Consequences of individual N-glycan deletions and of proteasomal inhibition on secretion of active BACE.

BACE is an aspartic protease involved in the production of a toxic peptide accumulating in the brain of Alzheimer's disease patients. After attainment of the native structure in the endoplasmic reticulum (ER), BACE is released into the secretory pathway. To better understand the mechanisms regulating protein biogenesis in the mammalian ER, we determined the fate of five variants of soluble BACE with 4, 3, 2, 1, or 0 N-linked glycans. The number of N-glycans displayed on BACE correlated directly with folding and secretion rates and with the yield of active BACE harvested from the cell culture media. Addition of a single N-glycan was sufficient to recruit the calnexin chaperone system and/or for oligosaccharide de-glucosylation by the ER-resident alpha-glucosidase II. Addition of 1-4 N-glycans progressively enhanced the dissociation rate from BiP and reduced the propensity of newly synthesized BACE to enter aberrant soluble and insoluble aggregates. Finally, inhibition of the proteasome increased the yield of active BACE. This shows that active protein normally targeted for destruction can be diverted for secretion, as if for BACE the quality control system would be acting too stringently in the ER lumen, thus causing loss of functional polypeptides.

Persistent glycoprotein misfolding activates the glucosidase II/UGT1-driven calnexin cycle to delay aggregation and loss of folding competence.

The UDP-glucose:glycoprotein glucosyltransferase (UGT) is a central player of glycoprotein quality control in the endoplasmic reticulum (ER). UGT reglucosylation of nonnative glycopolypeptides prevents their release from the calnexin cycle and secretion. Here, we compared the fate of a glycoprotein with a reversible, temperature-dependent folding defect in cells with and without UGT1. Upon persistent misfolding, tsO45 G was slowly released from calnexin and entered a second level of retention-based ER quality control by forming BiP/GRP78-associated disulfide-bonded aggregates. This correlated with loss in the ability to correct misfolding. Deletion of UGT1 did not affect the stringency of ER quality control. Rather, it accelerated release from calnexin and transfer to the second ER quality control level, but it did so after an unexpectedly long lag, showing that cycling in the calnexin chaperone system is not frenetic, as claimed by existing models, and is fully activated only upon persistent glycoprotein misfolding.