Towards the physiological role of ER quality control in an organismal context

ER-associated degradation (ERAD) plays a pivotal role in cells with a highly active secretory apparatus, such as B lymphocytes, beta cells in the pancreas or hepatocytes. However, it is not always clear whether the observed effect(s), such as pancreatic atrophy in Sel1L-/- mice, cartilage defect in Derlin-2-/- mice, or reduced susceptibility to arthritis in Hrd1+/- mice, are a direct consequence of impaired ERAD or rather a secondary effect, e.g. caused by an ongoing unfolded protein response (UPR), a situation that often applies to currently available mouse models. There is a growing body of data pointing towards a role of ERAD and/or individual ERAD components in other processes, which are only remotely or not at all related to its ‘classical’ role in degradation of misfolded ER luminal or transmembrane proteins. In this work I describe two mouse models, in which the genes that encode key quality control enzymes, Yod-1 and Ube2j1, were targeted for selective modification or ablation, respectively. These enzymes had previously been implicated in ERAD based on biochemical experiments. In neither dominantnegative Yod-1-transgenic nor Ube2j1-deficient mice a constitutive UPR is observed, qualifying them for the interrogation of possible UPR-independent functions of the genes at hand. Antigen-presenting cells from mice that overexpress a dominant-negative version of the deubiquitylase Yod1, present antigen more efficiently to CD8+ T cells via the cross-presentation pathway. Mice lacking the ERAD-associated E2 ubiquitin-conjugating enzyme Ube2j1 revealed an essential function for this gene to thrive early in life and, surprisingly, showed that ablation of Ube2j1 was responsible for male sterility by critically controlling one of the last steps of spermiogenesis. These findings shed light on novel or previously underexplored objects of ERAD in vivo and will help to create next generation experimental setups in vivo and in vitro to investigate the full spectrum of ERAD. They raise the important point that a narrow focus on a few model substrates may lead us to ignore other possible functions of the components involved, and at the very least show that there must be a considerable degree of redundancy in ensuring removal of unwanted proteins form the ER. By the same logic, it may well be that the unique properties of the few model substrates usually considered as misfolded (such as CFTR F508, alpha-1 antitrypsin variants or unpaired subunits of multimeric membrane glycoproteins complexes) may not be the generally applicable examples that have inspired so many broad generalizations. Only now are we entering a period where key components are being explored in an organismal setting, with at time surprising results. While many mutant mice, defective in a key quality control component fail to thrive or are even embryonic lethal, it is remarkable that they nonetheless execute many of the steps in early development without obvious defects. The same situation applies to components of the calnexin/calreticulin cycle, long considered an important aspect of glycoprotein quality control. Nonetheless, ablation of calnexin or calreticulin causes comparatively mild in vivo phenotypes when considered from the vantage point of early development, a period when protein synthesis is put to good use to build new cells and tissues