Faculty Bibliography

Polyethylene oxide (PEO) of molecular weights 5,000, 10,000, 18,500, and 100,000 g/mol was covalently grafted to surfaces of otherwise cell adhesive polyethylene terephthalate (PET) films. Analysis of these surfaces by measurement of contact angles and ESCA verified the presence of the grafted PEO. Protein adsorption assays of radiolabeled albumin and fibrinogen showed a marked reduction in adsorbed protein for the 18,500 and 100,000 molecular weight PEO coupled surfaces. Cell growth assays using human foreskin fibroblasts in culture showed that the higher-molecular-weight PEO surfaces supported cell growth to a much lower extent than the two lower-molecular-weight PEOs. Flow of whole blood over these surfaces and visualization of platelet adherence using epifluorescence video-microscopy showed very low platelet adherence only on the two higher-molecular-weight PEO coupled surfaces. Scanning electron microscopy corroborated these results. It was concluded that PEO of molecular weights neighboring 18,500 and higher was effective in reducing protein adsorption and cellular interactions on these surfaces.

We have found a novel adhesion receptor on the human endothelial cell for the peptide sequence Arg-Glu-Asp-Val (REDV), which is present in the III-CS domain of human plasma fibronectin, with a dissociation constant of 2.2 x 10(-6) M and 5.8 x 10(6) sites/cell. When a synthetic peptide containing this sequence was immobilized on otherwise cell nonadhesive substrates, endothelial cells attached and spread but fibroblasts, vascular smooth muscle cells, and platelets did not. Endothelial monolayers on REDV were nonthrombogenic: endothelial cells attached and spread upon other receptor-binding domains of fibronectin and laminin, but with lesser degrees of specificity or with a loss of nonthrombogenicity. This approach may provide a basis for a tissue engineered vascular graft where endothelial cell attachment is desired, but not the attachment of other blood vessel wall cells and blood platelets.

Laser-light scattering was used to observe and quantify the dynamics of human blood platelet aggregation in platelet-rich plasma (PRP). Aggregation was performed in a controlled shear environment by placing the PRP in the annular space between a rotating cylindrical rod and a stationary cylindrical tube. The instrument was capable of very sensitive continuous semi-quantitative measurements of chemically-induced microaggregation. As a demonstration of the technique, results are presented for ADP-induced aggregation at doses of 10, 1, and 0.1 microM and collagen-induced aggregation at a dose of 5 micrograms/ml, each at shear rates of 1,000 s-1 and 500 s-1. Extensive aggregation was observed in response to ADP at even the low dose of 0.1 microM, indicating a high sensitivity to microaggregates. The sensitivity of the ultimate size of the ADP-induced aggregates to ADP concentration was shear dependent. The formation of microaggregates by collagen stimulation was shown to be almost immediate, as contrasted with a 10-20 s typical lag when observed turbidometrically. Disaggregation was observed with 1 microM ADP, but this was only partial, as contrasted with the complete recovery of transmittance observed in the turbidometric technique. Electronic particle sizing and counting was employed to semiquantitatively verify the aggregate size distributions found from mathematical conversion of the laser-light scattering data.

A simple solution technique was used to incorporate polyethylene oxide (PEO, of 5000, 10,000, 18,500, and 100,000 g/mol) and other water-soluble polymers such as polyvinylpyrrolidone and polyethyl oxazoline into the surfaces of commonly used biomedical polymers such as polyethylene terephthalate, a polyurethane (Pellethane 2363-80AE), and polymethylmethacrylate. The presence of the water-soluble polymers on these surfaces was verified by using contact angle analysis and ESCA. Protein adsorption studies, fibroblast adhesion assays, and whole blood perfusions over these polymers showed that the surface modified with PEO 18,500 was the most effective in reducing all the tested biological interactions. It was concluded that PEO 18,500 had a chain length that was optimal, using this technique for surface incorporation, to reduce protein adsorption and hence prevent protein-mediated biological interactions.

The attachment, spreading, spreading rate, focal contact formation, and cytoskeletal organization of human umbilical vein endothelial cells (HUVECs) were investigated on substrates that had been covalently grafted with the cell adhesion peptides Arg-Gly-Asp (RGD) and Tyr-Ile-Gly-Ser-Arg (YIGSR). This approach was used to provide substrates that were adhesive to cells even in the absence of serum proteins and with no prior pretreatment of the surface with proteins of the cell adhesion molecule (CAM) family. This approach was used to dramatically enhance the cell-adhesiveness of substrates that were otherwise cell-nonadhesive and to improve control of cellular interactions with cell-adhesive materials by providing stably bound adhesion ligands. Glycophase glass was examined as a model cell-nonadhesive substrate prior to modification, and polyethylene terephthalate (PET) and polytetrafluoroethylene (PTFE) were examined as representative materials for biomedical applications. The peptides were surface-coupled by their N-terminal amine to surface hydroxyl moieties using tresyl chloride chemistry. Prior to peptide grafting, the PET and PTFE were surface hydroxylated to yield PET-OH and PTFE-OH. The PET-OH was less cell-adhesive and the PTFE-OH was much more cell-adhesive than the native polymers. Radioiodination of a C-terminal tyrosine residue was used to quantify the amount of peptide coupled to the surface, and these amounts were 12.1 pmol/cm2 on glycophase glass, 139 fmol/cm2 on PET-OH, and 31 fmol/cm2 on PTFE-OH. Although the glycophase glass did not support adhesion or spreading even in the presence of serum, the RGD- and YIGSR-grafted glycophase glass did support adhesion and spreading, even when the only serum protein that was included was albumin. Although PET and PTFE-OH supported adhesion when incubated in serum-supplemented medium, neither of these materials supported adhesion with only albumin present, indicating that cell adhesion is mediated by adsorbed CAM proteins. When these materials were peptide-grafted, however, extensive adhesion and spreading did occur even when only albumin was present. Since the peptide grafting is quite easily controlled and is temporally stable, while protein adsorption is quite difficult to precisely control and is temporally dynamic, peptide grafting may be advantageous over other approaches employed to improve long-term cell adhesion to biomaterials.

An endothelial cell monolayer with a single mechanically lysed cell was used as a model to examine the extent, kinetics, and nature of local calcium mobilization in the neighborhood of a wound. Individual endothelial cells from confluent monolayers were mechanically lysed with a minutien needle coupled to a micromanipulator while producing no observable mechanical trauma to the neighboring cells. Changes in calcium levels in individual cells surrounding the wound site were monitored by epifluorescence microphotometry with the calcium-sensitive fluorophore indo-1. Individual cells adjacent to the wound site showed a substantial increase in their intracellular calcium levels, almost as high as the calcium levels attained by ionophore controls. The magnitude of intracellular calcium mobilization in confluent monolayers decreased with distance from the wound site, and those cells located at a radius greater than seven cells from the wound site showed no change in their calcium levels. Thus, lysis of a single cell resulted in calcium mobilization in approximately 200 neighboring cells. The time necessary for intracellular calcium to reach maximum levels also increased with distance from the wound site. Calcium mobilization was partly intracellular and was inhibited by disrupting cell-cell coupling or by increasing gap junction resistance by heptanol. This mobilization was greatly attenuated in subconfluent endothelial monolayers, and it was not observed in fibroblasts or smooth muscle cells; furthermore, the effect was defective in monolayers intentionally contaminated with smooth muscle cells. This study examines the extent and possible mechanisms of local endothelial activation near a microscopic endothelial wound.

The role of the local synthesis of thrombin in platelet recruitment and thrombus stabilization in heparinized blood was examined in vitro. Mural thrombosis was visualized and measured in a thin, rectangular, collagen-coated capillary under controlled rheological conditions by using fluorescence digital videomicroscopy and fluorescence microphotometry. Thrombin activity was inhibited in heparinized blood by the synthetic competitive inhibitor, D-phenylalanyl-L-prolyl-L-arginyl chloromethylketone (FPRCH2Cl), resulting in a marked reduction in the rate of platelet accumulation on collagen surfaces, indicating a role for thrombin in platelet recruitment. Similar although lesser effects were observed by reducing thrombin synthesis with antibodies to factors II and X. To decouple the role of thrombin in platelet recruitment by direct stimulation of platelet activity from its role in thrombus stabilization via fibrin formation, thrombosis was measured in heparinized blood treated with the tetrapeptide glycyl-prolyl-arginyl-proline, which inhibits fibrin monomer assembly into fibrin. The ultimate level but not the initial rate of platelet accumulation was reduced markedly, indicating a role for fibrin in thrombus stabilization against hemodynamic forces. Scanning electron micrographs demonstrated fibrin stands in the heparinized control samples but not in the heparinized samples with glycyl-prolyl-arginyl-proline. These results demonstrate a role for the local action of thrombin synthesized on the surfaces of thrombi even under conditions when the thrombin exerts no bulk effect, such as under heparin anticoagulation. Furthermore, this role appears to be a result of both platelet recruitment and thrombus stabilization.

A series of 66 terpolymers of dl-lactide, glycolide, and epsilon-caprolactone was synthesized for the purpose of identifying those materials which exhibited rapid degradation in vitro. Polymers having half-lives from a few weeks to several months were identified. The morphology of each material was characterized by differential scanning calorimetry. A terpolymeric composition of 60% glycolide, 30% dl-lactide, and 10% epsilon-caprolactone, which exhibited a half-life of 17 days, was selected for further investigation. The hydrophilicity of this material was increased by performing the polymerization in the presence of a polyether prepolymer, Pluronic F-68, with the motivation of concomitantly reducing cell and tissue adhesion. An increase in the hydrophilicity of the material was apparent from contact angle measurements. Copolymerization with the prepolymer also resulted in a stronger and partly crystalline material which was mechanically stable at physiological temperature in water. A slight increase was observed in the half-life of the polymer relative to the base polymer due to the presence of the prepolymer.

The synthetic peptides Gly-Arg-Gly-Asp-Tyr and Gly-Tyr-Ile-Gly-Ser-Arg-Tyr, which contain Arg-Gly-Asp (RGD) and Tyr-Ile-Gly-Ser-Arg (YIGSR), the ligands for two important classes of cell adhesion receptors, were covalently coupled to a nonadhesive modified glass surface by the N-terminal Gly. The N-terminal Gly served as a spacer, and the C-terminal Y served as a site for radioiodination. These modified substrates supported the adhesion and spreading of cultured human foreskin fibroblasts (HFFs) independently of adsorbed proteins and, it was demonstrated that a covalently immobilized YIGSR-containing peptide has biological activity. The surface concentration of grafted peptide on the glass was measured by 125I radio-labeling and was 12.1 pmol/cm2. HFFs spread on both immobilized peptide substrates, but at much slower rates on grafted YIGSR glass surfaces than on the RGD-containing substrates. Cells formed focal contacts on the RGD-derivatized substrates in the presence or absence of serum. Focal contacts formed on the YIGSR-grafted surfaces only when serum was present in the medium and had morphologies different from those observed on the RGD-containing substrates. Serum influenced the organization of microfilaments and the extent of spreading of adherent cells, although adsorption of adhesion proteins was minimal on all substrates. This derivatization method produced chemically stable substrates which may be useful in studying receptor-mediated cell adhesion, as the quantity of peptide available at the surface may be precisely measured and controlled.