Inhibition of cancer invasion and metastasis: Mechanistic analysis of contortrostatin function at the molecular and cellular levels

PDF

Download

Open document

Flip pages

Download a page range

Download transcript

Copy asset link

Request this asset

Request accessible transcript

Transcript (if available)

Content INFORMATION TO USERS This manuscript has been reproduced from the microfilm m aster. UMI films the text directly from the original or copy submitted. Thus, som e thesis and dissertation copies are in typewriter face, while others may be from any type of com puter printer. The quality of this reproduction is dependent upon the quality of the cop y submitted. Broken or indistinct print, colored or poor quality illustrations and photographs, print bleedthrough, substandard margins, and improper alignment can adversely affect reproduction. In the unlikely event that the author did not send UMI a complete manuscript and there are missing pages, these will be noted. Also, if unauthorized copyright material had to be removed, a note will indicate the deletion. Oversize materials (e.g., maps, drawings, charts) are reproduced by sectioning the original, beginning at the upper left-hand com er and continuing from left to right in equal sections with small overlaps. Photographs included in the original manuscript have been reproduced xerographically in this copy. Higher quality 6” x 9” black and white photographic prints are available for any photographs or illustrations appearing in this copy for an additional charge. Contact UMI directly to order. ProQ uest Information and Learning 300 North Zeeb Road, Ann Arbor, M l 48106-1346 USA 800-521-0600 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. INHIBITION OF CANCER INVASION AND METASTASIS: MECHANISTIC ANALYSIS OF CONTORTROSTATIN FUNCTION AT THE MOLECULAR AND CELLULAR LEVELS by Matthew R. Ritter A Dissertation Presented to the FACULTY OF THE GRADUATE SCHOOL UNIVERSITY OF SOUTHERN CALIFORNIA In Partial Fulfillment of the Requirements for the Degree DOCTOR OF PHILOSOPHY (BIOCHEMISTRY AND MOLECULAR BIOLOGY) August 2000 Copyright 2000 Matthew Ray Ritter Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. U M I Number: 3018118 ___ ® UMI UMI Microform 3018118 Copyright 2001 by Bell & Howell Information and Learning Company. All rights reserved. This microform edition is protected against unauthorized copying under Title 17, United States Code. Bell & Howell Information and Learning Company 300 North Zeeb Road P.O. Box 1346 Ann Arbor, Ml 48106-1346 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. UNIVERSITY OF SOUTHERN CALIFORNIA TH E GRADUATE SCHOOL UNIVERSITY PARK LOS ANGELES, CALIFORNIA 90007 This dissertation, written by ......... Mat .thew ..Ray^.Ri.t ter,......... under the direction of h.Ls Dissertation Committee, and approved by all its members, has been presented to and accepted by The Graduate School, in partial fulfillment of re quirements for the degree of DOCTOR OF PHILOSOPHY Dean of Graduate Studies Date Axxgio.sJt 8 , 2 000 DISSERTATION COMMITTEE Chairperson rr: Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Matthew R. Ritter Francis S. Markland (chair) INHIBITION OF CANCER INVASION AND METASTASIS: MECHANISTIC ANALYSIS OF CONTORTROSTATIN FUNCTION AT THE MOLECULAR AND CELLULAR LEVELS Contortrostatin is a homodimeric, RDG-containing disintegrin isolated from the venom of the southern copperhead snake that acts as an inhibitor of cancer progression in animal models. Determination of the complete primary structure o f contortrostatin revealed that contortrostatin lacks two important cysteine residues at the N-terminus that are commonly found in most other disintegrin family members, which may enable the dimerization of this disintegrin. It was hypothesized that, by blocking specific integrins, contortrostatin could have a negative effect on the ability of tumor cells to degrade the extracellular matrix. Using zymography, it was demonstrated that contortrostatin had no effect on the secretion of gelatinases (MMPs) or plasminogen activators by MDA-MB-435 cells. Similar results were obtained using a radiolabeled matrix degradation assay in which contortrostatin treated cells were not inhibited in their ability to degrade the matrix. Studies into the effects of contortrostatin on integrin signaling unexpectedly revealed that contortrostatin stimulates tyrosine phosphorylation of FAK and CAS, but monomeric disintegrins do not. These events are shown to be mediated exclusively by the ocv(33 integrin. Contortrostatin treatment is also shown to result in the activation of ERK2. Evidence is provided that activation of ERK2 and the tyrosine phosphorylation events are mediated by different integrin receptors and possibly different signaling pathways. Contortrostatin-induced tyrosine phosphorylation events are correlated to disruptions in the actin cytoskeleton and focal adhesion structures since the av(33 integrin is required for Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. these disruptions to occur, and monomeric disintegrins do not induce such disruptions. These disruptions include collapse of actin stress fibers and altered subcellular localization of FAK. Contortrostatin is demonstrated to effectively inhibit tumor cell motility, which is believed to be an activity directly related to the structural disruptions induced by contortrostatin. It is concluded that contortrostatin acts as an inhibitor of cell motility by activating integrin signaling pathways in a spatially and temporally inappropriate manner, and that the inhibition of motility in part accounts for the ability of contortrostatin to inhibit cancer progression. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table of Contents List of Figures iii Chapter 1. Introduction. 1 Chapter 2. Primary structure of contortrostatin 13 Chapter 3. Effect of contortrostatin on extracellular 28 matrix degradation by tumor cells Chapter 4. Contortrostatin induces avp3-mediated 43 tyrosine phosphorylation events in tumor cells Chapter 5. Contortrostatin activates ERK2 and tyrosine 71 phosphorylation events via distinct pathways Chapter 6. Contortrostatin causes avp3-mediated structural 90 disruptions and inhibits tumor cell motility Chapter 7. Conclusion and General Discussion 113 References 135 ii Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure Figures Page 2.1 Chromatograms of RP-HPLC of nonreduced contortrostatin and reduced and alkylated contortrostatin 2.2 Primary of structure of contortrostatin and selected other members of the disintegrin family 2.3 SDS-PAGE analysis of the two isoforms of contortrostatin 3.1 Zymographic analysis of secreted matrix- degrading proteases by MDA-MB-435 cells 3.2 Time course of radiolabeled extracellular matrix degradation by MDA-MB-435 cells 3.3 Effect of contortrostatin and/or vitronectin on degradation of a radiolabeled extracellular matrix by MDA-MB-435 cells 4.1 Tyrosine phosphorylation in adherent T24 cells 4.2 Time course of contortrostatin-induced tyrosine phosphorylation events in suspended T24 cells 4.3 Phosphotyrosine levels in tumor cells after contortrostatin treatment 4.4 Tyrosine phosphorylation induced by contortrostatin in MDA-MB-435 cells 4.5 Effects of monomeric disintegrins on tyrosine phosphorylation in tumor cells 20 22 24 35 37 39 51 52 54 55 58 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 4.6 Contortrostatin-induced tyrosine phosphorylation 59 is mediated by the av(33 integrin 4.7 Contortrostatin treatment causes tyrosine 62 phosphorylation of CAS and FAK 4.8 Involvement of Src family kinases in 63 contortrostatin-induced tyrosine phosphorylation 4.9 Contortrostatin-induced tyrosine phosphorylation 64 is independent of cellular adhesion 5.1 ERK2 activation in adherent T24 cells 77 5.2 Dose response of contortrostatin-induced 78 signaling events in suspended T24 cells 5.3 Contortrostatin-induced tyrosine phosphorylation 80 is mediated by the avp3 integrin, but ERK2 activation is not 5.4 Contortrostatin activates ERK2 in av(33-negative 83 cells but fails to stim u late tyrosine phosphorylation 5.5 A hypothetical model depicting two mechanisms 84 leading to induction of tyrosine phosphorylation and activation of ERK2 6.1 Effect of contortrostatin on morphology and 99 cytoskeletal structures in T24 cells 6.2 FAK localization is altered by contortrostatin 100 treatment 6.3 Contortrostatin does not affect adhesion to a 101 reconstituted basement membrane iv Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 6.4 Flow cytometric analysis of av(33 expression 102 6.5 The avp3 integrin is required for contortrostatin 105 to cause cytoskeletal disruptions 6.6 Comparison of the effects of contortrostatin and 106 flavoridin treatment on cytoskeletal structures 6.7 Contortrostatin inhibits motility o f tumor cells 107 7.1 Schematic model depicting contortrostatin- 120 mediated crosslinking of av(33 and the resulting biochemical events 7.2 Fluorescently-labeled contortrostatin is 127 internalized by T24 cells 7.3 Cellular events required for both metastasis and 129 angiogenesis 7.4 Hypothetical model describing the effect of 131 contortrostatin on cytoskeletal and focal adhesion structures v Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Chapter 1 Introduction Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Cancer remains a critical problem facing society. The complex etiology and difficulties in detection have added to the impact of this disease on the world's populations. Much of the difficulty in treating cancer lies in the fact that each type of cancer generally has a different molecular defect that allows cellular proliferation to go unchecked. Thus a single therapeutic approach is unlikely to be effective in treating all cancers. Another major issue confronting cancer therapies is the difficulty in eliminating cancer cells while leaving healthy cells unharmed. For these reasons, massive research efforts are underway to develop targeted drugs that are active only at the desired site, exploiting unique features of the cancer cells and reducing or eliminating side effects. A major characteristic of malignancy is the ability of the tumor to invade surrounding tissue and to metastasize. Both of these processes rely on a three step mechanism which requires (1) adhesion to the extracellular matrix (ECM), (2) digestion of the surrounding matrix and (3) cellular movement over the matrix [Terranova et al., 1986]. All three of these steps have been shown to depend on integrin receptors [Seftor et al., 1992] which have become targets for therapy. These ubiquitously expressed cell surface receptors are non-covalently linked <x/p heterodimers. The a subunit is comprised of approximately 1000 amino acids and contains calcium-binding motifs that are critical to integrin function [Tuckwell et al., 1992]. Integrin P subunits are made up of approximately 750 amino acids and introduce additional variability through alternative splicing of the cytoplasmic regions. Different pairings of a and P subunits produce at least 20 different integrin heterodimers with distinct but overlapping binding specificities 2 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. [Hynes, 1992]. In turn, each ECM component may be recognized by several integrins. In general, integrins recognize amino acid sequences that contain a key acidic residue that is critical for binding [Pierschbacher and Ruoslahti, 1987]. The most common example of an integrin ligand sequence is the RGD (Arg-Gly- Asp) sequence which is found in a number of integrin-binding proteins [Ruoslahti and Pierschbacher, 1987]. Integrin-binding sequences that are unrelated to the RGD motif show structural and topological similarities to the RGD sequence suggesting that specific spatial elements are required for recognition [Main et al., 1992]. Some integrins are known to require activation in order to bind their corresponding ligands, a mechanism referred to as “inside-out” signaling. For example, the allbfB integrin is a constitutively expressed receptor on platelets but is only able to bind its primary ligand, fibrinogen, after a conformational change induced by platelet activation [Plow and Ginsberg, 1989]. In addition to their roles as adhesion molecules, an increasingly large body of work has described integrins as signal transducing molecules that regulate important cellular functions such as proliferation, gene expression, migration and apoptosis [Giancotti and Ruoslahti, 1999]. Integrins possess no intrinsic biochemical activity. Thus, in order to serve as signal transducers, they must recruit other molecules to perform these functions. Integrin cytoplasmic domains associate with a large number of adapter proteins and tyrosine kinases that carry out the signaling functions of these receptors [Sastry and Horwitz, 1993]. The integrins are named for this function as integrators of the ECM and cytoplasmic complexes. Through various adapter proteins, the integrins are indirectly linked 3 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. to the cytoskeleton and exert control over its functions. The small GTPase, Rho has been shown to mediate effects of integrins on the cytoskeleton, although the events downstream of Rho are complex and not well understood [Barry et al., 1997]. A reciprocal relationship exists in this system where integrins regulate the cytoskeleton and in turn, the cytoskeleton can influence integrin function, for example, integrin clustering and focal adhesion formation [Ridley and Hall, 1992]. Integrin ligation results in increased levels of tyrosine phosphorylation, events which are carried out by protein tyrosine kinases that propagate extracellular signals initiated by integrin receptors [Komberg et al., 1991]. Phosphorylated tyrosines can be recognized by proteins that contain the modular adapter domain known as the Src homology 2 (SH2) domain, named after a central player in integrin signaling, Src [Schaller et al., 1994]. This domain, along with a number of other modular domains, including SH3 and protein tyrosine binding (PTB) domains, are present in different combinations in proteins that lack catalytic activity but serve important functions as adapter proteins. Thus, extremely complex signaling networks exist that use tyrosine kinases and their corresponding recognition domains as “switches” to turn particular pathways “on” or “off.” Also included in these networks are phosphatases which play important roles in the regulation of pathways involving tyrosine phosphorylation events [Garton and Tonks, 1999; Harder et al., 1998]. The endpoints of these integrin- regulated pathways include control of cell migration, gene expression and cell survival. These cellular functions are also regulated by the receptor tyrosine kinase family which includes many of the growth factor receptors. In contrast 4 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. with integrins, these transmembrane proteins possess intrinsic kinase activity and therefore do not depend on other enzymes to initiate signals [Egan et al., 1993]. Integrins and receptor tyrosine kinases have been shown to cooperate since integrin input is necessary for optimal activation of receptors for insulin, epidermal growth factor (EGF) and vascular endothelial growth factor (VEGF) [Cybulsky et al., 1994; Jones et al., 1997; Vuori and Ruoslahti, 1994]. The mitogen activated protein (MAP) kinases are regulated by both receptor types and appear to reside at a major point of intersection of integrin- and growth factor receptor-mediated signaling pathways [Canagarajah et al., 1997; Chen et al., 1994]. Many of the early investigations into the function of integrins were carried out in platelets using the otIIb(33 integrin (GPIIb/nia). Integrin antagonists became useful reagents in the study of these receptors. An often used antagonist was a family of snake venom-derived proteins known as the disintegrins. These small disulfide-rich proteins potently inhibit platelet aggregation by blocking the binding of fibrinogen to allb(33 on activated platelets with ICS 0 values in the nanomolar range [Niewiarowski et al., 1994]. Subsequently, integrins of the (31 and (33 families were found to recognize disintegrins [Juliano et al., 1996; Trikha et al., 1994a]. Interaction is mediated primarily by the RGD sequence found in a conserved position in the disintegrins. In some disintegrins the RDG is replaced with KGD which results in a relatively higher specificity for allb(33. An important distinction among the disintegrins is the number of integrin-binding subunits present. This issue has been highlighted by the discovery of a number of 5 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. dimeric disintegrins. The first member of this group to be described was contortrostatin from Agkistrodon contortrix contortrix [Trikha et al., 1994b]. While all disintegrins described prior to contortrostatin were monomeric with either RDG or KGD sequences, this disintegrin is composed of two identical subunits, each with an RGD motif. The other members of this subfamily contain a single RGD or a variation with a substitution at the first position. For example, EMF10 from Eristocophis macmahoni venom has an RGD in the A subunit and MGD in the B subunit [Marcinkiewicz et al., 1999b]. This disintegrin is selective for the a5f$l integrin, a property reportedly associated with the MGDW motif in the B subunit. Another heterodimeric disintegrin, EC3 from Echis carinatus venom, lacks the RGD sequence altogether [Marcinkiewicz et al., 1999a]. In place of RGD, EC3 contains VGD and MLD and is specific for a4 and a 5 integrins. The only other homodimeric disintegrin described beside contortrostatin is ussuristatin 2 from the venom of Agkistrodon ussuriensis [Oshikawa and Terada, 1999]. This aUbP3-specific disintegrin inhibits platelet aggregation with an IC5 0 of 17-33 nM but does not have an inhibitory effect on adhesion to fibronectin. Thus the disintegrin family continues to grow with new subfamilies emerging which include both heterodimeric and homodimeric proteins. Since their early use as antagonists of the aIIbP3 integrin on platelets, disintegrins have been used to study the function of other integrins on various cell types. Contortrostatin has been shown to be a specific integrin-binding agent by performing solid phase cell adhesion assays using immobilized contortrostatin 6 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. and showing that a small RGD-containing peptide effectively inhibits cell binding to the disintegrin (Trikha et al., 1994a; Zhou et al., 2000d]. Further evidence for the idea that integrins are the exclusive mediators of cell adhesion to contortrostatin is provided by demonstrating that chelating divalent cations using EDTA inhibits adhesion to contortrostatin, since integrins depend on the presence of Ca2 + or Mn2 + for their adhesive function. Preliminary information regarding the species of integrins that are mediating adhesion to contortrostatin was provided in experiments using extracellular m atrix proteins as immobilized adhesion substrates. Contortrostatin was shown to block binding to fibronectin and vitronectin with IC5 0 values of 18 nM and 1.5 nM, respectively [Zhou et al., 2000d]. When laminin is used as the substrate, contortrostatin shows little inhibitory effect on adhesion [Trikha et al., 1994a], which indicates a lack of interaction of contortrostatin with laminin receptors. These results indicated that contortrostatin binds to at least one fibronectin receptor and at least one vitronectin receptor. The a 5 P 1 integrin was tentatively identified as the fibronectin receptor that is bound by contortrostatin by demonstrating that antibodies against this receptor can block cellular binding to immobilized contortrostatin [Trikha et al., 1994a]. Two vitronectin receptors have been shown to be binding sites for contortrostatin. Using a flow cytometry method, it was demonstrated that binding of the anti-av(33 mAb, 7E3, can be effectively blocked by contortrostatin indicating that contortrostatin is capable of binding ccv(33 [Zhou et al., 2000b; Zhou et al., 2000d]. Identification of the second vitronectin receptor required more involved methods using cells lines with specific integrin expression 7 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. profiles [Zhou et al., 2000c]. Incubating T24 cells with 7E3 alone failed to completely inhibit binding to vitronectin. However, binding to vitronectin could be completely blocked with a combination of 7E3 and an antibody against ocvp5 (P1F6) or with contortrostatin alone. Further, adhesion to vitronectin could be completely blocked by P1F6 or contortrostatin in a subpopulation of T24 cells that lack expression of avp3, thus indirectly demonstrating that contortrostatin binds the av(35 integrin. Collectively, these studies have led to the identification of four integrins to which contortrostatin binds: anb(33, a 5 p l, avp3 and avP5. Having established that contortrostatin is a specific integrin-binding protein capable of inhibiting cellular adhesion to extracellular matrix components, experiments designed to study the affect of contortrostatin treatment on cellular behavior were initiated. It was hypothesized that by blocking integrins on the surface of tumor cells, contortrostatin might have a negative effect on tumor cell invasiveness. As mentioned previously, in order for a tumor cell to metastasize, it must penetrate the laminin-rich basement membrane. W ith this in mind, the use of Matrigel, a reconstituted basement membrane, affords a biologically relevant barrier through which invading cells can penetrate. Using a Boy den chamber system [Repesh, 1989], the effect of contortrostatin on invasion through Matrigel can be specifically measured since contortrostatin does not affect adhesion to this substrate. Tumor cells are stimulated to invade the barrier by adding a chemoattractant to the system, and the affect of contortrostatin is measured by pretreating cells with the disintegrin and allowing them to penetrate the membrane. These experiments revealed that contortrostatin is an effective 8 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. inhibitor of invasion in vitro, and this system has been used successfully to demonstrate the inhibition of invasiveness by contortrostatin with a variety of tumor cells [Zhou et al., 2000c; Zhou et al., 2000d] and unpublished data). Integrin av(33 has been shown to have a critical role in the regulation of angiogenesis [Brooks et al., 1994a]. Blocking this integrin with specific antibodies or RGD peptides prevents the development of new blood vessels on the chick chorioallantoic membrane (CAM) [Brooks et al., 1994b; Friedlander et al., 1995]. In fact, several disintegrins have been shown to act as inhibitors of angiogenesis [Kang et al., 1999; Sheu et al., 1997; Yeh et al., 1998]. Our demonstration of the interaction of contortrostatin with av|33 indicated the possibility that contortrostatin could have anti-angiogenic properties. The first step in addressing this hypothesis was carried out using the Boyden chamber invasion assay described above, but in this case human umbilical vein endothelial cells (HUVECs) were treated with contortrostatin and their ability to penetrate a Matrigel barrier was measured. The utility of the this assay for our studies is evident since both tumor cell metastasis and angiogenesis depend directly on an ability of cells to migrate on, and penetrate the basement membrane. Similar to the results with tumor cells, contortrostatin was found to effectively inhibit the invasion of HUVECs through Matrigel [Zhou et al., 2000b]. The next step was to investigate the effect of contortrostatin on angiogenesis in vivo. Employing the CAM model, it was demonstrated that topical treatment with contortrostatin was effective at inhibiting angiogenesis induced by VEGF, bFGF, or tumors grown on the CAM [Zhou et al., 2000b]. The results of these studies provided a significant 9 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. basis for investigations into the potential ability of contortrostatin to block cancer progression in vivo through dual inhibition of both tumor cell invasion and metastasis and angiogenesis. Using an orthotopic xenographic breast cancer model in nude mice, it was demonstrated that contortrostatin inhibits the growth of MDA-MB-435 tumor masses by 74% when daily intratumor injections were made [Zhou et al., 2000d]. Additionally, contortrostain treatment resulted in the reduction of pulmonary macro-metastasis by 68% and micro-metastasis by 62%. Immunohistochemical quantitation of vascular endothelial cells in tumor tissue by Factor V m staining demonstrated a significant reduction in tumor angiogenesis following contortrostatin treatment of these animals. Inhibiting angiogenesis will reduce the supply of nutrients to a growing tumor and also limit critical avenues used by metastasizing cells. Thus it is possible that the observed effects of contortrostatin on pulmonary metastasis might be due solely to the inhibition of angiogenesis, although additional negative effects on metastasis could clearly be produced as a result of the anti-adhesive and anti-invasive activity of contortrostatin. The focus of my dissertation research was to determine the molecular mechanisms employed by contortrostatin and how these mechanisms impact its ability to inhibit invasion, metastasis and angiogenesis. The aforementioned studies raised the question of whether contortrostatin caused the observed effects by blocking adhesion of tumor and endothelial cells through simple antagonism of specific integrins, or by effecting integrin-mediated signaling pathways that control invasiveness and/or angiogenesis. Although the activities of 10 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. contortrostatin in vivo appear extremely complex since it inhibits both metastasis and angiogenesis, a unifying fact exists that may simplify the way we view the action of contortrostatin. Metastasis and angiogenesis might appear to be mechanistically unrelated events, but since they both depend directly on adhesion and migration on the basement membrane, and an ability to degrade the basement membrane barrier, contortrostatin can act as an inhibitor of both metastasis and angiogenesis by negatively affecting any or all of these steps. This idea is extended when considering the fact that a subset of integrin receptors have been implicated as regulators of these processes, primarily avfl3 and av(35, both of which are targets for contortrostatin. One of the major issues to be resolved at the outset of this research project was determining if there is a relationship between the unique homodimeric structure of contortrostatin and its function. Earlier studies indicated that contortrostatin had the ability to cause an increase in tyrosine phosphorylation of some proteins in platelets, whereas a monomeric disintegrin lacked this ability [Clark et al., 1994]. These preliminary findings provided an impetus to further study the role of the dimeric structure of contortrostatin and its contribution to the observed inhibitory effects on metastasis and angiogenesis. Described here are the results of these studies, beginning with an important description of the prim ary structure of contortrostatin. Investigations into the affect of contortrostatin on the ability of tumor cells to degrade the ECM are reported, followed by a description of unique signaling activities attributed to contortrostatin. Finally, representing the culmination of 11 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. this project, a mechanism is presented that introduces a novel approach to therapy for cancer. 12 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Chapter 2 Primary Structure of Contortrostatin 13 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. SUMMARY Previously, partial amino acid sequencing of contortrostatin revealed the presence of an RGD motif in the C-terminal region of the protein suggesting that it be included as part the disintegrin family [Trikha, 1994]. Analysis by SDS- PAGE and mass spectrometry indicate that contortrostatin is composed of two identical subunits each with a mass of approximately 6.7 kDa. Described here is the complete amino acid sequence of contortrostatin showing that, in addition to the characteristic RGD sequence, each 64-residue subunit contains 10 highly conserved cysteines common to the disintegrins. Each subunit shows high similarity to other members of the disintegrin family with the exception of what appears to be a truncation at the N-terminus, which results in the loss of two cysteine residues that might be important in allowing contortrostatin to homodimerize. Additionally, evidence is provided suggesting that the two isoforms of contortrostatin are translated from the same gene, since the two isoforms appear to be identical at the N- and C-termini. The findings described here are supported by cDNA cloning efforts conducted in our laboratory. This work both confirms the membership of contortrostatin in the disintegrin family and distinguishes it as the only RGD-containing homodimeric disintegrin yet described. 14 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. INTRODUCTION Disintegrin selectivity is determined by a number of variables in their structure including the three-dimensional geometry of the hallmark R/KGD motif [McLane et al., 1996] and the residues adjacent to this sequence [Pfaff et al., 1994], the arrangement of the disulfide bridges, and by amino acids elsewhere in the protein [Marcinkiewicz et al., 1997; Tselepis et al., 1997]. A new subclass of disintegrins has emerged which includes several dimeric proteins. The first to be described was contortrostatin [Trikha et al., 1994b]. More recently, several new dimeric disintegrins have been reported. EC3, from the venom of Echis carinatus [Marcinkiewicz et al., 1999a] is a heterodimer that interacts with a 4 and a5 integrins. This disintegrin has the RGD sequence replaced by VGD in one subunit (EC3A) and MLD in the other (EC3B). The three other members of this new group are EMF10 from Echis macmahoni, and CC5 and CC8 from Cerastes cerastes. While amino acid sequences have yet to be reported for CC5 and CC8, EMF10 is composed of one subunit with an RGD motif and one with MGD at this position. EMF10 is a selective inhibitor of the a5(3l integrin, and inhibits oc5|3l- m ediated binding to fibronectin with an IC5 0 of 1 — 4 nM. Other than contortrostatin, only one other homodimeric disintegrin has been described. This protein, ussuristatin 2, comes from the venom of Agkistrodon ussuristatin and is an inhibitor of platelet aggregation but does not inhibit cell adhesion to fibronectin [Oshikawa and Terada, 1999]. This disintegrin contains the KGD 15 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. sequence which is known to preferentially interact with aIIbP3 integrin, thus giving it specificity as a platelet aggregation inhibitor. Reported here is the primary structure of contortrostatin, a distinct member of the disintegrin family. Contortrostatin exists as a homodimer with two identical 64 residue disulfide-linked subunits combining to form a protein with a molecular weight of 13,510 Da. The 13 amino acid loop in which the RGD sequence is presented is identical to albolabrin, the disintegrin from Trimeresurus albolabris, although the two disintegrins appear to have different integrin binding characteristics [Juliano et al., 1996]. These findings identify contortrostatin as a member of the new bivalent subgroup of disintegrins and distinguish it as the only RGD-containing homodimeric disintegrin described to date. MATERIALS AND METHODS Reduction and derivitization - Contortrostatin was purified from crude venom by a four step HPLC procedure as described previously [Clark et al., 1994], Approximately 0.4mg of purified, lyophilized protein was dissolved in 0.6ml of 0.2M Tris buffer containing 6M guanidine HC1, pH 8.2, and incubated at room temperature for 16h. Freshly prepared dithiothreitol (DTT) was added to a final concentration of 15mM. The reaction was incubated at 37°C in the dark for 2-4h 16 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. and boiled 5min. The above step was then repeated. Iodoacetamide was added to a concentration of 30mM and the reaction was incubated at 37°C for 20h under nitrogen. This step was also repeated. The reduced and derivitized protein was separated from the reagent mixture by passing the reaction over a D-Salt Exocellulose Column (Pierce). The desalted protein was then examined by reverse-phase HPLC using an analytical C18 column (Vydac 218TP54) equilibrated with 4% acetonitrile/0.1% TFA and a linear gradient to 80% acetonitrile/0.1% TFA over 100 min. Fractions were collected, lyophilized and redissolved in water before visualization by SDS-PAGE under non-reducing conditions using a 16.5% gel. The major protein-containing fraction, as determined by SDS-PAGE, was subjected to amino acid sequencing. Amino terminal sequencing - This work was carried out as a collaboration with Drs. Jay Fox and John Shannon at the University of Virginia Biomolecular Research Center. The complete amino acid sequence was determined by dissolving the derivitized protein in 10% TFA and loading it onto a Biobrene- treated glass fiber filter followed by sequencing on an Applied Biosystems Procise 494 sequencer using the manufacturer’s standard program. The N- terminal 50 residues were determined using the undigested protein. A peptide containing an overlapping sequence encompassing the remaining C-terminai 14 residues was generated by combining lpg contortrostatin, 2|ig BSA, 2|J.g cytochrome C, 2p.g transferrin, 20|il 8M urea, 0.32M Tris and 0.3jig lysyl peptidase. The reaction was incubated overnight at 37°C. The resulting digest 17 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. was separated by reverse phase chromatography on a Vydac 150mm x 1mm C l8 column (218TP5115) run at 32°C and equilibrated with solvent A (0.1% TFA/water). Peptides were eluted using a 10 min gradient to 15% solvent B (0.09% TF A/70% acetonitrile/water) followed by a 90 min gradient to 80% solvent B at a flow rate of 50pl/min while monitoring at a wavelength of 215nM. Eluting peptides were collected and sequenced as described above. An independent N-terminal sequencing method was also employed. Here, native contortrostatin was subjected to electrophoresis in a 16.5% polyacrylamide gel using a Tris/Tricine buffer system under reducing conditions. Protein was electroblotted to polyvinylidene difluoride (PVDF) using lOmM CAPS in 10% methanol. Protein bands were visualized by staining the membrane with 0.1% Coomassie Blue R-250, 40% methanol/1% acetic acid. Bands were excised from the membrane and analyzed in a Applied Biosystems model 47XA Protein Sequencer at the UCS/Norris Microchemical Core Facility. Identification o f carboxy terminal residue - Identification of the C-terminal residue was accomplished through carboxypeptidase A digestion followed by automated amino acid analysis and was carried out by Dr. Steve Swenson. Digests were performed using lOOpg contortrostatin in 25mM Tris-Cl buffer, pH 7.5 containing 500mM NaCl. Carboxypeptidase A (0.1 unit) was added to the protein solution (25|il) and incubated at 25°C for 15 min. The reaction was stopped by boiling, immediately frozen, and submitted for amino acid analysis. 18 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. SDS-PAGE — To compare the relative sizes of the two isoforms of contortrostatin, the proteins were subjected to electrophoresis under non-reducing and reducing conditions. Contortrostatin samples (5p.g) were mixed with standard sample buffer with or without 10% fl-mercaptoethanol and boiled 5 min prior to loading onto a Tris/Tricine 10 — 20% precast gradient gel (Ready Gel, BioRad). Gels was stained with Coomassie blue. RESULTS Reduction and derivitization of contortrostatin yields a single major protein species as determined by RP-HPLC. After initial separation of linearized contortrostatin from the reagent mixture by passage over a desalting column, the resulting protein was loaded onto a C18 RP-HPLC column as described above. This run yielded one major protein peak eluting at 33.1 min and one minor peak at 34.5 min (Fig 2. IB), suggesting that the majority of protein was present as a single reduced and derivitized species. When nonreduced contortrostatin is run under that same conditions, a single sharp peak eluting at 32.7 min is observed (Fig 2.1 A), suggesting that the hydrophobicity of the protein is altered by reduction and alkylation. However, the fact that elution times are somewhat variable with contortrostatin under these conditions prevents firm conclusions 19 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 32.7 33.1 34.5 Fig 2.1 Chromatograms from reversed-phase high performance liquid chromatography of nonreduced contortrostatin (A) and reduced and alkylated contortrostatin (B). An analytial C18 column and a linear gradient from 4% to 80% acetonitrile over 100 min was used and the elution times are shown. Reduced and alkylated contortrostatin eluted as one major and one minor peak. The major peak was used for further analysis. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. regarding the properties of the derivitized material. The major fraction from the reduced and derivitized material was collected, lyophilized and subjected to amino acid sequencing. Amino acid sequence determination of the full length protein. The first 50 residues of contortrostatin were determined by direct N-terminal sequencing without digestion. The remaining 14 residues were determined after lysyl peptidase digestion and N-terminal sequencing of an overlapping peptide (Fig 2.2). Based on cDNA cloning of contortrostatin, it was expected that the last two residues at the C-terminus would be His-Ala since a stop codon is fo und immediately following the Ala codon [Zhou et al., 2000a]. The initial am ino acid sequencing data reported for contortrostatin indicated that the C-terminal residue was Phe, but the sequencing reaction is known to become increasingly inefficient after 20 — 25 residues. Since Phe was detected at the expected position (Phe63), when combined with the cDNA data, this confirmed the presence of Phe in the intact protein. Also predicted by the cDNA clone is the presence of Ala at the C- terminus. This was not found during sequence analysis using the overlapping fragment obtained following lysyl peptidase digestion. To determine conclusively the C-terminal residue, we performed carboxypeptidase A digestion of the intact protein followed by amino acid analysis. This experiment not only failed to detect any appreciable quantity of Ala, but established the presence o f His at the C-terminus (this differs from the cDNA sequence). The combination of this data 21 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Hamodimeric 4 CONTORTROSTATIN DAPANPCCDAATCKLTTGSOCADGLCCDOCKFMKEGTVCRRARGDDL D D YCNGISAGCP RNP FH ----------------------------------------------------------------------------------------------------- p. — ■ » ■ U s s u r i s t a t i n . EAGEECDCGAPANPCCDAATCKLRPGAOCAEGDCCEOCRFVKEGTVCREAKODWNDDSCTGOSADCPRNGF Heterodimerie EM F10 MNSANPCCDP1TCKBKKGEHCVSGPCCRNCKFLNPGTICKKGRQDNLNDYCTGVSSDCPRNPWKSEEED E M F 10B 2 ELLONSGNPCCDPVTCKPRRGEHCVSGPCCDNCKFiNAGTVCWPAltgDWNDDYCTGISSDCPRNPVFK EC3A NSVHPCCDPVKCEPREGEHCXSGPCCRNCKFLRAGTVCKRAVQDDVDDYCSGITPDCPRNRYKGKED EC3 B NSVHPCCDPVKCEPREGEHCISGPCCRNCKFLNAGTICKRAMLDGLNDYCTGKSSDCPRNRYKGKED Monomeric A P P L A G IN EAGEECDCGSPENPCCDAATCKLRPGAOCAEGLCCDOCKFMKEGTVC-RARQDDVNDYCNGISAGCPRNPFH TR IG R A M IN EAGEDCDCGSPANPCCDAATCKLIPGAOCGEGLCCDOCSFIEEGTVCRIARQDDLDDYCUGRSAGCPRNPFH ALBOLABRXN EAGEDCDCGS PANPCCDAATCKLLPGAQCGEGLCCDQCSFMKKGTTCRRARGDDLDDYCNGI SAGCPRNPLHA E LE G A N TIN EAGEECDCGSPENPCCDAATCKLRPGAOCADGLCCDOCRFKKKRTICRRARQDNPDDRCTGQSADCPRNGLYS K IS T R GKECDCSSPENPCCDAATCKLRPGAOCGEGLCCEOCKFSRAGKICRIPRODMPDDRCTGOSADCPRYH BARBUORIN EAGEECDCGSPENPCCDAATCKLRPGAOCADGLCCDOCRFMKKGTVCRVAKgDWNDDTCTGOSADCPRNGLYG F LA V Q R ID IN GEBCDCGSPSNPCCDAATCKLRPGAOCADGLCCDOCRFKKKTGXCRIARGDFPDDRCTGLSUDCPRWNDL E C H IS T A T IN ECESGPCCRNCKFLKEGTICKRARQDDMDDYCNGKTCDCPRNPHKGPAT Fig 2.2 Primary structure of contortrostatin and selected other members of the disintegrin family. Proteins are grouped in to subfamilies based on subunit content. The fragments used to generate the complete sequence of contortrostatin are indicated by arrows. Proteins are aligned by conserved cysteine residues and by the RGD motif or the substituted derivitive. 22 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. allows the assembly of the complete amino acid sequence of contortrostatin (Fig 2.2). The two isoforms of contortrostatin appear to be identical using SDS-PAGE and amino acid sequencing. The four step HPLC purification procedure yields two major fractions with inhibitory activity in the platelet aggregation assay [Trikha et al., 1994a; Trikha et al., 1994b]. Our laboratory has historically referred to the two proteins as Isoform 1 and Isoform 2 of contortrostatin. Initial studies to determine if functional differences exist between the two species concluded that they possess identical activity. This conclusion was based on inhibition of both platelet aggregation and tumor cell adhesion to fibronectin and vitronectin (data not shown). When the two isoforms are examined by SDS- PAGE under reducing and nonreducing conditions, there are no observable differences (Fig 2.3). Similarly, at the level of primary structure, no differences were detected. The studies above describing the complete sequencing of contortrostatin were conducted using Isoform 2. To complement these studies, we initiated independent studies using Isoform 1. It was determined that the two isoforms were identical in sequence through the first 15 residues at the N- terminus and both proteins have His as their C-terminal residue. These findings strongly suggests that the two isoforms are generated from the same gene, but does not explain why these two proteins elute at different times under cation exchange HPLC. One explanation for this might be that one or more of the 8 Asp residues detected after amino acid sequencing are products of deamination from 23 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. CN1 CN2 NR R NR R Fig 2.3 SDS-PAGE analysis of the two isoforms of contortrostatin (CN)* Proteins were run under either non-reducing (NR) or reducing (R) conditions on a Tris/Tricine 10% - 20% gradient gel and stained with Coomassie blue. The expected molecular weights based on mass spectrometry are indicated since proteins often migrate aberrantly using SDS-PAGE. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Asn residues. This could occur under the low pH conditions experienced by the protein during RP-HPLC. These deamination events could alter the charge on the protein such that it will interact with a cation exchange matrix differendy than the non-deaminated form and elute under different conditions. Since two major isoforms are consistendy observed, it is possible that two major deaminadon states exist, possibly resulting from the deaminadon of a certain number of Asn residues. DISCUSSION The partial amino acid sequencing conducted previously identified contortrostatin as a likely member of the disintegrin family based on the presence of the hallmark RGD motif and conserved cysteine residues. The work presented in this chapter confirms that contortrostatin is indeed a member of this growing family and further distinguishes it among the other members. Contortrostatin is shown to contain the same 13-member loop on which the RGD sequence is presented as the monomeric disintegrin albolabrin. Since albolabrin shows littie inhibition of cell adhesion to fibronectin or vitronectin and does not bind av(33 in solid phase binding assays [Juliano et al., 1996] this indicates that regions outside of the RGD m otif and/or the dimeric structure have significant impact on the integrin binding specificity of the disintegrins. Models generated from NMR studies suggest that the C-terminus of these molecules is located near the active 25 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. RGD loop [Marcinkiewicz et al., 1997], and may have profound effects on integrin-binding characteristics. Further study in this area should yield important information regarding how integrins recognize their ligands. With the emergence of other dimeric disintegrins, contortrostatin retains the distinction as the only homodimeric, RGD-containing member of this subfamily reported. To date, there is only one other homodimeric disintegrin described. This disintegrin, from the venom of another member of the Agkistrodon genus, A. ussuriensis, is comprised of 71 amino acids per subunit and is distinguished from contortrostatin primarily by the presence of KGD in place of RGD [Oshikawa and Terada, 1999], making it a specific antagonist of the aDbP3 integrin on platelets. It will be interesting to determine if this is a genuine homodimer since it appears to lack the two free cysteines that enable dimerization in the other members of this group. Further, it will also be of interest to determine whether this disintegrin will stimulate tyrosine phosphorylation events in platelets similar to those induced contortrostatin [Clark et al., 1994]. The publication of amino acid sequences for the other dimeric disintegrins allows for comparison of primary structures. Two of the recently reported dimeric disintegrins, EMF10 and EC3, share a distinctive characteristic with contortrostatin in that they lack two cysteine residues at the N-terminus which are found in most other disintegrins. It might be expected that the absence of these cysteines leaves other cysteines elsewhere in the protein available to form interchain disulfide bonds. In the case of EMF10, for which the disulfide bond pattern has been published [Calvete et al., 2000] this appears to be true. Cys2 and 26 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Cys3 of EMF10A are disulfide bound to Cys2 and Cys3 o f EMF10B, respectively. In monomeric disintegrins kistrin [Dennis et al., 1993] and flavoridin [Senn and Klaus, 1993], these corresponding residues are involved in disulfide bonds with the cysteines that are absent in EMF10. However, in albolabrin [Calvete et al., 1991], the residue corresponding to Cys3 in EMF10 is bound to the residue corresponding to Cys5. Thus, there is heterogeneity among the disintegrins with respect to disulfide bond arrangement and although there are strong similarities between contortrostatin and EMF10, it is possible that contortrostatin may have a disulfide pattern different than that of EMF10. Also of interest is a model proposed by the Niewiarowski group [Calvete et al., 2000] which suggests that the two integrin-binding regions of EMF10 are located at opposite ends of the dimer at a distance of 7.14 nm and that the RGD in EMF10A and the MGD in EMF10B are oriented 180° from each other. This model was constructed based on previous models of kistrin and flavoridin since these two disintegrins share the same disulfide bonding patterns as the EMF10 monomers. Dr. Juan Calvete, the primary author of the report describing the EMF10 structure, is currently collaborating with us to determine the arrangement of disulfide bonds in contortrostatin. We are also collaborating with Dr. Robert Bau of the USC Department of Chemistry on the determination of the X-ray crystallographic structure of contortrostatin. We anticipate that the combination o f these two approaches will lead to useful information about the three dimensional structure of contortrostatin and how its structure relates to its function. 27 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Chapter 3 Effect o f Contortrostatin on Extracellular Matrix Degradation by Tumor Cells 28 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. SUMMARY Integrins and their ligands can regulate the production of matrix degrading proteases by tumor cells. It was hypothesized that, as an integrin binding protein, contortrostatin might have an effect on the ability of tumor cells to degrade the extracellular matrix and thus affect invasiveness of the cells. To investigate this idea, we performed two independent series of experiments. Using the method of zymography, the first series of experiments was designed to identify and measure the relative quantities of matrix-degrading proteases produced by cancer cells and to determine the effect of contortrostatin on the secretion of these proteases. These experiments indicated that contortrostatin had no effect on the secretion of gelatinases or plasminogen activators by MDA-MB-435 human breast carcinoma cells. Another method was employed to address the contribution of membrane- associated proteases and the potential affect of contortrostatin on this activity'. A biologically produced, radiolabeled extracellular matrix was used as a substrate for proteases associated with the membrane of MDA-MB-435 cells and the degree of degradation was quantitated by measuring the number of counts released into the supernatant medium. The results of these experiments supported those from the zymography experiments since no differences were observed in the degradation of the matrix in the presence of contortrostatin. Thus contortrostatin appears to inhibit invasiveness of tumor cells by a mechanism unrelated to proteolytic matrix degradation, since it demonstrates no effect on gelatinase or plasminogen activator activity. 29 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. INTRODUCTION A large number of cancer-related deaths are due to invasion of primary tumors into surrounding tissues and to metastasis of tumor cells to distant sites. Invasion is dependent on degradation of the extracellular matrix (ECM) and cell migration through the compromised matrix. Metastasis involves a more complex series of events including separation from the primary tumor mass, local invasion and migration into a blood or lymph vessel, arrest at a distant site and migration out of the vessel and into the tissue at that site. Both intravasation and extravasation depend directly on the ability of metastasizing cells to degrade the basement membrane found beneath the endothelial cell layer in blood vessels [Terranova et al., 1986]. The basement membrane is composed largely of laminin and type IV collagen and thus is degraded effectively by collagenases. The matrix metalloproteinases (MMPs) are a group of zinc-dependent enzymes that, by definition, degrade at least one component of the ECM [Emonard and Grimaud, 1990]. These enzymes are proteolytically activated from inactive zymogens and collectively, are capable of degrading essentially all elements of the ECM. MMP-2 and MMP-9, also known as gelatinase A and gelatinase B, primarily use collagens as their substrates including type IV collagen, and have been widely implicated in cancer invasion and metastasis . A variety of cells have been shown to produce elevated levels of MMP-2 and MMP-9 including melanoma [Montgomery et al., 1993], ovarian [Shibata et al., 1998] and breast 30 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. cancer cells [Thompson et al., 1994]. In a recent study, 71 primary breast carcinomas were compared to 17 fibroadenomas for MMP-2 levels [Hanemaaijer et al., 2000]. It was found that active MMP-2 was present in 90% of the malignant carcinomas versus 35% in fibroadenomas. The activity of the MMPs is balanced by naturally occurring inhibitors including a 2 -macroglobulin and the tissue inhibitors of metalloproteinases (TTMPs) [Gomez et al., 1997]. Expression of recombinant TIMPs inhibits collagenolytic activity and invasion and metastasis [DeClerck et al., 1992], which highlights the important role of these enzymes. In a breast cancer model using MDA-MB-435 cells transfected with cDNA encoding TIMP-4 it was shown that these cells have reduced invasiveness in vitro and form significantly smaller tumors that have reduced ability to metastasize in orthotopic mouse models [Wang et al., 1997]. Thus reducing proteolysis of the ECM is a potentially useful approach to treat metastatic disease in breast and other cancers. A second class of enzymes has been identified as important in facilitating invasiveness by mediating ECM degradation. Urokinase and tissue type plasminogen activator are serine proteases whose activity is restricted to the conversion of plasminogen to plasmin. Plasmin however, is a broad-speciflcity protease that is capable of degrading several components o f the ECM, including fibronectin, laminin and type IV collagen [Liotta et al., 1981]. Plasmin also activates certain MMPs, suggesting that this enzyme may have a central role in a larger proteolytic cascade. Similar to MMPs, the plasminogen activators have been reported to have increased expression in transformed cells [Jones and 31 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. DeClerck, 1980] and it is likely that these two types of proteases cooperate to permit the invasive phenotype [DeClerck and Laug, 1996]. The integrin family of receptors has been associated with the regulation of MMP production. Seftor et al. reported that MMP-2 expression is regulated by av|33 and oc5|3l integrins on C8161 human melanoma cells and that perturbation of these integrins can effect their invasiveness in vitro [Seftor et al., 1993]. Another report describes the regulation of metalloprotease gene expression in fibroblasts by a5(3l and a4|3l integrins [Huhtala et al., 1995]. These cells express elevated levels of matrix-degrading metalloproteases when plated onto the 120 kDa RGD-containing region of fibronectin or an antibody against 05(31. These studies raised the possibility that blockage of integrins with contortrostatin could result in decreased expression of these matrix-degrading proteases and consequently reduce tumor cell invasiveness. Contortrostatin has been shown to bind the a5(3l integrin and to inhibit experimental metastasis of tail vein-injected M24met human melanoma cells in mice [Trikha et al., 1994a], Our laboratory has also demonstrated that contortrostatin is an effective inhibitor of MDA-MB- 435 breast cancer cell invasion using an in vitro model [Markland and Zhou, 2000]. One mechanism that may account for the inhibition of invasion and metastasis is a reduction in matrix-degrading protease production by the tumor cells caused by contortrostatin treatment. To directly address this hypothesis, the effect of contortrostatin on the ability of tumor cells to produce matrix degrading enzymes was analyzed using two independent techniques. 32 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. MATERIALS AND METHODS Zymography - Approximately 2 x 106 MDA-MB-435 cells were plated onto tissue culture dishes coated with Matrigel diluted 1:100 with serum-free medium (SFM). The cells were allowed to adhere for 24 h at 37°C prior to washing with serum- free medium. Cells were then treated with various agents in SFM and incubated for 18 h at 37°C. Supernatant conditioned medium was collected and concentrated 50-fold in a Microcon 3 concentrator with a molecular weight cutoff of 3 kDa (Amicon, Beverly, MA). Concentrated samples were then applied to 10% polyacrylamide gels containing 0.1% gelatin plus 100(j.l plasminogen (volume determined empirically). Following electrophoresis, the gels were washed in 2.5% Triton for lh and in water 4 x 5 min to remove SDS from the gel. Gels were then moved to renaturing buffer (50 mM Tris, 10 mM CaCl2 , 150 mM NaCl, pH 7.5) and incubated with agitation 18 — 20 h at 37°C and then stained with Coomassie blue. M atrix degradation assays - Radiolabled extracellular matrix was produced according to the protocol described previously [Jones and DeClerck, 1980] by first coating 6-well culture plates with 0.1% gelatin overnight. Rat smooth muscle cells (R22D) were then seeded into the coated wells at 2 x 106 per well. Sterile ascorbic acid was added to all wells (20 (i.1 of 5 mg/ml stock) daily and 2,3-3 H proline (NEN, Boston, MA) at lpCi/ml was added every other day. After 14 days of culture under these conditions, the cells were washed with sterile water 33 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. and then lysed with 0.025 M NH4 OH for 20 min. Wells were washed gently 3 times with sterile PBS. Treated or untreated test cells or medium without cells was then added to the radiolabled matrices. Samples (100 pi) were taken from the wells at various times and mixed with SafetySolve scintillation cocktail and counted in a scintillation counter to quantitate the degree of matrix degradation. RESULTS Zymography This method allows for simultaneous analysis of two important classes of matrix-degrading enzymes by including the respective substrates in the gels. Although substantial activity was observed, the presence of 1 fiM contortrostatin was found to have no effect on secreted gelatinases or plasminogen activators in MDA-MB-435 cells (Fig 3.1). Treatment with the 120 kDa fragment of fibronectin has been reported to stimulate production of matrix- degrading proteases in rabbit synovial fibroblasts [Huhtala et al., 1995], although we observed no changes in proteolytic activity upon treatment of MDA-MB-435 cells with this agent at 10 |J.g/ml. Proteolytic activity on these gels is visualized as clear bands on a dark staining background. These cells secrete both gelatinases and plasminogen activators since the lower bands observed comigrate with the tissue plasminogen activator (tPA) positive control and direct gelatin digestion 34 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. tPA — cntrl -------- FN120 CN CN/FN120 97kDa — MMP9 66kDa — Fig 3.1 Zymographic analysis of secreted matrix-degrading proteases by MDA-MB-435 cells. Conditioned medium from untreated cells (-), cells treated with 10 (ig/ml 120 kDa fibronectin fragment (FN120), 1 (jMcontortrostatin (CN) or both FN120 and CN simultaneously (CN/FN120) was concentrated and run on a poly aery amide gel containing gelatin and plasminogen. Proteolytic activity is visualized as clear bands. These results are representative of five independent experiments. 35 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. also appears to have occurred in the higher molecular weight band. Some plasminogen activation appears at molecular weights ranging from approximately 90 - 125 kDa in the positive control lane. The reasons for this observation are not clear, but may be the result of tPA aggregates present in the stock solution. The results shown are representative of several independent experiments conducted under various conditions. Also included in this series of experiments was the study of the effects of a monomeric disintegrin multisquamatin [Trikha et al., 1994b]. The results from all experiments lead to the conclusion that these disintegrins have no effect on the secretion of matrix-degrading enzymes in MDA-MB-435 cells. Matrix degradation Another approach was used to address the question of whether contortrostatin can affect the ability of tumor cells to digest the extracellular matrix. This assay is perhaps more relevant since it utilizes a biologically produced matrix. The results from these experiments support those obtained from the zymography studies, in that contortrostatin did not produce an effect on expression of matrix-degrading enzymes. Figure 3.2 shows that the presence of cells results in significant degradation of the matrix since serum-free medium alone has little effect. This time course experiment also shows that the degree of matrix degradation increases with time out to 24 h. The presence of contortrostatin, however does not have any inhibitory effect on the ability of the cells to degrade the matrix when cells are pretreated with contortrostatin for 20 min and remain in the presence of contortrostatin throughout the course of the experiment. 36 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. cpm 5000 a b c a bc abc a b c Fig 3.2 Time course of radiolabled extracellular matrix degradation by MDA-MB-435 cells. Values represent average counts per minute (cpm) in aliquots of equal volume from duplicate wells taken at the times indicated, (a) serum-free medium (SFM) (b) untreated cells in SFM (c) cells pretreated with 1|±M contortrostatin for 20 min in SFM. 37 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. The negative results obtained above raised the possibility that exogenous stimulation of protease activity was necessary in order for contortrostatin to show an inhibitory effect. To address this, we attempted to stimulate MDA-MB-435 cells with soluble vitronectin which has been shown to positively influence invasiveness and secretion of MMP-2 in human melanoma cells [Seftor et al., 1992]. We observed no enhancement of matrix degradation with vitronectin treatment nor any effect with contortrostatin treatment (Fig 3.3). These results are in agreement with previous findings in the zymography studies and support the conclusion that contortrostatin does not affect tumor cell invasion by inhibiting the production of matrix degrading enzymes. DISCUSSION By using two independent approaches to address the question of whether disintegrins will affect the production of matrix degrading proteases, we are able to conclude that contortrostatin does not significantly alter protease levels in MDA-MB-435 human breast cancer cells. Zymography is a valuable approach since it provides information regarding the species of matrix degrading protease present, as well as the relative quantities. This is evidenced by the molecular weights of the bands with proteolytic activity seen on the stained gels. From 38 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 4000 6 h o 3000 2000 1 0 0 0 lh 0 -L a b c d e f a b c d e f Fig 3.3 Effect of contortrostatin and/or vitronectin on degradation of a radiolabeled extracellular matrix by MDA-MB-435 cells. Values represent counts per minute (cpm) in aliquots of equal volume from wells taken at the times indicated, (a) serum-free medium (SFM) (b) 1 x 106 cells in SFM (c) 2 x 10s cells in SFM (d) 2 x 106 cells treated with 1 (iM contortostatin (e) 2 x 106 cells treated with 15 fig/ml vitronectin (f) 2 x 106 cells treated with both contortrostatin and vitronectin. 39 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. these experiments, MMP-9 appears to be the major gelatinase produced by MDA- MB-435 cells since the largest band of activity is approximately 86 kDa in size. MMP-9 is produced in a latent form of 92 kDa and is proteolytically activated to the 86 kDa form. MMP-9 seen in these experiments is present in one form only, which appears to be the activated from. The presence of an activated matrix- degrading enzyme is consistent with the invasive and metastatic characteristics of this cell line [Price et al., 1990]. However, there was no observed activity that would indicate the presence of MMP-2 in these experiments. MMP-2 is produced in the latent form at 72 kDa and the activated from migrates at 66 kDa. We did not see any bands in this size range during these experiments, but this observation does not exclude MMP-2 as a possible participant in the invasiveness of these cells. The design of the zymography experiments only allows for detection of secreted proteases and does not address the involvement of membrane-associated enzymes. MMP-2 is known to associate with the av(33 integrin at the surface of invasive tumor cells and angiogenic endothelial cells [Brooks et al., 1996], and thus would not be detected in these experiments. The association of MMP-2 with av{33 promotes invasion by concentrating proteolytic activity at the leading margin of the cell. MMP-2 in these studies was shown to be produced by the surrounding stroma and not by the invading cell itself. The interaction with ccvP3 is mediated by the C-terminus of MMP-2, since a fragment of MMP-2 which includes the C-terminal hemopexin-like domain (PEX) blocks binding of MMP-2 to avP3, reducing cell surface collagenolytic activity and inhibiting angiogenesis on the chick chorioallantoic membrane [Brooks et al., 1998]. Interestingly, a 40 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. naturally occurring form of PEX can be detected in vivo during tumor vascularization and in developmental retinal neovascularization, suggesting an important role for the regulation o f MMP-2 association with av|33. With the knowledge that contortrostatin binds the av|33 integrin, we directly explored the possibility that contortrostatin may block MMP-2 association with avp3 and contribute to the ability of contortrostatin to inhibit the invasivness of avP3- expressing cells. Although the results were not entirely conclusive, it appeared that contortrostatin had no effect on the association of these two molecules in a solid phase binding assay using purified proteins. The major confounding factor in these experiments was that MMP-2 showed significant binding to BSA and poly-lysine, which were used as negative controls, and thus any inhibitory effect on MMP-2 binding to av(33 by contortrostatin cannot conclusively be described as specific. Since MMP-2 lacks an RGD sequence, it might be expected that this enzyme binds to avf$3 at a different site than contortrostatin, and that contortrostatin would not affect the association of these two molecules. To further explore the role of membrane associated proteases, radiolabled matrix degradation assays were performed which is more biologically relevant since a cell-produced ECM is used as the experimental substrate. The conclusion from this series of experiments was similar to that of the zymography experiments: contortrostatin had no effect on the ability of breast cancer cells to break down the matrix. It has been demonstrated that, although tumor cells and endothelial cells localize MMP-2 to the leading edge, these cells do not produce the MMP-2 found 41 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. at their membrane. The enzyme is produced by the surrounding stroma and recruited by the invasive cell by binding to av(33 [Brooks et al., 1996]. The participation of exogenously-produced proteases was not examined in our cell- based studies and could be of importance in vivo. However, these experiments provide evidence that contortrostatin does not inhibit invasiveness by reducing the levels of matrix-degrading enzymes produced by tumor cells, a conclusion that has been corroborated in our laboratory by Dr. Stephanie Schmitmeier in U87 and T98G human glioma cells. These findings point to other another cellular process that could account for the anti-invasive and anti-angiogenic properties of contortrostatin. Since both invasion and angiogenesis directly depend on cell motility, inhibiting motility should be effective at halting these processes. These investigations into the ability of contortrostatin to affect tumor cell degradation of the extracellular matrix, and the negative results observed, directed us to examine the effect of contortrostatin on cellular motility. 42 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Chapter 4 Contortrostatin induces avP3-mediated tyrosine phosphorylation events in tumor cells 43 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. SUMMARY The effect of contortrostatin on integrin-mediated signaling in tumor cells was investigated by studying tyrosine phosphorylation events and activation of specific signaling molecules. It was observed that cells adhering to immobilized contortrostatin showed increased tyrosine phosphorylation with a pattern similar to cells adhering to vitronectin, as measured by anti-phosphotyrosine immunoblot. In separate studies using suspended cells, we found that at concentrations as low as 1 nM, soluble contortrostatin activates integrin signals leading to increased tyrosine phosphorylation of FAK and CAS, and that these signals are abolished by inhibiting Src family kinase activity. Using transfected 293 cells expressing specific integrins, we determined that contortrostatin-generated signals are mediated exclusively by the avp3 integrin. This observation was extended by showing that cells lacking av(J3 but expressing avps and a 5 p i, do not respond in this way to contortrostatin treatment. In cells expressing avP3, blocking contortrostatin binding with antibodies against avP3 completely abrogates contortrostatin signals. Monovalent disintegrins echistatin and flavoridin were incapable of affecting tyrosine phosphorylation alone, but when added simultaneously with contortrostatin, completely inhibited contortrostatin-initiated signals. It is proposed that the homodimeric nature of contortrostatin imparts the ability to crosslink avP3 integrins, causing Src activation and hyperphosphorylation of FAK and CAS. 44 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. INTRODUCTION Much of what is known about integrin signaling is the product of studies carried out in platelets [Huang et al., 1993], fibroblasts and epithelial cell lines [Giancotti and Ruoslahti, 1999; Schlaepfer and Hunter, 1998]. Significant progress has been made in the area of integrin-mediated signaling and the role of integrins in cell motility, but this area remains understudied in tumor cells. Focal adhesion kinase (FAK) is a non-receptor tyrosine kinase that localizes with integrins and is phosphorylated on tyrosine residues upon cellular adhesion to the extracellular matrix (ECM) [Burridge et al., 1992]. FAK plays an important role in allowing integrins, which lack catalytic activity, to convert extracellular stimuli into intracellular signals. Cell binding to the ECM causes integrin clustering and association of a number of cytoskeletally associated proteins into complexes known as focal adhesions [Burridge et al., 1988]. FAK associates with the cytoplasmic domain of the |3 integrin subunit and can undergo trans- autophosphorylation at specific tyrosine residues, creating a binding site for the Src family of protein tyrosine kinases [Schaller et al., 1994]. Src then can phosphorylate tyrosines in the FAK activation loop, resulting in full catalytic activity. The function of FAK is complex and is not yet fully appreciated, however a role of FAK in cell migration has recently been established. Overexpressing FAK in CHO cells leads to enhanced migration [Cary et al., 1996], and association of the adapter protein CAS via its SH3 domain with the proline-rich region of FAK (amino acids 712-718) mediates this enhancement 45 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. [Cary et al., 1998]. In addition to its SH3 domain, CAS contains multiple tyrosines that undergo phosphorylation in response to cell adhesion to the ECM, and this phosphorylation is dependent on FAK and Src [Vuori et al., 1996]. CAS is bound by SH2-domain containing proteins Crk and Nek. The CAS/Crk complex has been shown to serve as a “molecular switch” in the regulation of motility in carcinoma cells [Klemke et al., 1998]. Although the events downstream of the CAS/Crk complex are not well understood, it is known that this complex is involved in migration regulated by the GTPase Rac. Thus, FAK and CAS are important participants in integrin signaling and in the regulation of motility. Earlier investigations into the effects of disintegrins on platelet function demonstrated that, in addition to its ability to block aggregation, contortrostatin caused an increase in tyrosine phosphorylation of a subset of platelet proteins [Clark et al., 1994]. This activity distinguished contortrostatin from a monomeric disintegrin which blocked platelet aggregation but was unable to induce protein tyrosine phosphorylation. These findings led to the present investigations into the functional consequences of contortrostatin structure and the role of contortrostatin in regulating integrin-mediated signaling in tumor cells. It is shown in this chapter that contortrostatin has the unique ability to act as an integrin agonist by stimulating avp3-m ediated tyrosine phosphorylation of important signaling molecules in tumor cells, an activity not found in other disintegrins. It is hypothesized that this activity may be related to decreased cellular motility. 46 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. MATERIALS AND METHODS M a teria ls - MDA-MB-435 human mammary carcinoma cells were obtained from Janet Price (M.D. Anderson Cancer Center Houston, TX). T24 human bladder carcinoma cells were purchased from ATCC (Manassas, VA). 293 human embryonic kidney cells transfected with cDNA for |33 and $5 integrin subunits and parental 293 cells were provided by Jeffrey Smith (The Burnham Institute, La Jolla, CA)[Lin et al., 1997]. OVCAR-5 human ovarian carcinoma cells were from Thomas Hamilton (Fox Chase Cancer Center, Philadelphia, PA). Contortrostatin was purified from the venom of the southern copperhead snake (Agkistrodon contortrix contortrix) as described previously [Trikha et al., 1994a; Trikha et al., 1994b]. The monomeric disintegrins echistatin and flavoridin, and the general protease inhibitor cocktail used in lysis buffers were obtained from Sigma (St. Louis, MO). Vitronectin, fibronectin and Matrigel were purchased from Becton Dickinson (Bedford, MA). PP1 Src inhibitor was from Calbiochem (La Jolla, CA). Anti-phosphotyrosine monoclonal antibody (mAb) PY(99) was obtained from Santa Cruz Biotechnology (Santa Cruz, CA). Anti-FAK and anti- CAS mAbs were purchased from Transduction Laboratories (Lexington, KY) and 7E3 mAb was provided by Centocor (Malvern, PA). Cell culture, preparation and stimulation - T24 cells were maintained in RPMI 1640 medium containing 5% fetal bovine serum and MDA-MB-435, 293 and OVCAR-5 cells were maintained in Dulbecco’s modified Eagle’s medium 47 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. with 10% serum at 37°C in 5% C 0 2 . Cells were washed with phosphate-buffered saline (PBS) and starved in the appropriate serum-free medium for 6 h at 37°C. Cells were detached by brief treatment with 0.05% trypsin/0.02% EDTA in PBS and collected by centrifugation, resuspended in soybean trypsin inhibitor (1 mg/ml in serum-free medium), and washed in 2% bovine serum albumin/serum- free medium. Cells were maintained in suspension for 1 h in 2% bovine serum albumin/serum-free medium at 37°C with end-over-end agitation. Quiescent cells (3 x 106 /ml) were treated with disintegrins or other reagents while in suspension, or were allowed to adhere to fibronectin (20 fig/ml), vitronectin (6.75 jig/ml) contortrostatin (4.75 Jig/ml) or Matrigel (Becton Dickinson, Bedford, MA) diluted 1:100 with serum-free medium. Lysate preparation and immunoprecipitation — Suspended and adherent cells were washed twice with cold PBS and lysed in cold lysis buffer (50 mM Tris, pH 8.0, 150 mM NaCl, 1% Nonidet P-40, 0.5% deoxycholate, 0.1% SDS, protease inhibitor cocktail, 1 mM sodium pyrophosphate, 1 mM sodium orthovanadate, 50 mM sodium fluoride). After 10-15 min incubation on ice, insoluble material was rem oved by centrifugation at 14,000 RPM in a microcentrifuge for 15 min. Supernatants were collected and total protein concentrations standardized by the BCA protein assay (Pierce, Rockford, IL). Immunoprecipitation was carried out by incubating whole cell lysates (200 jig total protein) with 1.25 jig anti-FAK or anti-CAS mAb 4-6h at 4°C followed by 20 pi protein G-agarose overnight at 4°C. Immunoprecipitates were washed 4 times in lysis buffer without inhibitors and dissociated by adding SDS-PAGE 48 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. sample buffer and boiling 5 min. Whole cell lysates (30 Jig total protein) or im m unoprecipitates were resolved by 7.5% SD S-polyacrylam ide gel electrophoresis and transferred to nitrocellulose membranes. Immunoblotting — Membranes were blocked with 5% nonfat milk/Tris- buffered saline/0.1% Tween 20 (blocking buffer) 1 h at room temperature or overnight at 4°C. Primary antibody incubations were performed in blocking buffer for 1 h at room temperature. After washing in Tris-buffered saline/0.1% Tween 20, membranes were incubated with horseradish peroxidase-conjugated secondary antibody in blocking buffer 1 h at room temperature. Membranes were washed extensively. Immunoblots were developed using Super Signal® West Pico Chemiluminescent Substrate from Pierce. Densitometry was performed using UN-SCAN-ITtv software (Silk Scientific, Orem. UT)) RESULTS Cells adhering to immobilized contortrostatin show increased tyrosine phosphorylation of various proteins. Prior experiments had shown that tumor cells adhere well to immobilized contortrostatin [Markland and Zhou, 2000; Trikha et al., 1994a]. This adhesion was determined to be RGD-dependent since pretreatment of the cells with small RGD peptides blocked adhesion to 49 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. contortrostatin. In order to compare the intracellular effects after adhesion to the extracellular matrix proteins, fibronectin and vitronectin, to that of contortrostatin, we performed immunoblot analysis to measure phosphotyrosine levels after adhesion to these substrates. It was observed that adhesion to all three matrices resulted in enhanced tyrosine phosphorylation of similar sized proteins, with the most prominent bands at 120 — 140 kDa (Fig 4.1). Phosphorylation of these proteins increased steadily throughout the 180 min duration o f the experiment. However, adhesion to fibronectin resulted in substantial tyrosine phosphorylation of bands migrating to 70 - 75 kDa. These bands were significantly less pronounced in cells adhering to vitronectin or contortrostatin, suggesting that T24 cells use a vitronectin receptor when adhering to contortrostatin. Contortrostatin treatment of suspended tumor cells induces protein tyrosine phosphorylation - To investigate how tumor cells respond to contortrostatin while they are in suspension, T24 human bladder carcinoma cells were treated for various times with 500 nM soluble contortrostatin. The cells showed an increase in tyrosine phosphorylation of proteins with molecular weights from 120 — 140 kDa at 10 min and declining phosphorylation after 60 min. (Fig 4.2). Subtle changes were also observed in proteins of various sizes. Soluble fibronectin used at 20(lg/ml for 10 min failed to activate tyrosine phosphorylation. The reduction in tyrosine phosphorylation after 60 m in suggests the possibility that contortrostatin is internalized by the cells, an idea that is further explored in Chapter 7. 50 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. m in 0 10 30 60 180 Susp Fn Vn CN Fn CN Vn B min o io iso o 1 0 iso o 1 0 iso 219 kDa 1 0 8 67 Fig 4.1 Tyrosine phosphorylation in adherent T24 cells. Serum-starved T24 cells were allowed to adhere to the indicated substrate for the indicated time. (A) Kinetics of tyrosine phosphorylation are similar on all three substrates with evidence of some differences in overall pattern. (B) When equal amounts of total protein are loaded on the same gel direct comparisons came be made after immunoblot. An intense band appears at 180 min in cells on fibronectin (Fn) migrating to 70-75 kDa, but is essentially absent in cells adhering to contortrostatin (CN) or vitronectin (VN). A kDa Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. min o 10 Fn 30 60 180 219 108 67 Contortrostatin (500nM) Fig 4.2 Time course of contortrostatin-induced tyrosine phosphorylation events in suspended T24 cells. Serum-starved cells were treated with 500 nM soluble contortrostatin for the indicated times, and phosphotyrosine immunoblot was performed on whole cell lysates. To compare the effect of another integrin ligand, cells were treated with 20 pg/ml soluble fibronectin for 10 min. 52 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Using the time point of 10 min, the effect of dose was examined in T24 cells using a range of contortrostatin concentrations from 1 nM to 1000 nM. Maximal tyrosine phosphorylation levels of the 120 - 140 kDa proteins were observed at the lowest concentrations, while a decrease from basal tyrosine phosphorylation levels was observed for a protein of approximately 70 kDa (Fig 4.3). Another major band at 60 kDa showed increased phosphorylation at low concentrations of contortrostatin. Interestingly, all bands showed a reversal in their respective tyrosine phosphorylation patterns at higher concentrations. Although changes in tyrosine phosphorylation of the 60 and 70 kDa proteins was observed in other cell lines, this was not completely consistent, suggesting that changes in these bands may be cell type specific. However, the 120 - 140 kDa bands showed similar changes in phosphorylation consistently in several cell lines tested, including MDA-MB-435 human breast carcinoma cells (Fig 4.4). Based on these observations, future experiments were focused on the study of these 120 — 140 kDa proteins. Monomeric disintegrins, echistatin and flavoridin, do not induce tyrosine phosphorylation in tumor cells - In an effort to determine whether the ability to stimulate tyrosine phosphorylation in tumor cells was related to the homodimeric structure of contortrostatin, we tested two well-characterized monomeric disintegrins, echistatin [Gan et al., 1988] and flavoridin [Niewiarowski et al., 1994], for their ability to affect integrin signaling in suspended MDA-MB-435 cells. In sharp contrast to the observed effects of contortrostatin, the monomeric 53 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. CN (nM) 0 1 10 100 1000 120kDa — 65kDa — arbitrary units - b p-Tyr c 40 20- I Jl ■ ■ 01 ■ 111 I B Fig 4.3 Phosphotyrosine levels in tumor cells after contortrostatin treatment. T24 cells were treated in suspension with the indicated concentrations of soluble contortrostatin (CN) for 10 min. Whole cell lysates (30|j.g total protein) were subjected to SDS-PAGE and immunoblotting with antibody against phosphotyrosine (PY99) as described in Methods. Densitometry plots of the indicated bands are shown. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. CN (nM) 0 1 10 100 1000 120kDa— blot:P-Tyr Fig 4.4 Tyrosine phosphorylation induced by contortrostatin in MDA- MB-435 cells. MDA-MB-435 cells were treated with the indicated concentrations of contortrostatin (CN) for 10 min while in suspension and whole cell lysates were immunoblotted for phosphotyrosine. These cells showed changes in the 120-140 kDa bands only, leading to a focus on these proteins in future studies. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. disintegrins alone had no effect on tyrosine phosphorylation at concentrations up to 1 jiM. However, when echistatin or flavoridin (1 (iM) were added simultaneously with contortrostatin (10 nM), contortrostatin-induced tyrosine phosphorylation was completely abrogated (Fig 4.5). These findings indicate that the monomeric disintegrins competitively inhibit contortrostatin binding to specific signal-generating integrins. Contortrostatin-induced signaling events are mediated by the av(33 integrin The integrin-binding specificities of echistatin and flavoridin have been determined previously [Niewiarowski et al., 1994; Pfaff et al., 1994]. Both monomeric disintegrins interact with allbp3, av|33 and a 5 p l. We have shown that contortrostatin binds these same integrins [Markland and Zhou, 2000; Trikha et al., 1994a; Trikha et al., 1994b], as well as avP5 [Zhou et al., 2000c]. With the knowledge that allbp3 is not expressed on MDA-MB-435 cells, as shown by a lack of staining with a specific anti-aIIbP3 mAb 10E5, this implicated avP3, oc5pl or both, in contortrostatin-induced signaling events, based on the monomeric disintegrin versus contortrostatin observations presented above. In order to determine which receptor(s) was involved, we employed transfected 293 cell lines with specific integrin profiles [Lin et al., 1997]. The parental 293 cells express the a v subunit but have no detectable avp3 and only trace levels of avp5 expression. Cells transfected with cDNAs encoding the P3 or P5 integrin subunits show significant levels of the avP3 or avP5 heterodimers, respectively [Lin et al., 1997]. Integrin expression was confirmed in our laboratory by flow 56 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. cytometry. When these cell lines are treated with contortrostatin using established methods, only the avP 3-expressing cells show the robust induction of tyrosine phosphorylation observed in other cell types (Fig 4.6A). Importantly, the proteins undergoing tyrosine phosphorylation in the p3 transfected cells are in the 120- 140kDa range, the same molecular weights as those observed in other cell lines tested. These findings demonstrate the involvement of avP3 in contortrostatin signaling but do not directly address the potential contribution of a 5 p i, since echistatin and flavoridin are known to bind a 5 p l. This possibility was ruled out by studies using 7E3, a mAb generated against aIIbP3 that has equal affinity for avp3 [Tam et al., 1998]. Contortrostatin-induced tyrosine phosphorylation was completely blocked when MDA-MB-435 cells, which express avP3, avp5 and a 5 p i [Chandrasekaran et al., 1999; Wong et al., 1998], are treated simultaneously with 1 fiM 7E3 and 10 nM contortrostatin for 10 min (Fig 4.6B, lower panel). This result was duplicated in T24 cells (Fig 4.6B, upper panel), providing convincing evidence that avP3 is solely responsible for mediating contortrostatin- induced tyrosine phosphorylation. Contortrostatin binding to avP3 results in tyrosine phosphorylation of CAS and FAK - In order to identify the specific proteins that are tyrosine phosphorylated in response to contortrostatin treatment, lysates prepared from contortrostatin-treated cells were subjected to immunoprecipitation with CAS or FAK monoclonal antibodies followed by anti-phosphotyrosine immunoblotting. CAS and FAK were selected as likely candidates based on the similarity of their 57 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. CN (lOnM) echi (nM ) 0 1 100 1000 0 100 1000 120kDa— blot:P-Tyr CN (lOnM) flavo (nM ) 0 1 1 0 0 1 0 0 0 0 1 0 0 1 0 0 0 120kDa- v p blot:P-Tyr Fig 4.5 Effects of monomeric disintegrins on tyrosine phosphorylation in tumor cells. Suspended MDA-MB-435 cells were treated with the indicated concentrations of echistatin (upper panel) or flavoridin (lower panel) or were treated simultaneously with monomeric disintegrins and 10 nM contortrostatin (CN) as indicated for 10 min. Lysates were analyzed for phosphotyrosine content by immunoblot as described in Experimental Procedures. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. A |3 5 293 P3 293 CN(nM) 0 1 10 100 1000 0 1 10 100 1000 120kDa— B CN (lOnM) 7E3(nM) 0 0 1 10 100 1000 120kDa- CN (lOnM) 7E3(nM) 0 1000 120kDa— Fig 4.6 Contortrostatin-induced tyrosine phosphorylation is mediated by the av(33 integrin. A, Suspended 293 cells expressing either av|35 ((35 293) or av|33 ((33 293) were treated for 10 min with the indicated concentrations of contortrostatin (CN). (33 293 cells responded to contortrostatin treatment with increased tyrosine phosphorylation while (35 293 cells showed no response. Lysates were analyzed for phosphotyrosine content by immunoblot as described in Methods. B, T24 cells (upper panel) or MDA-MB-435 cells (lower panel) were treated simultaneously for 10 min with contortrostatin (CN) and the indicated concentrations of the anti-avp3 mAb 7E3. Whole cell lysates were analyzed by phosphotyrosine im m unoblot. 59 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. molecular weights (130 and 125 kDa, respectively) with those observed on the anti-phosphotyrosine immunoblots with whole cell lysates. It was found that CAS and FAK are both tyrosine phosphorylated in response to contortrostatin treatment (Fig 4.7), and immunoprecipitated CAS and FAK co-migrate with the major bands observed after anti-phosphotyrosine immunoblot with whole cell lysates, indicating that these are the major proteins phosphorylated by contortrostatin treatment. Src family kinase activity is necessary for contortrostatin-induced tyrosine phosphorylation events - The Src family of tyrosine kinases are known to play a central role in integrin signaling. In order to determine if the Src kinases participate in transmitting contortrostatin-induced signals from avp3, T24 cells were pretreated in suspension for 30 min with the Src family inhibitor, PP1 [Hanke et al., 1996], prior to stimulation with 10 nM contortrostatin. As shown in Figure 4.8 (upper panel), PP1 demonstrates a dose-dependent inhibition of tyrosine phosphorylation with complete elimination of contortrostatin-induced signals at a concentration of 100 fiM. Similar results were obtained following PP1 treatment of MDA-MB-435 cells, although complete inhibition was achieved at 10 pM PP1 (Fig 4.8, lower panel). These findings implicate the Src fa m ily kinases as being involved in integrin signaling stimulated by contortrostatin. Contortrostatin-induced tyrosine phosphorylation is independent of cellular adhesion - In order to determine if contortrostatin is able affect tyrosine 60 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. phosphorylation in adherent cells where integrin ligation and cytoskeletal structure exist, T24 cells were pretreated with contortrostatin before allowing them to adhere to Matrigel-coated plates. It should be noted that contortrostatin does not significantly inhibit cellular binding to laminin. Therefore, contortrostatin does not inhibit cellular binding to Matrigel, which is rich in laminin. Phosphotyrosine immunoblotting revealed that these cells showed modest increases in tyrosine phosphorylation after adhesion to Matrigel, but contortrostatin treatment of adherent cells causes a significant additional increase in these signals, including the 120-140 kDa bands shown to contain CAS and FAK (Fig 4.9). It was again observed that high concentrations of contortrostatin (1000 nM) generated reduced levels of tyrosine phosphorylation. In similar experiments, T24 cells were allowed to adhere to Matrigel for 30 min prior to treatment with contortrostatin. Following an additional 30 min incubation on M atrigel in the presence of contortrostatin, cells showed similar increased tyrosine phosphorylation o f the 120-140 kDa bands. Thus, contortrostatin- induction of tyrosine phosphorylation can occur in adherent cells, in the presence of stimuli from ECM proteins, as well as in non-adherent cells. 61 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. CN(nM) 0 1 1 0 100 2000 blot:P-Tyr blot:FAK FAK IP CN(nM) 0 1 10 100 2000 blot:P-Tyr blot:CAS CAS IP Fig 4.7 Contortrostatin treatment causes tyrosine phosphorylation of CAS and FAK. Lysates from MDA-MB-435 cells were immunoprecipitated with antibodies specific for CAS or FAK followed by anti-phosphotyrosine immunoblotting, or were immunoblotted with the same antibody used for immunoprecipitation to demonstrate equal loading. FAK (upper panel) and CAS Qower panel) are maximally tyrosine phosphorylated at lOnM contortrostatin (CN), which corresponds to the peak in tyrosine phosphorylation with varying contortrostatin concentrations (Figure 4.4). The reason for the appearance of two bands on the CAS blots is not known, but results are similar to those of the manufacturer. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. CN (lOnM) P P l Q l M ) 0 0 1 1 0 1 0 0 d m s o 120kDa— P P l ( l i M ) 120kDa— Fig 4.8 Involvement of Src family kinases in contortrostatin-induced tyrosine phosphorylation. T24 cells (upper panel) or MDA-MB-435 cells Gower panel) were pretreated with the indicated concentrations of the Src family inhibitor PP1 for 30 min prior to stimulation with lOnM contortrostatin (CN). Lysates were prepared and analyzed by anti-phosphotyrosine immunoblot. Stock solutions of PP1 are prepared in DMSO, but DMSO alone had no effect on contortrostatin-induced tyrosine phosphorylation (upper panel). 63 CN (lOnM) v* blot:P-Tyr Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. CN(nM 120kDa- 100 Matrigel C3 I blot:P-Tyr Fig 4.9 Contortrostatin-induced tyrosine phosphorylation is independent of cellular adhesion. T24 cells were pretreated with the indicated concentrations of contortrostatin (CN) for 5 min prior to addition to Matrigel-coated plates. Control cells were maintained in suspension. Lysates were prepared after incubating cells on Matrigel for 20 min and analyzed by anti-phosphotyrosine immunoblot (upper panel). The lower panel shows the relative intensity of the corresponding bands as determined by densitometry. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. DISCUSSION Since their discovery, disintegrins have been have been studied almost exclusively for their ability to block the function of various integrins. Disintegrins have been used extensively to investigate the function of allb(33 on platelets, and more recent work has been conducted with endothelial cells in the study of the role of avP3 in angiogenesis [Kang et al., 1999; Sheu et al., 1997; Yeh et al., 1998]. In these reports, the antiangiogenic effects of the disintegrins are characterized as a function of their ability to block avp3, an integrin which has been shown to be involved in induction of endothelial cell apoptosis [Brooks et al., 1994b]. The direct effects of disintegrins on integrin-mediated signaling remains largely unstudied, however. Echistatin has been shown to cause a decrease in FAK phosphorylation and disassembly of focal adhesions prior to melanoma cell detachment from fibronectin [Staiano et al., 1997]. In contrast to these descriptions of disintegrins as passive integrin-blocking agents, the present work shows that the disintegrin, contortrostatin, has a structure that enables it to function as an integrin agonist, initiating signals that are usually observed only after cellular binding to natural ECM ligands or artificial crosslinking with anti- integrin antibodies. Our studies suggests that contortrostatin actively regulates the function avP3 in tumor cells. The effects of contortrostatin on integrin signaling have been studied previously in platelets where it was found that the contortrostatin dimer and the monomeric disintegrin, multisquamatin, both inhibited a lib |3 3-mediated platelet aggregation and aggregation-dependent 65 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. tyrosine phosphorylation of numerous proteins including FAK [Clark et al., 1994]. A distinct set of platelet proteins have been shown to become tyrosine phosphorylated upon allb p 3 crosslinking with fibrinogen or al3b(33 antibodies [Huang et al., 1993]. Contortrostatin was shown to activate tyrosine phosphorylation of these same proteins, presumably by virtue of its dimeric structure. However, there are notable differences between the present study and the work performed in platelets, particularly the tyrosine phosphorylation status of FAK following contortrostatin treatment. This discrepancy is likely the result of differences in the regulatory mechanisms of FAK tyrosine phosphorylation in platelets and tumor cells. In platelets, FAK phosphorylation is dependent on platelet aggregation, and does not occur after fibrinogen binding to al3b|33 alone, indicating that events occurring during platelet-platelet aggregation, and not integrin crosslinking, are critical in regulating FAK phosphorylation [Lipfert et al., 1992]. In contrast, studies in fibroblasts show that FAK undergoes tyrosine phosphorylation after integrin aggregation with non-inhibitory mAbs in the absence of integrin ligation [Miyamoto et al., 1995a]. Thus, simple dimerization of integrins is sufficient to cause FAK phosphorylation in fibroblasts, and this mechanism is expected to function through av{33 crosslinking during contortrostatin-induced signaling in tumor cells. It is attractive to speculate that contortrostatin might have a geometry that allows it to specifically crosslink (33 integrins and not others. Perhaps the distance spanned by the two RGD motifs is spatially complimentary to a pair of allb(33 or av(33 integrins in close proximity, but other integrins are sterically prevented from being dimerized by 66 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. contortrostatin. In a recent report describing the structure of the heterodimeric disintegrin EMF-10, a molecular model is offered that suggests the two integrin- binding regions of this disintegrin are oriented 180° relative to each other and are separated by a distance of 71.4 angstroms [Calvete et al., 2000]. This model would seem incompatible with any dimerizing effect since the integrin-binding motifs are in opposite orientations. However, this model is based on the NMR structures of kistrin and flavoridin and thus may not accurately depict the true 3- dimensional structure of this heterodimer. Investigations into the structure of contortrostatin will provide valuable information into the possible mechanism of the interaction between this disintegrin and the integrins to which it binds. From the observations presented in this report, we propose that each subunit of contortrostatin binds to a separate avP3 integrin, bringing the integrins into close proximity allowing rra/ 1 5 -autophosphorylation of integrin-associated FAK at tyrosine 397, creating a binding site for Src [Schaller et al., 1994]. Binding of Src leads to further tyrosine phosphorylation of FAK [Schlaepfer and Hunter, 1996] and to Src-mediated phosphorylation of CAS [Vuori et al., 1996]. We observed that at higher concentrations of contortrostatin (1 pM), the levels of tyrosine phosphorylation decreased. This apparent paradox might be explained when talcing into account two possible binding orientations of the contortrostatin dimer. At low concentrations, each of the two subunits is bound to a different integrin on a single cell, bringing them into close proximity and allowing for the initiation of signaling cascades. At high concentrations, only one of the two subunits is bound to a single cell because of competition between subunits, and 67 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. integrin clustering will not occur. Under these conditions, contortrostatin acts as a monovalent ligand, and will not initiate tyrosine phosphorylation. Therefore, when contortrostatin concentrations are above saturating levels, the “monomeric” binding orientation will dominate and a reduction in tyrosine phosphorylation will be observed. Other investigators used soluble monomeric and multimeric vitronectin and examined their ability to differentially regulate tyrosine phosphorylation in bovine pulmonary artery endothelial cells [Bhattacharya et al., 1995]. M ultimeric vitronectin was shown to mediate enhanced tyrosine phosphorylation of several proteins, including FAK, yet monomeric vitronectin did not produce this effect. This finding was confirmed in our studies with MDA- MB-435 cells in which monomeric vitronectin failed to stimulate tyrosine phosphorylation at 10 (ig/ml (data not shown). These studies support our findings that contortrostatin possesses the ability to initiate avp3-mediated signaling by crosslinking integrins at the surface of tumor cells, resulting in dramatic stimulation of tyrosine phosphorylation of the important signaling molecules FAK and CAS. Attention has been directed to these two molecules recently with respect to their roles in cell motility, and their physiological importance has been highlighted in a report investigating the role of a protein tyrosine phosphatase, PTP-PEST [Angers-Loustau et al., 1999]. Fibroblasts lacking expression of PTP- PEST show severe defects in motility. Biochemical and immunocytochemical analysis revealed that this defect was due in part to a constitutive increase in tyrosine phosphorylation of CAS, FAK, and paxillin and to an increase in the number of focal adhesions present. CAS is a known PTP-PEST substrate and 68 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. paxillin has been shown to associate with PTP-PEST in vitro. It is not clear why FAK is hyperphosphorylated in these cells, but the effects of FAK on cell migration are known to be dependent on CAS [Cary et al., 1998]. Although PTP- PEST and contortrostatin produce their effects through entirely different mechanisms, the end result of CAS and FAK hyperphosphorylation is very similar. These similarities suggest a functional consequence of tumor cell treatment with contortrostatin where a disruption, through av|33-mediated hyperphosphorylation of CAS and FAK, occurs in the delicate and highly regulated machinery that drives cell motility. The involvement of av(33 and FAK in tumor cell motility is emphasized in a recent report showing that the presence of av(33 on human prostatic carcinoma cells generated a migratory phenotype that is modulated by pathways involving FAK [Zheng et al., 1999], In addition to its involvement in motility, av|J3 has been shown to be critical to other events in tumor progression, including localization of MMP-2 and degradation of the surrounding matrix [Brooks et al., 1996] and in tumor-induced angiogenesis [Brooks et al., 1994a; Brooks et al., 1994b]. The notion of inappropriate tyrosine phosphorylation and activation of signaling molecules resulting in inhibition of cell migration has been suggested previously [Claesson-Welsh et al., 1998]. Although these effects were reported to be RGD-independent, the finding that FAK phosphorylation and overall increase in cellular tyrosine phosphorylation is correlated with inhibition of motility supports the hypothesis that we offer here. Conclusive proof that the homodimeric structure of contortrostatin is what imparts this activity, and not some other unidentified properties, can be provided through 69 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. use of a monomeric form o f contortrostatin. At present, this form of contortrostatin is not available. In conclusion, this work identifies activity novel to the disintegrin family through which integrin signaling can be modulated in tumor cells. This activity appears to be unique to contortrostatin, as three other disintegrins have been found to lack the ability to stimulate tyrosine phosphorylation [Clark et al., 1994]. Our report identifies contortrostatin as a useful reagent for the further study of av(33 function, and identifies a novel integrin-mediated mechanism that may negatively effect tumor cell motility. We propose that the combined effects of blocking the binding of av(35, a5(3l and avP3 to the ECM and the initiation of inappropriate signals leading to hyperphosphorylation of critical signaling molecules can lead to immobilization of otherwise motile and invasive tumor cells, a subject that is addressed in Chapter 6. 70 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Chapter 5 Contortrostatin activates ERK2 and tyrosine phosphorylation events via distinct pathways 71 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. SUMMARY It is demonstrated in this chapter that cells adhering to contortrostatin show transient increases in activation of ERK2. The kinetics and degree of activation are similar to cells adhering to fibronectin or vitronectin. In Chapter 4 it was shown that contortrostatin induces tyrosine phosphorylation in tumor cells. Contortrostatin is shown here to stimulate activation of ERK2 in suspended cells, but this activation follows a different dose-response pattern than contortrostatin- induced tyrosine phosphorylation. With the knowledge that contortrostatin induces tyrosine phosphorylation via avp3, we explored the effects of the av(33- blocking antibody, 7E3, on contortrostatin-stimulated ERK2 activation. While 7E3 completely blocks the effect of contortrostatin on tyrosine phosphorylation, this antibody had no effect on the activation of ERK2. In cells lacking expression of av(33, tyrosine phosphorylation was unaffected by contortrostatin treatment, but ERK2 was activated. This is strong evidence that contortrostatin is regulating tyrosine phosphorylation events and ERK2 activation via separate pathways and through different integrin receptors. 72 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. INTRODUCTION Integrins not only provide cells with anchorage to the extracellular matrix, but also transduce signals that influence cell survival, division, differentiation and motility [Giancotti and Ruoslahti, 1999]. The MAP kinase family of serine/threonine kinases is an important group of molecules regulated by integrins [Morino et al., 1995]. The MAP kinases lie on the well-characterized Ras pathway and have been shown to be involved in regulating a multitude of cellular events. Growth factors initiate signaling to MAP kinase by binding to receptor tyrosine kinases at the cell surface. The receptors undergo dimerization and tyrosine phosphorylation at specific residues which create binding sites for SH2 domain-containing adapter protein Grb2, which can recruit SOS, a guanine nucleotide exchange factor that causes exchange of GDP for GTP on Ras [Egan et al., 1993]. GTP-bound Ras is the active form and binds directly to Raf, a serine/threonine kinase. Raf phosphorylates MEK, a dual specificity kinase which binds and phosphorylates ERK, a member of the MAP kinase family, on threonine 183 and tyrosine 185, resulting in its full activation [Canagarajah et al., 1997]. ERK is found in two isoforms, ERK1 (44 kDa) and ERK2 (42 kDa) both of which affect gene expression by directly phosphorylating transcription factors [Hill and Treisman, 1995]. ERK2 can also directly regulate the function of myosin light chain kinase (MLCK), an enzyme tightly associated with the control of cellular motility [Klemke et al., 1997]. Myosins are ATPases that are activated by actin and are capable of translational movement along actin filaments, which is 73 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. important in generating force needed for cell motility. Myosin II is largely responsible for performing this role in non-muscle cells and is composed of two heavy chains (200 kDa) and two sets of light chains (16-20 kDa) [Jay et al., 1995]. The function of myosin II is regulated by phosphorylation of the light chains by MLCK. Thus, ERK2 regulates cell motility indirectly through its ability to influence the activity of MLCK. Integrins are able to regulate the activity of the MAP kinases by sharing components of the Ras pathway [Chen et al., 1994]. Integrin dimerization causes frans-autophosphorylation of FAK on a tyrosine residue that is recognized by the SH2 domain of Src. Src phosphorylates another tyrosine residue on FAK which becomes a binding site for Grb2, which propagates downstream signals via the Ras cascade [Schlaepfer and Hunter, 1996]. Grb2 can also bind She, a protein tyrosine phosphorylated by FAK [Schlaepfer et al., 1998]. Thus, the MAP kinases are situated at a point of convergence of multiple signaling pathways where they participate in the regulation of important cellular processes. This chapter describes the influence of contortrostatin on the activation of the MAP kinase, ERK2, and provides evidence for a pathway leading to activation of this molecule that is distinct from the contortrostatin-mediated, av(33-dependent pathway involving tyrosine phosphorylation of FAK and CAS. Contortrostatin-induced activation of ERK2 may be a separate but cooperative event negatively affecting cell motility and/or cell growth, and sheds new light on the activities of the disintegrin family. 74 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. MATERIALS AND METHODS Materials and Methods— The activated form of ERK2 was detected using the Anti-ACTTVE® MAPK polyclonal antibody from Promega (Madison, WI). See Chapter 4 for methods and other materials used. RESULTS Tumor cell adhesion to immobilized contortrostatin results in enhanced activation of ERK2. To test how T24 cells respond when adhering to immobilized contortrostatin we compared ERK activation in these cells to cells adhering to fibronectin or vitronectin. It was observed that T24 cells express much higher levels of ERK2 compared to ERK1. Lengthy exposure of blots probed for activated ERK1/2 revealed a faint band corresponding to ERK1 (44 kDa) that was slightly larger than the major band (ERK2, 42 kDa). It was found that T24 cells responded with ERK2 activation in a similar manner irrespective of 75 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. the substrate to which they adhered (Fig 5.1). Cells bound to contortrostatin, as well as fibronectin and vitronectin, showed a peak in ERK2 activation at 30 min and declining activation at 180 min. These findings are consistent with other reports indicating that ERK2 activation following adhesion to extracellular matrix proteins is transient [Chen et al., 1994; Short et al., 1998] and suggests that cells are unable to distinguish contortrostatin from other RGD-containing ligands under these conditions. Contortrostatin treatm ent of suspended T24 cells activates tyrosine phosphorylation and ERK2 with distinct dose-response patterns. Quiescent cells were treated for 10 min while in suspension with various concentrations of contortrostatin and protein tyrosine phosphorylation content was measured by anti-phosphotyrosine immunoblot. As previously reported [Ritter et al., 2000b] it was observed that contortrostatin-treated cells displayed dramatic activation of tyrosine phosphorylation of proteins in the size range of 120 — 140 kDa (Fig 5.2). W ith higher concentrations of contortrostatin, a reduction in tyrosine phosphorylation levels was observed. ERK2 was also found to be modestly activated in response to contortrostatin treatment, but the dose response was distinct from that of the tyrosine phosphorylation events, suggesting that these two phenomena are regulated by different pathways. Contortrostatin-induced alterations in tyrosine phosphorylation, but not ERK2 activation, are mediated by av(33. Our previous work had shown that 76 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. min o 10 30 60 180 Susp A active ERK2 ERK2 B active ERK2 ERK2 active ERK2 ERK2 F ig 5 .1 ERK2 activation in adherent T24 cells. ERK2 is transiently activated in T24 cells adhering to immobilized contortrostatin (5 flg/ml) (A) as well as in cells adhering to fibronectin (20 |J.g/ml) (B) or vitronectin (7 |ig/ml) (C). Whole cell lysates were probed with an antibody that specifically recognizes the activated form of ERK1/2. ERK2 (42 kDa) is the isoform predominantly expressed in T24 cells. Time 0 represents cell lysates prior to adhesion to the various substrates. Lysates from cells held in suspension for 180 min (Susp) are shown. Corresponding loading control blots are shown in the lower halves of each panel. Experiments were repeated to confirm results. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. CN(nM ) 0 1 10 102 103 120 kDa 42 kDa— m m g m m M m m m m x s P-Tyr active ERK2 Fig 5.2 Dose response of contortrostatin-induced signaling events in suspended T24 cells. Whole cell lysates were immunoblotted for phosphotyrosine (P-Tyr) or activated ERK2 after treatment with the indicated concentrations of contortrostatin (CN) for 10 min at 37°C. Since protein levels are not expected to change during the short 10 min incubation, ERK loading control blots are excluded. A densitometry plot with arbitrary units is shown for ERK2. Results shown are representative of three similar experiments. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. the avP3 integrin is the exclusive mediator of tyrosine phosphorylation events stimulated by contortrostatin [Ritter et al., 2000b]. In order to determine if ocv|33 is also mediating contortrostatin-induced ERK2 activation, T24 cells were pretreated with various concentrations of 7E3, a monoclonal antibody generated against allbp3 that has equal reactivity to avp3 [Tam et al., 1998]. T24 cells do not express aIIbP3 as shown by a lack of staining with a 10E5, a specific antibody for ocIIbP3 (data not shown). We found that 7E3 at 1000 nM was able to completely block the tyrosine phosphorylation effects of 10 nM contortrostatin treatment (Fig 5.3). The existence of multiple pathways regulated by contortrostatin was again evidenced by the observation that 7E3 had no effect on ERK2 activation in the same lysates. The involvement of the Src family of tyrosine kinases was explored in these experiments through use of a specific Src family inhibitor, PP1 [Hanke et al., 1996], Inhibition of Src family kinase activity with 10 pM PP1 blocked the effects of 10 nM contortrostatin on tyrosine phosphorylation in these cells. From this result it can be concluded that contortrostatin-induced changes in tyrosine phosphorylation levels depend on the activity of Src family kinases. Interestingly, contortrostatin activation of ERK2 was also completely abrogated in the presence of PPL Thus, these results support earlier findings that the avP3 integrin mediates contortrostatin-induced tyrosine phosphorylation [Ritter et al., 2000b] and indicate that contortrostatin activation of ERK2 is independent of this receptor. Both pathways appear to require activation of the Src family kinases since inhibition of these enzymes completely 79 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. CN (lOnM) 7E3(nM) 0 0 1 10 102 103 PP1 120kDa— 42 kDa 80 40 0 . Mi l l P-Tyr active ERK2 Fig 5.3 Contortrostatin-induced tyrosine phosphorylation is mediated by the av(33 integrin, but ERK2 activation is not. Whole cell lysates were probed for phosphotyrosine (P-Tyr) content or activated ERK2. Cells were pretreated with anti-av(33 mAb (7E3) or the Src family kinase inhibitor (PP1) for 10 min prior to additional 10 min incubation at 37°C with 10 nM contortrostatin (CN). A densitometry plot with arbitrary units is shown for ERK2 activation. Results are representative of two similar experiments. 80 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. eliminates contortrostatin-induced effects on both tyrosine phosphorylation and ERK2 activation. Contortrostatin activates ERK2 in av(33-negative cells, but tyrosine phosphorylation is unaffected. Further investigations into the different pathways regulated by contortrostatin were performed in OVCAR-5 human ovarian cancer cells. These cells were found to lack expression of the avp3 integrin, but do express av(35, an integrin to which contortrostatin has been shown to bind [Zhou et al., 2000c]. This provided another opportunity to distinguish the role of avp3 from other integrins in mediating contortrostatin signals. When contortrostatin-treated OVCAR-5 cell lysates were immunoblotted for phosphotyrosine, little change was observed (Fig 5.4), clearly contrasting with the effects of contortrostatin on tyrosine phosphorylation in avP3-expressing cells (Fig 5.2). However, when ERK2 activation was measured, contortrostatin was found to cause a modest, dose-dependent activation of ERK2 in these cells. It has been reported that signaling mediated by the avP 5 integrin is dependent on the activity of protein kinase C (PKC) [Lewis et al., 1996]. With the knowledge that contortrostatin binds avP5, and since PKC is likely inactive under the serum-free conditions used here, we pretreated cells for 30 min with phorbol ester (PMA) to determine if PKC activation would effect contortrostatin- induced tyrosine phosphorylation or ERK2 activation. PMA pretreatment (10 nM) had no enhancing effect on tyrosine phosphorylation levels in the presence of contortrostatin (Fig 5.4). When the same lysates are examined for ERK2 81 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. activation, PMA treatment alone caused dramatically increased activation, but had no effect on ERK2 activation in contortrostatin-treated cells. Thus the ability of contortrostatin to affect tyrosine phosphorylation events and ERK2 activation appears to be independent of the activation status of PKC. 82 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. PMA CN (nM) 0 10 103 10 0 120kDa— P-Tyr 42 kDa active ERK2 Fig 5.4 Contortrostatin activates ERK2 in av|33-negative cells but fails to stimulate tyrosine phosphorylation (P-Tyr). OVCAR-5 cells, which lack expression of av(33, were treated with the indicated concentrations of contortrostatin (CN) for 10 nun at 37°C, or were pretreated with the PKC activator, PMA at 10 nM for 30 min prior to treatment with contortrostatin. A densitometry plot with arbitrary units is shown for ERK2 activation. Results are representative of two similar experiments. 83 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Low CN High CN av(33 Iff! av|35 ? ^ fjftv Fig 5.5 A hypothetical model depicting two mechanisms leading to induction of tyrosine phosphorylation and activation of ERK2. Upper panel: at low concentrations, contortrostatin (CN) is able to crosslink av(33 integrins and initiate a pathway leading to tyrosine phosphorylation (P-Tyr) events. At high concentrations, each subunit of the dimer competes for binding sites, leading to a distinct binding orientation that fails to initiate tyrosine phosphorylation. Lower panel: it is proposed that the av(35 (or possibly 05(31) integrin mediates contortrostatin-induced activation of ERK2 (P-ERK2). This model depicts ERK2 activation as being independent of integrin crosslinking, where increased activation coincides with increased dose. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. DISCUSSION The study of integrins as signaling molecules has gained the interest of researches in the fields of tumor biology and angiogenesis. Work in our laboratory on the antitumor and antiangiogenic properties of contortrostatin suggested the possibility that this disintegrin would have effects on integrin signaling, contributing to its inhibition of these processes. Indeed, contortrostatin does effect the levels of tyrosine phosphorylated proteins within tumor cells, and it has been shown that the av(33 integrin is solely responsible for mediating these changes [Ritter et al., 2000b]. In an earlier study conducted in platelets, contortrostatin treatment was shown to have ccllb(33-mediated effects on tyrosine phosphorylation distinct from that of a monomeric disintegrin [Clark et al., 1994]. The most notable difference was that contortrostatin treatment caused an increase in tyrosine phosphorylation of several proteins, while the monomeric disintegrin did not, suggesting that the unique homodimeric structure of contortrostatin imbues it with added function. Other studies have shown that contortrostatin activates the tyrosine phosphorylation of FAK and CAS in tumor cells [Ritter et al., 2000b]. Recent reports have demonstrated that the activity of ERK2 is regulated by integrin signaling [Chen et al., 1994; Morino et al., 1995]. Our study of this important signaling molecule revealed that contortrostatin can positively affect ERK2 activation, but appears to do so through a pathway different than the pathway regulating tyrosine phosphorylation. These findings further distinguish 85 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. contortrostatin as an active regulator of integrin function and show that it acts as more than a passive integrin-blocking agent. The downstream events that may be effected by contortrostatin are numerous and include disruptions in motility and cell cycle progression. Other work in our laboratory has shown that contortrostatin inhibits the migration of tumor cells [Markland and Zhou, 2000; Ritter et al., 2000a; Ritter et al., 2000b; Zhou et al., 2000c] and is an inhibitor of angiogenesis [Zhou et al., 2000b]. Activation of ERK2 is often associated with enhanced MLCK activity and increased migration [Klemke et al., 1997]. Although it is possible that contortrostatin-induced ERK2 activation has little or no physiological significance, our findings suggest that activation of ERK2 in this context may have a negative effect on migration. A possible reconciliation for this may be found when considering that precise temporal and spatial regulation of ERK2 activity might be necessary for its ability to function as a positive regulator of motility. A recent report supports this idea by demonstrating that activated ERK is targeted to focal adhesions by integrin engagement or v-Src [Fincham et al., 2000]. Inhibiting the activation of ERK prevents its localization to focal adhesions which interestingly depends on the activity of MLCK, a downstream effector of ERK. The authors of this report speculate that preventing proper localization of ERK can have a negative effect on cell motility, which echoes the ideas presented here, and provides a basis for future investigation into the effect of contortrostatin on this molecule. Many other cellular functions are controlled by the MAP kinases, including proliferation. The fact that contortrostatin acts as an inhibitor of angiogenesis and cancer progression raises 86 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. the likelihood that contortrostatin may also have negative effects on the ability of cells to divide, a hypothesis that remains to be investigated. Although we have shown that the a v p 3 integrin is responsible for mediating contortrostatin-induced tyrosine phosphorylation, the receptor(s) involved in transmitting the signal leading to activation of ERK2 remains unidentified. This diverging pathway may be controlled by another vitronectin receptor, av p 5 . Previous work has identified four integrins to which contortrostatin binds: ocIIb(33, a5|3l, av^3 and avP5 [Markland and Zhou, 2000; Trikha et al., 1994a; Trikha et al., 1994b; Zhou et al., 2000c]. Since the monomeric disintegrin echistatin inhibited contortrostatin-induced tyrosine phosphorylation effects by blocking avP3 [Ritter and Markland, 2000; Ritter et al., 2000b] but was unable to block ERK2 activation (data not shown), and since there are no reports of echistatin binding to avP5, it is possible that this integrin mediates contortrostatin-induced activation of ERK2. This idea is supported by the findings with OVCAR-5 cells which lack expression of avP3 but do express av P 5 . Contortrostatin influenced ERK2 activation but not tyrosine phosphorylation in these cells. The role of avP5 can be tested by using an antibody that blocks contortrostatin binding to this integrin, however we have not found an antibody that is capable of this. Epitope incompatibilities are the likely cause of the inability of these antibodies to block contortrostatin binding to avP5. A similar situation exists with respect to the a 5 p i integrin. We have previously reported that contortrostatin interacts with a 5 p i, enabling it to block adhesion to fibronectin [Trikha et al., 1994a]. However, multiple a 5 p i conformations appear 87 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. to exist, not all of which are competent to bind contortrostatin, since we have shown that some cells expressing this integrin bind contortrostatin (K562 human leukemia cells), while others do not (KSY-1 Kaposi’s sarcoma cells). Thus identifying the receptor(s) responsible for transmitting contortrostatin-initiated signals to ERK2 is not a trivial matter but remains an area of interest in our laboratory. The most likely resolution to this problem lies in the engineering of cells with the desired integrin expression profile that are both capable of binding contortrostatin and have intact (non-mutated) signaling pathways leading to ERK2 activation. It is well established that simple dimerization of integrins is sufficient to initiate tyrosine phosphorylation events [Miyamoto et al., 1995a], This has been accomplished with crosslinked anti-integrin antibodies [Huang et al., 1993] and multimeric integrin ligands [Bhattacharya et al., 1995]. When taking into account that contortrostatin is composed of two identical subunits, both containing the integrin-binding RGD motif [Trikha et al., 1994a], it is probable that the ability of contortrostatin to effect tyrosine phosphorylation in tumor cells is directly related to its ability to crosslink individual av(33 integrins. The reduction in tyrosine phosphorylation seen at higher contortrostatin concentrations is a consistent observation [Ritter et al., 2000b] and might be explained by imagining two binding orientations of the contortrostatin dimer, one in which each subunit is bound by an integrin, and the other where only one subunit is bound (Fig 5.5). This idea is addressed in Chapter 4 and revisited here. At low concentrations, each subunit would be given an opportunity to bind to an integrin, bringing two 88 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. integrins into close proximity and allowing for the initiation of a signaling cascade. At high concentrations, all contortrostatin subunits will, in effect, compete with each other for binding sites. These conditions will force many of the contortrostatin dimers into a binding orientation with only one subunit bound, and it would be expected that a reduction in the effects on tyrosine phosphorylation would be observed. Although it is likely that integrin dimerization is essential for contortrostatin-induced tyrosine phosphorylation, the role of dimerization in influencing ERK2 activation is not clear. It has been suggested that integrin ligation, in the absence of dimerization, can initiate signals as well [Miyamoto et al., 1995b], and it is possible that this mechanism is operating during contortrostatin activation of ERK2 (Fig 5.5). This work extends earlier findings that contortrostatin functions as more than a simple integrin antagonist, activating integrin signals leading to tyrosine phosphorylation of specific proteins. ERK2 is identified as an important signaling molecule activated by contortrostatin and provides a basis for further work into the downstream signaling effects of this unique disintegrin. Evidence is provided that, due to its ability to bind multiple integrins, contortrostatin influences more than one signaling pathway. The significance of the avP3 and avP5 integrins in angiogenesis and cancer progression suggest that contortrostatin may be an valuable tool for the further investigation of the roles of these receptors in these important biological processes. 89 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Chapter 6 Contortrostatin causes ocv[33-mediated structural disruptions and inhibits tumor c e ll m otility Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. SUMMARY Integrins play a major role in the regulation of cell motility. They physically link the extracellular environment to the cytoskeleton and participate in large protein complexes known as focal adhesions. In this chapter, it is demonstrated that treatment of tumor cells with the homodimeric disintegrin contortrostatin causes severe disruptions in the actin cytoskeleton and disassembly o f focal adhesion structures without affecting cellular adhesion to a reconstituted basement membrane. Included in this disruption is the tyrosine phosphorylation and altered subcellular localization of FAK. Through use of cells with specific integrin expression profiles, we demonstrate that these events are mediated exclusively by the a v P 3 integrin and are likely the result of contortrostatin-mediated crosslinking of this receptor at the cell surface since the monovalent disintegrin, flavoridin, does not induce such effects. Further, it is shown that contortrostatin potently inhibits motility in cells expressing the av|33 integrin. The results of this study describe a novel integrin-mediated mechanism by which cell motility can be inhibited and suggest an alternative approach to therapeutic intervention for cancer invasion and metastasis. 91 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. INTRODUCTION Integrins provide both physical anchorage to the extracellular matrix (ECM) and signaling information to the interior of the cell. Both of these functions are critical to the cellular process of motility and disruption of either can lead to reduced motility [Aznavoorian et al., 1996; Garton and Tonks, 1999; Wong et al., 1998]. The machinery that is responsible for driving motility is complex and highly regulated, but the action of the molecular components is not fully understood. Protein tyrosine kinases are major players in mediating integrin function and their substrates are well-studied components of the integrin signaling pathways [Giancotti and Ruoslahti, 1999]. Of the thousands of proteins in the cell, a small fraction of these undergo tyrosine phosphorylation. Many tyrosine phosphorylated proteins are found associated with integrin cytoplasmic domains in focal adhesion complexes [Burridge et al., 1988]. Focal adhesion kinase (FAK) is a nonreceptor tyrosine kinase that has been implicated in a wide range of integrin- and nonintegrin-regulated cellular processes [Rodriguez-Femandez, 1999]. FAK associates with integrin subunits near the cytoplasmic face of the cell membrane where, upon integrin clustering, it can undergo trans- autophosphorylation at a specific tyrosine residue [Schaller et al., 1994]. This creates a binding site for the SH2 domain of another tyrosine kinase, Src, which can then catalyse the phosphorylation of FAK at other sites [Cobb et al., 1994]. Another Src substrate is CAS (Crk associated substrate), an adapter protein that contains multiple tyrosines that can be phosphorylated [Vuori et al., 1996]. 92 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Although its function is not completely understood, CAS has been shown to be involved in the regulation of motility [Cary et al., 1998; Garton and Tonks, 1999; Klemke et al., 1998]. Integrin signaling events are initiated by cellular adhesion to ECM proteins which leads to two-dimensional relocation of integrin receptors in the plasma membrane and reorganization of the actin cytoskeleton [Burridge et al., 1992]. These events can trigger recruitment of other focal adhesion proteins as well as initiate divergent pathways such as the RAS/MAPK pathway by recruiting the Grb2 adapter protein [Schlaepfer et al., 1998]. Integrin clustering by the ECM has been simulated by crosslinking receptors with anti-integrin monoclonal antibodies [Komberg et al., 1991], soluble multivalent integrin ligands such as vitronectin [Bhattacharya et al., 1995], or antibody- or ligand-coated beads [Miyamoto et al., 1995b]. These methods induce integrin-mediated tyrosine phosphorylation events, including phosphorylation of FAK. Coinciding with the increased interest in integrin biology relating to cancer and angiogenesis, disintegrins have been used to study the effects of integrin antagonism in these processes [Kang et al., 1999; Sheu et al., 1997; Trikha et al., 1994a; Yeh et al., 1998; Zhou et al., 2000b]. Integrin av(33, a vitronectin receptor, has received much attention recently for its role in cancer, and particularly for its role in angiogenesis. This receptor has been shown to be necessary for angiogenesis and blockage of avP3 induces endothelial cell apoptosis in newly sprouting blood vessels [Brooks et al., 1994a; Brooks et al., 1994b]. Contributing to its regulation of angiogenesis and tumor invasiveness is 93 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. the ability of avP3 to bind the extracellular matrix-degrading protease, MMP2, localizing its activity to the leading edge of a migrating cell and facilitating penetration through the matrix [Brooks et al., 1996]. The avp3 integrin has been shown to be involved in tumor cell motility and metastasis and antagonists of this receptor inhibit these processes [Aznavoorian et al., 1996; Beviglia et al., 1995]. Antagonism of avP3 has also been shown to block melanoma cell binding to denatured collagen and to induce apoptosis in these cells [Petitclerc et al., 1999]. Small peptides containing the RGD sequence are commonly used integrin antagonists, but can be up to 10,000-fold less potent than disintegrins [Pfaff et al., 1994]. In the reports using disintegrins as inhibitors of angiogenesis [Kang et al., 1999; Sheu et al., 1997; Yeh et al., 1998], the monomeric disintegrins used are described as passive integrin antagonists, preventing avp3 from binding to the ECM and blocking normal signals mediated by this integrin. Contortrostatin is an avP3-binding disintegrin that is also an inhibitor of angiogenesis [Zhou et al., 2000b] and has a structure that is unique among the members of the disintegrin family. This protein is a member of a growing subfamily of dimeric disintegrins but is distinguished since it is a homodimer composed of two RGD-containing subunits [Trikha et al., 1994a; Zhou et al., 2000a] whereas the other members are heterodimeric [Marcinkiewicz et al., 1999a; Marcinkiewicz et al., 1999b] or lack the RGD motif [Oshikawa and Terada, 1999]. Here it is demonstrated that contortrostatin acts as an effective inhibitor of tumor cell motility, and data is provided suggesting that the mechanism employed by contortrostatin represents a novel means by which tumor cell motility, and thus tumor invasion and 94 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. metastasis, can be inhibited. Contortrostatin accomplishes this inhibition by combining passive integrin blockade with active alteration of ocv{33 signaling, likely through integrin crosslinking, which causes major disruption of the actin cytoskeleton and focal adhesion structure. MATERIALS AND METHODS M aterials — Rhodamine-labeled phalloidin was purchased from Sigma (St. Louis, MO). T24 cells lacking expression of the avP3 integrin were generated as described [Brooks et al., 1998]. See Chapter 4 for other materials used. Flow cytometry — Following cell harvest, cells were washed twice and resuspended in 1% BSA/PBS. Integrin expression was detected by incubating cells with primary antibodies for 30 min at room temperature. Cells were washed twice with 1% BSA/PBS prior to incubation with FITC-conjugated secondary antibody for 30 min at room temperature. Cells were washed three times prior to analysis on a FACScan flow cytometer (Becton Dickinson, Bedford, MA). Adhesion assays — Cellular adhesion to solid phase substrates was measured by immobilizing Matrigel (diluted 1:100) or vitronectin (10 fig/ml) into the wells of a 96-well culture plate by overnight incubation at 4°C. Unbound 95 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. protein was removed by washing with PBS and remaining surfaces were blocked by incubation with 1% BSA/PBS for 1 h at room temperature. Cells were harvested and washed four times with serum-free medium. After adjusting density to 5 x 10s cells/ml, cells were treated with various concentrations of contortrostatin and incubated 20 min at room temperature prior to addition to the coated wells. Cells were allowed to adhere 30 min at which time unbound cells were removed. The number of cells remaining was quantitated colorimetrically using the CellTiter 96 Nonradioactive Cell Proliferation Kit (Promega, Madison, WI) and data was processed using Prism software (Graph Pad, La Jolla, CA). Immunocytochemistry - Cells were harvested and plated onto glass coverslips coated with Matrigel diluted 1:100 with serum-free medium and blocked with 1% bovine serum albumin (BSA)/ phosphate buffered saline (PBS). Cells were incubated overnight at 37°C. Cells were then treated with disintegrins in serum-containing medium or serum-containing medium alone for 30 min at 37°C. Coverslips were washed with PBS, fixed with 4% formaldehyde/PBS for 10 min at 37°C, permiablized with 0.1% Triton X-100/PBS for 1 min and blocked with 1% BSA/PBS for 30 min at 37°C. Coverslips were then incubated with primary antibodies for 1 h at 37°C followed by fluorescein (FITC)-conjugated secondary antibody plus rhodamine-labeled phalloidin for 1 h. After washing with PBS, coverslips were mounted and images documented on a Olympus AX70 fluorescent microscope equipped with a SPOT digital camera. Motility Assays — Tumor cell motility was quantitated using a modified Boyden chamber [Repesh, 1989]. Transwell chambers with 12 pm pore size 96 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. (Coming Costar, Cambridge, MA) were coated with Matrigel diluted 1:100 with serum-free medium. Treated or untreated cells were added to the upper chamber, and the lower chamber was filled with HT1080 conditioned medium. Cells were incubated at 37°C for 10 h and after removal of non-migrating cells, the number of cells migrating to the bottom side of the coated membrane were fixed, stained and quantitated using digital image analysis (NIH Image). RESULTS Contortrostatin causes morphological changes and a breakdown of the actin cytoskeleton and focal adhesion structures. To explore the effects of contortrostatin on cell morphology and cytoskeletal structure, we examined tumor cells under phase contrast microscopy and performed immunocytochemical analysis, focusing our attention on the actin cytoskeleton and focal adhesion structures. In contortrostatin-treated T24 cells adhering to Matrigel, we observed a distinct alteration in cellular morphology where less cell spreading was observed and cells displayed sharp projections from the cell body (Fig 6.1a and b). We also observed dramatic disruptions in the appearance of the cytoskeleton following contortrostatin treatment in which actin stress fibers present in control cells appeared to collapse (Fig 6.1c and d). Accompanying the cytoskeletal 97 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. disruptions is a dramatic decrease in the size and number of focal adhesions as shown by phosphotyrosine staining (Fig 6.1e and f). The subcellular localization of the focal adhesion protein FAK was also examined. In control cells, FAK staining was punctate and displayed a distribution similar to that observed for phosphotyrosine (Figure 6.2c). In contrast, FAK staining in contortrostatin- treated cells was diffuse and lacked the punctate appearance of control cells (Figure 6.2d). The effects of contortrostatin on focal adhesions and the cytoskeleton is not due to reduced adhesiveness to the substratum. Figure 6.3 shows that while contortrostatin can effectively inhibit cell adhesion to vitronectin, this disintegrin has no effect on T24 cell adhesion to Matrigel, which is rich in the extracellular matrix component, laminin. This is a consistent observation made with all cell lines tested and is due to the lack of interaction of contortrostatin with laminin receptors. These results suggest that contortrostatin causes these cytoskeletal disruptions via an active mechanism involving induction of tyrosine phosphorylation events. Integrin av(33 is required for contortrostatin to disrupt the cytoskeleton and focal adhesions. To test the potential involvement of av(33 in the cytoskeletal disruptions observed after contortrostatin treatment, T24 cells that lack expression of this integrin were selected by fluorescence activated cell sorting (Figure 6.4b) and allowed to adhere to Matrigel during treatment. As shown in Figure 6.3, these cells partially retain the ability to adhere to vitronectin despite the absence of avP3. This is attributed to the presence of the vitronectin-binding integrin, 98 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. control contortrostatin phase contrast F-actin P-Tyr Fig 6.1 Effect of contortrostatin on morphology and cytoskeletal structres in T24 cells. Phase contrast microscopy of T24 cells treated with 0.5 (xM contortrostatin for 30 min (b) shows an altered morphology and decreased cell spreading compared to control cells (a). F-actin staining of control cells (c) shows large stress fibers while actin in contortrostatin-treated cells (d) appears to have collapsed into amorphous aggregates. Phosphotyrosine staining shows numerous well-defined focal adhesion structures (e). In contrast, focal adhesions are almost entirely absent in cells treated with contortrostatin (f). 99 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. control contortrostatin F-actin FAK Fig 6.2 FAK localization is altered by contortrostatin treatment. Specific staining for FAK in T24 control cells (c) yielded a pattern similar to that observed for phosphotyrosine (Fig 6.1) while FAK staining in contortrostatin-treated cells (d) was diffuse and lacked punctate appearance. Counterstains of fields (c) and (d) are shown in (a) and (b). 100 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 1.5-. Oi © O o c (B XI i _ o m x* < 1.0 - 0.5- 0.0 -1 — o ■ 10° T ” 101 “ 1 — 102 “ 1 103 contortrostatin (nM) Fig 6.3 Contortrostatin does not affect adhesion to a reconstituted basement membrane. Contortrostatin at the indicated concentrations has no effect on T24 cell adhesion to Matrigel (□ ), but completely blocks adhesion to vitronectin (■). T24 cells selected for the absence of av(}3 integrin expression are similarly unaffected in binding to Matrigel by contortrostatin (O), but are effectively inhibited in binding to vitronectin ( • ) due to the presence of an alternative vitronectin receptor, av|35. Experiments were performed in triplicate. Error bars indicate SEM. 101 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. B k g 7E3 10 1 0 B k g 7E 3 7E3 10 10* FL l 10 Fig 6.4 Flow cytometric analysis of av(33 expression, (a) T24 cells incubated with anti-avP3 mAb 7E3 show significant expression of this integrin. Background (Bkg) represents staining with secondary antibody only. T24 cells selected for a lack of avf$3 expression (b), and OVCAR-5 cells which do not express ccvP3 (c), show no staining with 7E3. 102 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. avp5, to which contortrostatin has recently been shown to bind [Zhou et al., 2000c]. Immunocytochemical analysis revealed that, unlike the av(53-expressing population, T24 cells lacking expression of avP3 showed no alterations in the appearance of the actin cytoskeleton (Fig 6.5a and b) or focal adhesions (Fig 6.5c and d) after treatment with contortrostatin. In an extension of these findings, OVCAR-5 human ovarian carcinoma cells, which also lack expression of av(33 (Figure 6.4c), similarly showed no cytoskeletal (Fig 6.5e and f) or focal adhesion alterations (Fig 6.5g and h) after contortrostatin treatment. These findings indicate that contortrostatin causes structural disruptions in these experiments exclusively though the av^3 integrin without affecting adhesiveness. Treatment with a monomeric disintegrin has no effect on cytoskeletal or focal adhesion structure in T24 cells. To determine the role of the homodimeric structure of contortrostatin in inducing cellular disruptions, we directly compared cells treated with the monomeric disintegrin flavoridin, to those treated with contortrostatin. It was observed that T24 cells treated with the monomer (Fig 6.6e and f) had an appearance similar to control cells (Figure 6.6a and b), while contortrostatin-treated cells showed the expected disruptions of the cytoskeleton (Fig 6.6c) and focal adhesions (Fig 6.6d). These results were obtained using flavoridin at twice the concentration of contortrostatin, strongly suggesting that the observed disruptions are directly related to the homodimeric structure of contortrostatin. Since flavoridin is known to bind the av(33 integrin [Pfaff et al., 103 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 1994], this reinforces our finding that the observed effects of contortrostatin in these experiments are not due to anti-adhesive effects. Contortrostatin inhibits motility of tumor cells. To determine the effect of contortrostatin treatment at the functional level, a Boyden chamber assay was used to determine the effect of this disintegrin on cell migration. MDA-MB-435 cells were treated with contortrostatin prior to addition to Matrigel-coated migration chambers and then allowed to migrate toward a gradient of H T1080 cell conditioned medium for 10 h. Contortrostatin was found to inhibit migration under these conditions by 90% (Fig 6.7a and b). The ability of contortrostatin to inhibit motility is not restricted to MDA-MB-435 cells. We have observed similar inhibitory effects in a variety of avP3-expressing tumor cells including T24 bladder carcinoma, KSY-1 Kaposis’s sarcoma, several glioma cell lines, and in human umbilical vein endothelial cells and human dermal microvascular endothelial cells [Zhou et al., 2000b; unpublished observations). 104 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. control contortrostatin F-ac P-Tyr F-actin P-Tyr Fig 6.5 The av(33 integrin is required for contortrostatin to cause cytoskeletal disruptions. T24 cells selected for the absence of av(33 (a-d) and OVCAR-5 human ovarian carcinoma cells which lack av(33 (e-h), were stained for F-actin and phosphotyrosine. In contrast to av(33- expressing cells these cells showed no changes in morphology, cytoskeletal or focal adhesion structure after 30 min treatment with 0.5 |lM contortrostatin (b, d, f, h) when compared to control cells (a, c, e, g). 105 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. F-actin P-Tyr V Fig 6.6 Comparison of the effets of contortrostatin and flavoridin treatment on cytoskeletal structures. Contortrostatin was compared directly to the monomeric disintegrin flavoridin for the ability to affect subcellular structures in T24 cells adhering to Matrigel. To equalize the number of integrin-binding RGD motifs present, cells were treated for 30 min with flavoridin at double the concentration of contortrostatin. While disruptions in F-actin (c) and phosphotyrosine (d) staining are seen after contortrostatin treatment, flavoridin-treated cells (e and f) have an appearance similar to control cells (a and b). 106 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. a W 0 0.01 0.1 contortrostatin (uM) 1 b 3000' g j pL( 2000. c3 ^ 2 in ■ f-< C/3 4t= o 1000. 0 0.01 0.1 1 contortrostatin (uVl) Fig 6.7 Contortrostatin inhibits motility of tumor cells. The homodimeric disintegrin contortrostatin, at the indicated concentrations, inhibits motility of MDA-MB-435 cells on Matrigel. (a) Images of stained cells after migration to the bottom side of the Matrigel-coated membrane, (b) Average number of migrated cells per high power field. Error bars indicate SEM. Experiments were performed four times to confirm results. 107 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. DISCUSSION Tyrosine phosphorylation of FAK has been associated with establishment of focal adhesions and promotion of motility [Cary et al., 1998; Gilmore and Romer, 1996; Sieg et al., 1999]. CAS phosphorylation has been shown to be critical to regulation of motility in several systems [Angers-Loustau et al., 1999; Cary et al., 1998; Garton and Tonks, 1999; Klemke et al., 1998]. Thus the idea that inducing phosphorylation of these molecules can negatively impact cytoskeletal structure and motility is in conflict with several reports, but highlights the regulatory complexities of integrin function. It might be expected that increasing tyrosine phosphorylation of focal adhesion proteins would result in greater numbers of focal adhesions present, and in fact this idea has been demonstrated previously [Angers-Loustau et al., 1999] where cells showed reduced migration as a result of decreased breakdown of focal adhesions. This mechanism is clearly not operating in our experiments with contortrostatin since we see dramatic reduction in the number of focal adhesions after treatment with contortrostatin. Changes in the cytoskeleton and reduction in the number of focal adhesions has been correlated with increased migratory behavior and invasiveness of transformed cells [Friedman et al., 1985; Keely et al., 1998]. Our results, however, indicate that the observed alterations in the actin cytoskeleton and focal adhesions after contortrostatin treatment are coincident with reduced motility, and suggest a distinct functional role of these alterations. With respect to FAK, it has recently been suggested that proper localization and function of this molecule 108 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. involves cycles of phosphorylation and de-phosphorylation [Sieg et al., 1999]. Contortrostatin might thus be envisioned as upsetting the physiological balance in this cycle where FAK is phosphorylated inappropriately, both temporally and spatially, causing FAK to adopt a subcellular localization that is incompatible with normal focal adhesion structure. In the absence of contortrostatin, FAK is localized to focal adhesions on the ventral surface of the adherent cell. Soluble contortrostatin binds and crosslinks avp3 integrins on the dorsal surface, nonproductively recruiting FAK and stimulating its phosphorylation. This might be expected to sequester FAK and prevent its normal localization, disrupting further focal adhesion assembly events. These studies raise the question of which tyrosine residues are phosphorylated on FAK after contortrostatin treatment. It is expected that crosslinking integrins with contortrostatin would lead to trans- autophosphorylation of FAK at tyrosine 397 which becomes a site recognized by the SH2 domain of Src [Schaller et al., 1994]. It has been shown that treatment of cells with a Src family kinase inhibitor dramatically reduces contortrostatin- induced tyrosine phosphorylation events [Ritter et al., 2000b], suggesting that tyrosine 397 on FAK is phosphorylated and a Src family kinase is recruited and activated, leading to further phosphorylation of FAK and CAS [Schlaepfer et al., 1994; Vuori et al., 1996]. Apparently, contortrostatin-induced disruptions extend beyond focal adhesions as it is observed that the actin cytoskeleton is grossly altered as well. Although the molecular details involved are not known, this is not unreasonable since focal adhesion proteins, such as vinculin, talin and tensin, are known to be associated with the actin cytoskeleton [Miyamoto et al., 1995b] and 109 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. disassembly of cytoskeletal anchorage points would be expected to disrupt the appearance of actin stress fibers. In fact, many cytoskeletally-associated proteins have been shown to localize to the cell surface where beads coated with integrin ligands or anti-integrin antibodies are bound [Miyamoto et al., 1995b]. Similarly, inappropriate localization of cytoskeletal molecules may account for the observed effects of contortrostatin on the actin cytoskeleton. The effects of contortrostatin could be expected to be more profound than those observed when using coated beads since many more integrin receptors would be engaged by the small, soluble contortrostatin and greater numbers of cytoskeletal molecules would be affected. Additionally, accumulation of focal adhesion proteins or filamentous actin would not be expected at sites of contortrostatin binding since these contortrostatin- integrin complexes are free to move about in the cell membrane whereas much larger beads simulate an insoluble substrate and allow for establishment of immobilized focal adhesion complexes at sites of contact. Biochemical analysis shows that the adapter protein CAS undergoes tyrosine phosphorylation in cells treated with contortrostatin. CAS localizes to the leading edges of advancing migratory cells, and disruption of this specific localization leads to decreased migratory capacity [Garton and Tonks, 1999]. It is possible that CAS localization is altered in contortrostatin-treated cells, although we were not able to dem onstrate this during these studies since the cells observed by immunofluorescent microscopy were not migratory. A strong link is established between our biochem ical tyrosine phosphorylation results (Chapter 4) and the effect of contortrostatin on the 110 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. cytoskeleton and focal adhesions since the av(33 integrin is required for both processes to occur. Additionally, the monomeric avP3-binding disintegrin flavoridin did not have a significant effect on the actin cytoskeleton or focal adhesions, an observation that correlates well with the biochemical findings. The lack of activity of the monomeric disintegrin strongly suggests that the unique structure of contortrostatin enables it to stimulate tyrosine phosphorylation events and disrupt cellular infrastructure. There are several reports in the literature describing methods of crosslinking integrins that result in activation of tyrosine phosphorylation [Bhattacharya et al., 1995; Komberg et al., 1991; Lipfert et al., 1992; Miyamoto et al., 1995b]. It is likely that contortrostatin stimulates tyrosine phosphorylation by crosslinking av(33 integrins at the cell surface, an activity that would clearly not be found in monomeric disintegrins. These data identify a novel integrin-mediated mechanism that can have profound effects on avP3-expressing cells. We have shown that contortrostatin causes significant inhibition of motility in tumor cells, and is a more potent inhibitor of m otility in avP3-expressing cells than monovalent integrin antagonists [Ritter et al., 2000b]. It is also possible that contortrostatin may have a negative effect on cell division since this process also depends on proper regulation of FAK and the actin cytoskeleton [Gilmore and Romer, 1996; Oktay et al., 1999; Yamakita et al., 1999]. These findings have potential significance to angiogenesis since the avP 3 integrin has been shown to have increased expression on angiogenic endothelial cells and to be critical to regulating this process [Brooks et al., 1994a]. Thus, by passively blocking the binding o f avP3 111 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. to its physiological ligand and actively inducing inappropriate signals leading to m assive structural disruptions within the angiogenic endothelial cell, contortrostatin is expected to be an effective inhibitor of angiogenesis. We have, in fact shown that contortrostatin inhibits angiogenesis in other studies [Zhou et al., 2000b]. These findings suggest an alternative approach to halt tumor cell invasion and metastasis as well as angiogenesis by active alteration of signaling pathways rather than blockage with signaling inhibitors, and identify contortrostatin as a prototype agent to further explore the potential of this approach. 112 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Chapter 7 Conclusion and General Discussion 113 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. This work has distinguished contortrostatin as a unique integrin-binding agent and identified a novel approach to the treatment of cancer. Integrins or their downstream signaling molecules have been targeted previously as points of intervention for therapy. Contortrostatin combines both passive and active disruption of integrin function to achieve its effect. The active disruption induced by contortrostatin appears to be extremely specific for avp3, a previously identified target shown to be critical for angiogenesis and invasion in several systems [Aznavoorian et al., 1996; Brooks et al., 1994a; Zheng et al., 1999]. The ultimate objective of this and other projects ongoing in our laboratory is to develop contortrostatin as a potential therapeutic treatment for cancer. This objective prevented thorough investigations into potentially important molecular interactions and mechanistic details regarding how contortrostatin accomplishes its observed effects. The opportunity will be taken here to discuss some of these important ideas and to propose studies that could fill the gaps in knowledge. Contortrostatin-treated cells expressing av[33 are shown to have increased phosphotyrosine content and disruptions in their cytoskeletal and focal adhesion structures, which has been correlated to decreased cell motility. All of these observations are hypothesized to rely directly on the ability of contortrostatin to initiate integrin-mediated signaling events by crosslinking avp3 at the cell surface. The fact that contortrostatin is a homodimer with two integrin binding regions, and the observation that FAK and CAS undergo Src-dependent tyrosine phosphorylation after treatment with contortrostatin, makes this contention reasonable. Additionally, the monomeric disintegrins, flavoridin and echistatin, 114 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. are shown to lack the ability to stimulate these signaling events, providing further support for this hypothesis. There is no direct evidence however, proving that integrin crosslinking takes place after contortrostatin treatment. It is possible that some unique feature of contortrostatin, other than the homodimeric structure, is responsible for initiating these events. To conclusively prove that contortrostatin is indeed crosslinking integrins at the cell surface, a monomeric form of contortrostatin would need to be generated. W hen cells are treated with monomeric contortrostatin, it would be expected that no increases in tyrosine phosphorylation would be observed. Binding to av|33 would be demonstrated by flow cytometry using monomeric contortrostatin to block binding of the anti- av(33 monoclonal antibody 7E3, an activity present in the native disintegrin and expected to be retained in the monomeric form. To further confirm that the monomer retains the ability to bind av(33, the monomer would be used to competitively inhibit the binding of dimeric contortrostatin to av{33 resulting in decreased tyrosine phosphorylation levels. Findings here would be anticipated to essentially duplicate results obtained using the monomeric disintegrins, flavoridin and echistatin (Fig 4.5), assuming that the binding affinity of monomeric contortrostatin is similar to that of the natural monomers. A monomeric form of contortrostatin could be generated using two different methods. The first method would employ selective reduction and derivitization of specific cysteines involved in interchain disulfide bonds. The difficulty of this approach lies in the fact that reducing only the disulfides responsible for dimerization, while preserving those involved in intrachain bonds, may not be achievable. One may anticipate multiple 115 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. resulting forms with different combinations of reduced disulfides, many of which may be monomers, which would need to be purified and shown to retain native integrin-binding activity. Alternatively, it might be expected that, due to potentially increased accessibility to reagents, the interchain disulfides may be reduced under less stringent conditions than the intrachain bonds, allowing for successful isolation of a monomeric form of contortrostatin. The second method of generating monomeric contortrostatin relies on the cDNA cloning efforts of Dr. Qing Zhou. Dr. Zhou was able to isolate a clone from a cDNA library of the venom gland from the southern copperhead snake which encodes the full length contortrostatin sequence, including a pro-domain and metalloprotease domain [Zhou et al„ 2000a]. In combination with the amino acid sequence data presented in Chapter 1, the complete cDNA sequence of the contortrostatin domain was determined. A monomeric form of contortrostatin can be created by expressing a mutant where the cysteines involved in interchain disulfides are substituted with a different residue, such as alanine. This approach would prevent dimerization while maintaining the structure of the individual subunits, and is expected to produce the highest likelihood of success. At present, however, only preliminary information is available regarding the specific disulfide bond patterns in dimeric disintegrins. No direct information is available for contortrostatin, but the disulfide pattern for EMF10, a heterodimeric disintegrin, has recently been published [Marcinkiewicz et al., 1999b]. The similarities between contortrostatin and EMF10 with respect to cysteine content, specifically the absence of two cysteines at the N-terminus, suggests the 116 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. possibility that these two dimeric disintegrins share the same disulfide bond pattern. This information might be utilized in the described method for engineering a monomeric form of contortrostatin. However, the existence of multiple disulfide bond patterns among the disintegrins also raises the possibility that contortrostatin and EMF10 have different disulfide arrangements, and would make it necessary to determine the specific pattern for contortrostatin. Work is currently underway in the laboratory o f Dr. Juan Calvete to determine the disulfide structure of contortrostatin which will clearly be useful in the generation of a monomeric form of this homodimeric disintegrin. An alternative means of demonstrating that contortrostatin mediates crosslinking of av|33 at the cell surface involves covalent linkage of integrins after contortrostatin treatment. A bifunctional crosslinking agent could be designed that, upon contortrostatin binding to two av(}3 heterodimers, would stably link one or both integrin subunits to the corresponding opposite subunit. After lysis, the cells could be analyzed by immunoblot for the presence of bands reacting with antibodies specific for the a v and/or [33 subunits. If contortrostatin does indeed bring av|33 integrins into close enough proximity to allow the chemical crosslinking agent to react, then one would observe bands on immunoblot with approximately double the expected molecular weight of the antigen being probed, indicating the presence of two covalently crosslinked integrin subunits. If cells are treated with monomeric disintegrins (ideally monomeric contortrostatin), no crosslinking events would be expected. One factor that could be critically important to the success of this approach is the 117 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. length of the chemical crosslinking agent, since it might be expected that a narrow range o f distances would be compatible with the crosslinking events described. These studies would need to be carried out in suspended cells since adherent cells would likely contain integrins clustered into focal adhesions, whereas cells in suspension should have a more homogeneous distribution of integrins at the cell surface, allowing for detection of contortrostatin-mediated events. A variation of this method involves covalently crosslinking the disintegrin to the integrin [Calvete et al, 1994]. If contortrostatin is simultaneously binding two av|33 integrins, using this approach it would be expected that immunoblotting with either anti-av or anti-P3 antibody would yield a band with a molecular weight equal to the sum of the disintegrin and the integrin subunit(s). Again, a monovalent disintegrin would be useful to test. The monomer would be expected to be complexed with a single integrin subunit and would have a migration on immunoblot reflecting its smaller size. Perhaps this method would be least problematic since crosslinker length would be less of an issue. Based on the data presented and established integrin signaling mechanisms, it is likely that contortrostatin stimulates tyrosine phosphorylation through integrin crosslinking. It has been demonstrated that FA K is phosphorylated shortly after treatment with contortrostatin, which is consistent with a crosslinking model. FAK contains 6 tyrosine residues that are known to be phosphorylated in vivo [Schlaepfer and Hunter, 1998]. Tyrosine 397 is known to be rranj-autophosphorylated following integrin clustering and becomes a binding site for the SH2 domain of Src. In our studies, inhibition of Src family kinases 118 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. with the pharmacological inhibitor PP1 inhibited the ability of contortrostatin to induce tyrosine phosphorylation of proteins in the 120— 140 kDa range, which includes FAK and CAS. This finding supports the idea that FAK Tyr397 is phosphorylated after contortrostatin-mediated crosslinking of av(33 at the cell surface, and that Src is recruited, resulting in further Src-mediated tyrosine phosphorylation of FAK and CAS (Fig 7.1). The conclusion that Src is involved in transmitting contortrostatin-induced signals cannot be made with certainty since PP1 inhibits the entire Src family of kinases and it is possible that other kinases may be inhibited as well. The uncertainty of Src involvement further brings into question whether Tyr397 is actually phosphorylated upon contortrostatin binding. To fully characterize the mechanism of contortrostatin- induced events, it is necessary to define which of the 6 tyrosines are undergoing phosphorylation. This issue is best addressed through genetic manipulation of the proteins in question. Ilic et al. have described the isolation of FAK-null fibroblasts which provide an ideal background on which to study the role of FAK and any desired mutant of FAK [Ilic et al., 1995]. These cells have been used to establish the function of the different tyrosines in fibroblasts. Our interest in these studies was related to the function of FAK in tumor cells and, although contortrostatin might be expected to function similarly in fibroblasts with respect to integrin signaling, FAK-null fibroblasts represent a less-than-ideal system. This problem can be circumvented by using a tumor cell of choice and overexpressing different mutant forms of FAK, for example Phe397, and measuring FAK tyrosine phosphorylation levels following contortrostatin 119 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Fig 7.1 Schematic model depicting contortrostatin-m ediated crosslinking of av(33 and the resulting biochemical events. It is hypothesized that contortrostatin binding to individual av(53 heterodimers brings the receptors into proximity allowing for fra/w-autophosphorylation of FAK at Tyr397. This creates a binding site for the SH2 domain of Src which can then phosphorylate FAK at other tyrosine residues. CAS is recruited into a trimolecular complex with FAX and Src and undergoes tyrosine phosphorylation catalyzed by Src. PP1 inhibits the kinase activity of Src and reduces Src-mediated phosphorylation of FAK and CAS. 120 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. treatment. The dominant-negative effect of the Phe397 mutant would be expected to result in a decrease in overall FAK phosphorylation if this residue undergoes phosphorylation in wild type FAK. The same approach can be taken with other FAK tyrosine residues to demonstrate whether or not each specific residue is phosphorylated after contortrostatin treatment. A valuable tool was recently made available to investigate this problem. A polyclonal antiserum has been produced that reportedly recognizes FAK that is phosphorylated on Tyr397, and not other forms of FAK (Upstate Biotechnology, Lake Placid, NY). This antibody, and perhaps others developed in the future that recognize specific residues, will simplify these types of investigations. The role of Src in this mechanism can be elucidated by overexpressing a mutant of Src that retains the ability to bind Tyr397 of FAK, but lacks kinase activity. This could establish whether Src is responsible for further phosphorylation of FAK and phosphorylation of CAS. Chapter 5 describes studies demonstrating that contortrostatin activates a signaling pathway leading to the activation of ERK2. This activation is shown to utilize a pathway different from the pathway involving contortrostatin-induced tyrosine phosphorylation. Discussed in Chapter 5 are approaches that can be used to identify the receptor responsible for contortrostatin-induced ERK2 activation. In contrast to FAK phosphorylation, which is one of the earliest integrin signaling events detectable, MAP kinase activation is a relatively late event, occurring further downstream of membrane-proximal events. In addition to the integrin receptor responsible, the events preceding contortrostatin-induced ERK2 activation remain to be conclusively identified. Integrin-mediated activation of 121 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. ERK2 is an established event shown to involve Src-catalyzed phosphorylation of Tyr925 on FAK which promotes binding of Grb2 and activation of the Ras pathway leading to ERK2 activation [Schlaepfer et al., 1998]. Thus, inhibiting Src activity would be expected to reduce contortrostatin-induced ERK2 activation. In our studies, use of the Src family kinase inhibitor PP1 resulted in essentially complete inhibition of ERK2 activation following contortrostatin treatment (Fig 5.3). This result suggests that contortrostatin-mediated activation of ERK2 follows the canonical integrin signaling pathway involving a FAK/Src/Grb2 complex. As mentioned previously, pharmacological agents raise the issue of specificity and point to genetic approaches as a more conclusive means of defining pathways. A dominant-negative mutant of Src would again prove useful to resolve this issue. The events downstream of contortrostatin- induced activation of ERK2 also remain to be elucidated and may represent important factors in the function of contortrostatin. ERK2 is known to directly influence the activity of transcription factors and to regulate the function of myosin light chain kinase (MLCK), an important regulator of cellular motility [Klemke et al., 1997]. It would be of particular interest to study the specific affect of contortrostatin-mediated activation of ERK2 on MLCK activity and cell motility. The observation that contortrostatin acts as an inhibitor of motility is in some conflict with the fact that ERK2 is activated after contortrostatin treatment, since activation of ERK2 is usually associated with enhanced motility. A recent report describes the targeting of ERK to focal adhesion structures, providing additional information regarding the role of ERK in regulating motility [Fincham 122 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. et al., 2000]. Perhaps contortrostatin treatment causes ERK2 activation in a temporally and spatially inappropriate manner, similar to that proposed for FAK, and produces a situation that is incompatible with normal motility. This hypothesis can be readily tested using methods established during the course of these studies by performing immunocytochemical analysis of cultured cells and determining the subcellular localization of ERK2 in the presence and absence of contortrostatin. It is possible that a change would be observed in which ERK2 would loose its localization to focal adhesion structures, similar to what was observed for FAK after contortrostatin treatment (Fig 6.2). The ability of contortrostatin to inhibit invasive ness in the Boy den chamber is well established and has been demonstrated with several cell lines. The avP3 integrin is required for contortrostatin-induced tyrosine phosphorylation, which has been correlated with inhibition of motility, providing a potential mechanism for this inhibitory effect. However, we have also observed a substantial negative effect on migration in cells lacking expression of avP3. This initially conflicting observation may be rationally explained when considering that contortrostatin-mediated activation of ERK2 occurs in the absence of avP 3. avP5-expressing OVCAR-5 ovarian carcinoma cells show reduced migration and increased ERK2 activation after contortrostatin treatment. This temporally and spatially inappropriate activation of ERK2 could alone account for the inhibition of motility, as described above. Additionally, simple integrin antagonism clearly plays a role in the ability of contortrostatin to inhibit motility since anti-integrin antibodies and monovalent disintegrins also inhibit 123 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. cell motility [Ritter et al., 2000b; Zhou et al., 2000c]. Antagonism of the av(35 integrin is effective at inhibiting OVCAR-5 motility since contortrostatin and an antibody against this integrin (P1F6) have a similar negative effect on motility [Zhou et al., 2000c]. This is in contrast to avp3-expressing cells, where contortrostatin reproducibly inhibits motility to a greater degree than monovalent av|33 antagonists [Ritter et al., 2000b; Zhou et al., 2000d], indicating the presence of additional activity. With respect to the contortrostatin project, the primary areas if interest in our laboratory are tumor biology and angiogenesis. Tumor cells were the major subject of investigations into the molecular mechanisms of contortrostatin function, which fits our objectives of understanding how contortrostatin inhibits tumor progression. However, the molecular nature of contortrostatin function in the major cell type involved in angiogenesis, endothelial cells, went relatively unstudied. This was not a matter of choice since the earliest studies were conducted with human umbilical vein endothelial cells (HUVECs), which proved to be prohibitively expensive to maintain for extended periods. To replace the HUVECs with a cell line that required less demanding culture conditions, we obtained the ECV304 spontaneously transformed human endothelial cell line. During the course of our studies, we were informed by the American Type Tissue Collection (ATTC) that the ECV304 cell line was misidentified and was in fact a bladder carcinoma cell line known as T24. This thrust our studies into in vitro angiogenesis back to tumor biology. In another attempt to procure a suitable endothelial cell line, we were able to obtain an immortalized human dermal 124 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. micro vascular endothelial cell line, HMEC-1, from Dr. Robert Swerlick at Emory University [Ades et al., 1992]. These cells readily formed network-like structures, or “tubes,” when cultured on Matrigel, which was a desirable characteristic for our investigations into the antiangiogenic effect o f contortrostatin. However, contortrostatin was found to have no effect on the ability of these cells to form tubes. Although these cells were shown to express the avp3 integrin, they did not respond as expected with increased tyrosine phosphorylation as observed with all other av(33-positive cells tested. Further analysis revealed that HMEC-1 cells have unusually high levels of tyrosine phosphorylation of the 120-140 kDa bands shown previously to contain FAK and CAS. Contortrostatin treatment had no effect on these bands, but when the cells were incubated with the Src family kinase inhibitor PP1, tyrosine phosphorylation of these bands was completely eliminated. It was therefore concluded that HMEC-1 cells have a mutation that results in constitutive activity of Src, or a signaling molecule upstream of Src, and since contortrostatin signaling has been shown to require the Src family kinases, this mutation prevents any inhibitory effect of contortrostatin on tube formation. Although our studies with endothelial cells were thwarted in various ways, it is believed that contortrostatin will have profound effects on av(33-expressing endothelial cells, perhaps similar to what we have observed for av(33-expressing tumor cells. Primary microvascular endothelial cells might provide the best opportunity to study contortrostatin and its effect on in vitro angiogenesis. 125 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. One of the most intuitively obvious approaches to the treatment of cancer is to slow or stop the growth of tumor cells. This approach can also be applied to disrupt tumor-induced angiogenesis as well, since the proliferation of angiogenic endothelial cells is critical to the growth of new blood vessels. The observation that contortrostatin inhibits the growth of tumors and angiogenesis in mouse models raises the possibility that contortrostatin is producing these effects through inhibition of cellular proliferation. Integrins have been shown by other investigators to have regulatory roles in cell growth [Howe et al., 1998] and it is conceivable that contortrostatin can have a negative effect on proliferation through integrin disruption. A preliminary experiment was performed in our laboratory to address this question when MDA-MB-435 human breast cancer cells were allowed to adhere to Matrigel in the presence of contortrostatin and their ability to proliferate was quantitated [Zhou et al., 2000d]. Over the course of 5 days, there was no observed change in cell density with contortrostatin treatment versus untreated cells. A lingering question regarding the dynamic function of contortrostatin is whether or not contortrostatin is internalized by cells. This important question was directly addressed by performing a pulse-chase experiment with fluorescently labeled contortrostatin. Labeling was done with the help of Dr. Steve Swenson. T24 cells adhering to Matrigel were incubated for 4 h in the presence of the labeled disintegrin at 500 nM, followed by 1 h with unlabeled contortrostatin. Although preliminary in nature, this experiment clearly revealed the presence of labeled contortrostatin in the cells (Fig 7.2). Thus, if contortrostatin is taken up and degraded by the cells within hours, a negative 126 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Fig 7.2 Fluorescently labeled contortrostatin is internalized by T24 cells. T24 cells adhering to Matrigel were incubated with 500 nM FiTC-labeled contortrostatin for 4 h. Cells were then washed and incubated with unlabeled contortrostatin for 30 min, washed again and incubated an additional 30 min with unlabeled contortrostatin. Cells were fixed in 4% formaldehyde/PBS and examined under fluorescent microscopy. Internalized contortrostatin is visualized as bright cytoplasmic spheres. Some diffuse fluorescence is visible, indicating the possibility of contortrostatin degradation and release from cytoplasmic vesicles. 127 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. effect on proliferation with a single treatment m ight be unexpected. An interesting recent study adds potential significance to contortrostatin internalization, which reports that RGD-containing peptides can enter cells and directly cause apoptosis through activation of caspase-3 [Buckley et al., 1999]. Although the in vivo significance of this report is questionable, it certainly suggests some potentially interesting studies that can be done with contortrostatin relating to its fate upon entry into the cell, but also complicates the study of the possible role of contortrostatin as an inhibitor of proliferation. The similarities between metastasis and angiogenesis at the molecular level suggest that contortrostatin may be inhibiting both processes through a similar mechanism. As mentioned in Chapter 1, both metastasis and angiogenesis depend on the ability of cells to traverse the basement membrane found beneath the blood vessel endothelial cell layer. This requires proper adhesion, which is requisite to migration on the membrane, and also requires that the basement membrane be degraded to remove the physical barrier. The sum of the studies presented here have allowed for the development of a working hypothesis that can account for the in vitro and in vivo actions of contortrostatin (Fig 7.3). The three required cellular steps for metastasis and angiogenesis have been studied individually during the course of this project. Chapter 3 describes two independent approaches aimed at determining the effect of contortrostatin on the ability of breast cancer cells to digest the extracellular matrix. These studies lead to the conclusion that contortrostatin has no effect on matrix degradation in these cells, a conclusion that was subsequently corroborated in our laboratory using two 128 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. METASTASIS/ANGIOGENESIS Fig 7.3 Cellular events required for both metastasis and angiogenesis. Metastatic tumor cells or angiogenic endothelial cells must be capable of adhesion to the underlying basement membrane since the physical force of adhesion is necessary for cell migration on the membrane. The barrier presented by the basement membrane must be breached in order for cells to migrate into or out of the vessel. Our studies demonstrate that contortrostatin has no effect on cell adhesion to a laminin-rich reconstituted basement membrane. Further, additional studies indicate that contortrostatin does not effect the ability of cells to digest the extracellular matrix. Therefore, motility is identified as the major cellular event affected by contortrostatin. 129 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. different human glioma cell lines. In Chapter 6 it is demonstrated that, although contortrostatin is an effective inhibitor of adhesion to vitronectin, it does not effect adhesion to a reconstituted basement membrane, which is composed largely of laminin. This indicates that contortrostatin exerts its effects in a manner independent of its ability to inhibit cell adhesion. These findings rule out two of the three possible ways that contortrostatin can accomplish inhibition of both metastasis and angiogenesis. Thus cell motility is singled out as the cellular step most likely affected by contortrostatin. Our efforts, therefore, became focused on determining how, at the molecular level, contortrostatin was causing a decrease in cell m otility. Our studies initially lead to a paradoxical situation where contortrostatin was shown to activate integrin signaling pathways that are commonly associated with promotion of motility. After confirming these biochemical results, we directed our attention to cellular infrastructure by exam ining the actin cytoskeleton and focal adhesion structures by immunofluorescent microscopy. A model was developed to describe the effects of contortrostatin on cellular infrastructure which integrates the observed biochemical effects (Fig 7.4). The model predicts that, by crosslinking av(33 integrins at the cell surface, contortrostatin initiates an integrin signaling pathway that involves focal adhesion proteins FAK, Src and CAS. Based on the immunocytochemical results, it can also be concluded that other focal adhesion proteins are affected as well, since focal adhesion structures essentially disappear after contortrostatin treatment. Whether contortrostatin signaling directly affects the actin cytoskeleton is not known. Perhaps contortrostatin affects the activity of 130 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Fig 7.4 Hypothetical model describing the effect of contortrostatin on cytoskeletal and focal adhesion structures. For simplicity, FAK is shown as the only focal adhesion component. In untreated cells (top), clustered integrins are ligated to the extracellular matrix and colocalize with tyrosine phosphorylated FAK in focal adhesions. The actin cytoskeleton is organized into stress fibers terminating at focal adhesions, and the cell has a well-spread morphology. Contortrostatin treatment (bottom) causes tyrosine phosphorylation of FAK, but does so in a spatially and temporally inappropriate fashion, causing abberant localization of FAK and producing a situation that is incompatible with normal cytoskeletal function (ie. motility). The altered morphology reflects the contortrostatin-induced collapse of the cytoskeleton. 131 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. the GTPases, Rho or Rac, which are known integrin-controlled regulators of the actin cytoskeleton [Keely et al., 1998]. It is conceivable that the collapse of the cytoskeleton is a secondary effect occurring after disassembly of the focal adhesions, which serve to anchor actin stress fibers and link them to the extracellular matrix. Regardless of the order of events, the paradoxical situation with respect to the activities of contortrostatin is most plausibly explained by the idea that, although contortrostatin activates well studied signaling pathways, it does so at the wrong time and wrong place. This can cause dysregulation and inappropriate subceilular localization of molecules critical to the proper function of the cytoskeleton. A cellular process as complex and dynamic as motility can clearly be expected to require tight regulation of the machinery driving it, and biochemical events as dramatic as those induced by contortrostatin can clearly be envisioned as being disruptive to this regulation. An interesting feature of contortrostatin is that it possesses characteristics of both a broad specificity integrin antagonist, and a highly specific integrin agonist. Contortrostatin has the ability to block the adhesive function of a5 p l, av|33 and av(35, all of which have been shown to have important roles in cancer and/or angiogenesis. Its interaction with allb p 3 on platelets represents a technical hurdle to the use of contortrostatin as a therapeutic for cancer since platelets act to sequester contortrostatin and prevent sufficient concentrations at the tumor site. This problem can be overcome by packaging contortrostatin to prevent platelet interaction, which will also conceal the protein from immune surveillance. Preliminary studies have been conducted exploring the potential of 132 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. encapsulating contortrostatin in liposomes of a specific size that allows them to exit the circulation at the tumor site, exploiting the “leaky” nature of the tumor vasculature. An alternative approach would be to eliminate platelet interaction by changing the structure of contortrostatin through introduction of mutations that eliminate aIIbP3 interaction but retain the other integrin-binding specificities. In contrast to this relative lack of specificity as an integrin antagonist, contortrostatin appears to be highly specific for ccvp3 as an initiator of integrin signaling and cytoskeletal disruption. The critical role of this integrin in angiogenesis immediately increases the importance of further mechanistic study of contortrostatin on angiogenic endothelial cells. This av(33-targeted mechanism should clearly participate in the anti-angiogenic and anti-tumor effects of contortrostatin. It would be of interest to determine if an agent that possesses only the avP3-mediated signaling activities of contortrostatin would produce the same kind of negative effect on angiogenesis, tumor growth and metastasis. One might expect that natural contortrostatin would perform better than a specific avP3 crosslinker in this regard since integrins other than avP3 are likely involved in tumor invasion and metastasis. Thus, contortrostatin could represent a unique and powerful combination of activities residing in a single molecule that can inhibit several critical steps in cancer progression. This line of reasoning suggests that native contortrostatin would have higher potency than engineered analogs that have narrower integrin-binding specificity, and that a successful method of targeting contortrostatin to the tumor site could yield the desired inhibition of 133 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. cancer progression, while avoiding unwanted platelet interaction and immune responses. By adding a new dimension to the activity of contortrostatin, this work has helped propel the ambitious hopes of developing a new way to fight cancer. Good research will always raise questions as well as answer them, and many new questions have arisen over the course of this project. Answering some of these questions will, at worst, add to our knowledge of biology - and at best, continue the march forward to a cure for cancer. 134 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. R eferences Ades EW, Candal FJ, Swerlick RA, George VG, Summers S, Bosse DC, Lawley TJ (1992): HMEC-1: Establishm ent of an Immortalized Human M icrovascular Endothelial Cell Line. The Journal of Investigtive Dermatology 99:683-690. Angers-Loustau A, Cote J-F, Charest A, Dowbenko D, Spencer S, Lasky LA, Tremblay ML (1999): Protein Tyrosine Phosphatase-PEST Regulates Focal Adhesion Disassembly, Migration, and Cytokinesis in Fibroblasts. J. Cell Biol. 144:1019-1031. Aznavoorian S, Stracke ML, Parsons J, McClanahan J, Liotta LA (1996): Integrin alphavbeta3 mediates chemotactic and haptotactic motility in human melanoma cells through different signaling pathways. J. Biol. Chem. 271:3247-54. Barry ST, Flinn HM, Humphries MJ, Critchley DR, Ridley AJ (1997): Requirement for Rho in integrin signalling. Cell Adhesion Commun 4:387-98. Beviglia L, Stewart GJ, Niewiarowski S (1995): Effect of four disintegrins on the adhesive and metastatic properties of B16F10 melanoma cells in a murine model. Oncology Research 7:7-20. Bhattacharya S, Fu C, Bhattacharya J, Greenberg S (1995): Soluble Ligands of the av(33 Integrin Mediate Enhanced Tyrosine Phosphorylation of Multiple Proteins in Adherent Bovine Pulmonary Artery Endothelial Cells. J. Biol. Chem. 270:16781-16787. Brooks PC, Clark RA, Cheresh DA (1994a): Requirement of Vascular Integrin av(33 for Angiogenesis. Science 264:569-571. Brooks PC, Montgomery AMP, Rosenfeld M, Reisfeld RA, Hu T, Klier G, Cheresh DA (1994b): Integrin av(33 Antagonists Promote Tumor Regression by Inducing Apoptosis of Angiogenic Blood Vessels. Cell 79:1157-1164. Brooks PC, Silletti S, von Schalscha TL, Friedlander M, Cheresh DA (1998): Disruption of angiogenesis by PEX, a noncatalytic metalloproteinase fragment with integrin binding activity. Cell 92:391-400. 135 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Brooks PC, Stromblad S, Sanders LC, von Schalscha TL, Aimes RT, Stetler- Stevenson WG, Quigley JP, Cheresh DA (1996): Localization of Matrix Metalloproteinase MMP-2 to the Surface of Invasive Cells by Interaction with Integrin avP3. Cell 85:683-693. Buckley CD, Pilling D, Henriquez NV, Parsonage G, Threlfall K, Scheel-Toellner D, Simmons DL, Akbar AN, Lord JM, Salmon M (1999): RGD peptides induce apoptosis by direct caspase-3 activation. Nature 397:534-9. Burridge K, Fath K, Kelly T, Nuckolls G, Turner C (1988): Focal adhesions: transmembrane junctions between the extracellular matrix and the cytoskeleton. A. Rev. Cell Biol. 4:487-525. Burridge K, Turner CE, Romer LH (1992): Tyrosine phosphorylation of paxillin and p p l25FA.K accompanies cell adhesion to extracellular matrix: A role in cytoskeletal assembly. J. Cell Biol. 119:893-903. Calvete JJ, Jurgens M, Marcinkiewicz C, Romero A, Schrader M, Niewiarowski S (2000): Disulphide-bond pattern and molecular modelling of the dimeric disintegrin EMF-10, a potent and selective integrin alpha5betal antagonist from Eristocophis macmahoni venom. Biochem. J. 345 Pt 3:573-81. Calvete JJ, Mclane MA, Stewart GT, Niewiarowski, S (1994) Characterizaton of the cross-linking site of disintegrins albolabrin, bitistatin, echistatin, and eristostatin on isolated human platelet integrin GPIIb/IIIa. BBRC 202:135-140 Calvete JJ, Schafer W, Soszka T, Lu WQ, Cook JJ, Jameson BA, Niewiarowski S (1991): Identification of the disulfide bond pattern in albolabrin, an RGD- containing peptide from the venom of Trimeresurus albolabris: significance for the expression of platelet aggregation inhibitory activity. Biochemistry 30:5225-9. Canagarajah BJ, Khokhlatchev A, Cobb MH, Goldsmith EJ (1997): Activation mechanism of the MAP kinase ERK2 by dual phosphorylation. Cell 90:859-69. Cary L, Chang J, Guan J-L (1996): Stimulation o f cell migration by overexpression of focal adhesion kinase and its association with Src and Fyn. J. Cell Sci. 109:1787-1794. Cary LA, Han DC, Polte TR, Hanks SK, Guan J-L (1998): Identification of p l3 0 < --as as a Mediator of Focal Adhesion Kinase-promoted Cell Migration. J. Cell Biol. 140:211-221. 136 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Chandrasekaran S, Guo N-h, Rodrigues RG, Kaiser J, Roberts DD (1999): Pro adhesive and Chemotactic Activities of Thrombospondin-1 for Breast Carcinoma Cells Are Mediated by a3 |3 l Integrin and Regulated by Insulin-like Growth Factor-1 and CD98. J. Biol. Chem. 274:11408-11416. Chen Q, Kinch MS, Lin TH, Burridge K, Juliano RL (1994): Integrin-mediated Cell Adhesion Activates Mitogen-activated Protein Kinases. J. Biol. Chem. 269:26602-26605. Claesson-Welsh L, Welsh M, Ito N, Anand-Apte B, Soker S, Zetter B, O'Reilly M, Folkman J (1998): Angiostatin induces endothelial cell apoptosis and activation of focal adhesion kinase independently of the integrin-binding motif RGD. Proc. Natl. Acad. Sci. USA 95:5579-5583. Clark EA, Trikha M, Markland FS, Brugge JS (1994): Structurally Distinct Disintegrins Contortrostatin and Multisquamatin Differentially Regulate Platelet Tyrosine Phosphorylation. J. Biol. Chem. 269:21940-21943. Cobb BS, Schaller MD, Leu TH, Parsons JT (1994): Stable association of pp60src and pp59fyn with the focal adhesion-associated protein tyrosine kinase, ppl25FAK. Mol. Cell. Biol. 14:147-55. Cybulsky AV, McTavish AJ, Cyr MD (1994): Extracellular matrix modulates epidermal growth factor receptor activation in rat glomerular epithelial cells. Journal of Clinical Investigation 94:68-78. DeClerck YA, Laug WE (1996): Cooperation between matrix metalloproteinases and the plasminogen activator-plasmin system in tumor progression. Enzyme & Protein 49:72-84. DeClerck YA, Perez N, Shimada H, Boone TC, Langley KE, Taylor SM, Alvarez OA, Carmichael DF (1992): Inhibition of invasion and metastasis in cells transfected with an inhibitor of metalloproteinases Dennis MS, Carter P, Lazarus RA (1993): Binding interactions of kistrin with platelet glycoprotein Hb-HIa: analysis by site-directed mutagenesis. Proteins 15:312-21. Egan SE, Giddings BW, Brooks MW, Buday L, Sizeland AM, Weinberg RA (1993): Association of Sos Ras exchange protein with Grb2 is implicated in tyrosine kinase signal transduction and transformation. Nature 363:45- 51. Emonard H, Grimaud JA (1990): Matrix metalloproteinases. A review. Cellular & Molecular Biology 36:131-53. 137 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Fincham VJ, James M, Frame MC, Winder SJ (2000): Active ERK/MAP kinase is targeted to newly forming cell-matrix adhesions by integrin engagement and v-Src. EMBO 19:2911-2923. Friedlander M, Brooks PC, Shaffer RW, Kincaid CM, Varner JA, Cheresh DA (1995): Definition of two angiogenic pathways by distinct alpha v integrins. Science 270:1500-2. Friedman E, Verderame M, Lipkin M, Pollack R (1985): Altered actin cytoskeletal patterns in two premalignant stages in human colon carcinoma development. Cancer Research 45:3236-42. Gan ZR, Gould RJ, Jacobs JW, Friedman PA, Polokoff MA (1988): Echistatin, a potent platelet aggregation inhibitor from the viper Echis carinatus. J. Biol. Chem. 263:19827-19832. Garton AJ, Tonks NK (1999): Regulation of Fibroblast Motility by the Protein Tyrosine Phosphatase PTP-PEST. J. Biol. Chem. 274:3811-3818. Giancotti FG, Ruoslahti E (1999): Integrin Signaling. Science 285:1028-1032. Gilmore AP, Romer LH (1996): Inhibition of focal adhesion kinase (FAK) signaling in focal adhesions decreases cell motility and proliferation. Mol. Biol. Cell 7:1209-24. Gomez DE, Alonso DF, Yoshiji H, Thorgeirsson UP (1997): Tissue inhibitors of metalloproteinases: structure, regulation and biological functions. Eur. J. Cell Biol. 74:111-22. Hanemaaijer R, Verheijen JH, Maguire TM, Visser H, Toet K, McDermott E, O'Higgins N, Duffy MJ (2000): Increased gelatinase-A and gelatinase-B activities in malignant vs. benign breast tumors. International J Cancer 86:204-7. Hanke JH, Gardner JP, Dow RL, Changelian PS, Brisette WH, W eringer EJ, Pollok BA, Connelly PA (1996): Discovery of a Novel, Potent, and Src Family-selective Tyrosine Kinase Inhibitor J. Biol. Chem. 271:695-701. Harder KW, M oller NP, Peacock JW, Jirik FR (1998): Protein-tyrosine phosphatase alpha regulates Src family kinases and alters cell-substratum adhesion. J. Biol. Chem. 273:31890-900. Hill CS, Treisman R (1995): Transcriptional regulation by extracellular signals: mechanisms and specificity. Cell 80:199-211. 138 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Howe A, Aplin AE, Alahari SK, Juliano RL (1998): Integrin signaling and cell growth control. Current Opinion in Cell Biology 10:220-31. Huang M-M, Lipfert L, Cunningham M, Brugge JS, Ginsberg MH, Shattil SJ (1993): Adhesive Ligand Binding to Integrin aDb(33 Stimulates Tyrosine Phosphorylation of Novel Protein Substrates before Phosphorylation of ppl25FAK. J. Cell Biol. 122:473-483. Huhtala P, Humphries MJ, McCarthy JB, Tremble PM, Werb Z, Damsky CH (1995): Cooperative signaling by alpha 5 beta 1 and alpha 4 beta 1 integrins regulates metalloproteinase gene expression in fibroblasts adhering to fibronectin. J. Cell Biol. 129:867-79. Hynes RO (1992): Integrins: versatility, modulation, and signaling in cell adhesion. Cell 69:11-25. Die D, Furuta Y, Kanazawa S, Takeda N, Sobue K, Nakatsuji N, Nomura S, Fujimoto J, Okada M, Yamamoto T (1995): Reduced cell motility and enhanced focal adhesion contact formation in cells from FAK-deficient mice. Nature 377:539-44. Jay PY, Pham PA, Wong SA, Elson EL (1995): A mechanical function of myosin H in cell motility. J. Cell Sci. 108:387-93. Jones PA, DeClerck YA (1980): Destruction of extracellular matrices containing glycoproteins, elastin, and collagen by metastatic human tumor cells. Cancer Research 40:3222-7. Jones PL, Crack J, Rabinovitch M (1997): Regulation of tenascin-C, a vascular smooth muscle cell survival factor that interacts with the alpha v beta 3 integrin to promote epidermal growth factor receptor phosphorylation and growth. J. Cell Biol. 139:279-93. Juliano D, Wang Y, Marcinkiewicz C, Rosenthal LA, Stewart G, Niewiarowski S (1996): Disintegrin Interaction with avf$3 Integrin on Human Umbilical Vein Endothelial Cells: Expression of Ligand-Induced Binding Site on £3 Summit. Exp. Cell Res. 225:132-142. Kang I-C, Lee Y-D, Kim D-S (1999): A Novel Disintegrin Salmosin Inhibits Tumor Angiogenesis. Cancer Research 59:3754-3760. Keely P, Parise L, Juliano R (1998): Integrins and GTPases in tumour cell growth, motility and invasion. Trends Cell Biol. 8:101-6. 139 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Klemke RL, Cai S, Giannini AL, Gallagher PJ, de Lanerolle P, Cheresh DA (1997): Regulation of Cell Motility by Mitogen-activated Protein Kinase. J. Cell Biol. 137:481-492. Klemke RL, Leng J, Molander R, Brooks PC, Vuori K, Cheresh DA (1998): CAS/Crk Coupling Serves as a "Molecular Switch" for Induction of Cell Migration. J. Cell Biol. 140:961-972. Komberg LJ, Earp HS, Turner CE, Prockop C, Juliano RL (1991): Signal transduction by integrins: Increased protein tyrosine phosphorylation caused by clustering of pi integrins. Proc. Natl. Acad. Sci. USA 88:8392- 8396. Lewis LM, Cheresh DA, Schwartz MA (1996): Protein Kinase C Regulates avP5— dependent Cytoskeletal Associations and Focal Ahesion Kinase Phosphorylation. J. Cell Biol. 134:1323-1332. Lin ECK, Ratnikov BI, Tsai PM, Carron CP, Myers DM, Barbas CF, Smith J (1997): Identification of a Region in the Integrin P3 Subunit That Confers Ligand Binding Specificity. J. Biol. Chem. 272:23912-23920. Liotta LA, Goldfarb RH, Brundage R, Siegal GP, Terranova V, Garbisa S (1981): Effect of plasminogen activator (urokinase), plasmin, and thrombin on glycoprotein and collagenous components of basement membrane. Cancer Research 41:4629-36. Lipfert L, Haimovich B, Schaller MD, Cobb BS, Parsons JT, Brugge JS (1992): Integrin-Dependent Phosphorylation and Activation o f the Protein Tyrosine Kinase ppl25^AK jn platelets. J. Cell Biol. 119:905-912. Main AL, Harvey TS, Baron M, Boyd J, Campbell ID (1992): The three- dimensional structure of the tenth type III module of fibronectin: an insight into RGD-mediated interactions. Cell 71:671-8. Marcinkiewicz C, Calvete JJ, Marcinkiewicz MM, Raida M, Vijay-Kumar S, Huang Z, Lobb RR, Niewiarowski S (1999a): EC3, a Novel Heterodimeric Disintegrin from Echis carinatus Venom, Inhibits a4 and a 5 Integrins in an RGD-independent Manner. J. Biol. Chem. 274:12468-12473. Marcinkiewicz C, Calvete JJ, Vijay-Kumar S, Marcinkiewicz MM, Raida M, Schick P, Lobb RR, Niewiarowski S (1999b): Structural and functional characterization of EMF10, a heterodimeric disintegrin from Eristocophis macmahoni venom that selectively inhibits alpha 5 beta 1 integrin. Biochemistry 38:13302-9. 140 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Marcinkiewicz C, Vijay-Kumar S, McLane MA, Niewiarowski S (1997): Significance of RGD Loop and C-Terminal Domain of Echistatin for Recognition of allbp3 and av|33 Integrins and Expression of Ligand- Induced Binding Site. Blood 90:1565-1575. Markland FS, Zhou Q (2000): Snake Venom Disintegrin: An Effective Inhibitor of Breast Cancer Growth and Dissemination. In Tu AT, Gaffield W (eds): “Natural and Synthetic Toxins: Biological Implications.” W ashington D.C.: American Chemical Society, pp 262-282. McLane MA, Vijay-Kumar S, Marcinkiewicz C, Calvete JJ, Niewiarowski S (1996): Importance of the Structure of the RGD-containing Loop in the Disintegrins Echistatin and Eristostatin for Recognition of ocIIbP3 Integrins. FEBS Letters 391:139-143. Miyamoto S, Akiyama SK, Yamada KM (1995a): Synergistic Roles for Receptor Occupancy and Aggregation in Integrin Transmembrane Function. Science 267:883-885. Miyamoto S, Teramoto H, Coso OA, Gutkind JS, Burbelo PD, Akiyama SK, Yamada KM (1995b): Integrin Function: M olecular Hierarchies of Cytoskeletal and Signaling Molecules. J. Cell Biol. 131:791-805. Montgomery AM, De Clerck YA, Langley KE, Reisfeld RA, Mueller BM (1993): Melanoma-mediated dissolution of extracellular matrix: contribution of urokinase-dependent and m etalloproteinase-dependent proteolytic pathways. Cancer Research 53:693-700. Morino N, Mimura T, Hamasaki K, Tobe K, Ueki K, Kikuchi K, Takehara K, Kadowaki T, Yazaki Y, Nojima Y (1995): Matrix/Integrin Interaction Activates the Mitogen-activated Protein Kinase, p44er^‘ 1 and p42er^"^. J. Biol. Chem. 270:269-273. Niewiarowski S, McLane MA, Kloczewiak M, Stewart G (1994): Disintegrins and Other Naturally Occurring Antagonists of Platelet Fibrinogen Receptors. Seminars in Hematology 31:289-300. Oktay M, Wary KK, Dans M, Birge RB, Giancotti FG (1999): Integrin-mediated activation of focal adhesion kinase is required for signaling to Jun NH2- terminal kinase and progression through the G1 phase of the cell cycle. J. Cell Biol. 145:1461-9. Oshikawa K, Terada S (1999): Ussuristatin 2, a novel KGD-bearing disintegrin from Agkistrodon ussuriensis venom. J. Biochem. 125:31-5. 141 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Petitclerc E, Stromblad S, von Schalscha TL, Mitjans F, Piulats J, Montgomery AM, Cheresh DA, Brooks PC (1999): Integrin alpha(v)beta3 promotes M21 melanoma growth in human skin by regulating tumor cell survival. Cancer Research 59:2724-30. Pfaff M, McLane MA, Beviglia L, Niewiarowski S, Timpl R (1994): Comparison of Disintegrins with Limited Variation in the RGD Loop in Their Binding to Purified Integrins allb|33, av|33 and a5|3l and in Cell Adhesion Inhibition. Cell Adhesion Commun. 2:491-501. Pierschbacher MD, Ruoslahti E (1987): Influence of stereochemistry of the sequence Arg-Gly-Asp-Xaa on binding specificity in cell adhesion Plow EF, Ginsberg MH (1989): Cellular adhesion: GPIIb-IIIa as a prototypic adhesion receptor. Progress in Hemostasis & Thrombosis 9:117-56. Price JE, Polyzos A, Zhang RD, Daniels LM (1990): Tumorigenicity and Metastasis of Human Breast Carcinoma Cell Lines in Nude Mice. Cancer Research 50:717-721. Repesh LA (1989): A New in vitro Assay for Quantitating Tumor Cell Invasion. Invasion and Metastasis 9:192-208. Ridley AJ, Hall A (1992): The small GTP-binding protein rho regulates the assembly of focal adhesions and actin stress fibers in response to growth factors. Cell 70:389-99. Ritter MR, M arkland FS (2000): Differential Regulation of Tyrosine Phosphorylation in Tumor Cells by Contortrostatin, a Homodimeric Disintegrin, and Monomeric Disintegrins Echistatin and Flavoridin. Toxicon 39:317-323. Ritter MR, Zhou Q, Markland FS (2000a): Contortrostatin, a Homodimeric Disintegrin, Actively Disrupts Focal Adhesion and Cytoskeletal Structure and Inhibits Cell Motility Through a Novel Mechanism. Cell Adhesion and Communication, in press Ritter MR, Zhou Q, Markland FS (2000b): Contortrostatin, A Snake Venom Disintegrin, Induces av(33-mediated Tyrosine Phosphorylation of CAS and FAK in Tumor Cells. J. Cell. Biochem. in press. Rodriguez-Femandez JL (1999): Why do so many stimuli induce tyrosine phosphorylation of FAK?. Bioessays 21:1069-75. Ruoslahti E, Pierschbacher MD (1987): New perspectives in cell adhesion: RGD and integrins. Science 238:491-7. 142 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Sastry SK, Horwitz AF (1993): Integrin cytoplasmic domains: mediators of cytoskeletal linkages and extra- and intracellular initiated transmembrane signaling. Current Opinion in Cell Biology 5:819-31. Schaller MD, Hildebrand JD, Shannon JD, Fox JW, Vines RR, Parsons JT (1994): Autophosphorylation of the focal adhesion kinase, p p l25FAK^ directs SH2-dependent binding of pp60src. Mol. Cell. Biol. 14:1680-1688. Schlaepfer DD, Hanks SK, Hunter T, van der Geer P (1994): Integrin-mediated signal transduction linked to Ras pathway by GRB2 binding to focal adhesion kinase. Nature 372:786-91. Schlaepfer DD, Hunter T (1996): Evidence for in vivo phosphorylation of Grb2 SH2-domain binding site on focal adhesion kinase by Src-family protein- tyrosine kinases. Mol. Cell. Biol. 16:5623-5633. Schlaepfer DD, H unter T (1998): Integrin signalling and tyrosine phosphorylation: just the FAKs? Trends Cell Biol. 8:151-157. Schlaepfer DD, Jones KC, Hunter T (1998): Multiple Grb2-Mediated Integrin- Stimulated Signaling Pathways to ERK2/Mitogen-Activated Protein Kinase: Summation of Both c-Src- and Focal Adhesion Kinase-Initiated Tyrosine Phosphorylation Events. Mol. Cell. Biol. 18:2571-2585. Seftor RE, Seftor EA, Gehlsen KR, Stetler-Stevenson WG, Brown PD, Ruoslahti E, Hendrix MJ (1992): Role of the alpha v beta 3 integrin in human melanoma cell invasion. Proc. Natl. Acad. Sci. USA 89:1557-61. Seftor RE, Seftor EA, Stetler-Stevenson WG, Hendrix MJ (1993): The 72 kDa type IV collagenase is modulated via differential expression of alpha v beta 3 and alpha 5 beta 1 integrins during human melanoma cell invasion. Cancer Research 53:3411-5. Senn H, Klaus W (1993): The nuclear magnetic resonance solution structure of flavoridin, an antagonist of the platelet GP nb-IUa receptor. J. Mol. Biol. 232:907-25. Sheu J-R, Yen M-H, Kan Y-C, Hung W-C, Chang P-T, Luk H-N (1997): Inhibition of angiogenesis in vitro and in vivo: comparison of the relative activities of triflavin, an Arg-Gly-Asp-containing peptide and anti-av(33 integrin monoclonal antibody. Biochim. Biophys. Acta 1336:445-454. 143 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Shibata K, Kikkawa F, Nawa A, Thant AA, Naruse K, Mizutani S, Hamaguchi M (1998): Both focal adhesion kinase and c-Ras are required for the enhanced matrix metalloproteinase 9 secretion by fibronectin in ovarian cancer cells. Cancer Research 58:900-3. Short SM, Talbott GA, Juliano RL (1998): Integrin-mediated Signaling Events in Human Endothelial Cells. Mol. Biol. Cell 9:1969-1980. Sieg DJ, Hauck CR, Schlaepfer DD (1999): Required role of focal adhesion kinase (FAK) for integrin-stimulated cell migration. J. Cell Sci. 112:2677- 91. Staiano N, Garbi C, Squillacioti C, Esposito S, Di Martino E, Belisario MA, Nitsch L, Di Natale P (1997): Echistatin induces decrease of p p l25FAK phosphorylation, disassembly of actin cytoskeleton and focal adhesions, and detachment of fibronectin-adherent melanoma cells. Eur. J. Cell Biol. 73:289-305. Tam SH, Sassoli PM, Jordan RE, Nakada MT (1998): Abciximab (ReoPro, chimeric 7E3 Fab) demonstrates equivalent affinity and functional blockade of GPIIb/IIIa and av(33 integrins. Circulation 98:1085-1091. Terranova VP, Hujanen ES, Martin GR (1986): Basement membrane and the invasive activity of metastatic tumor cells. Journal of the National Cancer Institute 77:311-6. Thompson EW, Yu M, Bueno J, Jin L, Maiti SN, Palao-Marco FL, Pulyaeva H, Tamborlane JW, Tirgari R, Wapnir I, et al. (1994): Collagen induced MMP-2 activation in human breast cancer. Breast Cancer Research & Treatment 31:357-70. Trikha M (1994): Purification and Characterization of Disintegrins As Inhibitors of Integrin-Mediated Platelet Aggregation and Tumor Cell Adhesion and Metastasis, doctoral dissertation, University of Southern California. Trikha M, De Clerck YA, Markland FS (1994a): Contortrostatin, a Snake Venom Disintegrin, Inhibits p i Integrin-mediated Human Metastatic Melanoma Cell Adhesion and Blocks Experimental Metastasis. Cancer Research 54:4993-4998. Trikha M, Rote W, Manley P, Lucchesi BR, Markland FS (1994b): Purification and Characterization of Platelet Aggregation Inhibitors From Snake Venoms. Thrombosis Research 73:39-52. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Tselepis VH, Green LJ, Humphries MJ (1997): An RGD to LDV M otif Conversion within the Disintegrin K istrin Generates an Integrin Antagonist That Retains Potency but Exhibits Altered Receptor Specificity. J. Biol. Chem. 272:21341-21348. Tuckwell DS, Brass A, Humphries MJ (1992): Homology modelling of integrin EF-hands. Evidence for widespread use of a conserved cation-binding site. Biochem. J. 285:325-31. Vuori K, Hirai H, Aizawa S, Ruoslahti E (1996): Induction of pl30cas Signaling Complex Formation upon Integrin-Mediated Cell Adhesion: a Role for Src Family Kinases. Mol. Cell. Biol. 16:2606-2613. Vuori K, Ruoslahti E (1994): Association of insulin receptor substrate-1 with integrins. Science 266:1576-8. Wang M, Liu YE, Greene J, Sheng S, Fuchs A, Rosen EM, Shi YE (1997): Inhibition of tumor growth and metastasis of human breast cancer cells transfected with tissue inhibitor of metalloproteinase 4. Oncogene 14:2767-74. Wong NC, Mueller BM, Barbas CF, Ruminski P, Quaranta V, Lin EC, Smith JW (1998): Alphav integrins mediate adhesion and migration of breast carcinoma cell lines. Clinical Exp. Metastasis 16:50-61. Yamakita Y, Totsukawa G, Yamashiro S, Fry D, Zhang X, Hanks SK, Matsumura F (1999): Dissociation of FAK/pl30(CAS)/c-Src complex during mitosis: role of mitosis-specific serine phosphorylation of FAK. J. Cell Biol. 144:315-24. Yeh CH, Peng H-C, Huang T-F (1998): Accutin, a New Disintegrin, Inhibits Angiogenesis In Vitro and In Vivo by Acting as Integrin av(33 Antagonist and Inducing Apoptosis. Blood 92:3268-3276. Zheng D-Q, Woodard AS, Fomaro M, Tallini G, Languino LR (1999): Prostatic Carcinoma Cell Migration via av[53 Integrin Is Modulated by a Focal Adhesion Kinase Pathway. Cancer Research 59:1655-1664. Zhou Q, Hu P, Ritter MR, Swenson SD, Argounova S, Epstein AL, Markland FS (2000a): Molecular cloning and functional expression of contortrostatin, a homodimeric disintegrin from southern copperhead snake venom. Archives of Biochemistry & Biophysics 375:278-88. Zhou Q, Nakada M, Arnold C, Shieh KY, Markland FS (2000b): Contortrostatin, a Dimeric Disintegrin from Agkistrodon contortrix contortrix, Inhibits Angiogenesis. Angiogenesis in press. 145 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Zhou Q, Nakada MT, Brooks PC, Swenson SD, Ritter MR, Argounova S, Arnold C, Markland FS (2000c): Contortrostatin, a homodimeric disintegrin, binds to integrin alphavbeta5. Biochemical & Biophysical Research Communications 267:350-5. Zhou Q, Sherwin RP, Parrish C, Richters V, G GS, Tsao-Wei D, Markland FS (2000d): Contortrostatin, a dimeric disintegrin from A g kistro d o n contortrix contortrix, inhibits breast cancer progression. Breast Cancer Research and Treatment in press. 146 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

Asset Metadata

Creator Ritter, Matthew Ray (author)

Core Title Inhibition of cancer invasion and metastasis: Mechanistic analysis of contortrostatin function at the molecular and cellular levels

Contributor Digitized by ProQuest (provenance)

School Graduate School

Degree Doctor of Philosophy

Degree Program Biochemistry and Molecular Biology

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag biology, cell,biology, molecular,OAI-PMH Harvest

Language English

Advisor Markland, Francis (committee chair), [illegible] (committee member)

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c16-85408

Unique identifier UC11328631

Identifier 3018118.pdf (filename),usctheses-c16-85408 (legacy record id)

Legacy Identifier 3018118.pdf

Dmrecord 85408

Document Type Dissertation

Rights Ritter, Matthew Ray

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the au...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus, Los Angeles, California 90089, USA