Attachment of human endothelial cells to polyester vascular grafts: pre-coating with adhesive protein assemblies and resistance to short-term shear stress.

Cardiovascular prosthetic bypass grafts do not endothelialize spontaneously in humans, and so they pose a thrombotic risk. Seeding with cells improves their performance, particularly in small-caliber applications. Knitted tubular polyethylene-terephthalate (PET) vascular prostheses (6 mm) with commercial type I collagen (PET/Co) were modified in the lumen by the adsorption of laminin (LM), by coating with a fibrin network (Fb) or a combination of Fb and fibronectin (Fb/FN). Primary human saphenous vein endothelial cells were seeded (1.50 × 10(5)/cm2), cultured for 72 h and exposed to laminar shear stress 15 dyn/cm(2) for 40 and 120 min. The control static grafts were excluded from shearing. The cell adherence after 4 h on PET/Co, PET/Co +LM, PET/Co +Fb and PET/Co +Fb/FN was 22%, 30%, 19% and 27% of seeding, respectively. Compared to the static grafts, the cell density on PET/Co and PET/Co +LM dropped to 61% and 50%, respectively, after 120 min of flow. The cells on PET/Co +Fb and PET/Co +Fb/FN did not show any detachment during 2 h of shear stress. Pre-coating the clinically-used PET/Co vascular prosthesis with LM or Fb/FN adhesive protein assemblies promotes the adherence of endothelium. Cell retention under flow is improved particularly on fibrin-containing (Fb and Fb/FN) surfaces.

Interaction of blood plasma with antifouling surfaces.

Nonspecific adsorption of proteins is a crucial problem in the detection of analytes in complex biological media by affinity sensors operating with label-free detection. We modified the gold surface of surface plasmon resonance (SPR) sensors with three types of promising antifouling coatings: self-assembled monolayers (SAM)s of alkanethiolates terminated with diethylene glycol and carboxylic groups, poly(ethylene glycol) (PEG) grafted onto the SAMs, and zwitterionic polymer brushes of poly(carboxybetaine methacrylate), poly(sulfobetaine methacrylate), and poly(phosphorylcholine methacrylate). Using SPR, we compared the efficacy of the coatings to reduce nonspecific adsorption from human blood plasma and from single-protein solutions of human serum albumin, immunoglobulin G, fibrinogen, and lysozyme. There was no direct relationship between values of water contact angles and plasma deposition on the coated surfaces. A rather high plasma deposition on SAMs was decreased by grafting PEG chains. Fouling on PEG was observed only from plasma fractions containing proteins with molecular mass higher than 350 000 Da. The adsorption kinetics from plasma collected from different healthy donors differed. Poly(carboxybetaine methacrylate) completely prevented the deposition from plasma, but the other more hydrophilic zwitterionic polymers prevented single-protein adsorption but did not prevent plasma deposition. The results suggest that neither wettability nor adsorption of the main plasma proteins was the main indicator of deposition from blood plasma.

Optical biosensors for real-time measurement of analytes in blood plasma.

The preparation of assemblies consisting of multiple molecular layers of bovine serum albumin (BSA), monoclonal antibodies against horseradish peroxidase (anti-HRP), and monoclonal antibodies against methotrexate (anti-MTT), as well as interaction of the assemblies with human blood plasma were observed using a grating coupler and Young interferometer (YI). The assemblies could be arranged according to decreasing amounts of nonspecific deposits bound irreversibly to them from blood plasma as follows-an adsorbed antibody monolayer saturated with adsorbed BSA, antibody multilayers linked with polycations, antibodies covalently immobilized on a BSA layer densely crosslinked with glutaraldehyde (GA), slightly crosslinked BSA double layer, slightly crosslinked antibody double layers. The occurrence of human serum albumin (HSA), human fibrinogen (Fg), IgG, and IgM in the plasma deposits was studied by binding the respective antibodies. IgG, IgM, and Fg were detected in plasma deposits on the immobilized assemblies while the composition of a plasma deposit on the unmodified sensor surface reflected roughly the plasma composition containing mainly adsorbed HSA and Fg. A crosslinked anti-HRP double layer was immobilized on a waveguiding branch of YI and a similar anti-MTT double layer was immobilized on the other branch. The sensor response to blood plasma was fairly decreased owing to a compensation of the respective optical changes in the two branches, in which a similar non-specific adsorption took place. The addition of HRP or MTT to plasma induced specific responses of the corresponding branches.

Albumin and heparin multilayer coatings for blood-contacting medical devices.

Three types of covalently crosslinked assemblies consisting of multiple (1) molecular layers of human serum albumin (HSA); (2) alternating layers of HSA and unfractionated heparin; and (3) alternating layers of HSA and partly depolymerized heparin fixed with one end to HSA were prepared on various surfaces. Adsorption of fibrinogen, IgG, and antithrombin (ATIII) from human citrated plasma on coated surfaces was evaluated by ELISA. Fibrinogen adsorption on coated ELISA plates was lower than that on bare polystyrene. There was no IgG adsorption on the HSA coating alone, but considerably high IgG adsorption was detected on the heparin-containing surface. The adsorption of ATIII increased with increasing heparin on the surface. The effect of multilayer coatings on platelets was tested by incubation of modified vascular prostheses with citrated blood. The most favorable interaction with platelets was observed on the HSA assembly. The interaction of platelets with the surface bearing unfractionated heparin was higher than that of the surface covered with partly depolymerized heparin. The long-term durability of the HSA-heparin coating was proven by a 21-day implantation of coated polyurethane plates in goat heart.

The detection of human beta 2-microglobulin by grating coupler immunosensor with three dimensional antibody networks.

Immunosensors for the detection of human beta 2-microglobulin (B2M) were prepared by immobilisation of covalently crosslinked assemblies containing various numbers of molecular layers of monoclonal antibody against B2M (anti-B2M) on the surface of a Ta2O5 grating coupler sensor. The immobilisation procedure consisted of repeated successive adsorption of anti-B2M and dextran sulfate (DS) followed by glutaraldehyde (GA) crosslinking of anti-B2M and washing out DS. The flexibility of the resulting anti-B2M networks was evaluated from the sensor response to the reversible expansion and contraction of the networks induced by changing pH of the ambient solution. A decreased GA concentration and the use of a higher-molecular-mass DS increased the network flexibility. The sensor sensitivity to B2M increased with increasing flexibility of the antibody networks and with increasing number of anti-B2M molecular layers, indicating that B2M can penetrate inside the antibody network.

Immobilisation of multilayer bioreceptor assemblies on solid substrates.

Multilayer assemblies were prepared by alternating adsorption of monolayers of monoclonal antibody against horse radish peroxidase (anti-HRP) and dextran sulfate (DS) on solid supports at acid pH. After crosslinking with glutaraldehyde, DS was washed out of the film with buffered physiological saline, while the antibody remained immobilised on the support. Assembly was monitored in situ on germanium supports by infrared multi-internal reflection spectroscopy. The binding capacity of the immobilised antibodies for HRP was measured by ELISA and by optical waveguide light mode spectroscopy. The activity of an immobilised anti-HRP bilayer was approximately twice that of a monolayer prepared by simple physiosorption. An addition of further anti-HRP layers could increase the activity only up to 2.5 of the monolayer activity independently of a number of layers in the assembly. The non-specific adsorption of proteins from human blood plasma was three times lower on the immobilised anti-HRP multilayer film than on the surface covered only with a physiosorbed anti-HRP monolayer.

Haemocompatibility of segmented polyurethanes investigated in vivo.

Samples of segmented polyurethanes differing in composition and in surface morphology were introduced into the left ventricles of hearts of goats for 72 h. After removal of the samples, their surface and the surface of the heart endothelium were evaluated visually with respect to the formation of thrombi. Differences in the interaction of the individual polyurethanes with blood were examined by XPS photoelectron spectroscopy, SEM, and by infrared (i.r.) reflexion spectroscopy. The results suggest ways for improving the haemocompatibility of the surfaces of polyurethanes used as parts of the total artificial heart.

Polyethylene/hydrophilic polymer blends for biomedical applications.

Polyethylene blends with poly(2-hydroxyethyl methacrylate) [poly(HEMA)] or poly(2,3-dihydroxypropyl methacrylate) [poly(DHPMA)] were prepared by swelling polyethylene with HEMA or 2,3-epoxypropyl methacrylate (EPMA) and by polymerization of the respective monomers. Poly(EPMA) in blends was hydrolysed to poly(DHPMA) with acetic acid. The blends had similar surface and bulk compositions. Swelling with water and surface wettability were proportional to the content of the hydrophilic component; at the same content the polyethylene/poly(DHPMA) blends appeared more hydrophilic than those of polyethylene/poly(HEMA). Thrombus formation in contact with blood examined ex vivo and in vivo was considerably slower on the blends than on unmodified polyethylene. The tests indicated optima in composition; the best biological response was achieved with the blends containing about 14% poly(HEMA) or 16% poly(DHPMA).