In the ERCOPE program, we considered how the ER controls the positioning of the endosomal pathway and vice versa. This yielded new biology.
Long distance transport of organelles runs through the action of kinesin- and dynein-based motor proteins that is tightly controlled and can be dramatically remodeled on demand. We defined an ER-associated pathway that establishes and maintains a bilateral architecture of the entire endosomal system, consisting of static perinuclear as well as dynamic peripheral endosomes. Abdication of this spatiotemporal divide attenuates endosomal maturation and delays receptor downregulation (Jongsma et al., Cell 2016). The program considered various aspects of ER control of endosomal positioning.
1. ER-Endosome interactions in endosome positioning, motility and beyond:
We identified the ER-associated E3 ubiquitin ligase RNF26 and its cytosolic substrate SQSTM1/p62 as master controllers of perinuclear endosomal positioning. This work has prompted new investigations on how ER-endosome interactions affect the timing and directionality of endosomal movement. Exploring enzymatic regulation of RNF26 activity, we have shown that deubiquitination of RNF26 substrate SQSTM1 by USP15 liberated perinuclear vesicles for further transport (Jongsma et al., Cell 2016). If there is an E3 enzyme, there also should be an E2 enzyme. We identified UBE2J1 as the E2 enzyme for RNF26. Inactivation of UBE2J1 resulted in the loss of the control of perinuclear endosomes that now scatter and move around the cytosol, a phenotype similar to inactivation of RNF26 (Cremer et al, Cell Rep 2021). Both RNF26 and UBE2J1 are ER embedded cytosolic proteins that then should move around by diffusion. This then would be conflicting with the perinuclear location of RNF26 and by default the endosomal pathway. We performed a BioID experiment to show that the inactive RING domain of RNF26 interact with vimentin, the protein that makes the intermediate filaments. RNF26 is released by activation and then spatially constraint by the interactions between active RNF26 and the different endosomes. Because UBE2J1 is also involved in ER-associated degradation (ERAD), we showed that the close spatial connection between the ER and endosomes are coupled (Cremer et al., EMBO J in revision). This work explains how a diffusible protein complex (RNF26/UBEJ1) still locates in the perinuclear space and why ER and endo-lysosomes have to be closely located.
2. ER master regulator proteins for intraorganellar interactions:
VAP-A and VAP-B are ER proteins that interact with other proteins that have so-called FFAT motifs. As a consequence, virtually any compartment including endosomes and mitochondria directly interacts with the ER in so-called membrane contact sites. We identified another three family members: MOSPD1, 2 and 3. These cluster in two subdomains, one containing VAP-A, VAP-B and MOSPD2 and the other MOSPD1 and 3. By BioID experiments the VAPome was defined; a comprehensive atlas of proteins in different organelles that interact with the ER (Cabukusta et al, Cell Rep 2020). These data are used as a platform for understanding the different biochemical reactions occurring between ER and other compartments.
3. Endosomes and the ER architecture:
If the ER and endosomes interact, the complex ER network is in part be formed moving endosomes. We have shown that moving late endosomes can drag along pieces of ER to alter the ER morphology. Limiting late endosomal movement then results in a simpler ER architecture (Spits et al., EMBO Rep 2021). The ER not only controls endosomes but endosomes also control the ER architecture.
