Leo Vroman

Adsorption of proteins out of plasma and solutions in narrow spaces

Attention of those studying the adsorption of proteins has, rather suddenly, become focused on extreme detail. New techniques are generating information on conformation and modes of attachment of adsorbed proteins (16). Circular dichroism, frustrated total reflection, electron scattering, and the use ofmonoclonal antibodies are providing insight in orientation and essential submolecular events that occur when a protein species adheres. Our own interest in these events originated with the study of blood, and hence of blood plasma, requiring us to identify the proteins being adsorbed out of this mixture in the course of a few minutes. Under such conditions, the techniques listed above do not identify the proteins adsorbed out of mixtures. For our purpose, antisera containing a set of antibodies to a specific human protein appear more useful to identify the protein than any monoclonal antibody that would provide information on specific groups remaining or becoming exposed as a specific protein is adsorbed. The following plasma proteins appear subjected to interchange with others after having been adsorbed. Fibrinogen. By combining ellipsometry and antisera, we found that on materials that activate clotting markedly, such as glass and anodized metals, intact normal plasma first deposits fibrinogen and thenwwithin 30 s if the plasma was undiluted--removes it again (7).

The effect of high molecular weight kininogen on surface-adsorbed fibrinogen

High molecular weight kininogen (HMWK) plays an important role in altering the association of plasma fibrinogen with surfaces. Plasma initially deposits fibrinogen onto most materials, but on hydrophilic surfaces within 10 min adsorbed plasma fibrinogen cannot be detected on the surface by anti-fibrinogen antisera. However, using HMWK-deficient plasma, fibrinogen remains immunologically identifiable. The interrelationship of adsorbed plasma fibrinogen with kininogen on hydrophilic surfaces is studied further using glass slides stained for protein with Coomassie Blue, and oxidized silicon crystal slices in an automated ellipsometer. On glass slides when plasma that is deficient in both low molecular weight kininogen (LMWK) and HMWK, is reconstituted with HMWK (0.04 Units/ml), fibrinogen is no longer detected on the surface. This finding is specific for HMWK, since, when the same plasma is reconstituted with LMWK (220 micrograms/ml), the amount of fibrinogen detected on the surface is unchanged. The alteration of surface-adsorbed fibrinogen by HMWK is not due to plasmin-induced fibrinolysis, since it occurs in plasminogen-free plasma. In the ellipsometer, surface adsorption of normal plasma is associated with a significantly less (p less than 0.0005) thick protein layer (1.99 +/- 0.08 degree change in azimuth) than plasmas deficient in HMWK (2.32 +/- 0.11). Using ellipsometry, HMWK in plasma is shown to shorten the time in which immunologically detectable surface-adsorbed fibrinogen was removed or altered. These studies in a whole plasma system present further evidence that HMWK specifically modifies the association of plasma fibrinogen with hydrophilic surfaces.

Three simple ways to detect antibody—Antigen complex on flat surfaces

Abstract Three simple ways of demonstrating the presence of antibody that has been adsorbed upon flat antigen covered surfaces are described. They are: (1) exposure to condensing water vapor to show increased wettability, (2) staining with Coomassie Brilliant Blue R and (3) observation of changing interference color on anodized tantalum. By premixing known and unknown amounts of antigen with antiserum and then observing interactions with antigen that has been preadsorbed on flat surfaces, crude but rapid immunoassays can be carried out.

The complexity of blood at simple interfaces

Abstract In the study of blood at interfaces, a combination of simple questions, techniques, and data is yielding increasingly intricate answers. We had shown earlier that the behavior of the plasma protein, fibrinogen, in blood at interfaces depends on the adsorbing substrate and determines subsequent platelet adhesion. In the present report we show that another plasma protein component, γ-globulin, when adsorbed out of pure solutions onto a hydrophobic rather than a hydrophilic substrate, causes certain white blood cells to adhere. On a wettable surface such as glass, adsorbed γ-globulins cause adhesion of such white cells only if the film had once been dried. Though whole plasma also deposits γ-globulins on these various surfaces, this film—in contrast to one produced by the purified protein solutions—does not markedly enhance white cell deposition. Recording ellipsometry and simpler techniques thus far failed to show differences in orientation of the adsorbed globulin that would account for the described white cell behavior. Our findings suggest that multiple interactions among plasma proteins at interfaces are responsible for the visible results of contact between blood and surfaces. In addition, findings by others indicate that the first phase of blood clotting is interwoven with these interactions, creating both positive and negative feedback systems within these events.

Rapid identification of proteins on flat surfaces, using antibody-coated metal oxide suspensions.

Suspensions of Fe3O4 (black), Fe2O3 (red), and Cr2O3 (green) were exposed to solutions of protein A, and then each to a different antiserum to one of the following human proteins: fibrinogen, high molecular weight kininogen (HMK), albumin or immunoglobulins (IgG). Test surfaces were patterns of human proteins adsorbed out of solutions or out of plasma, onto glass as well as onto polyvinylchloride slides. They were exposed to single or mixed suspensions of the treated oxides for about 30 s and rinsed. Adhesion of each oxide onto each matching protein of these patterned test surfaces resulted, thus identifying each protein by color.

The nature of substrate-attached materials in human fibroblast cultures: Localization of cell and fetal calf serum components

Abstract Human fibroblasts have been used as an in vitro model to examine the morphology and origin of substrate-attached materials. In cultures of subconfluent cells, no ‘tracks’ or ‘pools’ of material could be detected on substrata by anodic oxide interferometry or electron microscopy. However, a continuous layer of densely staining material was present on Falcon plastic tissue culture dishes never exposed to cells or culture medium. Exposure of substrata to culture medium caused the adsorption of fetal calf serum (FCS) components onto the substratum within a few minutes. Although antigenic FCS components remained on the substrata for several days, they were seldom adsorbed to the cells. The hypothesis was formulated that adhesion was mediated by FCS components on the substrata, but not by cellular materials deposited extracellularly. Support for this hypothesis was obtained by studying serum-dependent differences in cell adhesion. Fibroblasts subcultured in the presence of FCS components were usually separated from the substratum by a distance of at least 30 A. In the absence of FCS components, the cells were more closely adherent, in the range at which the near van der Walls forces were effective. Fibroblasts subcultured in the absence of serum components could be removed readily from the substratum, leaving lsfootprints’ of cell surface material behind. Although this material has been prepared similarly to ‘microexudates’ from other types of cultured cells, its relationship to those microexudates has not been determined.

Effects of protamine sulfate, polybrene and heparin on the behavior of plasma, plasma proteins, platelets and factor XII activity at interfaces

Abstract Using various techniques such as ellipsometry, and measurement of the ability of adsorbed plasma matter to correct the clotting time of factor XII deficient plasma and to cause adhesion of platelets suspended in defibrinogenized plasma, we found that on glass, anodized tantalum, oxidized silicon crystal slices (SiW) and on some other surfaces, plasma deposits fibrinogen and factor XII, the former causing platelets to adhere, the latter causing clotting. Within one minute, the plasma, if intact, converts the fibrinogen film and renders it less attractive to platelets. Then, in the presence of intact factor XII, some of the film is removed. In the present study, we found that on preadsorbed protamine sulfate (PS), plasma behaved normally; yet, no platelets adhered to these films and clotting was inhibited by them. When the PS films had been preexposed to heparin, they adsorbed more factor XII, no longer inhibited the clotting of normal intact plasma and no longer adsorbed fibrinogen out of it. Surfaces that had adsorbed Polybrene (PB) also inhibited clotting of plasma; they did not adsorb factor XII activity out of plasma but large numbers of platelets adhered, whether or not the PB films had been exposed to heparin. Yet, material adsorbed by PB or by heparinized PB was not fibrinogen. Various purified plasma proteins behaved differently at PB and PS surfaces and their preadsorbed films were affected variously by PB and PS. In solution, PS and heparin had little effect on the behavior of plasma at SiW surfaces, but PB reduced its ability to convert the deposited fibrinogen.

IN VITRO EVALUATION SYSTEMS: A BIOLOGICAL APPROACH

The primary problem becoming most apparent during the Workshop session on this subject was: What is the purpose and meaning of our in vitro tests? Many of us have recently become aware, I believe, of the danger in first developing a theory on the basis of a rather arbitrarily chosen test system, and then limiting ourselves to the same test system in efforts to prove the theory. Dr. Silberberg pointed out that the limitations imposed by specific techniques upon the way in which surfaces are to be prepared often appear unrelated to planned medical application. I stressed my own position: in our laboratory we promise no more than data that indicate what proteins are adsorbed under our conditions onto more or less flat material surfaces out of plasma, what reactions these adsorbed proteins may be subjected to by the plasma, and what effects the resulting film may have on the adhesion of platelets and white blood cells. We are not testing for thrombogenicity. Those who claim they do have a predictive test for in vivo behavior will have to correlate much in vitro and in vivo data for many biomaterials to prove their claim. This does not relieve any of the work load resting on those now occupied with true (namely, in vivo) thrombogenicity testing. Yet, the original purpose of in vitro testing must have been to provide such relief. Let us presume, then, that the primary purpose of studying the blood/ material interface reactions in vitro now is to understand how certain chains of events-especially those no longer than a few hours-are set in motion by certain surface properties. Removed as far away as possible from the living organism, we replace its complex and therefore confusing homeostatic system by a different, lifeless, and therefore more controllable kind of constancy. If, however, its isolation renders plasma subject to scrutiny so close that single molecular films can be observed, then the biomaterial substrate must be studied at least as closely. At this initial point of study we are already blocked by the nonbiological problem that sample surfaces, especially of many polymers, are nonuniform beyond description, as Dr. Yasuda, Dr. Brash, Dr. Chawla, Dr. Stromberg, and others pointed out at this session. The variability of these surfaces requires handling history as well as preparation of samples to be reproducible. Why can even the most minute surface discontinuities have such profound effects upon the behavior of blood that they will determine the fate of an implant? Perhaps it is because blood must have the ability to defend itself against surface accidents by means of its molecular contact-sensitive system. The sensitivity of this system provides a biologically amplified image of the material’s physical and chemical surface properties, as soon as the blood “sees” this surface. The roughness described by Dr. Merrill earlier as either being of a scale of about 1000 &-and significant to protein molecules-or of a scale of about 100 microns-and significant to cells-is only one example where