Experiments with factor 8 separated from fibrinogen by electrophoresis in free buffer film.

Abstract

Approaches to an understanding of the chemistry of Factor VIII (antihaemophilic factor) have been few and necessarily indirect since Factor VIII occurs in plasma in only trace amounts, possibly less than 1 mg./l., is very labile, and the biological assay methods are subject to large errors. There have been repeated suggestions that the actual active Factor‐VIII molecule might be rather small and that in plasma it is associated with another protein (Thelin and Wagner, 1961; Surgenor, 1964). Methods of purification of Factor VIII are essentially modifications of methods applicable to fibrinogen. The highest purification of Factor VIII is probably that of Michael and Tunnah (1963) but the protein of even their best preparation is at least 50 per cent fibrinogen. Methods directed towards removing fibrinogen specifically by conversion to fibrin or heating result in a loss of Factor‐VIII activity. It seemed possible, therefore, that if there is an association between Factor VIII and another protein, the partner might be fibrinogen.Cross‐linked gels provide the possibility of separating molecules according to their size. The degree of cross linking determines the size of molecules which cannot enter the lattice of the gel particles, and so travel fastest down a column of gel. According to the manufacturers (Pharmacia Ltd., Uppsala, Sweden) Sephadex G‐200 has a structure which should totally exclude molecules having a molecular weight above 200,000, so that fibrinogen (molecular weight about 340,000) should be separable under very mild conditions from any substances having molecular weights much below 200,000. Thelin and Wagner (1961) produced evidence that in high salt concentration (0.4 M‐NaCl) Factor VIII dissociated to a size comparable to that of albumin (molecular weight about 70,000). Molecules of this size should be readily and completely separable from fibrinogen on a Sephadex G‐200 column unless the two had some definite bond.In experiments reported below we tried to separate bovine Factor VIII from fibrinogen in commercial bovine antihaemophilic globulin using columns of Sephadex G‐200, but found the Factor‐VIII activity at the front of the fibrinogen peak even when the gel filtration took place in M‐NaCl or M‐urea in an effort to minimize protein‐protein interaction. We confirmed the observation of Lewis (1964) that Factor VIII in whole human plasma appears to be totally excluded from Sephadex G‐200. Although correlation with molecular weight would not be precise if the molecule were markedly asymmetrical this probably indicates that the Factor‐VIII molecule or complex has a molecular weight in excess of 200,000. These experiments do not distinguish between the possibilities that Factor VIII is itself a large molecule, or is a small molecule in association with fibrinogen. It seemed that discussion about the possible association between Factor VIII and fibrinogen could only be settled by repeating this type of molecular sieve experiment with Factor VIII prepared free from fibrinogen by some other method. Even the best preparations of Factor VIII made by methods depending on salt fractionation and chromatography still contained a large proportion of fibrinogen (Michael and Tunnah, 1963). Only electrophoretic methods seemed to offer the possibility of a complete separation of these two factors.