O-GLYCOSYLATION IN PROTEIN FOLDING

O-glycosylation modifications were largely thought of as post-translational a that may affect protein function, (Notch signaling being the most famous example. Yet puzzlingly, in the absence of these decorative modifications, the proteins (almost all of which are cell surface or secreted) failed to get to their final destinations. Thrombospondin Type I repeats (TSRs) are protein domains that can be modified first by the addition of an O-fucose (by POFUT2), and this modification can be extended by addition of a glucose (by B3GLCT). Using this model of O-glycosylation, I sought to understand why this post-translational modification affected protein localization.

I found that lack of O-glycosylation induces stress in the endoplasmic reticulum (ER), where TSRs are folded, suggesting that this modification may be affecting protein folding. Further work showed that O-glycosylation stabilizies and accelerates protein folding, and promotes ER exit of properly folded substrates (PMC4214879, PMC4318717). My work has demonstrated that POFUT2 is a quality control sensor that functions co-translationally to “mark” properly folded substrates with an O-fucose. This work formed the basis of a paradigm where specialized chaperones are recruited for the folding and quality control of particular protein domains.

To demonstrate these ideas, I developed several methods in the lab to isolate fucosylated proteins, separate mixtures of folded and unfolded proteins and observe protein refolding. I also developed metabolic labeling methods to demonstrate for the first time that O-glycosyation is co-translational (PMC4318717)! I was fortunate that the Haltiwanger laboratory had it’s own mass spectrometry setup where I learned how to map post-translational modifications, and also identified interactors of POFUT2. In the years since, others in the lab have demonstrated that the quality control and protein folding role is common to most O-glycosyltransferases, including POFUT1 that modifies Notch.