which then bind together with other collagen fibrils to form the

collagen fiber (Chung et al., 2004; Messent et al., 1998; Meyer and Morgenstern, 2003). The

triple-helical conformation of collagen makes interstitial collagens resistant to most proteinases

(Chung et al., 2004). In vertebrates, intracellular, lysosomal hydrolases known as cathepsins and

extracellular collagenases known as matrix metalloproteinases (MMPs), cleave the triple-helical

structure of collagens to enable degradation and remodeling of the ECM in health, development

and disease (Chung et al., 2004; Messent et al., 1998).

3

I. B. Cells of the Soft Connective Tissue

Connective tissue cells comprise a very large group of cells in soft and mineralizing tissues and

include for example, fibroblasts, chondrocytes, adipocytes, osteoblasts, osteocytes, odontoblasts

and smooth muscle cells, all of which are specialized for the secretion and degradation of

collagens and other ECM proteins (Alberts, 2002). Fibroblasts in particular are the cells

responsible for the maintenance and organization of soft connective tissues (Grinnell, 2003).

Previously, fibroblasts were believed to only synthesize the supporting ECM framework for

other cell types. More recently they are reported to function in a wide range of immune and

inflammatory responses (Fries et al., 1994; Williams, 1998).

In response to cytokines, fibroblasts are able to modify their production of ECM components,

which allow them to respond and cooperate with cells in complex responses such as wound

healing. Intercellular communication between fibroblasts and cells of the immune system occurs

frequently and contributes to the maintenance of homeostasis in mammalian tissues. For

example, in chronic inflammatory diseases, perturbations of fibroblast function can lead to the

development of disabling fibrotic disorders (Williams, 1998).

II. Homeostasis

Connective tissue remodeling results in a change in shape and/or structure of tissues whereas

turnover involves the degradation and replacement of ECM proteins with no alteration in tissue

structure. In both cases there is no impairment of tissue function. The functional differences

between turnover and remodeling are relatively minor and the two terms are often used

4

interchangeably. Notably, in both processes, the synthesis and degradation of ECM must be kept

in a steady state to maintain normal tissue form and function (Ten Cate and Deporter, 1974).

II. A. Function

Homeostasis involves a dynamic, functional balance between the production and breakdown of

ECM components (Knowles et al., 1991). In soft connective tissues, fibroblasts contribute to

tissue homeostasis and maintain equilibrium by balancing synthesis and degradation of collagen

in physiological remodeling (Knowles et al., 1991). In normal, healthy tissues, remodeling is

necessary for growth, development and for tissue repair in various pathological conditions

including inflammatory diseases (Cox and Erler, 2011; Knowles et al., 1991).

II. B. Regulatory Mechanisms in Collagen Homeostasis

The mechanisms that regulate collagen homeostasis are not completely defined. Currently, two

major routes for collagen degradation have been identified: an extracellular and an intracellular

pathway (Knowles et al., 1991). In soft connective tissues, three separate extracellular pathways

for ECM degradation have been identified: the MMP pathway, the plasmin pathway, and the

pathway involving polymorphonuclear leukocyte-derived serine proteinases (Birkedal-Hansen et

al., 1993). The polymorphonuclear leukocyte serine protease pathway mediates degradation of

the ECM at neutral pH (Brower and Harpel, 1982) through the release of serine proteases,

neutrophil elastase and cathepsin G (Birkedal-Hansen et al., 1993; Pham, 2006). These proteases

are able to cleave various proteins including collagens, laminin, fibronectin and proteoglycans

(Birkedal-Hansen et al., 1993).

5

Plasminogen, the precursor to plasmin, is found at high levels circulating in lymph and

interstitial fluids (Birkedal-Hansen et al., 1993). This broad spectrum enzyme plays an important

role in matrix remodeling, whereby activated plasmin rapidly cleaves Lys-Arg peptide bonds

exposed on the surface of a wide range of ECM macromolecules including fibrin and fibronectin

(Birkedal-Hansen et al., 1993). Although type I and II collagens cannot be cleaved by plasmins,

these enzymes can activate MMPs (Davis et al., 2001) and thereby increase the degradation rates

of collagen and other ECM molecules by MMPs.

II. B. i. Matrix Metalloproteinases

MMPs are zinc-dependent endopeptidases that can be activated at neutral pH. These enzymes are

secreted in a pro-form and are typically activated near the cell surface by other activated MMPs

or by serine proteinases. Their activity is tightly regulated by tissue inhibitors of

metalloproteinases (TIMPs) (Davis et al., 2001; Hwang et al., 2009; Messent et al., 1998) and by

cell receptors such as β1 integrins (Eckes et al., 1999). The MMP family consists of at least 23

known members, three of which are interstitial collagenases (Haas et al., 2000). MMP-1, -8 and

-13 are the only three known members of this family that can cleave and degrade triple helical

collagen fibrils, generating fragments approximately one quarter and three quarters of the total

length of the native molecule (Chung et al., 2004; Eyre, 2004; Messent et al., 1998). These

fragments denature at temperatures below physiological body temperatures, thereby allowing

normally insoluble collagen to denature into gelatin (Eyre, 2004; Messent et al., 1998). Gelatin in

turn is degraded further by gelatinases such as MMP-2 and MMP-9.

The role of collagenases is not just simply to degrade the collagen matrix; these enzymes also

function to control cell behavior during tissue remodeling. Collagenase cleavage fragments alter

6

cellular activity by expressing cryptic biological functions (Chung et al., 2004) such as the

promotion of epithelial cell migration during wound healing (Zhao et al., 1999) and the

enhancement of tumor necrosis factor-α (TNF-α) processing and release (Birkedal-Hansen,

1995; Santibanez-Gallerani et al., 2000).

III. Disorders of Homeostasis

III. A. Mechanisms of Fibrosis

When the balance between the synthesis and breakdown of collagen is disrupted, disorders of the

ECM such as fibrosis may develop. Inflammatory disease of the skin, liver, kidney, heart,

gingiva and lung can all be attributed to overabundance and deregulated structural organization

of ECM proteins, whether through reduced degradation, increased synthesis or inappropriate

molecular organization (Fries et al., 1994). For example in osteoarthritis there is a shift towards

increased proteolytic damage by MMPs that lead to the destruction of joint cartilages (Eyre,

2004). Similarly in periodontal tissues, the gingival crevicular fluid of patients with progressive

periodontitis exhibits much higher active collagenase activity compared with healthy and

gingivitis patients, emphasizing the role of active collagenase in destructive periodontitis

(Kinane, 2000; Overall et al., 1991). Collagenase disrupts the fibrous meshwork of the ECM and

increases matrix porosity (Berrier and Yamada, 2007), facilitating infiltration by inflammatory

cells, as is seen in early stages of gingivitis (Kinane, 2000).

Conversely, collagen degradation may be inhibited in various fibrotic disorders. An in vitro

study which examined collagen-coated bead internalization by human gingival fibroblasts noted

a large reduction in bead internalization in cells obtained from fibrotic lesions compared with

7

cells from normal tissues (McCulloch and Knowles, 1993). Methylglyoxal has been detected in

increased concentrations in the gingival crevicular fluids of patients with chronic periodontitis

infections (Kashket et al., 2003). This glucose metabolite has been shown to block the binding

step of phagocytosis though modifications of the Gly-Phe-Hyp-Gly-Glu-Arg binding motif in

collagen (Chong et al., 2007). Methylglyoxal can also be found at high levels in patients with

diabetes where individuals often suffer from cardiac fibrosis. Recent data suggest that

methylglyoxal not only inhibits cell-matrix attachments but also promotes the development of

intercellular adhesions and myofibroblast differentiation (Yuen et al., 2010).

Excessive production of matrix proteins is also associated with several diseases and fibrotic

conditions (Diegelmann and Evans, 2004). In radiation-induced fibrosis, type I collagen

production is increased 14-fold compared with healthy tissues (Remy et al., 1991). Similarly, in

fibroblast cultures from individuals with progressive systemic scleroderma, there was a marked

increase in type I and II collagen synthesis (Krieg et al., 1985). In oral sub-mucous fibrosis, a

potential mechanism for increased collagen production has been suggested whereby transforming

growth factor β (TGF-β)-induces synthesis of connective tissue growth factor-CCN2, which in

turn is mediated by c-Jun NH2-terminal kinase (JNK), activin receptor-like kinase 5, and the p38

mitogen activated kinase (MAPK) (Chang et al., 2012).

III. B. Drug-Induced Gingival Overgrowth

The gingiva and periodontal ligament are excellent models for the study of ECM homeostasis in

soft connective tissues because they exhibit a remarkably rapid rate of collagen turnover (Sodek,

1977). Further, drug-induced gingival overgrowth offers a good example of the complexities of

homeostatic imbalance when small perturbations of collagen turnover occur as a result of certain

8

drugs that affect collagen remodeling. Drug-induced gingival overgrowth is associated with three

different classes of drugs: phenytoin, an antiepileptic; cyclosporin A (CsA), an

immunosuppressant; and the calcium channel blockers nifedipine and diltiazem (Kataoka et al.,

2005). The severity and prevalence of drug-induced overgrowth differs between medications and

there are also very large inter-patient variations of overgrowth in response to drugs. What is most

perplexing about these disorders is that the pathological mechanisms of gingival overgrowth are

not well-defined. Currently, these disorders appear to be induced by disruptions in collagen

homeostasis (Kataoka et al., 2005).

III. B. i. Cyclosporin A-Induced Gingival Overgrowth

In the context of CsA-induced gingival overgrowth (Diagram 1), several studies have described

potential mechanisms by which gingival fibrosis is induced. It has been proposed that MMPs,

specifically those that degrade collagen and gelatin, and TIMPs, are implicated in CsA-induced

fibrosis (Gagliano et al., 2004; Hyland et al., 2003; Murai et al., 2011; Tipton et al., 1991).

Currently, the literature to support this view is inconclusive. MMP levels are reported as being

either unchanged (Gagliano et al., 2004; Murai et al., 2011; Tipton et al., 1991) or decreased

(Hyland et al., 2003) in patients who take CsA.

Similarly, TIMP levels have been reported as

unchanged (Gagliano et al., 2004; Hyland et al., 2003) or increased (Tipton et al., 1991) by CsA.

In fibroblasts, CsA inhibits the release of calcium ions from intracellular stores, such as the

endoplasmic reticulum (Arora et al., 2001) and uptake of calcium ions by mitochondria

(Gagliano et al., 2004), thereby blocking integrin activation and inhibiting collagen phagocytosis.

Further, the lysosomal enzymes cathepsins B and L, which digest collagen intracellularly, have

been identified as possible targets for inhibition by CsA. Recently it has been shown that CsA

9

impairs cathepsin B and L activity and down-regulates their synthesis (Murai et al., 2011; Omori

et al., 2007). Although the exact mechanisms of drug-induced fibrosis are unknown, collagen

degradation by fibroblasts in the gingiva is clearly inhibited (Arora et al., 2001; Chan et al.,

2007).

Diagram 1. A severe case of Cyclosporin A-induced gingival overgrowth.

Image courtesy of: http://www.exodontia.info/Drug-Induced_Gingival_Hyperplasia.html

IV. Phagocytosis

IV. A. Mechanism of Action

The intracellular or phagocytic pathway of collagen degradation is distinct from the extracellular,

MMP-dependent pathway. In certain tissues such as the periodontium, MMPs are thought to be

responsible for most ECM degradation that occurs in inflammatory lesions (Sodek and Overall,

1988) whereas phagocytosis by fibroblasts is believed to be responsible for the majority of

collagen turnover in healthy tissues (Segal et al., 2001; Ten Cate and Deporter, 1974).

Phagocytosis is a receptor-driven process that is dependent on remodeling of the actin

10

cytoskeleton and shares some of the cytoskeletal regulatory components involved in cell

migration and cell adhesion (Arora et al., 2008; Segal et al., 2001).

As outlined in Diagram 2, the process of collagen degradation can be broken down into five

major stages that involve: A) the initial recognition of the fibril by membrane-bound collagen

receptors; B) partial digestion of collagen by cell-surface MMPs; C) envelopment of the collagen

fibril by the phagocytic cup, and subsequent internalization into the phagosome; D) fusion of the

phagosome with the lysosome to form the phagolysosome, followed by a rapid decrease in pH to

allow for E) lysosomal digestion by cysteine proteases, such as cathepsins B and L (Everts et al.,

1996; Yajima, 1986).

Diagram 2. Intracellular pathway of collagen degradation: Phagocytosis

Image courtesy of: Dr. C.A.G. McCulloch

11

Internalization of the collagen fibril in phagocytosis is mediated by the initial binding of collagen

by cognate receptors such as the α2β1 integrin (Arora et al., 2000). Further, in addition to the

importance of the α1β1 integrin in the binding step of phagocytosis, this adhesion receptor may

also regulate the internalization step (Lee et al., 1996). During phagocytosis, actin assembly and

integrin activation are coordinated through activation of members of the Rho family of small

gaunosine triphosphatases (GTPase) in response to extracellular signals including collagen-

binding (Arora et al., 2008; Cougoule et al., 2004; Del Pozo et al., 2002). The small GTPase

Rap1 is known to activate β1 integrins (Bos, 2005; de Bruyn et al., 2002) and is also required for

collagen phagocytosis. Activation and recruitment of Rap1 to collagen binding sites are

dependent on non-muscle myosin II-A filament assembly (Arora et al., 2008), which is

implicated in cytoskeletal regulation and phagocytosis (Kang et al., 2002). Following

phagocytosis, the α2 integrin subunit is rapidly recycled or synthesized (Lee et al., 1996). While

little is known of the regulation of collagen degradation by phagocytosis, this process is thought

be stimulated by TGF-β (van der Zee et al., 1995).

As indicated above, digestion of phagocytosed particles in lysosomal compartments is largely

dependent on the activity of cysteine proteinases such as cathepsin B (Everts et al., 1996). These

enzymes cleave collagens in helical and non-helical regions that are distinct from MMP cleavage

sites and cleavage requires an acidic environment with an optimal pH range of 4.0-5.0 (Burleigh

et al., 1974). Acidification of the phagolysosome is necessary for collagen degradation but not

for the fusion events that lead to phagosomal maturation (Arora et al., 2000). Phagosomes

normally fuse with lysosomes downstream of collagen binding and internalization, which lead to

the production of phagolysosomes, wherein the ingested materials are degraded (Oh and

Swanson, 1996; Saftig and Klumperman, 2009). In collagen-containing vacuoles, the abundance

12

of lysosomal associated membrane protein 2 (LAMP-2), a late endosomal maturation marker that

is enriched in endosomes, peaks at 120 minutes after internalization of collagen-coated beads,

indicating that maximal fusion of the phagosome and lysosome occurs at about 120 minutes after

initial collagen binding (Arora et al., 2000). The phagolysosome is generally considered the

endpoint of the phagocytic pathway; however, there is evidence suggesting that the

phagolysosome continues to undergo maturational changes after phagosome-lysosome fusion

and that internalized particles can be further fragmented (Oh and Swanson, 1996). Notably, the

vacuolar-type H+-adenosine triphosphatase (V-ATPase) inhibitor bafilomycin A1 affects transit

of proteins from early to late endosomes (Clague et al., 1994) and late endosomes to lysosomes

(van Weert et al., 1995), suggesting that phagosomal maturation is a distinct pathway from

endosomal maturation.

V. Receptors

V. A. Collagen binding: The rate limiting step in matrix remodeling

The cellular recognition and binding to localized domains on collagen fibrils are significant early

events in the phagocytic pathway. Collagen binding is the rate-limiting step in phagocytic

degradation of collagen by fibroblasts (Chong et al., 2007; Knowles et al., 1991). Previous work

has demonstrated that in gingival fibroblasts the rate of phagocytosis of fibronectin- (McKeown

et al., 1990) and collagen-coated (Knowles et al., 1991) latex beads is dependent on attachment

to cell-surface receptors. Recognition and attachment systems in fibroblasts include cell surface

receptors with high avidity for collagen (Knowles et al., 1991) such as integrins, specifically the

α2β1 integrin, which is the principal adhesion receptor for type I fibrillar collagen (Chong et al.,

2007; Dickeson et al., 1999)

13

Many collagen receptors have been identified in mammalian tissues including integrin

heterodimers that comprise β1 integrins, discoidin domain receptors, glycoprotein VI, leukocyte-

associated immunoglobulin (IG)-like receptor 1, and the mannose receptor family which includes

endo180 and urokinase receptors (Leitinger and Hohenester, 2007). The β1 integrins are the best

defined and most studied mammalian receptors for cell adhesion to collagen (Hynes, 2002).

V. B. β1 Integrins

V. B. i. Structure and Function

The term integrin first came into use in the 1980’s to describe cell surface receptors that

“integrated” the cytoskeleton of one cell with that of an extracellular matrix protein (Schnapp,

2006). Integrins are heterodimeric cell adhesion receptors consisting of two, non-covalently

associated α and β subunits. There have been 18 α and 8 β subunits identified in mammals,

which pair to form a total of 24 integrins. The β1 subunit in particular, pairs with 12 different α

subunits (Gu and Taniguchi, 2004; Schnapp, 2006; Sun et al., 2009). Each pairing of an α

subunit with the β1 integrin leads to the formation of receptors with generally specific and non-

redundant functions. Indeed, the specific α and β integrin subunits contribute to ligand specificity

(Hynes, 2002) as outlined in Diagram 3. Aside from the β4 integrins, all β subunits have a large

extracellular domain, a single pass trans-membrane domain and a short cytoplasmic tail (Gu and

Taniguchi, 2004; Schnapp, 2006). The extracellular, N-terminal domains of α and β subunits

associate to form the integrin headpiece, which contains the ECM binding site (Gu and

Taniguchi, 2004). The C-terminal cytoplasmic region links to the actin cytoskeleton through

binding to proteins such as talin, filamin, and vinculin, and also mediates interactions with

cytoplasmic signaling molecules (Critchley, 2000; Gu and Taniguchi, 2004). Ligation of actin

14

filaments to integrins regulates the mobility of collagen receptors, thereby enhancing the

recruitment and increasing the interaction of integrins with collagen during cell spreading (Arora

et al., 2003).

Diagram 3. The β1 integrin family of extracellular matrix receptors.

Image courtesy of: (Lal et al., 2009)

The integrin family of cell adhesion receptors is essential for development, cell proliferation and

wound healing (Gu and Taniguchi, 2004; Schnapp, 2006). The functional activity of integrin

receptors are affected by subtle changes in the cellular environment, post-translational

modifications, their organization and orientation at biological membranes (Alberts, 2002), and

the presence of divalent cations, such as Ca2+

and Mg2+

(Schnapp, 2006). These various

molecules and modifications to integrins, promote or suppress ligand binding, change ligand

specificity, or stabilize integrin structure (Schnapp, 2006). For instance, Mg2+

promotes cell

adhesion whereas ligand affinity may be decreased by Ca2+

or increased by Mn2+

(Schnapp,

2006).

15

V. B. ii. Integrin Signaling and Regulation of Activity

In addition to their roles in enabling cell adhesion to the ECM, integrins also act as trans-

membrane linkages that connect the ECM to the cytoskeleton (Hynes, 2002; Hynes, 2009). The

cytoplasmic tails of integrins do not contain any intrinsic catalytic activity but instead act as

organizing centers and scaffolding molecules for other signaling components and cytoskeletal

proteins (Schnapp, 2006). The linkage provided by integrins facilitates the triggering of a large

variety of signal transduction events that occur in a bidirectional mode. Through inside-out

signaling, intracellular signals control cell motility and cell-ECM or cell-cell adhesion, whereas

outside-in signals relay extracellular messages into cells (Hynes, 2009; van der Flier and

Sonnenberg, 2001). Many integrins when bound by ligand can activate the focal adhesion kinase

(FAK), which subsequently activates a number of classical signaling pathways such as MAPKs,

phosphatidylinositide 3 kinase (PI3K), and protein kinase C (Schnapp, 2006). These signals

regulate cell behavior including proliferation, survival, polarity, migration and differentiation

(Hynes, 2009; Schnapp, 2006).

As shown in Diagram 4, integrins exist in different conformational states but at rest, most

integrins exist in a low affinity state (Schnapp, 2006). In response to ligand binding, signaling

pathways are activated that cause a conformational change in integrins to allow high affinity

binding (Calderwood, 2004). This change in integrin conformation is known as affinity

modulation (Schnapp, 2006). Ligand binding may also be strengthened through avidity

modulation, which involves clustering of individual integrin molecules at a discrete, ligand

binding site within the plasma membrane (Calderwood, 2004; Schnapp, 2006).

16

Diagram 4. β1 integrin activation in response to collagen binding.

Image courtesy of: (Niland and Eble, 2012)

V. B. iii. Glycosylation

Glycosylation is one of the most common types of post-translational modifications (Hagglund et

al., 2007) and indeed the majority of secreted and cell surface proteins are glycosylated (Gu and

Taniguchi, 2004). Glycosylation denotes a class of structurally diverse modifications that

comprise O and N-linked glycans with carbohydrate functional groups ranging from

monosaccharides to large, branched oligosaccharide chains (Gu and Taniguchi, 2004; Hagglund

et al., 2007). Carbohydrates on the cell surface are frequently the first molecules to be

17

encountered and recognized by other cells, antibodies and invading viruses and bacteria.

Accordingly these molecules contribute to a variety of interactions between the cell and the

extracellular environment (Gu and Taniguchi, 2004).

β1 integrins are glycosylated by N-glycans (Gu and Taniguchi, 2004). The β1 integrin subunit is

initially synthesized as an 87 kDa polypeptide core and partially glycosylated in the endoplasmic

reticulum (ER) to form the immature β1 (β1 precursor) (Sun et al., 2009). The β1 precursor is

transported to the Golgi complex where it is further glycosylated to produce the mature β1

(Akiyama et al., 1989; Bellis, 2004; Sun et al., 2009). This conversion from the precursor to the

mature form is denoted as the maturation of β1 integrins (Sun et al., 2009). Once the mature β1

pairs with an α subunit, the heterodimer is transported to the cell surface (Sun et al., 2009).

While it is unknown how this process is regulated, the β1 maturation process represents a

mechanism by which the function of β1 integrins may be controlled (Gu and Taniguchi, 2004;

Sun et al., 2009). The adhesion of integrins to specific peptide sequences of a particular ligand is

based on binding to α and β receptor subunits; however, strength of binding may be modulated

by glycosylation (Zheng et al., 1994). For example, cell-ECM or cell-cell adhesion and other

cellular processes mediated by integrins may be modulated by β1 maturation (Gu and Taniguchi,

2004; Sun et al., 2009). Zheng et al. (1994) showed that N-glycosylation is necessary for the

association of the α5 and β1 subunits and for normal binding of the adhesion receptor to

fibronectin. TGF-β accelerates β1 maturation, thereby increasing cell surface levels of β1

integrins and enhancing cell adhesion (Bellis et al., 1999). Talin and lipoprotein receptor-related

protein-1 are required for the proper maturation process of the β1 integrin (Salicioni et al., 2004;

Sun et al., 2009). Sphingosine that is generated in the Golgi complex through the action of the

human alkaline ceramidase 2 (ACER2), but not other ceramidases, inhibits β1 integrin

18

maturation, leading to defects in cell-ECM adhesion (Sun et al., 2009). In contrast, RNA

interference-mediated inhibition of ACER2 promotes β1 maturation and cell adhesion (Sun et al.,

2009).

V. C. DDR1

V. C. i. Structure and function

The discoidin domain receptors (DDRs) are a more recently discovered family of non-integrin-

type collagen receptors. Two gene products have been identified in mammals so far, and these

include DDR1 and DDR2, which constitute a subfamily of tyrosine kinase receptors (RTK)

(Leitinger, 2011; Vogel, 1999; Vogel et al., 1997). Both DDRs are different from other RTKs in

that they are activated only by binding to collagens in their native triple-helical form (Vogel et

al., 1997). DDR1 in particular is activated by all collagens tested so far, including collagens type

I-VI and VIII (Leitinger, 2011). Further, while RTK activation of most cell surface receptors is

generally very rapid, tyrosine phosphorylation of DDR is delayed, peaking at 90 minutes and

lasting upwards of 18 hours after cell attachment to collagen (Vogel et al., 1997).

DDR1 functions as a collagen sensor and is activated by several types of collagen, subsequently

triggering ECM degradation and turnover (Franco et al., 2002). In smooth muscles cells, the

absence of DDR1 is associated with decreased cell proliferation, migration, and MMP and

collagen synthesis (Franco et al., 2002). DDR1 and non-muscle myosin IIA interact to regulate

adhesive contacts with collagen since DDR1 promotes cells spreading and is important for the

assembly of non-muscle myosin IIA heavy chain into filaments on cells plated on collagen

(Huang et al., 2009).

19

DDR1 and -2 contain an extracellular discoidin domain that is composed of approximately 160

amino acids and contains two discoidin domains, which are members of the coagulation factor

V/VII type C superfamily (Ferguson, 2012). No other receptor tyrosine kinases contain

extracellular discoidin domains, but these domains are found in a number of other proteins

involved in cell adhesion (Ferguson, 2012). The extracellular region shares homology with a

region identified in the protein discoidin I from the slime mold Dictyostelium discoideum, where

it functions as galactose-binding lectin (Matsuyama et al., 2003).

Recombinant preparations of DDR1 have been used to identify loops 1 and 3 as the sites of

ligand binding and receptor activation (Vogel et al., 2006). The juxtamembrane regions of DDRs

are much longer than other RTKs and it has been speculated that similar to several growth factor

receptors or Eph receptor subfamilies, this region has an auto-inhibitory function that accounts

for the delayed kinetics of DDR1 phosphorylation (Vogel, 1999).

Unlike other RTKs whereby ligand binding precedes receptor dimerization, it is thought that

DDR1 exists as pre-formed dimers on the cell surface. Collagen binding activates signaling

cascades by altering the conformation of these dimers instead of by inducing receptor

oligomerization (Lemmon and Schlessinger, 2010). Recent reviews present conflicting notions

regarding the role of DDR1 cell surface dimers. According to Valiathan et al. (2012), DDR1

receptor dimerization of the ectodomains is a requirement for collagen binding. To support this

view, ligand-independent DDR1 dimers were found to originate within the biosynthetic pathway

and to be stable at the cell surface even in the absence of collagen (Abdulhussein et al., 2008;

Mihai et al., 2009; Noordeen et al., 2006). Conversely, Ferguson et al. (2012) presented evidence

indicating that short triple-helical collagen peptides are sufficient to activate DDR1, suggesting

20

that receptor clustering induced by collagen is not an essential aspect of DDR1 activation

(Ferguson, 2012).

V. C. ii. DDR1 Isoforms

DDR1 maps to human chromosome 6 and is composed of 17 exons, which are alternatively

spliced to produce five transcript variants, giving rise to five different isoforms (Valiathan et al.,

2012). Conversely, DDR2 has 19 exons, which encode for only a single transcript (Valiathan et

al., 2012). The DDR1 a- and b-receptor isoforms lack 37 and 6 amino acids respectively,

whereas the c-isoform is the longest at 919 residues. DDR1d is truncated and is missing the

entire kinase region. DDR1e is “kinase dead” and lacks sections of the juxtamembrane region

and the ATP binding site (Alves et al., 2001; Playford et al., 1996; Valiathan et al., 2012). The

most commonly expressed isoforms are DDR1 a and b. The relative expression ratios of these

isoforms and their post-translational modifications appear to be controlled by complex but poorly

understood regulatory mechanisms (Vogel et al., 2006).

21

Diagram 5. Structure of DDR1 transcript variants and isoforms.

Image courtesy of: (Alves et al., 2001)

V. C. iii. DDR1 Signaling Pathways

The activation of DDR1 appears to regulate a wide range of processes, suggesting that DDR1

may exert its effects through several pathways (Ruiz and Jarai, 2011). However, in general, the

downstream signaling pathways of DDR1 are not well defined (Vogel, 1999). It has been

proposed that DDR1 interacts with several cofactors including ShcA, Nck2, SHP-2, signal

transducer and activator of transcription 5 (STAT5), nuclear factor kappa-light-chain-enhancer

of activated B cells (NF-кB) and p38 MAPK (Ruiz and Jarai, 2011). Src regulates DDR1

activation and downstream signaling through ERK1/2 and influences cell migration (Lu et al.,

2011).

22

Downstream signaling of DDR is dependent on which DDR isoform is expressed. For instance,

upon collagen-induced tyrosine phosphorylation, DDR1b associates with the phosphotyrosine-

binding domain of the ShcA adaptor protein through its LLXNPXY motif (Vogel et al., 1997).

DDR1b in macrophages activates the TNF receptor associated factor 6 (TRAF6) complex,

triggering the p38 mitogen activated protein kinase and NF-κB pathways (Abdulhussein et al.,

2004; Baumgartner et al., 1998; Vogel et al., 1997). Conversely, fibroblast-growth factor

receptor substrate-2 binds to the juxtamembrane region of DDR1a (Vogel, 2001). SHP-2 is

necessary for suppression of STAT1 and STAT3 tyrosine phosphorylation, cell migration and

branching tubulogenesis induced by DDR1 a- or b- isoforms (Wang et al., 2006).

V. C. iv. DDR1 and Fibrosis

DDR1 has been implicated in various fibrotic conditions as indicated by experiments using

knock-out mice and DDR1 silencing and over-expression systems. Flamant and colleagues

(Flamant et al., 2006) reported reduced fibrosis and immune cell infiltration in DDR1 knock-out

mice compared with wild-type animals, which exhibited increased fibrosis and macrophage and

leukocyte infiltration and increased microalbumin, an indicator of renal disease. Similarly, Gross

and co-workers (Gross et al., 2010) demonstrated that loss of DDR1 expression in mouse kidney

delayed renal fibrosis and inflammation by reducing TGF-β and connective tissue growth factor.

Compared with wild-type controls, DDR1 knock-out mice exhibit reduced collagen levels,

myofibroblast differentiation and immune cell infiltration in bleomycin-induced lung fibrosis

(Avivi-Green et al., 2006). The absence of DDR1 decreased smooth muscle cell proliferation and

collagen synthesis, thereby reducing intimal thickening of the arterial wall (Franco et al., 2002).

In cirrhotic liver, transcriptional profiling using DNA array and RT-PCR have detected elevated

23

DDR1 mRNA expression by more than two-fold, which is manifest in hepatocytes, leukocytes

and biliary epithelial cells (Song et al., 2011). In the same study it was found that there was

increased abundance of soluble and membrane-associated DDR1 fragments of varying size,

which were detected in cirrhotic liver but not in liver. Further, increased DDR1a expression

affected cell behavior in hepatocytes, which included elevations of cell adhesion and

immobilization of cells on type I collagen and fibronectin (Song et al., 2011).

VI. DDR1 Interactions

VI. A. DDR1 and Metalloproteinases

As indicated above, cell attachment to type I collagen induces DDR1 phosphorylation and

activation as well as the production of a C-terminal fragment of DDR1 (Vogel, 2002). The

generation of the C-terminal fragment can be blocked by metalloproteinase inhibitors, indicating

that DDR1 might represent a novel target for limited proteolysis by a family of enzymes known

as disintegrin metalloproteinases or secretases (Slack et al., 2006; Vogel, 2002). Two inhibitors

of transporter-associated-with-antigen-processing subunit 1 and TIMP-3 caused dose-dependent

reductions in collagen-induced shedding of the DDR1 ectodomain but did not block generation

or phosphorylation of the C-terminal fragments. These data suggest that liberation of the DDR1

C-terminal fragments may be mediated by a distinct collagen-regulated protease (Dejmek et al.,

2003; Franco et al., 2002). Conceivably, DDR1 ectodomain shedding may be part of an

inhibitory feedback loop that serves to regulate the ability of DDR1 to promote adhesion and

migration of cells on collagen (Slack et al., 2006).

24

DDR1 is widely expressed in fast-growing invasive tumors of breast (Johnson et al., 1993),

ovary (Laval et al., 1994), esophagus (Nemoto et al., 1997) and brain (Ram et al., 2006).

Hepatocellular carcinoma cells over-expressing either DDR1 a- or b-isoforms showed a

significant increases of MMP-2 and -9 expression, indicating that increased tumor invasiveness

may be linked to increased DDR1 and may be mediated by MMP-2 or -9 (Park et al., 2007). To

support this finding Hou and co-workers (Hou et al., 2001) found that MMP-2 and -9 levels were

down-regulated in DDR1-null smooth muscle cells. Ruiz and Jari (2011) showed that in normal

human lung fibroblasts, type I collagen is able to induce DDR1 and MMP-10 expression through

integrin-independent activation of DDR2. These authors also found that type I collagen-induced

DDR1 gene expression requires recruitment of phospho-Janus kinase 2 (JAK2) to DDR2 and

extracellular signal regulated kinase (ERK) 1/2 activation, and involves the recruitment of the

nuclear factor PEA3 to the DDR1 promoter region (Ruiz and Jarai, 2011).

VI. B. DDR1 and β1 Integrins

DDR1 and β1 integrins are two completely different types of collagen receptors. DDR1 can be

activated in the presence of β1 integrin blocking antibodies, indicating that DDRs can participate

in signaling responses independent of integrins (Vogel et al., 2000). Notably, some of the

downstream signaling pathways activated by DDR1 appear to intersect with β1 integrin-activated

pathways (Valiathan et al., 2012).

In MDCK cells, DDR1 blocks integrin→FAK→Cdc42-mediated cell spreading (Yeh et al.,

2009) and integrin-STAT1/3-mediated cell migration (Wang et al., 2006). A recent study by Suh

and Han (2011) examined the effect of collagen I on mouse embryonic stem cell self-renewal

and related signal pathways. They found that binding of type I collagen to β1 integrins activated

25

integrin-linked kinase, Notch and Gli-1. In contrast, type I collagen binding to DDR1 activated

Ras, PI3K/Akt and ERK. Both activated Gli-1 and ERK enhance Bmi-1 expression, which leads

to cell cycle progression and increased cell proliferation (Suh and Han, 2011).

Coordinated DDR1 and β1 integrin signaling can induce upregulation of N-cadherin expression,

cell scattering and epithelial to mesenchymal transition in response to collagen I in pancreatic

cancer cells (Shintani et al., 2008). In this experimental system, DDR1 activates Pyk2 while

integrin β1 activates FAK. The signaling scaffold protein p130Cas, which binds to FAK and

Pyk2, also binds DDR1 in pancreatic cancer cells. The Pyk/FAK-p130Cas complex activates

JNK1, which then upregulates N-cadherin through c-Jun, leading to cell scattering (Shintani et

al., 2008). Based on these examples, it is conceivable that DDR1 and β1 integrins may play

cooperating or antagonizing roles in their interactions with type I collagen and the cellular

processes that they control. Currently it is not known how these two different receptor systems

interact to control cell adhesion to collagen and their potential involvement in the generation of

fibrotic lesions.

26

Statement of the Problem

Collagen remodeling by fibroblasts is crucial for physiological matrix turnover and is disrupted

in inflammation, fibrosis and malignancy. In physiological conditions, the balance between

secretion and degradation of the collagen matrix is maintained. However, disruptions of collagen

phagocytosis and the resultant imbalances of matrix homeostasis have important clinical

consequences in various tissues, specifically the gingiva in which the rate of collagen turnover is

one of the fastest of any connective tissue. The rate limiting step of phagocytosis is the binding

of fibrillar collagen to specific receptors, which include β1 integrins and the discoidin domain

receptor 1 (DDR1). The receptor tyrosine kinase, DDR1 is endogenously expressed in fibroblasts

along with the β1 integrin. Previous data suggest that these two receptors interact, but, the

functional nature of these interactions has not been well defined. One of the central steps in

phagocytosis, which determines tissue homeostasis is collagen binding to cells. My preliminary

data indicates that DDR1 over-expression in fibroblasts enhances collagen binding through β1

integrins. Currently, the mechanism by which this interaction is enhanced is not well defined.

Hypothesis

The Discoidin Domain Receptor 1 enhances β1 integrin binding to type I collagen.

Objectives

1. Determine whether DDR1 expression affects collagen remodeling.

2. Examine whether DDR1 regulates cell adhesions.

3. Determine the role of DDR1 in mediating β1 integrin cell surface expression and activation.

4. Assess a potential role for DDR1 in regulating β1 integrin glycosylation.

5. Examine whether DDR1 plays a role in drug-induced fibrosis of gingival connective tissues.

27

Chapter 2

Introduction

Homeostasis of connective tissue in many organ systems is maintained through balanced

synthesis and degradation of matrix proteins but is disrupted in inflammatory and fibrotic

diseases. A critical, defining process that contributes to connective tissue homeostasis is collagen

degradation, which in physiological remodeling is mediated by collagen phagocytosis (Everts et

al., 1996). Collagen phagocytosis by fibroblasts is a receptor-driven process in which cellular

recognition and binding to localized domains on collagen fibrils are crucial regulatory events in

the phagocytic pathway (Chong et al., 2007; Knowles et al., 1991). Collagen recognition and

attachment systems in fibroblasts include cell surface receptors with high affinity for collagen

such as integrins (Knowles et al., 1991), specifically the α2β1 integrin, which is an important

adhesion receptor for type I fibrillar collagen (Chong et al., 2007; Dickeson et al., 1999). The

α2β1 integrin is also an important regulatory molecule for the binding step of collagen

phagocytosis (Arora et al., 2000; Lee et al., 1996)

The functional activity of β1 integrin receptors is affected by a broad range of regulatory

molecules and processes including the concentration of divalent cations such as Ca2+

and Mg2+

(Schnapp, 2006), collagen structure and folding, and the clustering, allosteric modifications,

post-translational modifications, organization and arrangement of integrins at cell membranes

(Alberts, 2002). N-linked glycosylation in particular, is a poorly defined, post-translational

regulatory mechanism for control of β1 integrin function (Bellis, 2004). Variations of β1 integrin

glycosylation may influence receptor conformation (Bellis, 2004) surface expression (Akiyama

et al., 1989; Hotchin and Watt, 1992), and receptor-mediated functional activity including cell

28

adhesion and spreading on collagen (von Lampe et al., 1993).(Diskin et al., 2009) Since β1

integrin binding is affected by differences of glycosylation, it is likely that the downstream

signaling processes that regulate cell adhesion are also affected, which includes the recruitment

of actin binding proteins such as talin, paxillin and vinculin to focal adhesion complexes

(Critchley, 2000; Keselowsky et al., 2004). While variations of normal glycosylation of the β1

integrin have been identified in tumor cells (Bellis, 2004), the role of integrin glycosylation in

the regulation of collagen adhesion and phagocytic function has not been described.

In addition to fibrillar collagen-binding integrins, discoidin domain receptors (DDRs) are a

separate family of collagen receptors that exhibit tyrosine kinase activity after ligand binding

(Leitinger, 2011) but differ from other receptor tyrosine kinases in that collagens are the primary

ligand for DDR activation (Vogel et al., 1997). DDR1 in particular is activated by many types of

collagens and appears to act as a sensor that triggers ECM degradation and turnover (Franco et

al., 2002; Leitinger, 2011). The biological importance of DDR1 in physiological matrix turnover

is supported by experiments using genetic disruption that manifest as a variety of fibrotic

conditions of kidney (Flamant et al., 2006; Gross et al., 2010), liver (Song et al., 2011), and lung

(Avivi-Green et al., 2006). Consequently, normal expression of DDR1 may be important for

connective tissue homeostasis.

DDR1 is tyrosine phosphorylated and activated by cell binding to collagen even in the presence

of β1 integrin blocking antibodies, indicating that DDR1 can participate in signaling responses

independent of β1 integrins (Vogel et al., 2000). Curiously, downstream signaling pathways

activated by DDR1 can also intersect with β1 integrin-activated pathways (Valiathan et al.,

2012). For example, activation of DDR1 inhibits integrinFAKCdc42-mediated cell

spreading (Yeh et al., 2009) and integrinSTAT1/3-mediated cell migration (Wang et al., 2006).

29

After stimulation with type I collagen, β1 integrin activates Gli-1 whereas DDR1 activation

stimulates ERK; combined activation of these two proteins enhances Bmi-1, which drives cell

proliferation (Suh and Han, 2011). In pancreatic cancer cells, coordinated DDR1 and β1 integrin

signals can induce N-cadherin trafficking to the cell membrane, cell scattering and epithelial to

mesenchymal transition in response to type I collagen (Shintani et al., 2008). Collectively these

data indicate that β1 integrin and DDR1 may interact to regulate adhesion to collagen. However,

the nature, locus and post-translational modifications that are involved in these potential

regulatory processes are not defined and the functional interactions between DDRs and integrins

in control of collagen phagocytosis are not understood. We have examined here the association

and functional relationships between DDR1 and β1 integrin and the potential role of DDR1 in

modulating collagen adhesions through changes in post-translational modifications of β1

integrin.

30

Materials & Methods

Reagents

Rabbit monoclonal anti-DDR1 (D1G6) XP® antibody was obtained from Cell Signaling

Technology (Danvers, MA, USA). Rabbit polyclonal anti-DDR1 (C-20, sc-532) was obtained

from Santa Cruz Biotechnology (Santa Cruz, CA, USA). Rabbit polyclonal anti-integrin β1,

(cytosolic), mouse monoclonal anti-glyceraldehyde-3-phosphate dehydrogenase (GAPDH; 6C5),

and mouse monoclonal anti-paxillin (5H11) antibody were obtained from Millipore (Billerica,

MA, USA). Mouse monoclonal anti-vinculin (hVIN-1), mouse monoclonal anti-talin (8d4),

jararhagin snake venom from Bothrops jararaca, cyclosporin A from Tolypocladium inflatum,

dimethyl sulphoxide Hybri-Max™, and protease inhibitor cocktail were obtained from Sigma-

Aldrich (St. Louis, MO, USA). Rat monoclonal anti-mouse CD29 (9EG7) antibody, hamster

monoclonal anti-mouse CD29 (HM β1-1), purified rat monoclonal anti-mouse CD29 (KMI6),

and type I rat tail collagen were purchased from BD Biosciences (Mississauga, ON, Canada).

Rabbit monoclonal anti-vimentin (EPR3776) was purchased from Epitomics (Burlingame, CA,

USA). FITC-conjugated streptavidin and Cy3-conjugated streptavidin were purchased from

Cedarlane Labs (Burlington, ON, Canada). Mouse monoclonal (2A 8F4) secondary antibody to

rat IgG2a-heavy chain (biotin) was purchased from Abcam (Cambridge MA, USA). FITC mouse

monoclonal anti-rat IgG2a (MRG2a-83), Alexa Fluor® 647Armenian hamster IgG anti-mouse

CD49b (HMα2) were purchased from BioLegend (San Diego, CA, USA). Goat anti-mouse and

goat anti-rabbit IgG (H + L)-HRP conjugates were purchased from Bio-Rad (Hercules, CA,

USA). Fluoresbrite® Microparticles (1 μm crimson and FITC beads) were purchased from

Polysciences (Warrington, PA, USA). Endoglycosidase-H was purchased from Roche

31

(Mannheim, Germany). Versene solution was purchased from Invitrogen (Grand Island, NY,

USA).

Cell Culture

Mouse NIH-3T3 and NIH-3T3-stably transfected with DDR1 (b-isoform) were provided by

Wolfgang Vogel (University of Toronto, Toronto, Canada). Integrin β1-deficient GD25 cells

were provided by Dr. Reinhard Fässler (Max-Planck Institute for Biochemistry, Munich,

Germany). Cells were cultured at 37°C in complete DME medium containing 10% fetal bovine

serum and antibiotics (Penicillin G 124 units/mL, Gentamicin SO4, 50 μg/mL, Fungizone 0.25

μg/mL). Cells were maintained in a humidified incubator under 95% air and 5% CO2 conditions,

and were passaged with 0.05% trypsin with 0.53 mM EDTA (Gibco, Burlington, ON).

Immunoblotting

Whole cell extracts were prepared on ice by rinsing with cold Ca2- and Mg

2+-free PBS before

lysis with 1% TNT buffer (20 mM Tris pH 7.5, 150 mM NaCl, 1% Triton X-100, protease

inhibitor cocktail, 200 mM NaVO3, 20 mg/mL PMSF). The collected lysates were kept on ice

and sheared with a 27 gauge needle 4-5 times prior to constant rotation in TNT at 4°C for 30

minutes. The homogenate was centrifuged at 14,000 g for 10 minutes at 4°C and the supernatant

was retained for biochemical analysis of protein content via bicinchoninic acid (BCA) analysis.

Equal amounts of protein (10 μg) were separated on 9% sodium dodecyl sulfate polyacrylamide

gel electrophoresis (SDS-PAGE) gels, transferred to membranes and probed with appropriate

antibodies. Immunoblots were quantified by scanning densitometry and ImageJ software.

32

Flow Cytometry

Cells were seeded onto collagen coated (1 mg/mL type I rat tail) or uncoated tissue culture

plastic for indicated lengths of time. To measure β1 integrin activation, cells were

immunostained with the neoepitope antibody, 9EG7 (Lenter et al., 1993), that recognizes

activated β1 integrin. Prior to immunostaining, cells were quickly (~20 sec) harvested in ice-cold

Versene and fixed in ice-cold Versene containing 1% paraformaldehyde, a procedure that

preserves the integrin activation state in attached fibroblasts (Kim et al., 2010a; Kim et al.,

2010b). Bound 9EG7 was stained with FITC-conjugated anti-rat IgG2a antibody and

fluorescence of single cells was analyzed by flow cytometry (Beckman-Coulter). To measure

total β1 integrin surface expression cells which were not fixed or permeabilized were

immunostained with KMI6 antibody and fluorescence was measured using flow cytometry. To

measure α2 integrin surface expression cells which were not fixed or permeabilized were

immunostained with CD49b antibody and fluorescence was measured using flow cytometry.

Bead Binding

To measure matrix protein binding to cells, carboxylate-modified fluorescent polystyrene beads

(1 μm diameter; excitation/emission: 505/515 nm and excitation/emission: 625/664 nm) were

coated with either type I fibrillar collagen (3.66 mg/mL; polymerization induced by incubation of

beads at pH 7.4) or fibronectin (10 μg/mL), or with BSA (0.66 mg/mL) as a non-specific binding

control. Cells were seeded overnight onto tissue culture plastic. Collagen- or fibronectin-coated

FITC beads were loaded onto the dorsal surface of cells together with BSA-coated crimson beads

(12 of each bead type/cell, unless otherwise indicated) and incubated for 1 hour at 37°C, unless

otherwise specified. After incubation, unbound beads were washed with Ca2- and Mg

2+- free

33

PBS prior to being trypsinized for preparation of single cell suspensions. Cells were pelleted and

resuspended in PBS and mean bead binding was analyzed using flow cytometry (Knowles et al.,

1991).

In an experiment of similar design, cells were seeded overnight onto tissue culture treated 8-well

chamber slides (20,000 cells/well). Collagen-coated FITC beads were loaded onto the dorsal

surface of cells together with BSA-coated crimson beads (60 of each bead type/cell) and

incubated for 1 hour at 37°C in the presence or absence of β1 integrin blocking antibody (CD29;

62.5 μg/mL) or pre-immune serum as a control. As an indicator of the efficacy of the CD29-

blocking antibody activity, cells rounded up but were not completely lifted from the culture dish

prior to termination of the experiment. Following bead binding cells were gently rinsed 1x with

PBS to remove non-adherent and loosely attached beads and blocking antibody, and fixed with

4% paraformaldehyde for 20 minutes at room temperature. After removal of the fixative, nuclei

were stained with DAPI. By fluorescence microscopy, three separate fields of view were

randomly selected and a minimum of 20 DAPI-stained nuclei and associated beads were

counted. The mean ± S.E.M. of the percentage of the cell population with bound beads was

calculated.

Gel Contraction Assays

Type 1 rat tail collagen (1.36 mg/mL) was added to a solution consisting of DMEM, 0.24 M

NaHCO3, FBS, 10X antibiotic, and 0.1 N NaOH. Cells were added to the solution to produce a

final concentration of 150,000 cells/mL. Aliquots of the gel-cell solution (0.2 mL) were pipetted

onto the center of each well of a 24 well non-tissue culture plate, ensuring that no gel contacted

the side of the well. The cell-gel solution was allowed to polymerize for 5 mins at room

34

temperature before adding 1 mL of growth medium to each well. We used a floating gel

contraction assay to measure collagen remodeling by cells (Nakagawa et al., 1989); after

polymerization, gels containing cells were released from the base of the dish before incubation at

37°C. Measurements were made of the collagen gel diameters with a dissecting microscope and

an interocular grid at time 0 hrs and approximately every 10-12 hrs for a total of 72 hours. To

study collagen contraction by cells, an attached gel contraction assay was used (Nakagawa et al.,

1989). After collagen polymerization, gels containing cells were incubated at 37°C for 3 days

before releasing from the base of the dish with a pipette. Measurements of gel diameter were

made at 0 min and every 30 mins after until contraction stopped. Data obtained from three

separate experiments were plotted and linear regression was used to estimate the rate of collagen

contraction. The mean slope for each experiment was determined by best fit of the linear

regression.

Cell Migration Assay

Cells in monolayers were seeded onto type I rat tail collagen-coated plates (0.1 μg/mL) (Chou et

al., 1996). A scratch in the monolayer was created using a 200 μl pipette tip to study migration

into the denuded space (Coomber and Gotlieb, 1990; Zahm et al., 1997). Phase-contrast images

were obtained at regular intervals after scratching until the cell denuded area was closed by cell

migration. Images obtained in the same field of focus were compared to quantify cell migration

rate.

Immunostaining

Cells were plated overnight on type I fibrillar collagen-coated (1 mg/mL collagen) (Chou et al.,

1996) glass coverslips (MatTek dishes; MatTek Corp., Ashland, MA, USA) to enable spreading

35

and formation of focal adhesions. Cells were fixed with 4% paraformaldehyde in PBS for 10

min, permeabilized with 0.2% Triton X-100 in PBS, and blocked for 1 hr in 0.2% BSA in PBS.

Cells were incubated with primary antibody and FITC-conjugated secondary antibody for 1 hr

each. Immunostained cells were visualized by total internal reflection fluorescence (TIRF)

microscopy (Leica, Heidelberg, Germany; 100 x oil immersion lens). Focal adhesion proteins in

contact with the collagen substrate (optical penetration depth <110 nm) were analyzed. The mean

number of focal adhesions in the periphery (< 5 μm from the cell membrane) and in the cell body

(> 5 μm from the cell membrane) were quantified with Leica MetaMorph® Microscopy

Automation & Image Analysis Software (20 cells/condition; Leica Microsystems,. Wetzlar,

Germany).

Integrin Cleavage Experiments

Cells in monolayers were grown to 90% confluence on tissue culture plastic or non-tissue culture

plastic coated with polymerized type I collagen (Chou et al., 1996). Cells were treated with 10

μg/mL of jararhagin in DMEM growth medium for 1 hr at 37°C (Klein et al., 2011). After 1 hr

cells were immediately placed on ice and the jararhagin-containing medium was removed. Cells

were prepared by rinsing with cold Ca2+

- and Mg2+

- free PBS before lysis in 1% TNT buffer (20

mM Tris pH 7.5, 150 mM NaCl, 1% Triton X-100, protease inhibitor cocktail, 200 mM NaVO3,

20 mg/mL PMSF). Protein concentrations were measured using a BCA assay prior to

immunoblot analysis.

Endo-H Deglycosylation Experiments

Cells in monolayers were grown to 90% confluence on uncoated tissue culture plastic or non

tissue culture plastic coated with 200 μg/mL of type I rat tail collagen neutralized to pH 7.5.

36

Cells were rinsed with cold Ca2+

-, Mg2+

- free PBS prior to lysis with extraction buffer (1%

Triton X-100, 0.125% Tween-20, 0.5% deoxycholate, 50mM HEPES, 0.5M NaCl pH 7.5,

protease inhibitor cocktail, 200 mM NaVO3, 20 mg/mL PMSF). Protein concentrations were

measured using a BCA assay and 10 μg/μL of glycoprotein sample were combined with 5X

reaction buffer (250 mM sodium phosphate buffer), denaturation solution (0.02% SDS, 0.1 M β-

ME), PMSF and protease inhibitors. Samples were heated at 100°C for 3 min and

endoglycosidase-H was added to attain a final enzyme concentration of 0.25mU/μL, followed by

incubation at 37°C for 24h. The reaction was terminated by ice-cold acetone/trichloroacetic acid

precipitation. The pelleted samples were air dried prior to the addition of sample loading buffer

and subsequent immunoblot analysis (Akiyama et al., 1989; Salicioni et al., 2004).

Statistics

For all data, mean and standard errors of means were computed. Where appropriate, comparisons

between two samples were made by Student’s t-test (unpaired) with statistical significance set at

a type I error rate of p<0.05. All experiments were performed in triplicate unless otherwise

stated.

37

Results

Cell characterization

We examined the role of DDR1 in regulating β1 integrin interactions with collagen in cultured

cells. NIH3T3 wild-type and DDR1 (b isoform) over-expressing fibroblasts, and mouse GD25

cells derived from differentiated ES cells that carry a null mutation in both β1 integrin alleles

were examined. Prior to experiments cell lysates were prepared and immunoblotted for total

DDR1 and β1 integrin proteins. DDR1 was readily detected in NIH3T3 wild-type and GD25

cells (Fig. 1A). β1 integrins were expressed in NIH3T3 wild-type and DDR1 over-expressing

cells, but not in GD25 cells (Fig. 1B). Immunoblots of GAPDH protein were used as a loading

control.

Collagen binding to β1 integrins is enhanced by DDR1 over-expression

We first determined whether DDR1 over-expression affects cell adhesion to collagen. Wild-type

3T3 and DDR1 over-expressing cells, or GD25 cells, were plated overnight on tissue-culture

plastic. Collagen-coated FITC fluorescent polystyrene beads (1 μm, 12 beads/ cell) were loaded

onto the dorsal surface of cells and incubated for 1 hour prior to analysis by flow cytometry to

determine the percentage of the population with bound beads. Binding of BSA-coated crimson

beads was measured simultaneously as a non-specific binding control. Collagen-coated bead

binding to DDR1 over-expressing cells was enhanced by 4.23-fold (p < 0.001) and 5.18-fold (p <

0.001) compared with wild-type and GD25 cells, respectively (Fig. 2A). In an experiment of

similar design, we incubated wild-type or over-expressing cells with collagen-coated beads for 2,

4 or 24 hours prior to analysis by flow cytometry. Collagen-coated bead binding to DDR1 over-

38

expressing cells was enhanced by 1.54-fold at 2 hrs (p < 0.01), 1.74-fold at 4 hrs (p < 0.0001)

and 1.39-fold at 24 hrs (p < 0.0001) compared with wild-type cells (Fig. 2B).

We conducted experiments to determine whether the increased collagen binding observed in

cells over-expressing DDR1 was due in part to the additive adhesive effect of high levels of

DDR1 expression. Since the primary ligand of DDR1 is fibrillar collagen (Vogel et al., 1997),

we coated beads with fibronectin (McKeown et al., 1990) to examine whether the increased

collagen binding was due to enhanced β1 integrin adhesion independent of the integrin

subunits. Notably, an important adhesion receptor for fibronectin is the α5β1 integrin (Zhang et

al., 1993). Fibronectin-coated beads were loaded onto to the dorsal surface of cells and incubated

for 1 hour prior to analysis of bead binding. Similar to the data for collagen bead binding,

fibronectin-coated bead binding to DDR1 over-expressing cells was enhanced by 1.56-fold

compared with wild-type cells (p < 0.01) (Fig. 2C).

To determine the relative contribution of binding that was attributable to DDR1 over-expression

versus enhancement of β1 integrin activity, cells were plated overnight in 8-well chamber slides.

Binding of collagen-coated FITC beads (1 μm diameter; 60 beads/cell) in the presence of pre-

immune serum or CD29-blocking antibody (62.5 μg/mL) was analyzed after 1 hour incubation.

The mean percentage of cells with bound beads was determined by counting beads and DAPI-

stained nuclei. Compared with cells treated with pre-immune serum (controls), collagen coated-

bead binding to cells was reduced by 2.42-fold in wild-type cells (p < 0.001) and by 1.32-fold in

DDR1 over-expressing cells (p < 0.0001) treated with CD29-blocking antibody. BSA-coated

crimson beads (1 μm diameter; 60 beads/cell) were measured simultaneously as non-specific

binding controls (Fig. 2D).

39

Since α2β1 is the major integrin receptor for type I fibrillar collagen (Arora et al., 2000; Lee et

al., 1996), we wanted to determine whether DDR1 enhances cell surface expression of the α2

integrin subunit. Total surface α2 integrin expression of cells plated overnight on tissue culture

plastic was measured by immunostaining with CD49b antibody and flow cytometry. Analysis of

cell staining indicated that expression of total cell surface α2 integrins was enhanced by 1.91-

fold (p < 0.0001) in DDR1 over-expressing cells compared with wild-type cells (Fig. 2E).

DDR1 over-expression enhances collagen remodeling and contraction

We determined the effect of DDR1 over-expression on remodeling and contraction of collagen

gels in two related assays. First, collagen remodeling was measured using a floating gel

contraction assay (Nakagawa et al., 1989), which examines the ability of cells in compliant gels

to extrude fluid from the gels and reorganize the collagen fibrils through migratory and

reorganizational activities. For this assay cells were incubated in collagen gels during

polymerization and the gels were then floated in growth media. Gel contraction was measured

over 3 days. DDR1 over-expression enhanced floating collagen gel contraction by 3.99-fold (p <

0.01) compared with wild-type cells and increased gel contraction by 15.08-fold (p < 0.01)

compared with GD25 cells (Fig 3A).

Collagen gel contraction that is attributable to cell-mediated contractile behavior was measured

using an anchored gel contraction assay (Chung et al., 2004). Cells were incubated in collagen

gels that were then attached to the bottom of a 12-well plate. After 3 days of attachment cells

build up considerable tension in the gels. After release of the gels from the base of the dish, gel

contraction was measured over several hours. We determined that DDR1 over-expression

40

enhanced anchored collagen gel contraction by 3.55-fold (p < 0.01) compared with wild-type

cells and by 3.97-fold (p < 0.01) compared with GD25 cells (Fig. 3B).

DDR1 over-expression and cell migration on collagen

We examined whether DDR1 over-expression impacts cell migration using an in vitro scratch

assay. Wild-type or DDR1 over-expressing cells were grown to confluence on a collagen-coated

substrate prior to scratching and differential interference contrast images of cells were obtained.

Representative images obtained at 0 hours or 12 hours after scratching showed cell migration

into the scratch and a reduction of the width of the denuded cell monolayer (Fig. 4A). Cells over-

expressing DDR1 showed no significant difference in migration rate compared with wild type

cells (Fig. 4B, C; p = 0.21).

DDR1 over-expression enhances activated β1 integrin staining in focal adhesions

DDR1 does not localize to focal adhesion contacts (Vogel et al., 2000), however we sought to

determine whether over-expression of DDR1 influences the activation and recruitment of β1

integrins to focal contacts. Cells were plated overnight on collagen and stained with 9EG7, a

neo-epitope antibody that binds to activated β1 integrins (Lenter et al., 1993) prior to analysis by

TIRF microscopy. Representative images of 9EG7 staining indicate that DDR1 over-expressing

cells contained more activated β1 integrins compared with wild-type cells (Fig. 5A). Analysis of

the periphery of cells stained with 9EG7 by TIRF microscopy and subsequent computation of

cell areas indicated that there was no significant difference in cell area between wild-type and

DDR1 over-expressing cells (Fig. 5B; p = 0.48). Further analysis of focal adhesion staining

showed that the number of activated β1 integrin-containing focal adhesions was enhanced by

1.75-fold (p < 0.001) in DDR1 over-expressing cells compared with wild-type cells. Further, the

41

length of 9EG7-stained focal adhesions was increased by 1.33-fold (p < 0.0001) in DDR1 over-

expressing cells compared with wild-type cells (Fig. 5C). Focal adhesion area stained by 9EG7

was not significantly different between the two cell types (Fig. 5D; p = 0.064).

DDR1 over-expression enhances talin staining in focal adhesions

Talin is an adaptor protein recruited to focal contacts following the binding of β1 integrins to

collagen (Craig and Johnson, 1996) and other matrix ligands. We measured talin expression in

the focal adhesions of wild-type and DDR1 over-expressing cells plated overnight on collagen

by immunostaining and TIRF microscopy. Consistent with the results obtained from activated β1

integrin staining, representative images indicate that DDR1 over-expressing cells exhibited more

talin-stained focal adhesions compared with wild-type cells (Fig. 6A). Analysis of talin staining

in the cell periphery indicated that the cell area of DDR1 over-expressing cells was slightly

(0.82-fold) smaller (p < 0.05) compared with wild-type cells (Fig. 6B). Further, DDR1 over-

expressing cells exhibited 2.54-fold (Fig. 6C; p < 0.0001) more focal adhesions, which were

1.27-fold (Fig. 6D; p < 0.0001) longer, and had an area which was 1.35-fold greater (Fig. 6E; p <

0.0001) compared with wild-type cells.

DDR1 over-expression enhances paxillin staining in focal adhesions

Paxillin is recruited to focal adhesions after talin (Jiang et al., 2003). Accordingly, we measured

paxillin expression in focal adhesions in wild-type and DDR1 over-expressing cells.

Immunostaining and analysis of images obtained by TIRF microscopy indicated that DDR1

over-expressing cells demonstrated more paxillin-containing focal adhesions than wild-type cells

(Fig. 7A). Analysis of paxillin staining in the cell periphery indicated that the cell area measured

by paxillin staining was 1.19-fold (p < 0.05) greater in DDR1 over-expressing cells compared

42

with wild-type cells (Fig. 7B). DDR1 over-expressing cells exhibited 3.12-fold (Fig. 7C; p <

0.0001) more focal adhesions, which were 1.28-fold (p < 0.0001; Fig. 7D) longer, and were 1.30-

fold more numerous (Fig. 7E p < 0.0001) than wild-type cells.

DDR1 over-expression enhances vinculin staining in focal adhesions

Vinculin is an adaptor protein that links β1 integrins to the actin cytoskeleton and is recruited to

focal adhesions after the recruitment of talin and paxillin (Craig and Johnson, 1996; Critchley,

2000). We measured vinculin immunostaining in cells plated overnight on collagen by

immunofluorescence and TIRF microscopy. Representative images indicate that DDR1 over-

expressing cells exhibited more vinculin containing focal adhesions than wild-type cells (Fig.

8A). Analysis of vinculin staining in the cell periphery indicated that there was no significant

difference in total cell area between the wild-type and DDR1 over-expressing cells (Fig. 8B;

p=0.78). Consistent with our data obtained from activated β1 integrin, talin and paxillin staining,

the number of vinculin containing focal adhesions was 2.64-fold greater in DDR1 over-

expressing cells compared with wild-type cells (Fig. 8C; p < 0.0001). Further, vinculin-stained

focal adhesions in DDR1 over-expressing cells were 1.50-fold longer (Fig. 8D; p < 0.0001) and

with a 1.66-fold larger area (Fig. 8E; p < 0.0001) compared with wild-type cells.

DDR1 over-expression enhances total β1 integrin surface expression and

activation

We examined the effect of DDR1 over-expression on β1 integrin expression on the cell surface.

Total surface β1 integrin expression of cells plated overnight on tissue culture plastic was

measured by immunostaining with KMI6 antibody and flow cytometry. Analysis of cell staining

indicated that expression of total cell surface β1 integrins was enhanced by 3.20-fold (p < 0.01)

43

in DDR1 over-expressing cells compared with wild-type cells (Fig. 9A). Previously we measured

activated β1 integrin expression contained in focal adhesions, which are discrete sites associated

with tight attachment to the collagen substrate (Critchley, 2000). By flow cytometry and

immunostaining, we measured total activated β1 integrin expression on the entire cell surface. In

cells plated overnight on type I collagen, analysis of total 9EG7 staining indicated that DDR1

over-expression increased total activated β1 integrins on the cell surface by 1.55-fold (p < 0.05)

compared with wild-type cells (Fig. 9B). Analysis of 9EG7 staining of cells plated overnight on

tissue culture plastic indicated that total activated β1 integrin surface expression was 1.32-fold (p

< 0.001) greater in DDR1 over-expressing cells compared with wild-type cells (Fig. 9C). We

deprived cells of attachment to matrix ligands by preparation of cell suspensions and intentional

maintenance in suspension for 2 hours. Analysis of 9EG7 staining of cells in cells subjected to

prolonged detachment from matrix proteins indicated that total activated β1 integrin surface

expression was 1.62-fold (p < 0.01) greater in DDR1 over-expressing cells compared with wild-

type cells (Fig. 9D).

DDR1 over-expression enhances enzymatic cleavage of surface β1 integrins

Cells plated on tissue culture plastic or collagen were treated with the disintegrin jararhagin to

induce cleavage of β1integrins at the cell surface (Klein et al., 2011). After treatment, whole cell

lysates of wild-type and DDR1 over-expressing cells were immunoblotted for total β1 integrin

protein expression (Fig. 10). Immunoblots of GAPDH protein were used to estimate equality of

lane loading. Independent of the substrate, DDR1 over-expression increased cleavage of the β1

integrin subunit (indicated upper and lower molecular weight bands) after jararhagin.

44

DDR1 over-expression alters post translational modifications of β1 integrins

In all immunoblots examined we found a consistent difference of the apparent molecular mass of

β1 integrins in wild-type cells compared with DDR1-over-expressing cells. In wild-type cells β1

integrins appeared as a single, 130 kDa band; occasionally we detected a faint, ~125 kDa band

(Fig. 1B). In contrast, β1 integrins of DDR1 over-expressing cells consistently exhibited two

discrete bands: a faint, upper band migrating at ~130 kDa, and a denser lower band migrating at

~125 kDa. We considered that this variation was due to differences in post-translational

modifications of the β1 integrin subunit (Bellis, 2004) that might include variations of

glycosylation. Accordingly, whole cell lysates of wild-type and DDR1 over-expressing cells

plated on collagen or tissue culture plastic were treated with endoglycosidase-H and

immunoblotted for the β1 integrin (Fig. 11). Immunoblots of GAPDH protein were used as a

loading control. Independent of the substrate on which cells were plated, cells that over-

expressed DDR1 exhibited increased glycosylation of the β1 integrin subunit, which is suggested

by a shift of the lower molecular weight band (~125 kDa) to a band of ~110 kDa after treatment

with endoglycosidase-H.

45

Discussion

The principal finding of this study is that DDR1 enhances β1 integrin interactions with fibrillar

collagen. Over-expression of DDR1 increased cell surface β1 integrin expression and activation,

and enhanced collagen binding to β1 integrins, collagen remodeling and reorganization, and the

recruitment of integrins and the adaptor proteins, talin, paxillin and vinculin to focal adhesion

complexes. Collectively, these findings offer novel insights into the role of DDR1 in regulating

β1 integrin function, possibly involving post-translational modifications and trafficking of the β1

integrin to the cell surface.

Collagen binding to β1 integrins is enhanced by DDR1 over-expression

Collagen binding to receptors, in particular the α2β1 integrin, is the rate-limiting step in the

phagocytic pathway of collagen degradation by fibroblasts (Arora et al., 2000). We measured

fibrillar collagen binding using a collagen-coated bead binding assay (Knowles et al., 1991) and

found that over a 24 hour time course, DDR1 over-expression enhanced collagen binding. While

binding was increased for all time points measured, there was a relatively larger increase of

collagen bead binding at early times after bead incubation in cells with DDR1 over-expression,

indicating that DDR1 may, by itself, contribute to increased binding of collagen while also

contributing to enhanced binding because of its impact on the β1 integrin. Notably, we found in

the time courses that once this early increase of binding occurred, subsequent increases

attributable to DDR1 over-expression were diminished. We conjecture that these subsequent

increases of adhesion may be attributed to enhanced β1 integrin function. Indeed, since the

primary ligand of DDR1 is fibrillar collagen (Leitinger and Hohenester, 2007) and as an

important adhesion receptor for fibronectin is the α5β1 integrin (Hynes, 2002; Zhang et al.,

46

1993), we tested this conjecture using fibronectin-coated bead binding. We found that fibronectin

bead binding in DDR1 over-expressing cells was enhanced by similar proportions to those

observed at later time points in the time course experiment using collagen-coated beads. Further,

when we inhibited β1 integrin function with blocking antibody, the reductions in collagen

binding observed in DDR1 over-expressing cells closely matched the levels by which binding

was initially increased. Together these observations indicate that DDR1 over-expression

enhances collagen binding to β1 integrins independent of DDR1-mediated collagen adhesion.

DDR1 over-expression enhances collagen remodeling and contraction

Since we found that DDR1 over-expression enhanced collagen binding, we explored its role in

β1 integrin-mediated remodeling and reorganization of collagen, processes which are necessary

for the maintenance of connective tissue homeostasis (Leitinger, 2011). Data from floating

collagen gel assays, which measure remodeling of collagen fibrils through migratory and

reorganizational activities of cells (Grinnell, 2003), indicated that collagen gel contraction was

enhanced in cells over-expressing DDR1. These findings indicate that DDR1 enhances β1

integrin functional interactions with fibrillar collagen, which is in contrast to earlier reports

showing that DDR1 inhibits β1 integrin function(Yeh et al., 2009). The difference between the

current data and the findings of Yeh and co-workers could be a reflection of the relatively higher

levels of DDR1 expression that were obtained in our cells.

DDR1 enhances activated β1 integrin and adaptor protein staining in focal

adhesions

DDR1 does not localize to focal adhesion contacts (Vogel et al., 2000). However, since collagen

binding was enhanced by DDR1 over-expression, we sought to determine whether the activation

47

and recruitment of β1 integrins and the actin binding proteins talin, paxillin and vinculin to focal

contacts, were also influenced. Temporal analyses of the recruitment of these proteins to cell

adhesions show that after β1 integrin activation and ligand binding (Liu et al., 2000), talin (Craig

and Johnson, 1996), paxillin (Jiang et al., 2003) and vinculin (Critchley, 2000) are recruited

sequentially to focal complexes. These complexes link collagen-bound β1 integrins to the actin

cytoskeleton and enable downstream signaling (Critchley, 2000; Liu et al., 2000). TIRF imaging

indicated that the number, length and area of immunostained focal adhesion proteins were

generally larger in cells over-expressing DDR1 than wild type cells. Further, the observed

differences in the size and number of immunostained focal adhesions were increased sequentially

with each subsequent protein that was recruited during the adhesion maturation process,

consistent with our data showing that DDR1 promotes binding and interactions with collagen.

DDR1 over-expression and cell migration on collagen

Cell migration is dependent upon cell-generated forces that facilitate the formation of stable

attachments to the collagen matrix and that enable locomotion (Lauffenburger and Horwitz,

1996). However, if cell attachments are highly adherent, migration may be inhibited because the

trailing edge of cells cannot detach from the substratum to enable forward translocation (Yuen et

al., 2010). Since DDR1 over-expression enhanced collagen binding to β1 integrins, we examined

whether DDR1 over-expression impacts cell migration using an in vitro scratch assay. Compared

with wild type cells, cells over-expressing DDR1 showed no significant difference in migration

rate. Notably, we found that when the cell area was measured using peripheral immunostaining

for activated 1 integrin, vinculin, talin or paxillin, there were no marked differences attributable

to DDR1 over-expression compared with wild type cells. As unidirectional migration is

dependent on advancement of the cell periphery at the leading edge, the lack of effect of DDR1

48

over-expression on increasing expression of cell attachment proteins in peripherally-located

attachments indicate that DDR1 does not affect all cell attachments equally across the ventral

surface of the cell. In contrast, we found that cell adhesions were enlarged in the central part of

cells over-expressing DDR1. We speculate that as DDR1 over-expression selectively enhances

cell adhesion to collagen in the central part of cells, while not significantly affecting adhesion in

the cell periphery, cell migration is not detectably affected by DDR1 over-expression.

DDR1 over-expression enhances total β1 integrin surface expression and

activation

We found that DDR1 over-expression enhanced recruitment of activated β1 integrins and actin

binding proteins in focal adhesions, which are discrete sites associated with tight cellular

attachment to collagen substrates (Critchley, 2000). We followed up these analyses by examining

β1 integrin expression on the whole cell surface. Analysis of suspended cells by flow cytometry

indicated that expression of total and activated β1 integrins on the cell surface were increased in

equal proportion in DDR1 over-expressing cells compared with wild-type cells. Curiously, β1

integrin activation did not appear to be affected by cell detachment from collagen, since cells that

were maintained in prolonged suspension expressed comparable levels of cell surface activated

β1 integrins as did cells that were quickly detached from collagen and immediately analyzed.

These data indicate that DDR1 may promote recruitment of constitutively activated β1 integrin

to the cell surface, independent of ligand binding. These findings are consistent with our

observations that DDR1 over-expression enhances the adhesion of fibrillar collagen to β1

integrins. Notably, high levels of β1 integrin(Andrews et al., 2001; Klein et al., 1991) and DDR1

(Heinzelmann-Schwarz et al., 2004; Quan et al., 2011; Yang et al., 2010) expression have been

implicated in the development of tumor cells and DDR1 expression levels are associated with

49

increased MMP-1, -2 -9 and -10 expression (Ferri et al., 2004; Park et al., 2007; Ruiz and Jarai,

2011), proteins which are expressed by cells that actively remodel the ECM. Accordingly we

speculate that DDR1 may play a similar role in modulating β1 integrin expression and function

to facilitate cell adhesion-dependent remodeling of the ECM.

DDR1 regulation of trafficking and post-translational modification of β1 integrin

We determined whether DDR1 affects β1 integrin trafficking to the cell surface by exploiting the

susceptibility of cell surface β1 integrin to cleavage by the disintegrin jararhagin (Klein et al.,

2011). Independent of whether cells were plated on collagen or tissue culture plastic, DDR1

over-expression increased cleavage of the β1 integrin subunit after treatment with jararhagin.

This finding confirms our observation that DDR1 over-expression enhances cell surface β1

integrin expression. Further, in all immunoblots examined, we found consistent differences of

the apparent molecular mass of β1 integrins in wild-type cells compared with DDR1-over-

expressing cells. We considered that this variation may be due to differences in post-translational

modifications of the β1 integrin subunit (Bellis, 2004; Diskin et al., 2009; Gu and Taniguchi,

2004), which may include variations of glycosylation. Independent of the substrate on which

cells were plated, cells that over-expressed DDR1 exhibited increased N-linked glycosylation of

the β1 integrin subunit. While glycosylation is a largely understudied regulatory mechanism for

β1 integrin function (Bellis, 2004), our findings may explain earlier data. For example, variations

of β1 integrin glycosylation have been implicated in regulation of integrin trafficking to the cell

surface (Hotchin and Watt, 1992) as well as integrin-mediated cell functions including migration

(Janik et al., 2010), adhesion (von Lampe et al., 1993; Zheng et al., 1994), and spreading (Diskin

et al., 2009).

50

In summary, our findings indicate that DDR1 over-expression enhances fibrillar collagen binding

to β1 integrins. The effect of DDR1 over-expression on β1 integrin cell surface expression and

activation may be attributed to DDR1-regulated post-translational modifications of the β1

integrin. These modifications in turn may translate into enhanced adhesion formation, collagen

remodeling and reorganization, processes that are essential for connective tissue homeostasis.

51

Figure Legends

Figure 1. Cell characterization. A) Whole cell lysates of mouse NIH3T3 wild-type or

DDR1 (b isoform) over-expressing fibroblasts, or β1 integrin-null mouse GD25 cells were

prepared and immunoblotted for total DDR1 expression. B) Immunoblotting for total β1 integrin

protein expression in the same mouse cells described in Fig. 1A.

Figure 2. Collagen binding to β1 integrins is enhanced by DDR1 over-

expression. A) Binding of collagen-coated FITC beads (1 μm diameter; 12 beads/cell) to cells

was analyzed by flow cytometry after 1 hour incubation. Collagen coated-bead binding to DDR1

over-expressing cells was enhanced at 1 hour compared with either wild-type (p < 0.001) or

GD25 cells (p < 0.001). Cell binding to BSA-coated crimson beads (1 μm; 12 beads/cell) was

measured simultaneously as a non-specific binding control. Data represent mean+S.E.M. (n=3

independent experiments). B) Flow cytometric analysis to determine collagen-coated bead

binding (1 μm diameter beads; 12 beads/cell) to DDR1 over-expressing cells. Binding was

enhanced at 2 hrs (p < 0.01), 4 hrs (p < 0.001) and 24 hrs (p < 0.001) after incubation with beads

compared with wild-type cells. Data represent mean + S.E.M. (n = 3 independent experiments).

C) Binding of fibronectin-coated beads (1 μm diameter; 12 beads/cell) to DDR1 over-expressing

cells was enhanced compared with wild-type cells as indicated by analysis with flow cytometry

(p < 0.01). Data represent mean + S.E.M. (n = 3 independent experiments). D) Cells were plated

overnight in 8-well chamber slides. Binding of collagen-coated FITC beads (1 μm; 60 beads/cell)

in the presence of pre-immune serum or CD29 blocking antibody (62.5 μg/mL) was analyzed

after 1 hour incubation. Mean percentage of cell population with bound beads was determined by

52

counting beads and DAPI-stained nuclei. Collagen coated-bead binding to cells was reduced by

CD29 blocking antibody in wild-type (p < 0.001) and DDR1 over-expressing (p < 0.0001) cells.

BSA-coated crimson beads (1 μm; 60 beads/cell) were measured simultaneously as non-specific

binding controls. Three separate fields of view were randomly selected and a minimum of 20

DAPI-stained nuclei and associated beads were counted. Data represent mean + S.E.M. (n = 4

independent experiments). E) Cells were plated overnight on tissue culture plastic. Total α2

integrin surface expression was measured by immunostaining of non-fixed and non-

permeabilized cells with CD49b antibody, followed by flow cytometry. Analysis of cell staining

indicated that DDR1 over-expression increased total α2 integrin expression on the cell surface

compared with wild-type cells (p < 0.0001). Data represent mean + S.E.M. of fluorescence

channel number of cells stained with fluorescence-conjugated antibodies to (n = 3 independent

experiments).

Figure 3. DDR1 over-expression enhances collagen remodeling and contraction.

A) Collagen remodeling was measured over 3 days using a floating gel contraction assay. DDR1

over-expression enhanced floating collagen gel contraction compared with wild-type cells (p <

0.01) or GD25 cells (p < 0.01). Data represent mean + S.E.M. (n = 3 independent experiments).

B) Collagen contraction was measured using an anchored gel contraction assay. DDR1 over-

expression significantly enhanced anchored collagen gel contraction compared with either wild-

type (p < 0.01) or GD25 cells (p < 0.01). Data represent mean + S.E.M. (n = 3 independent

experiments).

Figure 4. DDR1 over-expression does not affect cell migration on collagen. A)

Representative, differential interference contrast images of wild-type or DDR1 over-expressing

53

cells plated on collagen. Cells were examined in in vitro scratch assays at 0 hours and 12 hours

after scratching. B) Best linear fit of reduction of scratch width (mm) over time. The reduction in

scratch width was not significantly different between the wild-type and DDR1 over-expressing

cells (p = 0.2098). C) Comparison of migration rates between wild-type and DDR1 over-

expressing cells. There was no significant difference in migration of DDR1 over-expressing cells

on collagen compared with wild-type cells (p = 0.2098). All data represent mean + S.E.M. (n = 3

independent experiments).

Figure 5. DDR1 over-expression enhances activated β1 integrin staining in focal

adhesions. A) Cells plated overnight on collagen were immunostained for activated β1

integrins using 9EG7 antibody and imaged by TIRF microscopy. Representative images indicate

that DDR1 over-expressing cells expressed more activated β1 integrins compared with wild-type

cells. Total cell area estimated from peripheral 9EG7 staining and 9EG7 staining in focal

adhesions was analyzed in DDR1 over-expressing cells and wild-type cells. Data are indicated in

following histograms. B) No significant difference in total cell area (p = 0.4807); C) increased

number of activated β1 integrin-containing focal adhesions (p = 0.001); D) greater focal

adhesion length (p < 0.0001); and E) no significant difference in focal adhesion area (p =

0.064). MetaMorph was used to quantify mean + S.E.M. of cell area and focal adhesions in cells

(n = 20 cells analyzed for each cell type).

Figure 6. DDR1 over-expression enhances talin staining in focal adhesions. A)

Cells plated overnight on collagen were immunostained for talin and imaged by TIRF

microscopy. Representative images indicate that DDR1 over-expressing cells contained more

54

talin-containing focal adhesions than wild-type cells. Total cell area estimated from peripheral

talin staining and talin staining in focal adhesions was analyzed in DDR1 over-expressing cells

and wild-type cells. Data indicate that in DDR1 over-expressing cells compared to wild type

cells: B) total cell area was significantly greater (p < 0.05); C) there were increased numbers of

talin-containing focal adhesions (p < 0.0001); D) focal adhesions were longer (p < 0.0001); and

E) talin-stained focal adhesions exhibited increased area (p < 0.0001). MetaMorph was used to

quantify mean + S.E.M. of cell area and focal adhesions in cells (n = 20 cells analyzed for each

cell type).

Figure 7. DDR1 over-expression enhances paxillin staining in focal adhesions.

A) Cells plated overnight on collagen were immunostained for paxillin and imaged by TIRF

microscopy. Representative images indicate that DDR1 over-expressing cells contained more

paxillin-stained focal adhesions than wild-type cells. Analysis of paxillin staining in the cell

periphery for cell area and in focal adhesions indicated that for DDR1 over-expressing cells: B)

total cell area was significantly greater (p < 0.05); C) there were increased numbers of paxillin-

containing focal adhesions (p < 0.0001); D) focal adhesion lengths were longer (p < 0.0001);

and E) paxillin-stained focal adhesion area was larger (p < 0.0001). MetaMorph was used to

quantify mean + S.E.M. of cell area and focal adhesions in cells (n = 20 cells analyzed for each

cell type).

Figure 8. DDR1 over-expression enhances vinculin staining in focal adhesions.

A) Cells plated overnight on collagen were immunostained for vinculin and imaged by TIRF

microscopy. Representative images indicate that DDR1 over-expressing cells contained more

55

vinculin-stained focal adhesions than wild-type cells. Analysis of peripheral vinculin staining

and vinculin staining in focal adhesions in DDR1 over-expressing cells compared with wild type

cells indicated: B) no significant difference in total cell area (p = 0.7783); C) increased number

of focal adhesions (p < 0.0001); D) longer focal adhesions (p<0.0001); and E) increased area of

vinculin-stained focal adhesions (p < 0.0001). MetaMorph was used to quantify mean + S.E.M.

of cell area and focal adhesions (n = 20 cells analyzed for each cell type).

Figure 9. DDR1 over-expression enhances total β1 integrin surface expression

and activation. A) Cells were plated overnight on tissue culture plastic. Total β1 integrin

surface expression was measured by immunostaining non-fixed and non-permeabilized cells with

KMI6 antibody, followed by flow cytometry. Analysis of cell staining indicated that DDR1 over-

expression increased total β1 integrin expression on the cell surface compared with wild-type

cells (p < 0.01). Total activated β1 integrin surface expression on cells was measured by

immunostaining with 9EG7 antibody and flow cytometry of non-fixed and unpermeabilized

cells. B) Analysis of cell staining indicated that in cells plated overnight on type I collagen,

DDR1 over-expression significantly increased total activated β1 integrin surface expression

compared with wild-type cells (p < 0.05). C) Analysis of cells plated overnight on tissue culture

plastic indicated that total activated β1 integrin surface expression was significantly greater in

DDR1 over-expressing cells compared with wild-type cells (p < 0.001). D) Analysis of 9EG7

staining of cells in suspension indicated that total activated β1 integrin surface expression was

significantly greater in DDR1 over-expressing cells compared with wild-type cells (p < 0.01).

All data represent mean + S.E.M. (n = 3 independent experiments).

56

Figure 10. DDR1 over-expression enhances enzymatic cleavage of β1 integrin at

the cell surface. Following 1 hr treatment with jararhagin to induce cleavage of surface β1

integrins, whole cell lysates of wild-type and DDR1 over-expressing cells plated on collagen or

tissue culture plastic were prepared and immunoblotted for total β1 integrin protein and GAPDH

expression. Independent of the substrate used for cell plating, DDR1 over-expression led to

increased cleavage of the β1 integrin subunit (upper and lower molecular weight bands)

following treatment with jararhagin. This is a representative image selected from 3 independent

experiments.

Figure 11. DDR1 over-expression alters post translational modification of β1

integrins. Whole cell lysates of wild-type and DDR1 over-expressing cells plated on collagen

or tissue culture plastic were treated with endoglycosidase H for 24 hours prior to

immunoblotting for total β1 integrin protein and GAPDH expression. Independent of the

substrate used for cell plating, DDR1 over-expression led to an increase in glycosylation of the

β1 integrin subunit as indicated by the band shift of the lower molecular weight band (denoted by

the arrows: glycosylated, deglycosylated) following treatment with endoglycosidase H.

This is a representative image selected from 3 independent experiments.

Figure 1

A

DDR1

GAPDH

130 kDa

36 kDa

3T3

WT

3T3

DDR1 OE

GD25

B

β1 Integrin

GAPDH

130 kDa

36 kDa

3T3

WT

3T3

DDR1 OE

GD25

57

DDR WT DDR OE GD250

8

4

12

16 *** *** CollagenBSA

Mea

n P

opul

atio

n w

ith

Bou

nd B

eads

(%)

A

Figure 2

B

Mea

n P

opul

atio

n w

ith

Bou

nd B

eads

(%)

2 4 24

DDR WTDDR OE

Time (hrs)

0

60

30

90

*****

*** C

Mea

n P

opul

atio

n w

ith

Bou

nd B

eads

(%)

DDR WT DDR OE

**

0

8

4

12

D ****

***

CollagenBSA

WT Ctrl OE BlockOE CtrlWT Block

Mea

n P

opul

atio

n w

ith

Bou

nd B

eads

(%)

0

60

30

90 E ****

DDR WT DDR OE0

1000

500

1500

2000

Fluo

resc

ence

Cha

nnel

#:

CD

49b

Sta

inin

g

58

DDR WT DDR OE GD250

40

20

60

80

100** **

Mea

n S

lope

of F

loat

ing

Gel

D

iam

eter

(µm

/hr)

A

Figure 3

B

Mea

n S

lope

of A

ttach

ed G

el

Dia

met

er (µ

m/h

r)

DDR WT DDR OE GD250

400

200

600

800

1000 ** **

59

A

Figure 4

B

Scr

atch

Wid

th (m

m)

0

0.2

0.1

0.3

0.4

0.5

C

DDR WT DDR OE

Mea

n S

lope

of S

crat

ch W

idth

(µm

/hr)

0

20

10

30

40

DDR OEDDR WT

DDR1 WT

DDR1 OE

0 hrs post scratch 12 hrs post scratch

0 2 4 6 8 10 12

DDR OEDDR WT

Time (hrs)

60

Num

ber o

f Foc

al A

dhes

ions

DDR WT DDR OE

A

Figure 5

B

D

WT OE

9EG7 9EG7

C

E

Cel

l Are

a (µ

m2 )

0

2000

1000

3000

4000

DDR WT DDR OE

DDR WT DDR OE

****

Foca

l Adh

esio

n Le

ngth

(µm

)

DDR WT DDR OE0

2

1

3

0

200

100

300

Foca

l Adh

esio

n A

rea

(µm

2 )

0

1

2

***

61

Num

ber o

f Foc

al A

dhes

ions

DDR WT DDR OE

A

Figure 6

B

D

WT OE

Talin Talin

C

E

Cel

l Are

a (µ

m2 )

0

2000

1000

3000

4000

DDR WT DDR OE

DDR WT DDR OE

***

Foca

l Adh

esio

n Le

ngth

(µm

)

DDR WT DDR OE0

2

1

3

0

200

100

300

Foca

l Adh

esio

n A

rea

(µm

2 )

0

1

2

****

*

***

62

Num

ber o

f Foc

al A

dhes

ions

DDR WT DDR OE

A

Figure 7

B

D

WT OE

Paxillin Paxillin

C

E

Cel

l Are

a (µ

m2 )

0

2000

1000

3000

4000

DDR WT DDR OE

DDR WT DDR OE

***

Foca

l Adh

esio

n Le

ngth

(µm

)

DDR WT DDR OE0

2

1

3

0

200

100

300

Foca

l Adh

esio

n A

rea

(µm

2 )

0

1

2

***

*****

63

Num

ber o

f Foc

al A

dhes

ions

DDR WT DDR OE

A

Figure 8

B

D

WT OE

Vinculin Vinculin

C

E

Cel

l Are

a (µ

m2 )

0

2000

1000

3000

4000

DDR WT DDR OE

DDR WT DDR OE

***

Foca

l Adh

esio

n Le

ngth

(µm

)

DDR WT DDR OE0

2

1

3

0

200

100

300

Foca

l Adh

esio

n A

rea

(µm

2 )

0

1

2

***

****

********

64

DDR WT DDR OE

A

Figure 9

DC

B

Fluo

resc

ence

Cha

nnel

#: K

MI6

Sta

inin

g

0

6

3

9

12

DDR WT DDR OE

DDR WT DDR OE

***

DDR WT DDR OE

****

0

6

3

9

12

0

6

3

9

12

0

6

3

9

12

Fluo

resc

ence

Cha

nnel

#:

9EG

7 S

tain

ing

Col

lage

n

Fluo

resc

ence

Cha

nnel

#:

9EG

7 S

tain

ing

Tiss

ue C

ultu

re P

last

ic

Fluo

resc

ence

Cha

nnel

#:

9EG

7 S

tain

ing

Sus

pens

ion

**

**

***

*

65

Figure 10

- + - + - + - + --+ +

WT OEDDR1Collagen

Jararhagin

β1 integrin

GAPDH

130 kDa

36 kDa

66

Figure 11

- + - + - + - + --+ +

WT OEDDR1Collagen

Endo H

β1 integrin

GAPDH

130 kDa

36 kDa

glycosylated deglycosylated

67

68

Future Directions and Conclusions

Our data provide new insights into the role of DDR1 in regulating β1integrin maturation,

trafficking and function. However, there are several unresolved issues. The current results focus

on the role of DDR1 in collagen remodeling and adhesion using a DDR1 over-expression

system. Although we have employed the use of β1 integrin blocking antibodies to examine

collagen binding, a useful approach would be to analyze receptor-collagen interactions in cells

with knockdown of β1 integrin expression and knockdown of DDR1 expression using the same

approach. These experimental methods would be achieved with β1 integrin or DDR1-targeted

siRNA knockdown and then conducting the collagen remodeling and adhesion assays described

above. We expect that knockdown of β1 integrins and DDR1 would inhibit cell binding and

matrix remodeling.

Partially glycosylated β1 integrin subunits form a pool within the ER (Akiyama et al., 1989).

Although we found that DDR1 enhances glycosylation of β1 integrins, our data do not indicate

the potential role of DDR1 in regulating trafficking of partially glycosylated β1 subunits from the

ER to the Golgi. Processing of N-linked glycans is controlled by six different N-

acetylglucosamine-glycosyltransferases (GnTs) located in the cis-Golgi and medial Golgi (Vagin

et al., 2009). Conceivably, DDR1 is involved in the up-regulation of these enzymes.

Accordingly, PCR analysis of GnT transcription may clarify how DDR1 regulates integrin

maturation and cell surface expression. Further, since we found that DDR1 modulates β1 integrin

function by increasing post-translational modifications, specifically through the addition of N-

linked glycans, it would be useful to learn how integrin function is affected by altered expression

of these oligosaccharides. Prior to experimental analysis of cell-collagen interactions, treatment

69

of cells with 1-deoxymannojirimycin, an inhibitor of α-mannosidase II, which prevents N-linked

oligosaccharide processing (Gu and Taniguchi, 2004), would provide insight into the functional

role that these glycans play in cell adhesion and collagen remodeling.

We analyzed cell surface expression of the α2 integrin, the primary α integrin for collagen

binding. We found that DDR1 over-expression increased expression of the α2 subunit on the cell

surface. Since we believe that DDR1 over-expression enhances collagen binding and remodeling

through β1 integrins, we need to determine how DDR1 may regulate pairing of the β1 subunit

with other α-subunits and their presentation on the cell surface. Further analysis by

immunostaining other α-subunits that pair with β1 to form type I collagen receptors, such as α10

or α11 and for fibronectin, the α5 fibronectin subunit, by fluorescence microscopy or flow

cytometry would be of value. We expect that DDR1 over-expression would enhance the pairing

and surface expression of these other β1 integrins as well.

Finally, our data provide some insight into the processes that are important for connective tissue

homeostasis. Analysis of our supplemental experiments indicated that subpopulations of HGFs

treated with CsA exhibited reduced collagen-coated bead binding. Additionally, when cells over-

expressing DDR1 were treated with CsA, we noted that collagen-coated bead binding and

collagen remodeling were also reduced. Since DDR1 over-expression has been implicated in

cancers and multiple fibrotic conditions, and since CsA-induced gingival overgrowth is a

condition in which the collagen phagocytic pathway is disrupted, we expect that DDR1

expression would be increased in the subpopulations of cells that are sensitive to CsA. In studies

using DDR1 knock-out animal models, we expect that these mice are protected against CsA-

induced gingival overgrowth.

70

In conclusion, DDR1-regulated post-translational modifications of the β1 integrin subunit may

regulate β1 integrin function in collagen remodeling and cellular adhesion. An in-depth

understanding of how β1 integrin maturation and function are controlled by DDR1 could provide

insight into how these important collagen receptors interact and may suggest novel strategies for

treatment of fibrotic conditions.

71

Appendix

Cyclosporin A (CsA) is an immunosuppressant commonly used in organ transplantation to

prevent rejection by lowering the activity of T cells (Chan et al., 2007). An adverse effect of CsA

treatment is gingival overgrowth. In CsA-induced gingival overgrowth, collagen degradation by

fibroblasts is inhibited, resulting in fibrosis (Arora et al., 2001; Chan et al., 2007). Since fibrillar

collagen is the primary ligand for DDR1, we hypothesized that CsA perturbs DDR1 function. In

the experiments described below, we examined the impact of CsA on DDR1 function in the

context of collagen binding and remodeling.

72

Appendix Legends

Appendix 1. Collagen binding is reduced in HGF subpopulations and DDR1

over-expressing cells treated with CsA. A) Human gingival fibroblasts obtained from

gingival explants from individual patients were plated overnight on tissue culture plastic in

media containing CsA (10 μg/mL) or vehicle (DMSO). Binding of collagen-coated FITC beads

(1μm; 12 beads/cell) to HGFs was analyzed after 1 hour incubation by flow cytometry. Collagen-

coated bead binding to cells was reduced in subpopulations of cells treated with CsA (patient

numbers 1, 2, 5 and 7: p<0.05, 0.01, 0.05, and 0.01, respectively) compared with untreated cells.

B) Cells were plated overnight on tissue culture plastic in media containing CsA (10 μg/mL) or

vehicle (DMSO). Binding of collagen-coated FITC beads (1μm; 12 beads/cell) was significantly

reduced in DDR1 over-expressing cells (p<0.01) and GD25 cells (p<0.01) treated with CsA. All

data represent mean+S.E.M. (n = 3 independent experiments).

Appendix 2. CsA treatment reduces collagen remodeling in DDR1 over-

expressing cells. A) Cells inoculated in type I collagen gels (1.36 mg/mL, neutralized)

containing collagen-coated FITC beads (50 beads/cell) were incubated overnight in medium

containing CsA (10 μg/mL) or vehicle (DMSO). Collagen gels containing cells were dissolved

by collagenase prior to analysis of bead binding by flow cytometry. Treatment with CsA reduced

collagen-coated bead binding in DDR1 over-expressing cells (p<0.01). B) Collagen remodeling

was measured using a floating gel contraction assay. Compared with untreated cells (vehicle,

DMSO), contraction of type I collagen gels (1.36 mg/mL, neutralized) was reduced in DDR1

over-expressing (p < 0.05) and GD25 cells (p < 0.05) treated with CsA (10 μg/mL). C) Collagen

73

contraction was measured using an anchored gel contraction assay. CsA treatment had no effect

on collagen contraction in all cell types analyzed. All data represent mean + S.E.M. (n = 3

independent experiments).

Appendix 3. CsA treatment does not affect cell migration on collagen. Cell

migration across collagen-coated plastic (1 mg/mL) was measured over 12 hours using a scratch

wound assay. A) Slope of scratch width (mm) over 12 hours. The reduction in scratch width was

not significant in wild-type (p = 0.564) or DDR1 over-expressing (p = 0.545) cells treated with

CsA (10 μg/mL) or vehicle (DMSO). B) Comparison of migration rates in wild-type and DDR1

over-expressing cells treated with CsA. There was no significant difference in migration of wild-

type (p = 0.564) or DDR1 over-expressing (p = 0.545) cells treated with CsA (10 μg/mL) or

vehicle (DMSO). All data represent mean + S.E.M. (n = 3 independent experiments).

Appendix 1

1 2 3 764 50

20

10

30

*

Mea

n P

opul

atio

n w

ith B

ound

Bea

ds (%

)

**

***

DMSOCsA

Patient Number

DDR WT DDR OE GD250

40

20

60

80

100 **

Mea

n P

opul

atio

n w

ith

Bou

nd B

eads

(%)

DMSOCsA

**

A

B

74

A

Appendix 2

B

DDR WT DDR OE GD250

40

20

60

80

100DMSOCsA

**

Mea

n P

opul

atio

n w

ith

Bou

nd B

eads

(%)

DDR WT DDR OE GD250

40

20

60

80

100

Mea

n S

lope

of F

loat

ing

Gel

D

iam

eter

(µm

/hr)

C

Mea

n S

lope

of A

ttach

ed G

el

Dia

met

er (µ

m/h

r)

DDR WT DDR OE GD250

400

200

600

800

1000

*

*

DMSOCsA

DMSOCsA

75

A

Appendix 3

B

DDR WT DDR OE

Mea

n S

lope

of S

crat

ch W

idth

(µm

/hr)

0

20

10

30

40

CsADMSO

Scr

atch

Wid

th (m

m)

0

0.2

DDR WT DMSO

DDR OE CsA0.4

0.6

0 2 4 6 8 10 12

DDR OE DMSODDR WT CsA

Time (hrs)

76

77

References

Abdulhussein, R., D.H. Koo, and W.F. Vogel. 2008. Identification of disulfide-linked dimers of

the receptor tyrosine kinase DDR1. The Journal of biological chemistry. 283:12026-

12033.

Abdulhussein, R., C. McFadden, P. Fuentes-Prior, and W.F. Vogel. 2004. Exploring the

collagen-binding site of the DDR1 tyrosine kinase receptor. The Journal of biological

chemistry. 279:31462-31470.

Akiyama, S.K., S.S. Yamada, and K.M. Yamada. 1989. Analysis of the role of glycosylation of

the human fibronectin receptor. The Journal of biological chemistry. 264:18011-18018.

Alberts, B.e.a. 2002. Molecular Biology of the Cell. Garland Science, New York, New York.

Alves, F., S. Saupe, M. Ledwon, F. Schaub, W. Hiddemann, and W.F. Vogel. 2001.

Identification of two novel, kinase-deficient variants of discoidin domain receptor 1:

differential expression in human colon cancer cell lines. FASEB journal : official

publication of the Federation of American Societies for Experimental Biology. 15:1321-

1323.

Andrews, E.J., J.H. Wang, D.C. Winter, W.E. Laug, and H.P. Redmond. 2001. Tumor cell

adhesion to endothelial cells is increased by endotoxin via an upregulation of beta-1

integrin expression. The Journal of surgical research. 97:14-19.

Arora, P.D., M.A. Conti, S. Ravid, D.B. Sacks, A. Kapus, R.S. Adelstein, A.R. Bresnick, and

C.A. McCulloch. 2008. Rap1 activation in collagen phagocytosis is dependent on

nonmuscle myosin II-A. Molecular biology of the cell. 19:5032-5046.

Arora, P.D., L. Fan, J. Sodek, A. Kapus, and C.A. McCulloch. 2003. Differential binding to

dorsal and ventral cell surfaces of fibroblasts: effect on collagen phagocytosis.

Experimental cell research. 286:366-380.

Arora, P.D., M.F. Manolson, G.P. Downey, J. Sodek, and C.A. McCulloch. 2000. A novel model

system for characterization of phagosomal maturation, acidification, and intracellular

collagen degradation in fibroblasts. The Journal of biological chemistry. 275:35432-

35441.

78

Arora, P.D., L. Silvestri, B. Ganss, J. Sodek, and C.A. McCulloch. 2001. Mechanism of

cyclosporin-induced inhibition of intracellular collagen degradation. The Journal of

biological chemistry. 276:14100-14109.

Avivi-Green, C., M. Singal, and W.F. Vogel. 2006. Discoidin domain receptor 1-deficient mice

are resistant to bleomycin-induced lung fibrosis. American journal of respiratory and

critical care medicine. 174:420-427.

Baumgartner, S., K. Hofmann, R. Chiquet-Ehrismann, and P. Bucher. 1998. The discoidin

domain family revisited: new members from prokaryotes and a homology-based fold

prediction. Protein science : a publication of the Protein Society. 7:1626-1631.

Bellis, S.L. 2004. Variant glycosylation: an underappreciated regulatory mechanism for beta1

integrins. Biochimica et biophysica acta. 1663:52-60.

Bellis, S.L., E. Newman, and E.A. Friedman. 1999. Steps in integrin beta1-chain glycosylation

mediated by TGFbeta1 signaling through Ras. Journal of cellular physiology. 181:33-44.

Berrier, A.L., and K.M. Yamada. 2007. Cell-matrix adhesion. Journal of cellular physiology.

213:565-573.

Birkedal-Hansen, H. 1995. Proteolytic remodeling of extracellular matrix. Current opinion in

cell biology. 7:728-735.

Birkedal-Hansen, H., W.G. Moore, M.K. Bodden, L.J. Windsor, B. Birkedal-Hansen, A.

DeCarlo, and J.A. Engler. 1993. Matrix metalloproteinases: a review. Critical reviews in

oral biology and medicine : an official publication of the American Association of Oral

Biologists. 4:197-250.

Bos, J.L. 2005. Linking Rap to cell adhesion. Current opinion in cell biology. 17:123-128.

Brower, M.S., and P.C. Harpel. 1982. Proteolytic cleavage and inactivation of alpha 2-plasmin

inhibitor and C1 inactivator by human polymorphonuclear leukocyte elastase. The

Journal of biological chemistry. 257:9849-9854.

Burleigh, M.C., A.J. Barrett, and G.S. Lazarus. 1974. Cathepsin B1. A lysosomal enzyme that

degrades native collagen. The Biochemical journal. 137:387-398.

79

Calderwood, D.A. 2004. Integrin activation. Journal of cell science. 117:657-666.

Chan, M.W., P.D. Arora, and C.A. McCulloch. 2007. Cyclosporin inhibition of collagen

remodeling is mediated by gelsolin. American journal of physiology. Cell physiology.

293:C1049-1058.

Chang, J.Z., W.H. Yang, Y.T. Deng, H.M. Chen, and M.Y. Kuo. 2012. EGCG blocks TGFbeta1-

induced CCN2 by suppressing JNK and p38 in buccal fibroblasts. Clinical oral

investigations.

Chong, S.A., W. Lee, P.D. Arora, C. Laschinger, E.W. Young, C.A. Simmons, M. Manolson, J.

Sodek, and C.A. McCulloch. 2007. Methylglyoxal inhibits the binding step of collagen

phagocytosis. The Journal of biological chemistry. 282:8510-8520.

Chou, D.H., W. Lee, and C.A. McCulloch. 1996. TNF-alpha inactivation of collagen receptors:

implications for fibroblast function and fibrosis. Journal of immunology. 156:4354-4362.

Chung, L., D. Dinakarpandian, N. Yoshida, J.L. Lauer-Fields, G.B. Fields, R. Visse, and H.

Nagase. 2004. Collagenase unwinds triple-helical collagen prior to peptide bond

hydrolysis. The EMBO journal. 23:3020-3030.

Clague, M.J., S. Urbe, F. Aniento, and J. Gruenberg. 1994. Vacuolar ATPase activity is required

for endosomal carrier vesicle formation. The Journal of biological chemistry. 269:21-24.

Coomber, B.L., and A.I. Gotlieb. 1990. In vitro endothelial wound repair. Interaction of cell

migration and proliferation. Arteriosclerosis. 10:215-222.

Cougoule, C., A. Wiedemann, J. Lim, and E. Caron. 2004. Phagocytosis, an alternative model

system for the study of cell adhesion. Seminars in cell & developmental biology. 15:679-

689.

Cox, T.R., and J.T. Erler. 2011. Remodeling and homeostasis of the extracellular matrix:

implications for fibrotic diseases and cancer. Disease models & mechanisms. 4:165-178.

Craig, S.W., and R.P. Johnson. 1996. Assembly of focal adhesions: progress, paradigms, and

portents. Current opinion in cell biology. 8:74-85.

80

Critchley, D.R. 2000. Focal adhesions–the cytoskeletal connection. Current opinion in cell

biology. 12:133-139.

Davis, G.E., K.A. Pintar Allen, R. Salazar, and S.A. Maxwell. 2001. Matrix metalloproteinase-1

and -9 activation by plasmin regulates a novel endothelial cell-mediated mechanism of

collagen gel contraction and capillary tube regression in three-dimensional collagen

matrices. Journal of cell science. 114:917-930.

de Bruyn, K.M., S. Rangarajan, K.A. Reedquist, C.G. Figdor, and J.L. Bos. 2002. The small

GTPase Rap1 is required for Mn(2+)- and antibody-induced LFA-1- and VLA-4-

mediated cell adhesion. The Journal of biological chemistry. 277:29468-29476.

Dejmek, J., K. Dib, M. Jonsson, and T. Andersson. 2003. Wnt-5a and G-protein signaling are

required for collagen-induced DDR1 receptor activation and normal mammary cell

adhesion. International journal of cancer. Journal international du cancer. 103:344-351.

Del Pozo, M.A., W.B. Kiosses, N.B. Alderson, N. Meller, K.M. Hahn, and M.A. Schwartz.

2002. Integrins regulate GTP-Rac localized effector interactions through dissociation of

Rho-GDI. Nature cell biology. 4:232-239.

Dickeson, S.K., N.L. Mathis, M. Rahman, J.M. Bergelson, and S.A. Santoro. 1999. Determinants

of ligand binding specificity of the alpha(1)beta(1) and alpha(2)beta(1) integrins. The

Journal of biological chemistry. 274:32182-32191.

Diegelmann, R.F., and M.C. Evans. 2004. Wound healing: an overview of acute, fibrotic and

delayed healing. Frontiers in bioscience : a journal and virtual library. 9:283-289.

Diskin, S., Z. Cao, H. Leffler, and N. Panjwani. 2009. The role of integrin glycosylation in

galectin-8-mediated trabecular meshwork cell adhesion and spreading. Glycobiology.

19:29-37.

Eckes, B., D. Kessler, M. Aumailley, and T. Krieg. 1999. Interactions of fibroblasts with the

extracellular matrix: implications for the understanding of fibrosis. Springer seminars in

immunopathology. 21:415-429.

Everts, V., E. van der Zee, L. Creemers, and W. Beertsen. 1996. Phagocytosis and intracellular

digestion of collagen, its role in turnover and remodelling. The Histochemical journal.

28:229-245.

81

Eyre, D.R. 2004. Collagens and cartilage matrix homeostasis. Clinical orthopaedics and related

research:S118-122.

Ferguson, K.M. 2012. Discoidin discoveries. Structure. 20:568-570.

Ferri, N., N.O. Carragher, and E.W. Raines. 2004. Role of discoidin domain receptors 1 and 2 in

human smooth muscle cell-mediated collagen remodeling: potential implications in

atherosclerosis and lymphangioleiomyomatosis. The American journal of pathology.

164:1575-1585.

Flamant, M., S. Placier, A. Rodenas, C.A. Curat, W.F. Vogel, C. Chatziantoniou, and J.C.

Dussaule. 2006. Discoidin domain receptor 1 null mice are protected against

hypertension-induced renal disease. Journal of the American Society of Nephrology :

JASN. 17:3374-3381.

Franco, C.D., G. Hou, and M.P. Bendeck. 2002. Collagens, integrins, and the discoidin domain

receptors in arterial occlusive disease. Trends in cardiovascular medicine. 12:143-148.

Fries, K.M., T. Blieden, R.J. Looney, G.D. Sempowski, M.R. Silvera, R.A. Willis, and R.P.

Phipps. 1994. Evidence of fibroblast heterogeneity and the role of fibroblast

subpopulations in fibrosis. Clinical immunology and immunopathology. 72:283-292.

Gagliano, N., C. Moscheni, C. Dellavia, C. Torri, G. Stabellini, V.F. Ferrario, and M. Gioia.

2004. Effect of cyclosporin A on human gingival fibroblast collagen turnover in relation

to the development of gingival overgrowth: an in vitro study. Biomedicine &

pharmacotherapy = Biomedecine & pharmacotherapie. 58:231-238.

Grinnell, F. 2003. Fibroblast biology in three-dimensional collagen matrices. Trends in cell

biology. 13:264-269.

Gross, O., R. Girgert, B. Beirowski, M. Kretzler, H.G. Kang, J. Kruegel, N. Miosge, A.C. Busse,

S. Segerer, W.F. Vogel, G.A. Muller, and M. Weber. 2010. Loss of collagen-receptor

DDR1 delays renal fibrosis in hereditary type IV collagen disease. Matrix biology :

journal of the International Society for Matrix Biology. 29:346-356.

Gu, J., and N. Taniguchi. 2004. Regulation of integrin functions by N-glycans. Glycoconjugate

journal. 21:9-15.

82

Haas, T.L., M. Milkiewicz, S.J. Davis, A.L. Zhou, S. Egginton, M.D. Brown, J.A. Madri, and O.

Hudlicka. 2000. Matrix metalloproteinase activity is required for activity-induced

angiogenesis in rat skeletal muscle. American journal of physiology. Heart and

circulatory physiology. 279:H1540-1547.

Hagglund, P., R. Matthiesen, F. Elortza, P. Hojrup, P. Roepstorff, O.N. Jensen, and J.

Bunkenborg. 2007. An enzymatic deglycosylation scheme enabling identification of core

fucosylated N-glycans and O-glycosylation site mapping of human plasma proteins.

Journal of proteome research. 6:3021-3031.

Heinzelmann-Schwarz, V.A., M. Gardiner-Garden, S.M. Henshall, J. Scurry, R.A. Scolyer, M.J.

Davies, M. Heinzelmann, L.H. Kalish, A. Bali, J.G. Kench, L.S. Edwards, P.M. Vanden

Bergh, N.F. Hacker, R.L. Sutherland, and P.M. O'Brien. 2004. Overexpression of the cell

adhesion molecules DDR1, Claudin 3, and Ep-CAM in metaplastic ovarian epithelium

and ovarian cancer. Clinical cancer research : an official journal of the American

Association for Cancer Research. 10:4427-4436.

Hotchin, N.A., and F.M. Watt. 1992. Transcriptional and post-translational regulation of beta 1

integrin expression during keratinocyte terminal differentiation. The Journal of biological

chemistry. 267:14852-14858.

Hou, G., W. Vogel, and M.P. Bendeck. 2001. The discoidin domain receptor tyrosine kinase

DDR1 in arterial wound repair. The Journal of clinical investigation. 107:727-735.

Huang, Y., P. Arora, C.A. McCulloch, and W.F. Vogel. 2009. The collagen receptor DDR1

regulates cell spreading and motility by associating with myosin IIA. Journal of cell

science. 122:1637-1646.

Hwang, Q., S. Cheifetz, C.M. Overall, C.A. McCulloch, and J. Sodek. 2009. Bone sialoprotein

does not interact with pro-gelatinase A (MMP-2) or mediate MMP-2 activation. BMC

cancer. 9:121.

Hyland, P.L., P.S. Traynor, T.T. Myrillas, J.J. Marley, G.J. Linden, P. Winter, N. Leadbetter,

T.E. Cawston, and C.R. Irwin. 2003. The effects of cyclosporin on the collagenolytic

activity of gingival fibroblasts. Journal of periodontology. 74:437-445.

Hynes, R.O. 2002. Integrins: bidirectional, allosteric signaling machines. Cell. 110:673-687.

Hynes, R.O. 2009. The extracellular matrix: not just pretty fibrils. Science. 326:1216-1219.

83

Janik, M.E., A. Litynska, and P. Vereecken. 2010. Cell migration-the role of integrin

glycosylation. Biochimica et biophysica acta. 1800:545-555.

Jiang, G., G. Giannone, D.R. Critchley, E. Fukumoto, and M.P. Sheetz. 2003. Two-piconewton

slip bond between fibronectin and the cytoskeleton depends on talin. Nature. 424:334-

337.

Johnson, J.D., J.C. Edman, and W.J. Rutter. 1993. A receptor tyrosine kinase found in breast

carcinoma cells has an extracellular discoidin I-like domain. Proceedings of the National

Academy of Sciences of the United States of America. 90:10891.

Kang, R., H. Kae, H. Ip, G.B. Spiegelman, and G. Weeks. 2002. Evidence for a role for the

Dictyostelium Rap1 in cell viability and the response to osmotic stress. Journal of cell

science. 115:3675-3682.

Kashket, S., M.F. Maiden, A.D. Haffajee, and E.R. Kashket. 2003. Accumulation of

methylglyoxal in the gingival crevicular fluid of chronic periodontitis patients. Journal of

clinical periodontology. 30:364-367.

Kataoka, M., J. Kido, Y. Shinohara, and T. Nagata. 2005. Drug-induced gingival overgrowth--a

review. Biological & pharmaceutical bulletin. 28:1817-1821.

Keselowsky, B.G., D.M. Collard, and A.J. Garcia. 2004. Surface chemistry modulates focal

adhesion composition and signaling through changes in integrin binding. Biomaterials.

25:5947-5954.

Kim, H., F. Nakamura, W. Lee, C. Hong, D. Perez-Sala, and C.A. McCulloch. 2010a. Regulation

of cell adhesion to collagen via beta1 integrins is dependent on interactions of filamin A

with vimentin and protein kinase C epsilon. Experimental cell research. 316:1829-1844.

Kim, H., F. Nakamura, W. Lee, Y. Shifrin, P. Arora, and C.A. McCulloch. 2010b. Filamin A is

required for vimentin-mediated cell adhesion and spreading. American journal of

physiology. Cell physiology. 298:C221-236.

Kim, S.H., J. Turnbull, and S. Guimond. 2011. Extracellular matrix and cell signalling: the

dynamic cooperation of integrin, proteoglycan and growth factor receptor. The Journal of

endocrinology. 209:139-151.

84

Kinane, D.F. 2000. Regulators of tissue destruction and homeostasis as diagnostic aids in

periodontology. Periodontology 2000. 24:215-225.

Klein, A., J.S. Capitanio, D.A. Maria, and I.R. Ruiz. 2011. Gene expression in SK-Mel-28

human melanoma cells treated with the snake venom jararhagin. Toxicon : official

journal of the International Society on Toxinology. 57:1-8.

Klein, C.E., D. Dressel, T. Steinmayer, C. Mauch, B. Eckes, T. Krieg, R.B. Bankert, and L.

Weber. 1991. Integrin alpha 2 beta 1 is upregulated in fibroblasts and highly aggressive

melanoma cells in three-dimensional collagen lattices and mediates the reorganization of

collagen I fibrils. The Journal of cell biology. 115:1427-1436.

Knowles, G.C., M. McKeown, J. Sodek, and C.A. McCulloch. 1991. Mechanism of collagen

phagocytosis by human gingival fibroblasts: importance of collagen structure in cell

recognition and internalization. Journal of cell science. 98 ( Pt 4):551-558.

Krieg, T., J.S. Perlish, R. Fleischmajer, and O. Braun-Falco. 1985. Collagen synthesis in

scleroderma: selection of fibroblast populations during subcultures. Archives of

dermatological research. 277:373-376.

Lal, H., S.K. Verma, D.M. Foster, H.B. Golden, J.C. Reneau, L.E. Watson, H. Singh, and D.E.

Dostal. 2009. Integrins and proximal signaling mechanisms in cardiovascular disease.

Frontiers in bioscience : a journal and virtual library. 14:2307-2334.

Lauffenburger, D.A., and A.F. Horwitz. 1996. Cell migration: a physically integrated molecular

process. Cell. 84:359-369.

Laval, S., R. Butler, A.N. Shelling, A.M. Hanby, R. Poulsom, and T.S. Ganesan. 1994. Isolation

and characterization of an epithelial-specific receptor tyrosine kinase from an ovarian

cancer cell line. Cell growth & differentiation : the molecular biology journal of the

American Association for Cancer Research. 5:1173-1183.

Lee, W., J. Sodek, and C.A. McCulloch. 1996. Role of integrins in regulation of collagen

phagocytosis by human fibroblasts. Journal of cellular physiology. 168:695-704.

Leitinger, B. 2011. Transmembrane collagen receptors. Annual review of cell and developmental

biology. 27:265-290.

85

Leitinger, B., and E. Hohenester. 2007. Mammalian collagen receptors. Matrix biology : journal

of the International Society for Matrix Biology. 26:146-155.

Lemmon, M.A., and J. Schlessinger. 2010. Cell signaling by receptor tyrosine kinases. Cell.

141:1117-1134.

Lenter, M., H. Uhlig, A. Hamann, P. Jeno, B. Imhof, and D. Vestweber. 1993. A monoclonal

antibody against an activation epitope on mouse integrin chain beta 1 blocks adhesion of

lymphocytes to the endothelial integrin alpha 6 beta 1. Proceedings of the National

Academy of Sciences of the United States of America. 90:9051-9055.

Liu, S., D.A. Calderwood, and M.H. Ginsberg. 2000. Integrin cytoplasmic domain-binding

proteins. Journal of cell science. 113 ( Pt 20):3563-3571.

Lu, K.K., D. Trcka, and M.P. Bendeck. 2011. Collagen stimulates discoidin domain receptor 1-

mediated migration of smooth muscle cells through Src. Cardiovascular pathology : the

official journal of the Society for Cardiovascular Pathology. 20:71-76.

Matsuyama, W., M. Faure, and T. Yoshimura. 2003. Activation of discoidin domain receptor 1

facilitates the maturation of human monocyte-derived dendritic cells through the TNF

receptor associated factor 6/TGF-beta-activated protein kinase 1 binding protein 1

beta/p38 alpha mitogen-activated protein kinase signaling cascade. Journal of

immunology. 171:3520-3532.

McCulloch, C.A., and G.C. Knowles. 1993. Deficiencies in collagen phagocytosis by human

fibroblasts in vitro: a mechanism for fibrosis? Journal of cellular physiology. 155:461-

471.

McKeown, M., G. Knowles, and C.A. McCulloch. 1990. Role of the cellular attachment domain

of fibronectin in the phagocytosis of beads by human gingival fibroblasts in vitro. Cell

and tissue research. 262:523-530.

Messent, A.J., D.S. Tuckwell, V. Knauper, M.J. Humphries, G. Murphy, and J. Gavrilovic. 1998.

Effects of collagenase-cleavage of type I collagen on alpha2beta1 integrin-mediated cell

adhesion. Journal of cell science. 111 ( Pt 8):1127-1135.

Meyer, M., and B. Morgenstern. 2003. Characterization of gelatine and acid soluble collagen by

size exclusion chromatography coupled with multi angle light scattering (SEC-MALS).

Biomacromolecules. 4:1727-1732.

86

Mihai, C., M. Chotani, T.S. Elton, and G. Agarwal. 2009. Mapping of DDR1 distribution and

oligomerization on the cell surface by FRET microscopy. Journal of molecular biology.

385:432-445.

Murai, O., K. Naruishi, S. Ogihara, N. Suwa, S. Kanazawa, T. Yaegashi, Y. Takeda, and K.

Kunimatsu. 2011. Cathepsin B, D, and L regulation in cyclosporin A-mediated gingival

hyperplasia of a patient with sarcoidosis. Clinical laboratory. 57:535-541.

Nakagawa, S., P. Pawelek, and F. Grinnell. 1989. Extracellular matrix organization modulates

fibroblast growth and growth factor responsiveness. Experimental cell research. 182:572-

582.

Narayanan, A.S., and R.C. Page. 1983. Biosynthesis and regulation of type V collagen in diploid

human fibroblasts. The Journal of biological chemistry. 258:11694-11699.

Nemoto, T., K. Ohashi, T. Akashi, J.D. Johnson, and K. Hirokawa. 1997. Overexpression of

protein tyrosine kinases in human esophageal cancer. Pathobiology : journal of

immunopathology, molecular and cellular biology. 65:195-203.

Niland, S., and J.A. Eble. 2012. Integrin-mediated cell-matrix interaction in physiological and

pathological blood vessel formation. Journal of oncology. 2012:125278.

Noordeen, N.A., F. Carafoli, E. Hohenester, M.A. Horton, and B. Leitinger. 2006. A

transmembrane leucine zipper is required for activation of the dimeric receptor tyrosine

kinase DDR1. The Journal of biological chemistry. 281:22744-22751.

Oh, Y.K., and J.A. Swanson. 1996. Different fates of phagocytosed particles after delivery into

macrophage lysosomes. The Journal of cell biology. 132:585-593.

Omori, K., K. Naruishi, T. Yamaguchi, S.A. Li, M. Yamaguchi-Morimoto, K. Matsuura, H.

Arai, K. Takei, and S. Takashiba. 2007. cAMP-response element binding protein (CREB)

regulates cyclosporine-A-mediated down-regulation of cathepsin B and L synthesis. Cell

and tissue research. 330:75-82.

Overall, C.M., J.L. Wrana, and J. Sodek. 1991. Transcriptional and post-transcriptional

regulation of 72-kDa gelatinase/type IV collagenase by transforming growth factor-beta 1

in human fibroblasts. Comparisons with collagenase and tissue inhibitor of matrix

metalloproteinase gene expression. The Journal of biological chemistry. 266:14064-

14071.

87

Park, H.S., K.R. Kim, H.J. Lee, H.N. Choi, D.K. Kim, B.T. Kim, and W.S. Moon. 2007.

Overexpression of discoidin domain receptor 1 increases the migration and invasion of

hepatocellular carcinoma cells in association with matrix metalloproteinase. Oncology

reports. 18:1435-1441.

Perez-Tamayo, R. 1978. Pathology of collagen degradation. A review. The American journal of

pathology. 92:508-566.

Pham, C.T. 2006. Neutrophil serine proteases: specific regulators of inflammation. Nature

reviews. Immunology. 6:541-550.

Playford, M.P., R.J. Butler, X.C. Wang, R.M. Katso, I.E. Cooke, and T.S. Ganesan. 1996. The

genomic structure of discoidin receptor tyrosine kinase. Genome research. 6:620-627.

Quan, J., T. Yahata, S. Adachi, K. Yoshihara, and K. Tanaka. 2011. Identification of Receptor

Tyrosine Kinase, Discoidin Domain Receptor 1 (DDR1), as a Potential Biomarker for

Serous Ovarian Cancer. International journal of molecular sciences. 12:971-982.

Ram, R., G. Lorente, K. Nikolich, R. Urfer, E. Foehr, and U. Nagavarapu. 2006. Discoidin

domain receptor-1a (DDR1a) promotes glioma cell invasion and adhesion in association

with matrix metalloproteinase-2. Journal of neuro-oncology. 76:239-248.

Remy, J., J. Wegrowski, F. Crechet, M. Martin, and F. Daburon. 1991. Long-term

overproduction of collagen in radiation-induced fibrosis. Radiation research. 125:14-19.

Ruiz, P.A., and G. Jarai. 2011. Collagen I induces discoidin domain receptor (DDR) 1 expression

through DDR2 and a JAK2-ERK1/2-mediated mechanism in primary human lung

fibroblasts. The Journal of biological chemistry. 286:12912-12923.

Saftig, P., and J. Klumperman. 2009. Lysosome biogenesis and lysosomal membrane proteins:

trafficking meets function. Nature reviews. Molecular cell biology. 10:623-635.

Salicioni, A.M., A. Gaultier, C. Brownlee, M.K. Cheezum, and S.L. Gonias. 2004. Low density

lipoprotein receptor-related protein-1 promotes beta1 integrin maturation and transport to

the cell surface. The Journal of biological chemistry. 279:10005-10012.

Santibanez-Gallerani, A.S., A.E. Barber, S.J. Williams, S. Davis, Y. Zhao, and G.T. Shires.

2000. Matrix metalloproteinase inhibition protects hepatic integrity in hemorrhagic

88

shock. Journal of gastrointestinal surgery : official journal of the Society for Surgery of

the Alimentary Tract. 4:536-541.

Schnapp, L.M. 2006. ADHESION, CELL–MATRIX | Integrins. In Encyclopedia of Respiratory

Medicine. J.L. Geoffrey and D.S. Steven, editors. Academic Press, Oxford. 47-53.

Segal, G., W. Lee, P.D. Arora, M. McKee, G. Downey, and C. McCulloch. 2001. Involvement of

actin filaments and integrins in the binding step in collagen phagocytosis by human

fibroblasts. Journal of cell science. 114:119-129.

Shintani, Y., Y. Fukumoto, N. Chaika, R. Svoboda, M.J. Wheelock, and K.R. Johnson. 2008.

Collagen I-mediated up-regulation of N-cadherin requires cooperative signals from

integrins and discoidin domain receptor 1. The Journal of cell biology. 180:1277-1289.

Slack, B.E., M.S. Siniaia, and J.K. Blusztajn. 2006. Collagen type I selectively activates

ectodomain shedding of the discoidin domain receptor 1: involvement of Src tyrosine

kinase. Journal of cellular biochemistry. 98:672-684.

Sodek, J. 1977. A comparison of the rates of synthesis and turnover of collagen and non-collagen

proteins in adult rat periodontal tissues and skin using a microassay. Archives of oral

biology. 22:655-665.

Sodek, J., and C. Overall. 1988. Matrix degradation in hard and soft connective tissues. The

biological mechanisms of tooth eruption and root resorption. 2:303-312.

Song, S., N.A. Shackel, X.M. Wang, K. Ajami, G.W. McCaughan, and M.D. Gorrell. 2011.

Discoidin domain receptor 1: isoform expression and potential functions in cirrhotic

human liver. The American journal of pathology. 178:1134-1144.

Suh, H.N., and H.J. Han. 2011. Collagen I regulates the self-renewal of mouse embryonic stem

cells through alpha2beta1 integrin- and DDR1-dependent Bmi-1. Journal of cellular

physiology. 226:3422-3432.

Sun, W., W. Hu, R. Xu, J. Jin, Z.M. Szulc, G. Zhang, S.H. Galadari, L.M. Obeid, and C. Mao.

2009. Alkaline ceramidase 2 regulates beta1 integrin maturation and cell adhesion.

FASEB journal : official publication of the Federation of American Societies for

Experimental Biology. 23:656-666.

89

Ten Cate, A.R., and D.A. Deporter. 1974. The role of the fibroblast in collagen turnover in the

functioning periodontal ligament of the mouse. Archives of oral biology. 19:339-340.

Tipton, D.A., G.P. Stricklin, and M.K. Dabbous. 1991. Fibroblast heterogeneity in collagenolytic

response to cyclosporine. Journal of cellular biochemistry. 46:152-165.

Vagin, O., J.A. Kraut, and G. Sachs. 2009. Role of N-glycosylation in trafficking of apical

membrane proteins in epithelia. American journal of physiology. Renal physiology.

296:F459-469.

Valiathan, R.R., M. Marco, B. Leitinger, C.G. Kleer, and R. Fridman. 2012. Discoidin domain

receptor tyrosine kinases: new players in cancer progression. Cancer metastasis reviews.

31:295-321.

van der Flier, A., and A. Sonnenberg. 2001. Function and interactions of integrins. Cell and

tissue research. 305:285-298.

van der Zee, E., V. Everts, K. Hoeben, and W. Beertsen. 1995. Cytokines modulate phagocytosis

and intracellular digestion of collagen fibrils by fibroblasts in rabbit periosteal explants.

Inverse effects on procollagenase production and collagen phagocytosis. Journal of cell

science. 108 ( Pt 10):3307-3315.

van Weert, A.W., K.W. Dunn, H.J. Geuze, F.R. Maxfield, and W. Stoorvogel. 1995. Transport

from late endosomes to lysosomes, but not sorting of integral membrane proteins in

endosomes, depends on the vacuolar proton pump. The Journal of cell biology. 130:821-

834.

Vogel, W. 1999. Discoidin domain receptors: structural relations and functional implications.

FASEB journal : official publication of the Federation of American Societies for

Experimental Biology. 13 Suppl:S77-82.

Vogel, W., C. Brakebusch, R. Fassler, F. Alves, F. Ruggiero, and T. Pawson. 2000. Discoidin

domain receptor 1 is activated independently of beta(1) integrin. The Journal of

biological chemistry. 275:5779-5784.

Vogel, W., G.D. Gish, F. Alves, and T. Pawson. 1997. The discoidin domain receptor tyrosine

kinases are activated by collagen. Molecular cell. 1:13-23.

90

Vogel, W.F. 2001. Collagen-receptor signaling in health and disease. European journal of

dermatology : EJD. 11:506-514.

Vogel, W.F. 2002. Ligand-induced shedding of discoidin domain receptor 1. FEBS letters.

514:175-180.

Vogel, W.F., R. Abdulhussein, and C.E. Ford. 2006. Sensing extracellular matrix: an update on

discoidin domain receptor function. Cellular signalling. 18:1108-1116.

von Lampe, B., A. Stallmach, and E.O. Riecken. 1993. Altered glycosylation of integrin

adhesion molecules in colorectal cancer cells and decreased adhesion to the extracellular

matrix. Gut. 34:829-836.

Wang, C.Z., H.W. Su, Y.C. Hsu, M.R. Shen, and M.J. Tang. 2006. 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. Molecular biology of the cell.

17:2839-2852.

Williams, I.R. 1998. Fibroblasts. In Encyclopedia of Immunology (Second Edition). J.D. Editor-

in-Chief: Peter, editor. Elsevier, Oxford. 905-909.

Woessner, J.F., Jr. 1993. Introduction to serial reviews: the extracellular matrix. FASEB journal :

official publication of the Federation of American Societies for Experimental Biology.

7:735-736.

Yajima, T. 1986. Acid phosphatase activity and intracellular collagen degradation by fibroblasts

in vitro. Cell and tissue research. 245:253-260.

Yang, S.H., H.A. Baek, H.J. Lee, H.S. Park, K.Y. Jang, M.J. Kang, D.G. Lee, Y.C. Lee, W.S.

Moon, and M.J. Chung. 2010. Discoidin domain receptor 1 is associated with poor

prognosis of non-small cell lung carcinomas. Oncology reports. 24:311-319.

Yeh, Y.C., C.Z. Wang, and M.J. Tang. 2009. Discoidin domain receptor 1 activation suppresses

alpha2beta1 integrin-dependent cell spreading through inhibition of Cdc42 activity.

Journal of cellular physiology. 218:146-156.

Yuen, A., C. Laschinger, I. Talior, W. Lee, M. Chan, J. Birek, E.W. Young, K. Sivagurunathan,

E. Won, C.A. Simmons, and C.A. McCulloch. 2010. Methylglyoxal-modified collagen

91

promotes myofibroblast differentiation. Matrix biology : journal of the International

Society for Matrix Biology. 29:537-548.

Zahm, J.M., H. Kaplan, A.L. Herard, F. Doriot, D. Pierrot, P. Somelette, and E. Puchelle. 1997.

Cell migration and proliferation during the in vitro wound repair of the respiratory

epithelium. Cell motility and the cytoskeleton. 37:33-43.

Zhang, Z., A.O. Morla, K. Vuori, J.S. Bauer, R.L. Juliano, and E. Ruoslahti. 1993. The alpha v

beta 1 integrin functions as a fibronectin receptor but does not support fibronectin matrix

assembly and cell migration on fibronectin. The Journal of cell biology. 122:235-242.

Zhao, W., M.H. Byrne, B.F. Boyce, and S.M. Krane. 1999. Bone resorption induced by

parathyroid hormone is strikingly diminished in collagenase-resistant mutant mice. The

Journal of clinical investigation. 103:517-524.

Zheng, M., H. Fang, and S. Hakomori. 1994. Functional role of N-glycosylation in alpha 5 beta 1

integrin receptor. De-N-glycosylation induces dissociation or altered association of alpha

5 and beta 1 subunits and concomitant loss of fibronectin binding activity. The Journal of

biological chemistry. 269:12325-12331.