CPYΔ1 is a poor substrate for ERAD because it lacks a vital sugar.
The team used a shortened CPY protein as a model unfolded protein. CPYΔ1 lacked the terminal 154 amino acids of the protein. Surprisingly, it was retained in the ER, rather than being targeted for ERAD. Although CPY contains four N-glycosylation sites, the team found that if they disrupted the last site—the one missing in CPYΔ1—the protein was retained in the ER.To further test whether a single glycan could act as a signal for degradation, the team created two novel mutations in the ERAD substrate proteinase A*. When they mutated the first of two N-glycosylation sites in the protein, it was retained in the ER, like CPYΔ1; but mutating the second site had no effect.
The team has started to further narrow down the features of the glycosylation sites that act as ERAD-entry signals. They have preliminary evidence that it is a bipartite signal, with a polypeptide sequence acting in conjunction with the sugar. If they move both together they can transfer the ERAD-targeting signal, independent of the rest of the protein. Remarkably, the glycosylation sites identified by Spear and Ng do not appear to have a function in the correctly folded protein, suggesting that they evolved for the purpose of targeting the protein for degradation when it misfolds. 
