PlA Polymorphism of Integrin β3 Differentially Modulates Cellular Migration on Extracellular Matrix Proteins
Objective— Cell migration is central to multiple physiological and pathologic processes and involves interactions between integrins on the cell surface and the extracellular matrix. The Leu33Pro (PlA) polymorphism of integrin β3 has been reported to be associated with a greater rate of restenosis after angioplasty, a process involving endothelial and smooth muscle cell migration. We have addressed the possibility that the Leu33Pro polymorphism could modify the migratory behavior of Chinese hamster ovary (CHO) cells expressing the β3-containing integrin complexes.
Methods and Results— Haptotactic migratory responses of CHO αIIbβ3 cells to fibronectin and vitronectin were not statistically different between the Leu33 and Pro33 cells. However, CHO cells with the Pro33 (PlA2) polymorphism had an enhanced haptotactic migratory response to fibrinogen and von Willebrand Factor. This enhanced migration (1) could be blocked by the αIIbβ3-complex–specific neutralizing mAb 10E5, by 7E3, a neutralizing mAb specific for the β3 integrin, and by the αIIbβ3-blocking peptide Integrilin; (2) was not observed with a CHO cell line expressing an activating β3 Cys435 to Ala mutation; and (3) was attributable to increased activity of mitogen-activated protein kinase and cyclooxygenase. CHO cell lines expressing the Pro33 isoform of αvβ3 had an enhanced haptotactic migratory response to vitronectin and osteopontin but not fibrinogen.
Conclusions— The Leu33Pro polymorphism alters the migratory behavior of cells on extracellular matrix substrates, and the α subunit influences the substrate specificity of this genetic effect.
β3 integrins are centrally involved in many adverse outcomes in patients with coronary heart disease.1 Activation of the αIIbβ3 integrin mediates fibrinogen binding and outside-in signaling, culminating in platelet aggregation,2 and αvβ3 mediates smooth muscle cell (SMC) and endothelial cell (EC) migration.3 β3 is polymorphic at residue 33 (Leu33Pro), with the frequency of alleles encoding Leu33 (often called the PlA1 antigen) and Pro33 (often called the PlA2 antigen) being ≈75% and ≈25%, respectively.4 The Pro33 polymorphism of integrin β3 has been associated with adverse cardiovascular events in some studies but not others.5 The apparently inconsistent results may be attributable in part to the variability in study design and clinical phenotypes that have been investigated. This seems to be especially true for studies of acute coronary syndromes (ACSs), where controls have usually been poorly matched with cases and little uniformity has existed among criteria for case selection.
In studies of outcomes after percutaneous transluminal angioplasty (PCTA) with or without stent placement, the cases (patients experiencing restenosis or the need for urgent revascularization) and controls are derived from the same starting population such that less heterogeneity exists between the two groups. All 5 prospective PCTA studies found the Pro33 polymorphism was overrepresented in the cases compared with the controls.5 Although statistical significance was not achieved in two of these studies, this nevertheless represents a much greater consistency of findings than in the studies of patients with ACSs. Like ACSs, adverse outcomes after PCTA occur in the presence of atherosclerosis and involve exposure of the subendothelium, formation of a platelet thrombus, and platelet clot retraction.6 In this context, the Pro33 isoform of αIIbβ3 demonstrates a lower threshold for activation,7 increased platelet aggregability,8 and clot retraction9 and even shorter bleeding times.10 However, restenosis after PCTA additionally involves migration of SMC and EC, as well as neointimal proliferation.11 Although loss of function in β3 knockout mice does not alter neointimal formation12 and there are no differences in atherosclerosis between humans lacking αIIbβ3 only or both αIIbβ3 and αvβ3,13 it is not known whether a gain-of-function alteration of β3 affects cell migration.
Increasing quantities of fibrinogen and fibrin are detected in progressively severe atherosclerotic plaques,14 and fibrinogen is a ligand for many cell types that have been implicated in atherosclerotic plaque development, such as platelets, EC, SMC, macrophages, and leukocytes.15 Furthermore, ligands for the αvβ3 integrin such as fibrinogen, vitronectin, and osteopontin are also overexpressed at sites of atherosclerotic and restenotic lesions and have been implicated in SMC migration.3 We now report studies of the influence of the Leu33Pro polymorphism on the migratory behavior of cells ectopically expressing αIIbβ3 and αvβ3 on ligands implicated in SMC migration and abundantly expressed at sites of atherosclerotic and angioplasty-induced lesions. We find that the Leu33Pro polymorphism modulates cell migration and that this migration is ligand- and integrin-dependent. We additionally report that this modulation is dependent on activation of the mitogen-activated protein (MAP) kinase extracellular signal-regulated kinase (ERK) 1/2 and cyclooxygenase (COX).
Reagents and Cell Culture
Fibrinogen was from Enzyme Research Laboratories Inc. von Willebrand factor (vWF) was a gift from Dr J.L. Moake (Baylor College of Medicine, Houston, Tex). Anti-β3 mAb (SZ21; Leu33-specific) and anti-αIIbβ3 mAb (P2) were from Immunotech. Anti-αvβ3 mAb (LM609) was from Chemicon International Inc. Blocking mAbs 7E3 (specific for β3) and 10E5 (specific for αIIbβ3) were gifts from Dr Barry Coller (The Rockefeller University, New York, NY). The αIIbβ3/αvβ3-specific peptide antagonist Integrilin was a gift from Dr D.R. Phillips (COR Therapeutics Inc, South San Francisco, Calif). The MAP kinase kinase (MEK) inhibitor U0126 and aspirin (acetylsalicylic acid, ASA) were from Promega and Sigma, respectively.
The CHO DG44 (β3 having a Cys435Ala substitution) cell line was provided by Dr Peter Newman.16 Generation of the Leu33 and Pro33 αIIbβ3 cell lines has been described.9 A second set of CHO cell lines called LK/pc (parental), Leu33b, and Pro33b was generated in a similar fashion. The αvβ3 cell lines were generated in a similar fashion, with a 3.3-kb cDNA for αv cloned into the pcDNA3.1 vector (Invitrogen, Carlsbad, Calif). αIIbβ3- or αvβ3-expressing cells were sorted by flow cytometry to obtain equivalent and high levels of expression using P2 or LM609, respectively. Flow cytometry to assess αIIbβ3 or αvβ3 expression was performed within 24 hours of each experiment. If expression varied by >20%, the cells were resorted to equivalency.
Cell Migration Assay
Haptotactic cell migration assays were performed as described earlier by Klemke et al17 using modified Boyden chambers (Transwell; Costar) in 24-well tissue culture dishes. The upper chambers contained polycarbonate membranes 10 μm thick with 8-μm pores. The undersurface only was coated with the matrix ligand diluted in PBS overnight at 4°C at the following final concentrations (in μg/mL): fibrinogen 12.5, vitronectin 10, fibronectin 50, and osteopontin 5. The final concentration for vWF was 77% of normal human plasma vWF. Cells at 0.25×106/mL in 0.1 mL αMEM/0.5% BSA were added to the upper chamber, and the migration assay was allowed to proceed for 2 hours at 37°C and 5% CO2. Cells that had migrated to the lower surface were fixed in 4% buffered paraformaldehyde and stained and counted. In a systematic fashion, 10 microscopic fields were counted for each filter at approximately the same position on all filters. Each experiment was performed in triplicate, and an average was determined. Based on the surface area of the filter, ≈25% of the cells added to the upper chamber were found to migrate at the 2-hour time point. No cells were observed in the buffer or on the base of the lower migration chamber at the end of the experiment. When both surfaces of the transwell were coated with fibrinogen, we did not observe any αIIbβ3 cells migrating to the lower surface of the membrane at 2 hours, indicating no random cell migration. All analyses were performed a minimum of 2 times.
Presentation of Migration Data
To account for minor differences in receptor expression, the cell migration data were normalized by dividing the number of cells by the mean fluorescence intensity (determined by flow cytometry using either the P2 or LM609 mAbs). The absolute numbers of migrating cells varied considerably from experiment to experiment, but there was consistency in the relative ratio of Leu33 to Pro33 cell migration. Therefore, the normalized differences in cell migration among the various cell lines were primarily expressed as a fold difference compared with the data observed with the Leu33-expressing cells.
Numerical data are presented as mean±SEM. t test or one-way ANOVA followed by the Newman-Keuls’ multiple range test was used to analyze for statistically significant differences (P≤0.05).
CHO αIIbβ3 Pro33 Cells Have an Enhanced Migratory Response to Fibrinogen
Flow cytometric analysis using the αIIbβ3-complex–specific mAb P2 demonstrated equivalent expression of the Leu33 or Pro33 isoforms of αIIbβ3 in our stable cell lines (data not shown). CHO cells expressing αIIbβ3 (Leu33 or Pro33) adhered to and spread on fibrinogen, whereas adhesion of control LK cells was minimal (not shown). In haptotactic migration assays, significantly more (1.5-fold; P<0.05) Pro33 cell migration on a fibrinogen substrate was observed than Leu33 cell migration (Figure 1A). Leu33 or Pro33 migration toward fibrinogen was β3-dependent, because it could be completely blocked by pretreatment with mAb 7E3 (Figure 1B). Importantly, the αIIbβ3-specific inhibitors 10E5 and Integrilin abolished the difference in the extent of migration between Leu33 and Pro33 cells (Figure 1C), indicating that the αIIbβ3 integrin mediated this difference and not the low level of expression of the chimeric hamster-human αvβ3 integrin (the latter was expressed to <16% of the levels of αIIbβ3). In this and all migration experiments, the parental cell line showed virtually no migration (usually ≈1% of Leu33, but never >4% of Leu33) and is not shown.
When the data were examined without normalization (Figure 1D), both Leu33 and Pro33 cells showed greater migration with greater (>3-fold) expression of αIIbβ3. For both the low and high expressing cells, Pro33 showed greater migration than Leu33 cells. Consistent with our data, Huttenlocher et al18 found that cells with high levels of integrin surface expression migrated faster than cells with lower expression. Therefore, we took the conservative approach of normalizing for levels of αIIbβ3 expression.
CHO Cells Expressing a Constitutively Active GPIIb-IIIa Do Not Exhibit Enhanced Migration on Fibrinogen
We considered whether the increased cell migration of the Pro33 isoform of β3 was typical of any β3 variant known to exhibit increased adhesion to immobilized fibrinogen. CHO cells stably expressing αIIbβ3 with the Cys435Ala substitution of β3 (which disrupts the long-range Cys5-Cys435 disulfide bond) result in greater adhesion and spreading to immobilized fibrinogen than the Leu33 form of wild-type αIIbβ3.16 These cells showed no difference in haptotactic migration on fibrinogen compared with wild-type Leu33 CHO cells (Figure 1E). Thus, amino acid substitutions that enhance adhesion and spreading do not necessarily increase cell migration.
We also assessed the time course and substrate concentrations on cell migration. Increased Pro33 migration was observed over a range of fibrinogen concentrations (6.25, 12.5, and 25 μg/mL, Figure 2A) as well as at multiple time points (2, 4, and 6 hours, Figure 2B).
These data were confirmed with a second set of cell lines generated as described in the Methods, called CHO LK/pc, Leu33b, and Pro33b for the parental, Leu33, and Pro33 expressing cells, respectively. Flow cytometry demonstrated equivalent expression and distinguished the Leu33 and Pro33 isoforms (data not shown). This second set of cell lines confirmed that compared with Leu33 cells, Pro33 cells demonstrated enhanced migration (1.4-fold) on fibrinogen (Figure 3A).
Enhanced Migration of CHO Pro33 Cells Is Dependent on Activation of ERK1/2 and COX
Integrin-dependent migration of CHO and other cell types has been shown to be dependent on activation of the MAP kinase and COX pathways.19 ERK 1/2 is a MAP kinase that induces COX-2 expression, resulting in increased metabolism of arachidonic acid to prostaglandins, which is required for cell migration, and COX-2 is required for αvβ3-dependent migration of ECs on vitronectin.20 We investigated the effect of MAP kinase and COX blockade in this system using U0126, an inhibitor of MEK, and ASA, respectively. CHO cells expressing the Pro33 isoform of αIIbβ3 were selectively blocked by both U0126 and ASA (Figure 3B) as well as with the mAb10E5 (data not shown), whereas the Leu33-expressing cells were unaffected by U0126 and ASA. Similar data were obtained with both independent sets of cell lines (only one is shown). These experiments suggest that increased migration of αIIbβ3-transfected CHO cells bearing the Pro33 polymorphism is dependent on an increased activity of MAP kinase and COX.
CHO αIIbβ3 Pro33 Cells Have an Enhanced Migratory Response to vWF But Not Fibronectin or Vitronectin
Significantly more (2.6-fold) Pro33 cell haptotactic migration on a vWF substrate was observed than Leu33 cell migration (Figure 4). Leu33 or Pro33 cells readily attached to and spread on vWF, whereas attachment of LK cells was minimal (not shown). Although LK cells attached well to fibronectin or vitronectin (presumably through endogenous CHO α5β1 and αvβ5, respectively), there was no significant difference between Leu33 and Pro33 cell migration on these substrates. Migration of LK cells was equal to Leu33 cells on fibronectin, presumably because of the similar expression of endogenous α5β1. These data indicate that the increased migration of the Pro33 variant of β3 on fibrinogen or vWF was substrate specific.
CHO αvβ3 Pro33 Cells Have an Enhanced Migratory Response to Vitronectin and Osteopontin But Not Fibrinogen
The other major β3 integrin, αvβ3, is crucial for EC and SMC migration, processes that underlie atherogenesis and restenosis.3 We investigated whether the Leu33Pro polymorphism altered αvβ3-dependent cell migration in transfected CHO cells. Cell lines expressing equivalent levels of the Leu33 and Pro33 isoforms of αvβ3 were generated, and flow cytometric analysis using LM609 demonstrated equivalent expression of αvβ3 in these cell lines (data not shown). In haptotactic migration assays, compared with Leu33-expressing αvβ3 cells (Leu33v), significantly more Pro33-expressing cell (Pro33v) migration was observed on vitronectin (1.4 fold; P<0.05) and osteopontin (1.3 fold; P<0.05) (Figure 5A). There was a nonsignificant decrease in Pro33v cell migration on fibrinogen. Migration on vitronectin was αvβ3-dependent, because it could be completely blocked by pretreatment of the cells with the anti-β3 mAb 7E3 (αvβ3 is the only β3 integrin in these cells), whereas a control antibody RMV7 did not block migration (Figure 5B). Similar specificity was shown on osteopontin (data not shown).
Increased migration of Pro33v cells was shown to be dose-dependent, because it could be observed over a range of vitronectin concentrations (Figure 6A). Although a time-dependent effect of increased Pro33v migration on vitronectin was also observed (2, 4, and 6 hours, Figure 6B), the magnitude of the difference between Leu33v and Pro33v cell migration at 4 and 6 hours was smaller than that at 2 hours. Finally, we studied the migration of all 6 cell lines (vector only, Leu33, and Pro33 for each integrin [αIIbβ3 or αvβ3]) on both substrates (fibrinogen or vitronectin) in the same experiment and observed the same results (not shown).
There is a well-established genetic influence on the development of myocardial infarction as well as several of the pathophysiologic components of myocardial infarction, such as atherogenesis and thrombosis.5 Our findings extend the genetic influence to another component of adverse events related to coronary heart disease, cell migration. Using haptotactic migration assays, we have found that cells expressing the Pro33 variant of integrin αIIbβ3 show a greater migratory response toward fibrinogen and vWF, and when this variant is part of the αvβ3 complex, there was greater migratory response toward vitronectin and osteopontin. This Pro33 effect on enhanced migration was substrate and integrin α subunit specific and depended on the COX and ERK1/2 signaling pathways. These genetic influences on integrin migratory responses have potential implications for SMC migration in atherosclerosis and restenosis, wound healing, and platelet-mediated clot retraction.
Pro33 Selectively Enhances Cell Migration
CHO cell migration on fibrinogen mediated by transfected αIIbβ3 has been demonstrated previously,18 but other substrates and the Leu33Pro polymorphism were not considered. Using a similar system of stably transfected CHO cells, we have shown that Pro33 cells exhibit greater migration than Leu33 cells (Figure 1A) in an αIIbβ3-dependent manner (Figure 1B). The possibility that the Pro33 CHO cells underwent clonal variation favoring increased migration is unlikely, because the same effect was observed in a second set of independently generated cell lines (Figure 3A). Furthermore, we found that CHO αvβ3 Pro33v cells exhibited greater migration on vitronectin and osteopontin than Leu33v cells in a β3-dependent manner (Figure 5). We speculate that enhanced platelet aggregation and clot retraction in the case of Pro33-αIIbβ3 and EC and SMC migration in the case of Pro33-αvβ3 could contribute to the pathophysiology of restenosis in the setting of PCTA.
Substrate and α-Subunit Specificity
The enhanced migration of Pro33-αIIbβ3 cells on fibrinogen cannot be attributed solely to a more favorable interaction with the γ-chain dodecapeptide of fibrinogen, because this latter sequence is not present in vWF. Likewise, the effect cannot be dependent only on RGDS, because these sequences are also present in fibronectin and vitronectin, which did not support enhanced migration of Pro33 cells. The Pro33 conformation most likely does not produce a novel ligand binding site on αIIbβ3, because migration could be blocked by antibodies and peptides that routinely block the Leu33 isoform. Our data do not exclude the possibility that the Pro33 conformation of αIIbβ3 reacts with a novel site of fibrinogen and vWF that is not on fibronectin or vitronectin. Our findings additionally suggest that the α integrin subunit contributes to the selectivity for enhanced migration on a substrate. Compared with Leu33 cells, only αIIbβ3 and not αvβ3 Pro33 cells had an enhanced migration on fibrinogen. Similarly, only CHO Pro33 cells expressing αvβ3, and not αIIbβ3, had an enhanced migration on vitronectin, compared with Leu33 cells. It is not clear how the integrin α subunit would confer such substrate selectivity for enhanced migration, but presumably there are unique conformations of the various possible combinations of integrin pairs that affect either the migration-dependent affinity state or postreceptor occupancy signaling potential on some substrates but not others.
Molecular Mechanisms of Migration
Cell migration is a complex and incompletely understood process wherein the leading edge of a cell must extend, attach, and contract, and this must be coordinated with the timely posterior release of the cell (reviewed by Huttenlocher et al21). Integrin ligation and dissociation with the substratum are central actions in this process, and both αIIbβ3 and αvβ3 have been used as models for studying integrin behavior in cell migration.18,22⇓ Reduced αIIbβ3-mediated haptotactic migration has been observed with mAb LIBS6-induced activation of αIIbβ3 and αIIb cytoplasmic domain mutations that cause constitutively increased soluble fibrinogen binding.18 Somewhat similarly, a T562N mutation in the extracellular cysteine-rich repeat region of integrin β3 leads to activation of αvβ3 defined as spontaneous binding of soluble fibrinogen and increased adhesion but reduced αvβ3-mediated migration to immobilized fibrinogen and vitronectin.23 The Cys435Ala β3 mutation examined in Figure 1E resembled these mutations by spontaneous soluble fibrinogen binding,16 but we observed no significant effect on migration. Although these findings suggest an inverse relationship between integrin affinity/avidity and efficient migration, this is not true for all β3 amino acid substitutions, because an αvβ3-expressing D723R mutation in the cytoplasmic tail of β3 exhibited enhanced migration toward vitronectin, as well as enhanced metastatic potential, despite increased adhesion to immobilized vitronectin and fibrinogen.24 This latter example resembles the Pro33 variant, which also shows increased migration despite enhanced adhesion to substrates.9 Taken together, these studies emphasize the need to describe the specificity of “function” in a “gain-of-function” mutation, because alterations in one function (eg, increased adhesion to immobilized substrate) do not necessarily predict the effect on another function (eg, cell migration). Such predictions may be possible as various integrin domains become better defined.
ERK1/2 and COX Dependence of Pro33 Effect
Cell de-adherence during migration involves disassembly of focal adhesions, followed by reorganization of the actin cytoskeleton, and is associated with activation of the MAP kinase pathway.25 We found that MAP kinase inhibition can also abolish the difference between Leu33 and Pro33 αIIbβ3 cell migration on fibrinogen (Figure 3B). This finding is consistent with our data that Pro33 cells exhibit greater ERK phosphorylation than Leu33 cells when adhering to immobilized fibrinogen (unpublished data). Because ERK1/2 phosphorylates and activates cytosolic phospholipase A2 to augment arachidonic acid synthesis, we studied the effect of COX inhibition on cell migration. Aspirin (and U0126) had no effect on the migration of Leu33 αIIbβ3 cells on fibrinogen but completely abolished the Pro33 increase in migration (Figure 3B). Thus, two post-receptor occupancy-signaling systems (ERK1/2 and COX) seem to regulate the greater migration seen in cells expressing the Pro33 variant of β3.
We have shown that the Leu33Pro polymorphism alters the migratory behavior of cells in a substrate and α-subunit specific manner. Unfortunately, the crystal structure of β3 was unclear around amino acid 33.26 Once the structures of this region of β3 and the complete αIIbβ3 complex are available, we may have information to help explain this functional interaction between the β3 variation and the α subunit. The clinical consequences of these findings will also have to wait pharmacogenetic studies. The rationale for aspirin therapy after coronary angioplasty and stent placement has been the antithrombotic effects of aspirin. Pro33-positive patients would be expected to derive particular benefit for inhibiting platelet function,7,27–29⇓⇓⇓ but our data raise the possibility that aspirin may benefit Pro33-positive patients more than Pro33-negative patients by providing an added benefit of inhibiting the prostenotic effect of vascular cell migration. In addition, pharmacologic blockade of both αIIbβ3 and αvβ3 by c7E3 Fab (abciximab; Reopro) has been shown to reduce the risk of ischemic complications after PCTA.1,30⇓ In patients undergoing percutaneous coronary intervention, heterozygous Leu33/Pro33 platelets were inhibited by abciximab to a lesser extent than Leu33/Leu33 platelets in an in vitro platelet function assay.31 This effect may be different in young healthy subjects.7 However, no information is available to indicate whether these commonly used therapies are modulated by the Leu33Pro polymorphism for outcomes after PCTA. Increased cell migratory potential in Pro33-positive patients warrants additional pharmacogenetic studies in the setting of coronary vessel revascularization.
Received August 15, 2002; revision accepted October 14, 2002.
- ↵Shattil SJ, Kashiwagi H, Pampori N. Integrin signaling: the platelet paradigm. Blood. 1998; 91: 2645–2657.
- ↵Williams MS, Bray PF. Genetics of arterial prothrombotic risk states. Exp Biol Med (Maywood). 2001; 226: 409–419.
- ↵Michelson AD, Furman MI, Goldschmidt-Clermont P, Mascelli MA, Hendrix C, Coleman L, Hamlington J, Barnard MR, Kickler T, Christie DJ, Kundu S, Bray PF. Platelet GP IIIa PlA polymorphisms display different sensitivities to agonists. Circulation. 2000; 101: 1013–1018.
- ↵Feng D, Lindpaintner K, Larson MG, Rao VS, O’Donnell CJ, Lipinska I, Schmitz C, Sutherland PA, Silbershatz H, D’Agostino RB, Muller JE, Myers RH, Levy D, Tofler GH. Increased platelet aggregability associated with platelet GPIIIa PlA2 polymorphism: the Framingham Offspring Study. Arterioscler Thromb Vasc Biol. 1999; 19: 1142–1147.
- ↵Undas A, Brummel K, Musial J, Mann KG, Szczeklik A. PlA2 polymorphism of β3 integrins is associated with enhanced thrombin generation and impaired antithrombotic action of aspirin at the site of microvascular injury. Circulation. 2001; 104: 2666–2672.
- ↵Casscells W. Migration of smooth muscle and endothelial cells: critical events in restenosis. Circulation. 1992; 86: 723–729.
- ↵Smyth SS, Reis ED, Zhang W, Fallon JT, Gordon RE, Coller BS. β3-integrin-deficient mice but not P-selectin-deficient mice develop intimal hyperplasia after vascular injury: correlation with leukocyte recruitment to adherent platelets 1 hour after injury. Circulation. 2001; 103: 2501–2507.
- ↵Shpilberg O, Rabi I, Schiller K, Walden R, Harats D, Tyrrell KS, Coller B, Seligsohn U. Patients with Glanzmann thrombasthenia lacking platelet glycoprotein αIIbβ3 (GPIIb/IIIa) and αvβ3 receptors are not protected from atherosclerosis. Circulation. 2002; 105: 1044–1048.
- ↵Bini A, Fenoglio JJ Jr, Mesa-Tejada R, Kudryk B, Kaplan KL. Identification and distribution of fibrinogen, fibrin, and fibrin(ogen) degradation products in atherosclerosis: use of monoclonal antibodies. Arteriosclerosis. 1989; 9: 109–121.
- ↵Suh TT, Holmback K, Jensen NJ, Daugherty CC, Small K, Simon DI, Potter S, Degen JL. Resolution of spontaneous bleeding events but failure of pregnancy in fibrinogen-deficient mice. Genes Dev. 1995; 9: 2020–2033.
- ↵Liu C-Y, Sun Q-H, Wang R, Paddock CM, Newman PJ. Disruption of the long-range GPIIIa Cys5-Cys435 disulfide bond results in the production of a constitutively active GPIIb-IIIa integrin complex. Blood. 1997; 90: 573a.
- ↵Klemke RL, Yebra M, Bayna EM, Cheresh DA. Receptor tyrosine kinase signaling required for integrin αvβ5-directed cell motility but not adhesion on vitronectin. J Cell Biol. 1994; 127: 859–866.
- ↵Huttenlocher A, Ginsberg MH, Horwitz AF. Modulation of cell migration by integrin-mediated cytoskeletal linkages and ligand-binding affinity. J Cell Biol. 1996; 134: 1551–1562.
- ↵Stockton RA, Jacobson BS. Modulation of cell-substrate adhesion by arachidonic acid: lipoxygenase regulates cell spreading and ERK1/2-inducible cyclooxygenase regulates cell migration in NIH-3T3 fibroblasts. Mol Biol Cell. 2001; 12: 1937–1956.
- ↵Filardo EJ, Brooks PC, Deming SL, Damsky C, Cheresh DA. Requirement of the NPXY motif in the integrin β3 subunit cytoplasmic tail for melanoma cell migration in vitro and in vivo. J Cell Biol. 1995; 130: 441–450.
- ↵Kashiwagi H, Tomiyama Y, Honda S, Kosugi S, Shiraga M, Kiyoi T, Kato Y, Kurata Y, Matsuzawa Y. Modulation of cell adhesion and migration by constitutive activation of integrin αvβ3. Thromb Haemost. 2001; 86: 1188a.
- ↵Felding-Habermann B, O’Toole TE, Smith JW, Fransvea E, Ruggeri ZM, Ginsberg MH, Hughes PE, Pampori N, Shattil SJ, Saven A, Mueller BM. Integrin activation controls metastasis in human breast cancer. Proc Natl Acad Sci U S A. 2001; 98: 1853–1858.
- ↵Xie H, Pallero MA, Gupta K, Chang P, Ware MF, Witke W, Kwiatkowski DJ, Lauffenburger DA, Murphy-Ullrich JE, Wells A. EGF receptor regulation of cell motility: EGF induces disassembly of focal adhesions independently of the motility-associated PLCγ signaling pathway. J Cell Sci. 1998; 111: 615–624.
- ↵Xiong JP, Stehle T, Zhang R, Joachimiak A, Frech M, Goodman SL, Arnaout MA. Crystal structure of the extracellular segment of integrin αvβ3 in complex with an Arg-Gly-Asp ligand. Science. 2002; 296: 151–155.