Although most biomedical devices are non-toxic, disturbed acute and chronic inflammation and the lack of integration in tissues is a concern. At the time of biomaterial insertion, protein adsorption onto material surfaces precedes cell adhesion and is believed to alter unfavorably the acute inflammatory response and the subsequent tissue healing. The wound healing may encapsulate the biomaterial in a fibrous tissue. The process depends probably on the surface physical and chemical characteristics, and the accumulation of blood plasma proteins such as fibrinogen, immunoglobulins (Ig:s) and complement. Platelets and neutrophil granulocytes, which both possess inflammatory capabilities, are the first cells to appear at a surface during contact with blood. In the present thesis, model biomaterial surfaces were prepared, and the in vitro deposition of plasma proteins and the subsequent behavior of neutrophils and platelets evaluated.
Complement activation at artificial surfaces during contact with blood is generally believed to proceed via the alternative pathway, i.e. through a direct covalent binding of the factor 3 (C3) thioester to nucleophilic surface groups (e.g. -OH and -NH2). The serum protein deposition onto a hydroxylated potent complement activator surface, mercaptoglycerol on gold, was studied by a combination of null-ellipsometry and polyclonal antibodies. It was observed that deposited C3 did not withstand elution with sodium dodecyl sulfate (SDS), and the binding was unaffected by reduction with hydroxylamine. Opposite results have been reported for biological surfaces and our findings call for a revision of the current activation model at artificial surfaces where instead the classical pathway of complement may be highly relevant.
The effects of immobilized and partially denatured IgG on the neutrophil respiratory burst at hydrophilic and hydrophobic model surfaces were studied by lurninol-arnplitied chemiluminescence in serum containing media. IgG supported frustrated phagocytosis and generation of extracellular reactive oxygen species (ROS) on both types of surfaces, although the kinetics were different. The response was particularly potent on IgG at hydrophobic surfaces, and the finding that the respiratory burst was only moderately quenched by the blocking of complement receptors (CR:s) or Fcγ (IgG) receptors, indicates a role for intracellular cross-talk. The IgG-triggered response depended on the presence of both C3 and C1q in serum and was inhibited by disruption of the intracellular actin dynamics. Classical complement activation may also be initiated by immobilized IgM. When the activation by spontaneously adsorbed IgG and IgM on methylated hydrophobic silicon was compared, both Ig:s deposited C3 from serum, but only the activation at IgG was C1q- and Ca2+-dependent. Depletion of C1q from serum lowered the neutrophil respiratory burst and the formation of intracellular filamentous (F) actin upon adhesion to IgG-surfaces. Hence, IgG- but not IgM-coated hydrophobic surfaces activate the classical pathway via the C1 complex.
Surface-bound IgG is also a potent platelet agonist via the Fcγ receptor. Neutrophil and platdet ROS generation, aggregation, and release of adenosine triphosphate in response to spontaneously adsorbed and covalendy immobilized IgG show that platelets enhance the neutrophil respiratory burst under both stirred and non-stirred serum free conditions. Blocking of the neutrophil Fcγ receptors was not sufficient to inhibit the amplification. Platelets supported neutrophil adhesion in a contact-dependent way, and the effect was mediated by intact platelets or platelet-derived fragments/microparticles. The response was, in contrast to complement dependent activation in serum, unaffected by the disruption of the actin cytoskeleton, or by blocking of neutrophil CR3 or platelet glycoprotein IIb/IIIa, suggesting an integrin- and fibrinogen-independent mechanism. Antibodies against platelet P-selectin (CD62) and P-selectin glycoprotein ligand-1 (PSGL-1 or CD162), but not L-selectin (CD62L), inhibited partly the neutrophil-platelet interaction, especially under shear. Accordingly, we suggest that during stimulation of the cells with immobilized IgG, platdet P-selectin interacts with neutrophil PSGL-1.
The majority of previous adsorption studies has dealt with blood plasma proteins. However, the concentration of released cytosolic proteins may locally reach high levels upon a tissue injury. Actin is one of the most abundant proteins in the eukaryotic cytoplasm, and may tentatively accumulate at interfaces. Actin was immobilized to gold and aminated silicon surfaces and polymerized into F-actin by adjusting the osmotic conditions. Upon incubation in human serum, the actin surfaces adsorbed serum proteins, amongst them C3 and C1q. However, the complement deposition was apparendy not a result of true or prolonged complement activation, and immobilized actin evoked only a low ROS-generation, aggregation, spreading and adhesion of neutrophils and platelets (similar to low-activating albumin-surfaces). Yet, F-actin on gold recruited platelets in a C1q-dependent manner, indicating an immunoregulatory capacity of surface-bound actin.
The results in the present thesis are relevant for a better understanding of the basic mechanisms that determine the fate of artificial devices in contact with human body fluids.
Linköping: Linköpings universitet , 2002. , 47 p.
2002-10-04, Hörsal Planck, Fysikhuset, Linköpings Universitet, Linköping, 10:15 (Swedish)