Characterization of precursor and secreted forms of human angiotensinogen.

To define the basis of the heterogeneity of angiotensinogen, we have characterized the immunoreactivity of high molecular weight (HMW) and low molecular weight (LMW) plasma angiotensinogen, the angiotensinogen precursor synthesized by cell-free translation, and angiotensinogen secreted by human hepatoma (Hep G2) cells. Angiotensinogen precursor synthesized by rabbit reticulocyte lysate primed with RNA prepared from liver or Hep G2 cells was compared with angiotensinogen secreted by Hep G2 cells by using immunoprecipitation and sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE). So as to assess the contribution of N-glycosylation of angiotensinogen, Hep G2 cells were incubated in the presence of tunicamycin. Glycosylation of secreted angiotensinogen was further characterized by using chromatography on concanavalin A-Sepharose, digestion with neuraminidase, and treatment with trifluoromethane sulfonic acid. In Sephadex G-200 column chromatography, HMW plasma angiotensinogen eluted just after the column void volume and was clearly separated from LMW angiotensinogen which eluted just before bovine serum albumin. Both HMW and LMW plasma angiotensinogen were shown to bind to monoclonal and polyclonal antibodies raised against pure LMW angiotensinogen. Only one angiotensinogen precursor (mol wt 50,000) was identified by cell-free translation which, after cleavage by renin, was reduced to mol wt 45,600. Angiotensinogen secreted by Hep G2 cells showed electrophoretic heterogeneity (mol wt 53,100-65,400). Tunicamycin-treated Hep G2 cells secreted five discrete forms of angiotensinogen, a predominant form of mol wt 46,200, with other forms (mol wt 46,800, 48,100, 49,200, and 49,600) representing 10% of secreted angiotensinogen. All five forms showed a similar reduction in molecular weight after cleavage by renin. The predominant 46,200-mol wt protein represented nonglycosylated angiotensinogen in that, after cleavage by renin, it had an electrophoretic mobility (mol wt 45,600) identical to the desangiotensin I-angiotensinogen resulting from renin cleavage of the angiotensinogen precursor. The other higher molecular weight forms of angiotensinogen secreted by tunicamycin-treated Hep G2 cells were shown to represent O-glycosylated angiotensinogen in that they were reduced to 46,200 mol wt by treatment with trifluoromethane sulfonic acid. Dexamethasone (10(-7) and 10(-6)M) stimulated angiotensinogen secretion by Hep G2 cells two- to fourfold, both in the absence and presence of tunicamycin. However, a small stimulatory effect of mestranol (10(-7) M) was evident only in the presence of tunicamycin. Neither dexamethasone nor mestranol influenced the electrophoretic pattern (SDS-PAGE) of angiotensinogen secreted by Hep G2 cells. However, when incubation media were chromatographed on Sephadex G-200 with subsequent immunoprecipitation of the column fractions, both dexamethasone and mestranol were shown to stimulate the secretion of HMW angiotensinogen (eluting just after the column void volume) which, on SDS-PAGE, migrated in a position identical to LMW angiotensinogen. From these studies, we conclude that all forms of human angiotensinogen are derived from a single precursor. The heterogeneity of secreted angiotensinogen represents differences in posttranslational processing of angiotensinogen. This processing includes both N- and O-glycosylation, and also the formation of HMW complexes (HMW angiotensinogen) through association either with other angiotensinogen molecules or with some other protein(s) whose secretion by hepatocytes is stimulated by glucocorticoids and estrogens.