Faculty Bibliography

We have recently reported the attachment and spreading of human umbilical vein endothelial cells (HUVECs) upon substrates containing covalently grafted Arg-Glu-Asp-Val (REDV) peptide (Hubbell, J. A., Massia, S. P., Desai, N. P., and Drumheller, P. D. (1991) Bio/Technology 9, 568-572). This peptide has been reported to be the minimal active sequence within the CS5 site of the alternatively spliced type III connecting segment (IIICS) region of fibronectin, and the integrin alpha 4 beta 1 has been identified as the receptor on melanoma cells for this site. The integrin alpha 4 beta 1 has also been identified as the receptor for the CS1 site in the IIICS region on cells of neural crest origin, melanoma cells, lymphocytes, and hematopoietic stem cells. In this study, we demonstrate that this integrin also serves as a receptor on HUVECs for the peptide REDV from the CS5 site. The alpha 4 subunit was shown to be expressed upon HUVEC membranes by whole-cell enzyme-linked immunosorbent assay. Antifunctional antibodies directed against integrin subunits alpha 4 and beta 1 inhibited cell adhesion on REDV-grafted substrates, but not on RGD-grafted substrates. The alpha 4 subunit localized into fibrillar structures within spread cells on the REDV-grafted substrates, but not within spread cells on RGD-grafted substrates. Two proteins (144 and 120 kDa) were isolated from HUVEC extracts by REDV ligand affinity chromatography and were demonstrated by immunoprecipitation and Western blot to be the integrin subunits alpha 4 (144 kDa) and beta 1 (120 kDa); furthermore, the immunoprecipitation analyses demonstrated that the subunits formed a complex. HUVEC binding to REDV-grafted substrates was inhibited by both soluble REDV and RGD, demonstrating that adhesion was biospecific and that the REDV peptide is RGD-like. In this report we demonstrate for the first time that alpha 4 is present in the endothelial cell membrane, in contrast to previous reports by others, and that integrin alpha 4 beta 1 is the receptor for REDV-mediated adhesion to the IIICS region of region of plasma fibronectin.

The role of thromboxane A2 (TxA2) in platelet deposition onto collagen was studied in flowing whole heparinized human blood in vitro by using a cyclooxygenase inhibitor, aspirin, and a TxA2 receptor antagonist, GR32191B. Previous studies have demonstrated a role for TxA2 in platelet aggregation in citrated plasma, and for platelet deposition in flowing citrated human and rabbit blood, but not in flowing heparinized rabbit blood. In contrast with the literature regarding rabbit blood, aspirin was demonstrated to be effective in reducing platelet accumulation in heparinized human blood, as was GR32191B. The temporal pattern of the platelet deposition, which reached an asymptote with TxA2 inhibition at 2.0 minutes but did not do so without inhibition, suggested that TxA2 plays a role in thrombus stabilization. The reduction in platelet deposition (which included some aggregation) seen with aspirin and GR32191B in the first 1.5 minutes of perfusion indicated that some inhibition of platelet recruitment occurred. Scanning electron microscopy revealed that less fibrin was present in thrombi derived from GR32191B-treated heparinized blood than in thrombi derived from control heparinized blood. No fibrin formation was observed in citrated blood with or without TxA2 inhibition. It is proposed that, in addition to its role as a mediator of platelet recruitment, TxA2 is involved in the stabilization of platelet-platelet interactions in the thrombus, perhaps by enhancing local fibrin formation or binding.

Diamines covalently coupled to glass substrates promoted human foreskin fibroblast adhesion in the absence of serum. These diamine-derivatized substrates were produced by coupling ethylene diamine, N-methylaminoethylamine, and N,N-dimethylaminoethylamine (NNDMAEA), to sulfonyl chloride-activated glass. Electron spectroscopy for chemical analysis demonstrated that the diamines were coupled via their primary amine ends to produce a surface-bound secondary amine linked to a free amino moiety via a two-carbon spacer. NNDMAEA-modified substrates containing free tertiary amines supported the highest degree of cell spreading (73 +/- 7% actively spreading cells) and the most extensive cytoskeletal organization. Both the free tertiary and surface-bound secondary amines were shown to be required for cell spreading. Lysine- and arginine-grafted substrates supported cell spreading and cytoskeletal organization similar to that on NNDMAEA-modified substrates. Although some stress fibers were observed within spread cells on these substrates, focal contacts did not form. Heparinase treatment did not inhibit cell attachment or spreading to the diamine-derivatized substrates, however chondroitinase ABC inhibited cell attachment and spreading on all substrates; heparinase inhibited spreading on lysine- and arginine-derivatized substrates to a lesser extent. These results imply that cell attachment to these substrates was mediated primarily by cell surface chondroitin sulfate proteoglycans. This study demonstrates that covalently grafted NNDMAEA, lysine, and arginine can mimic the adhesion-promoting activity of the glycosaminoglycan-binding domains of cell adhesion proteins. This study also demonstrates that the interaction with these proteoglycans depends in a very sensitive manner on the particular structure of the immobilized amine.

The effect of the ADP receptor antagonists ATP and adenosine 5'-(beta, gamma-methylene)triphosphate (AMP-PCP), and the ADP-utilizing enzyme systems creatine phosphokinase/creatine phosphate (CPK/CP) and pyruvate kinase/phosphoenol pyruvate (PK/PEP) on platelet deposition onto type I collagen was examined. An in vitro perfusion system was used, which allowed continuous visualization of the deposition of fluorescently labelled platelets. This system also provide well-controlled rheology, precise quantification of deposition, and allowed the use of heparinized whole human blood (3 u/ml). Heparinization at this level permits the local generation of thrombin near surface platelet aggregates. The contribution of ADP is thus studied with the combined effects of thrombin, thromboxane A2, and other aggregating agents present. Results from these studies indicate that ATP was capable of inhibiting deposition by 60% at 1 microM and 90% at 5 microM (whole blood conc.). AMP-PCP inhibited deposition in a dose dependent manner with a Ki of approximately 80 microM and a maximum inhibition of 60%. Inhibition by CPK/CP was measured at 20, 40, and 60 u/ml, with approximately 45% inhibition achieved for the latter two concentrations. PK/PEP at 60 u/ml resulted in 70% inhibition. These results support a role for ADP in mediating platelet recruitment in thrombus growth on collagen. Previous work utilizing animal bleeding times supports this conclusion; the present study demonstrates that this role is not dependent upon endothelial or vasoconstrictive effects. Intraplatelet cAMP levels were raised with respect to controls upon exposure to ATP at 8.3 microM (p less than 0.025), and 15 microM (p less than 0.005), as well as AMP-PCP at 42-500 microM (p less than 0.005).(ABSTRACT TRUNCATED AT 250 WORDS)

A graft copolymer having poly(L-lysine) (PLL) as the backbone and monomethoxy poly(ethylene glycol) (MPEG) as pendent chains was synthesized. This polycationic copolymer was used to form microcapsules with sodium alginate, a polyanion. Microcapsules and model surfaces formed with PLL-graft-MPEG demonstrated reduced protein adsorption, complement binding and cell adhesion in vitro compared to materials with unmodified PLL. Microcapsules with PLL-g-MPEG on the surface were seen to be much more biocompatible than the widely used alginate/PLL/alginate microcapsule in a mouse intraperitoneal implant model. The graft copolymers demonstrated a lower affinity for alginate and increased microcapsule permeability more than PLL. To correct this, pentalayered alginate/PLL/alginate/PLL-g-MPEG/alginate microcapsules were fabricated, and these demonstrated both appropriate permselectivity and enhanced biocompatibility.

Polyethylene terephthalate films surface modified with polyethylene oxide of mol wt 18,500 g/mol (18.5 k) by a previously described technique, were implanted in the peritoneal cavity of mice, along with their respective untreated controls, for periods of 1-28 d. The implants were retrieved and examined for tissue reactivity and cellular adherence. The control polyethylene terephthalate surfaces showed an initial inflammatory reaction followed by an extensive fibrotic response with a mean thickness of 60 microns at 28 d. By contrast, polyethylene oxide-modified polyethylene terephthalate showed only a mild inflammatory response and no fibrotic encapsulation throughout the implantation period: at 28 d a cellular monolayer was observed. Apparently either the polyethylene oxide-modified surface was stimulating less inflammation, which was in turn stimulating less fibroblastic overgrowth, or the cellular adhesion to the polyethylene oxide-modified surface was too weak to support cellular multilayers.

Polyethylene terephthalate films were surface-modified with polyethylene oxide (18,500 g/mol) using a solution technique described previously. These films were investigated for their resistance to bacterial adhesion. Three bacterial strains most commonly associated with implant infections, Staphylococcus epidermidis, Staphylococcus aureus and Pseudomonas aeruginosa, were cultured in tryptic soya broth, human plasma and human serum on the polymeric substrates. Significant reductions (between 70 and 95%) in adherent bacteria were observed on the polyethylene oxide-modified substrates compared to the untreated control polyethylene terephthalate. Surface modification with polyethylene oxide may reduce the risk of implant-associated infections. Plasma fibrinogen was observed to play an important role in the adhesion of all three of these species on both the polyethylene oxide-modified and control polyethylene terephthalate materials.

The synthetic peptide Gly-Arg-Gly-Asp-Tyr (GRGDY), which contains the RGD sequence of several adhesion molecules, was covalently grafted to the surface of otherwise poorly adhesive glass substrates and was used to determine the minimal number of ligand-receptor interactions required for complete spreading of human foreskin fibroblasts. Well-defined adhesion substrates were prepared with GRGDY between 10(-3) fmol/cm2 and 10(4) fmol/cm2. As the adhesion ligand surface concentration was varied, several distinct morphologies of adherent cells were observed and categorized. The population of fully spread cells at 4 h reached a maximum at 1 fmol/cm2, with no further increases up to 10(4) fmol/cm2. Although maximal cell spreading was obtained at 1 fmol/cm2, focal contacts and stress fibers failed to form at RGD surface concentrations below 10 fmol/cm2. The minimal peptide spacings obtained in this work correspond to 440 nm for spreading and 140 nm for focal contact formation, and are much larger than those reported in previous studies with adsorbed adhesion proteins, adsorbed RGD-albumin conjugates, or peptide-grafted polyacrylamide gels. Vitronectin receptor antiserum specific for integrin alpha V beta 3 blocked cell adhesion and spreading on substrates containing 100 fmol/cm2 of surface-bound GRGDY, while fibronectin receptor antiserum specific for alpha 5 beta 1 did not. Furthermore, alpha V beta 3 was observed to cluster into focal contacts in spread cells, but alpha 5 beta 1 did not. It was thus concluded that a peptide-to-peptide spacing of 440 nm was required for alpha V beta 3-mediated cellular spreading, while 140 nm was required for alpha V beta 3-mediated focal contact formation and normal stress fiber organization in human foreskin fibroblasts; these spacings represent much fewer ligands than were previously thought to be required.