page 897, Luedeke et al. show that, like the plasma membrane, the ER can be compartmentalized into distinct membrane domains. In budding yeast, this separation restricts the diffusion of ER membrane proteins between the bud and the mother cell.
The mRNAs of many bud-specific membrane proteins are actively transported into the bud. To keep these mRNAs and proteins in the bud, they must be kept from diffusing back into the mother cell through the ER, which traverses the bud neck. Using FRAP analyses, the authors show that proteins are kept in the bud via a diffusion barrier at the bud neck, where rough ER is replaced by a band of smooth ER.
The lumen of the ER was continuous, as shown by EM and by the exchange of lumenal ER proteins. But ER membrane proteins did not exchange—those in the bud stayed in the bud.
At the plasma membrane, mother and bud are separated by a ring of septin, which the authors show also controls ER separation. Septin mutants lost the diffusion barrier and the smooth ER band. Downstream of septin, the loss of Bud6 impaired the diffusion barrier, but did not affect smooth ER formation. Perhaps Bud6 organizes ER lipids into raft domains to create the blockade. For now, however, the contributions of lipids to this system are unknown.
In addition to its contribution to polarity, the unusual ER at the bud neck might control cytokinesis. For instance, if bud neck ER contains a higher concentration of calcium channels, it might control the localized calcium release that initiates contraction of the actomyosin ring. The domain might also pinch the ER in two at or before cytokinesis. The task now will be to determine whether ER compartmentalization exists in higher eukaryotes.