Isolation and interrelationships of the multiple molecular tissue-type and urokinase-type plasminogen activator forms produced by cultured human umbilical vein endothelial cells.

Primary and early subcultures (1st- to 3rd passage) of human umbilical vein endothelial cells produce tissue-type plasminogen activator (t-PA) antigen, consisting only of a major Mr 110,000 t-PA form. Later subcultures (greater than 4th passage) produce increasing amounts of t-PA antigen, consisting of a major Mr 110,000 and a minor Mr 68,000 form as well as increasing amounts of urokinase-type plasminogen activator (u-PA) antigen, consisting of a minor Mr 95,000 and major Mr 54,000 form. All of the major plasminogen activator forms were purified to homogeneity from 72 h serum-free conditioned media (3 liters, 1-1.8 x 10(9) cells) by a combination of immunoaffinity and gel filtration chromatography. Typically, 4th to 6th passage cultures produced/secreted t-PA-type proteins consisting of an inactive Mr 110,000 (220 IU/mg) and active Mr 68,000 (76,500 IU/mg) form representing about 39 and 8%, respectively, of the total starting sodium dodecyl sulfate stable t-PA activity, and u-PA-type proteins consisting of an inactive Mr 95,000 (700 IU/mg) and active Mr 54,000 (81,000 IU/mg) form representing about 9 and 38%, respectively, of the total starting sodium dodecyl sulfate stable u-PA activity. The isolated Mr 68,000 t-PA and Mr 54,000 u-PA proteins, exist only as two-chain forms in the absence of aprotinin and as mixtures of single- and two-chain proteins in the presence of aprotinin. Treatment with nucleophilic agents completely dissociated the Mr 110,000 t-PA and Mr 95,000 u-PA proteins into their respective Mr 68,000 t-PA and Mr 54,000 u-PA activity forms and a common Mr 46,000 protein, confirming the enzyme-inhibitor complex nature of these inactive plasminogen activator forms.

Isolation and characterization of a urokinase-type plasminogen activator (Mr = 54,000) from cultured human endothelial cells indistinguishable from urinary urokinase.

A urokinase-type plasminogen activator secreted by subcultured normal human umbilical vein endothelial cells was purified and compared to urinary urokinase (Mr = 54,000). The enzyme was isolated from serum-free conditioned medium in the presence of 0.1% (v/v) Triton X-100 by p-aminobenzamidine-agarose affinity chromatography, followed by Sephacryl S-200 gel filtration, followed by immunoadsorption chromatography on affinity purified specific anti-urokinase IgG-Sepharose CL-4B. This plasminogen activator form was obtained from the culture medium with a yield of about 47% and specific activity of about 93,000 IU/mg of protein, and represented approximately 18% of the total multiple molecular plasminogen activator activity forms present in endothelial cell conditioned medium. Sodium dodecyl sulfate-polyacrylamide gel electrophoresis of the purified enzyme showed a single band of plasminogen activator activity with an estimated molecular weight of about 54,000 that was completely inhibited by diisopropyl fluorophosphate (DFP) as well as a single band of radioactivity with similar molecular weight for both the isolated L-[4,5-3H]leucine and [3H]DFP-labeled enzyme. The radiolabeled protein focused as a single major band with a pI value of pH 8.5. The endothelial cell activator and urokinase appeared to be identical in terms of sodium dodecyl sulfate-polyacrylamide gel electrophoresis, location of the [3H]DFP-labeled active site in the Mr = 33,000 heavy chain and [3H]DFP-labeled active site tryptic peptide, and two-dimensional 125I-labeled tryptic peptide maps. In quenching experiments of the fibrinolytic activities using affinity purified specific anti-urokinase IgG the endothelial cell-derived activator and urokinase appeared to be immunochemically identical, but unrelated to tissue plasminogen activator. These results indicate that the Mr = 54,000 urokinase-type plasminogen activator from cultured normal human endothelial cells is similar to, or identical with, Mr = 54,000 urinary urokinase.

Nature of the vitamin K-dependent CO2 fixation in microsomal membranes.

Vitamin K is a component of a membrane-bound enzyme complex which catalyzes the posttranslational carboxylation of peptide-bound glutamate to form the gamma-carboxyglutamate (Gla) residues of prothrombin. The reaction requires reduced vitamin K, bicarbonate, oxygen, and a carboxylase, and does not require ATP. In a Triton X-100 solubilized carboxylase system, it was found that the naphthoquinone ring structure is essential for activity, as is the 2-methyl group. Menaquinone homologs from MK-1 to MK-4 all had carboxylase activity, whereas menadione was inactive. However, dithiothreitol and other thiols form thioethers with menadione, which restores considerable carboxylation activity to the provitamin. Hydrogenation of the beta-gamma double bond in phylloquinone reduced its activity only slightly. The active species of "CO2" utilized in this carboxylation is CO2 and not bicarbonate. Ribosomes contain Gla residues and are labeled with CO2 when whole microsomes are incubated with CO2 in the presence of NADH and vitamin K. About 25% of the activity is releasable with puromycin, suggesting that Gla residues are formed on both the nascent chains and the structural proteins of ribosomes. The deoxycholate-solubilized carboxylase system can be dialyzed to yield membranous vesicles with enhanced carboxylase activity. The warfarin-binding protein from normal rats, but not that from warfarin-resistant rats, further enhances the carboxylase activity of these reformed vesicles.

Studies on native ribosomal subunits from rat liver. Purification and characterization of a ribosome dissociation factor.

A population of free, native ribosomal 40S subunits, that do not react with 60S subunits to form 80S ribosomes, has been identified in the postmicrosomal fraction of rat liver homogenates. A protein (IF-3) has been purified from high salt (0.88 M KCI) extracts of native 40S subunits by gradient centrifugation and by ammonium sulfate fractionation; it prevents the reassociation of subunits and to a limited extent dissociates ribosomes to subunits. The activity is measured by ultracentrifugation of the reaction products on linear sucrose gradients, or with an assay developed in this laboratory that couples dissociation with the 60S-specific peptidyltransferase reaction; the latter procedure measures the amount of 60S subunits released from ribosomes or remaining in incubations in the presence of IF-3. Dissociation factor activity is recovered from most of the particles that are resolved by zonal centrifugation of the total "native subunits" obtained from the postmicrosomal fraction; the highest concentration of IF-3, however, appears to be associated with native 40S subunits. The purified dissociation factor IF-3 is composed of about ten polypeptides and the molecular weight is estimated to be between 500 000 and 700 000, on the basis of glycerol and cesium chloride gradient centrifugation. When purified 40S subunits react with IF-3 or when 80S ribosomes are dissociated by IF-3, a product is formed which is dependent on the concentration of the protein factor and has the characteristics of a 40SIF-3 complex; centrifugation of the complex on sucrose and cesium chloride gradients suggests that the complex consists of 1 equiv of each of the two components. Although dissociation factor IF-3 appears to react in a specific manner with free or ribosome-associated 40S subunits, the reaction with subunits differs in several respects from that with ribosomes. The dissociation factor also appears to interact with 60S subunits but multiple complexes are formed, some with more than 1 IF-3 equiv per 60S particle. The IF-3 converts 40S dimers (55S particles) to the 40S-IF-3 complex and dissociates free, native 80S particles present in the postmicrosomal fraction, but it does not affect polysome-associated ribosomes engaged in protein synthesis.