Human immunodeficiency virus binds and enters cells via the Envelope glycoprotein gp160 at its surface. In infected cells, gp160 is found not only on the plasma membrane but also in the endoplasmic reticulum (ER). Our aim was to establish rate-determining steps in the maturation process of gp160, using a radioactive pulse-chase approach. We found that gp160 has an intricate folding process: disulfide bonds start to form during synthesis but undergo extensive isomerization until the correct native conformation is reached. Removal of the leader peptide critically depends on formation of at least some disulfide bonds in subunit gp120 during folding. Envelope folds extremely slowly and therefore resides in the ER longer than other proteins, but the yield of properly folded molecules is high and degradation is undetectable. The large quantity of gp160 in the ER hence is a result of its slow transit through this compartment. We show here that newly synthesized HIV-1 Envelope glycoprotein apparently follows a slow but high-yield folding path in which co- and post-translational formation of disulfide bonds in gp120, disulfide isomerization and conformation dependent removal of the leader sequence are determining and intertwined events.

No abstract

No abstract

The involvement of N-glycans in the folding of influenza virus hemagglutinin (HA) was analyzed in two CHO-derived glycosylation mutants exhibiting a thermosensitive defect for secretion of human placental alkaline phosphatase. Truncated Man(5)GlcNAc(2)oligosaccharides with one or three glucose residues are attached to proteins of the MadIA214 and B3F7AP2-1 mutant cells, respectively. Newly synthesized proteins retained in these cells carry a Man(4)trimmed glycan generated by a mannosidase different from the ER mannosidases I and II and suggesting a recycling through the Golgi complex. The glucosidase inhibitor castanospermine affects the binding of HA folding intermediates to the lectin-like chaperone calnexin in B3F7AP2-1 but not in MadIA214 cells. We demonstrated that calnexin interacts in vivo with truncated Man(5)derivatives. In MadIA214 cells, this is only possible when Man(5)GlcNAc(2)on protein becomes reglucosylated. The pattern of intermediates seen during the folding of HA in the MadIA214 and B3F7AP2-1 mutant cell lines is different than in control cells. We also observed a variable occupancy of the seven glycosylation-sites. However, even under conditions that restore glycosylation of all sites, the folding intermediates of HA in the mutant cells still remain heterogeneous. Our results demonstrate that addition of truncated N-glycans interferes extensively with the folding of newly synthesized proteins in vivo.

No abstract