JAB1/CSN5 inhibits the activity of Luman/CREB3 by promoting its degradation.

Luman/CREB3 (also called LZIP) is an endoplasmic reticulum (ER)-bound transcription factor that has been implicated in the ER stress response. In this study, we used the region of Luman containing the basic DNA-binding domain as bait in a yeast two-hybrid screen and identified the Jun activation domain-binding protein 1 (JAB1) or the COP9 signalosome complex unit 5 (CSN5) as an interacting protein. We confirmed their direct binding by glutathione S-transferase pull-down assays, and verified the existence of such interaction in the cellular environment by mammalian two-hybrid and co-immunoprecipitation assays. Deletion mapping studies revealed that the MPN domain in JAB1 was essential and sufficient for the binding. JAB1 also colocalized with Luman in transfected cells. More interestingly, the nuclear form of Luman was shown to promote the translocation of JAB1 into the nucleus. We found that overexpression of JAB1 shortened the half-life of Luman by 67%, and repressed its transactivation function on GAL4 and unfolded protein response element (UPRE)-containing promoters. We therefore propose that JAB1 is a novel binding partner of Luman, which negatively regulates the activity of Luman by promoting its degradation.

Luman is capable of binding and activating transcription from the unfolded protein response element.

Luman (or LZIP, CREB3) is a transcription factor with an endoplasmic reticulum (ER)-transmembrane domain. Due to its structural similarities with ATF6, it is thought that Luman might also be involved in cellular stress responses. Here we report that Luman can bind and activate transcription from the consensus unfolded protein response element (UPRE). Mutations that disrupted the binding of Luman to the UPREs impaired its ability to activate transcription from these sites. Overexpression of Luman stimulated transcription of EDEM, a downstream effector of the mammalian unfolded protein response involved in ER-associated degradation (ERAD). Unlike ATF6, however, Luman was not activated by proteolytic cleavage in response to endoplasmic reticulum stressors such as tunicamycin and thapsigargin. These results suggest that the activation of ERAD by Luman is likely through a pathway different from the common ER stress response, and that additional factor(s) are required for the activation of this Luman-mediated pathway.