In most eukaryotic cells, coatomer protein (COP)II-coated vesicles, which carry cargo from the endoplasmic reticulum (ER) to the Golgi, bud from discrete sites called transitional ER (tER) sites. Although the existence of these sites has been documented for some time, how they are generated and maintained has remained elusive. However, work from Ben Glick and colleagues in the yeast Pichia pastoris has begun to yield some insights.

Unlike its better-studied relative Saccharomyces cerevisiae, which lacks tER sites, P. pastoris is similar to mammalian cells in that it has well-organized tER structures from which COPII vesicles bud. It has therefore been a useful model organism in which to study tER organization. Previous work from this group had indicated that tER sites assemble by self-organization and, to understand this process better, the authors focused on Sec12.

Sec12 is a tER-localized protein and it is the guanine nucleotide-exchange factor for the GTPase Sar1, which is required to initiate the budding of COPII-coated vesicles. They found that Sec12-containing tER sites are stable entities, but that individual molecules of Sec12 in the tER sites rapidly exchange with Sec12 in other ER regions. In addition, they showed that the association of Sec12 with tER sites is saturable, which indicates that, rather than being targeted by self-association, Sec12 localizes to these sites through an as-yet-unidentified determinant.

In S. cerevisiae, the Sec12 protein localizes throughout the ER and S. cerevisiae Sec12 is similarly localized when it is expressed in P. pastoris. The authors could therefore identify the targeting determinant for tER sites in P. pastoris Sec12 by doing a series of domain-swap experiments using Sec12 from the two yeast species. These experiments showed that both the cytosolic and lumenal domains of P. pastoris Sec12 are important for its tER localization. In a key experiment to test whether Sec12 localization determines the site of COPII-vesicle budding, the authors found that delocalized Sec12 affected neither the geography of vesicle budding nor the integrity of tER sites. In other words, Sec12 does not have to be at tER sites to catalyse vesicle budding. This result is consistent with what has been noted for mammalian Sec12, which is also not concentrated at tER sites.

Nonetheless, these data support the idea that a Sec12-independent scaffold exists at tER sites. Future work will be necessary to identify components of this scaffold, which the authors speculate could include proteins that target Sec12, as well as the peripheral ER membrane protein Sec16, to tER sites. A molecular definition of tER sites will also help determine whether these sites are crucial for regulating the spatially restricted budding of COPII vesicles.