Improved procedure for the purification of thromboplastin apoprotein from porcine brain

Thromboplastin (factor Ill) is a membrane-bound lipoprotein which initiates the extrinsic coagulation pathway when released from a latent state following tissue injury. This activation is brought about by the protein component (apoprotein 111) serving as a receptor for clotting factors VII and X. Apoprotein Ill, as well as the lipid component of thromboplastin, are both necessary for the expression of activity. Most purification schemes involve the removal of lipid from apoprotein I11 at an early stage, a process which exposes hydrophobic patches of the latter to the surrounding aqueous environment and leads to a rearrangement of the native state of the protein. Upon re-lipidation it was found that such rearrangement required reversal to obtain a functional complex for clotting. To decide whether alterations in the functional state of thromboplastin are related to changes in the conformation of its protein component, an improved purification scheme was devised giving reasonably pure apoprotein with high specific activity and yield. Porcine brain was chosen as a source of apoprotein Ill since brain tissue is rich in this protein (Glas & Astrap, 1970) and compared with the bovine the pig is closer to humans in terms of studies relating hypercoagulability and thrombosis. A number of methods have been employed for purifying thromboplastin from bovine and human tissue (Nemerson & Pitlick, 1970; Bjnrklid & Storm, 1977) but, to date, their application to porcine brain has produced either high yields of relatively impure thromboplastin of low solubility in water or low yields of highly pure protein. To overcome these problems, a successful purification scheme was devised for porcine apoprotein Ill, which combined a number of procedures selected from schemes used for other species. The initial stages were a modification of that used by Nemerson & Pitlick (19701, i.e. dehydration and delipidation of brain tissue followed by detergent solubilization and differential ammonium sulphate precipitation of apoprotein 111. The use of phenyl-Sepharose 4B as a hydrophobic column stage was adapted from a procedure of Carson & Konigsberg (198 1). In a typical purification, 50 g of acetone brain powder was repeatedly delipidated ( 5 x ) by extraction with heptane/ butanol (2:l,v/v) using 20 ml/g of powder each time. The resulting dried powder was homogenized with 0.5 M-NaCI (40 ml/g of powder) in a Waring blender, centrifuged at 10 000 g for 25 min and the precipitate solubilized by extraction with 0.25% (w/v) sodium deoxycholate solution (20 ml/g of powder). The insoluble protein was removed by centrifugation at 10 000 g for 25 min. The supernatant was brought to 30% saturation with respect to ammonium sulphate and the protein allowed to precipitate overnight. The precipitate was removed by centrifugation at 30 000 g for 120 min and the supernatant was then brought to 60% saturation with ammonium sulphate. After 6 h, the precipitate was removed by centrifugation at 30000 g for 90 min and it was then dissolved in a minimum amount of distilled water before dialysis against a large volume of 0.05 M-imidazole/HCI buffer, pH 7.2, containing 0.1 M-NaCI. The apoprotein solution was then applied to a DEAE-Sephadex column, equilibrated with the same buffer and the fractions with apoprotein 111 activity were eluted with 0.05 M-imidazole/ HCI buffer, pH 7, containing 0.5 M-NaCI. The pooled fractions were then brought to 30% saturation with ammonium sulphate and applied to a phenyl-Sepharose 4B column. The column was washed first with 0.05 M-imidazole/HCI buffer, pH 7, containing 0.5 M-NaCI and saturated to 30"/o with ammonium sulphate, followed by 0.05 M-imidazole/HCI buffer, pH 7, containing 0.1 M-NaCI. Finally, a gradient of O-l%(w/v) Triton X-100 in the same buffer was used to elute apoprotein 111. The fractions containing thromboplastin activity were concentrated by vacuum dialysis against 0.05 M-imidazole/HCl buffer, pH 7, containing 0.1 M-NaCI, before being applied to a Sephacryl 200 column. Upon elution with this buffer the fractions containing thromboplastin activity were pooled and stored at 20°C. Throughout the purification, thromboplastin activity was monitored by the one-stage prothrombin time-test (Nemerson, 19681, after re-lipidation with a commercial preparation of soyabean phospholipids (Sigma Chemical Co., Ltd). It was found that, after elution from the hydrophobic column, little activity was obtained using the re-lipidation method of Howell & Rezvan (1980). However, the activity could be restored by re-lipidation in the presence of cadmium ions, which is in agreement with observations of Carson & Konigsberg (1980). This could be explained by reversal of conformational changes caused by exposure of thromboplastin apoprotein to the hydrophobic column, especially as it has been shown that cadmium ions can induce a-helix formation (Maeda et al., 1986), which is known to be required for protein-lipid interactions. Since we have confirmed that calcium cannot replace cadmium, in this respect, it seems that cadmium also affects specifically protein-lipid interactions in porcine thromboplastin. The above purification scheme yielded thromboplastin apoprotein of 360 000 Da, as determined by gel filtration on a calibrated Sepharose 6B column, while the monomer was found to be 59 000 Da by SDS/polyacrylamide-gel electrophoresis. This scheme provided a 17O0-fold purification of porcine apoprotein Ill and the yield of 43% was sufficient to enable conformational changes during transitions from an inactive to an active state to be investigated. Preliminary analysis by circular dichroism confirmed that the presence of 20-30% a-helix in the protein gave optimal activity upon relipidation. Although a change in circular dichroism spectra in the presence of cadmium suggested some alteration to the secondary conformation, the alternative explanation of