Endoplasmic reticulum (ER) quality control mechanisms target terminally misfolded proteins for ER-associated degradation (ERAD). Misfolded glycophosphatidylinositol-anchored proteins (GPI-APs) are, however, generally poor ERAD substrates and are targeted mainly to the vacuole/lysosome for degradation, leading to predictions that a GPI anchor sterically obstructs ERAD. Here we analyzed the degradation of the misfolded GPI-AP Gas1* in yeast. We could efficiently route Gas1* to Hrd1-dependent ERAD and provide evidence that it contains a GPI anchor, ruling out that a GPI anchor obstructs ERAD. Instead, we show that the normally decreased susceptibility of Gas1* to ERAD is caused by canonical remodeling of its GPI anchor, which occurs in all GPI-APs and provides a protein-independent ER export signal. Thus, GPI anchor remodeling is independent of protein folding and leads to efficient ER export of even misfolded species. Our data imply that ER quality control is limited for the entire class of GPI-APs, many of them being clinically relevant.
© 2016 Sikorska et al.