A Ubiquitin-Binding Rhomboid Protease Aimed at ERADication

Abstract

Proteins are degraded from the ER by endoplasmic reticulum-associated degradation (ERAD). In a recent issue of Molecular Cell, Fleig et al. (2012) describe a role for a ubiquitin-binding rhomboid protease, RHBDL4, in degradation of select ERAD substrates. These findings and the significance of rhomboids and other intramembrane proteases are discussed.

Intracellular protein degradation occurs predominantly through two routes. Intralumenal degradation takes place primarily in lysosome-derived structures either by trafficking through the secretory and endocytic pathways or as a consequence of autophagy; both of these involve ubiquitination. The second mechanism is through, often exquisitely regulated, degradation by the ubiquitin proteasome system (UPS). One of the major sites from which proteins are targeted for degradation is the endoplasmic reticulum (ER). Up to a third of proteins are translocated into the ER, where they are modified (e.g. N-glycosylation, signal peptide cleavage, proline isomerization, disulfide bond formation), fold into native conformations with the assistance of chaperones, and assemble into multi-protein complexes. Proteins that successfully navigate these hurdles are either secreted, or take up residence in the secretory pathway, in the endocytic-lysosomal pathway, or at the plasma membrane. Managing this complexity requires efficient means for degrading misfolded or excess proteins. It was originally thought that this “quality control” occurred through intralumenal proteases in the ER or another pre-Golgi compartment. We now know that such degradation (ER-associated degradation or ERAD) relies on ubiquitination, dislocation to the cytosol, and degradation by the 26S proteasome. However, in a recent issue of Molecular Cell, Fleig and colleagues (2012) revive interest in the concept of intralumenal protein degradation in the ER by providing evidence that a rhomboid protease located in the ER can cleave select proteins in their transmembrane domains leading to, apparently lumenal, proteolytic intermediates and thereby contributing to UPS-mediated ERAD (Figure 1).

Figure 1. Rhomboid-like proteins in ERAD.

Degradation of proteins from the ER requires their ubiquitination, generally with chains of ubiquitin, dislocation from the membrane, and proteasomal degradation. (Left) The degradation of many substrates via the classical ERAD pathway involves the ER membrane protein Derlin-1, which is associated with several ERAD ubiquitin ligases (not shown). Derlin-1 also interacts with the p97 complex and promotes substrate dislocation from the ER membrane. The transmembrane domains of Derlin-1 have homology to the rhomboid proteases, making it a rhomboid pseudo-protease. (Center) The iRhoms are inactive homologs of the rhomboid proteases. Initial studies of mammalian iRhoms suggest that they affect substrate trafficking by enhancing the ERAD of membrane-tethered EGF ligand. Mouse iRhom2 has also been shown to promote the exit of TNF-converting enzyme from the ER to the Golgi where it is activated by furin-mediated proteolysis. (Right) RHBDL4 is an active rhomboid protease in the ER that cleaves select membrane proteins with atypical, positively charged transmembrane domains. The proteolytic fragments are subsequently degraded by the proteasome as in classical ERAD pathways. RHBDL4, which presumably recognizes ubiquitinated substrates via its UIM, also interacts with p97, which likely promotes dislocation of RHBDL4-cleaved proteins from the ER membrane. A common theme for RHBDL4 and rhomboid pseudo-proteases in the ER seems to be in the regulation of substrate flux through the secretory pathway.

The importance of ERAD is underscored by the ER stress that occurs when ERAD is insufficient. Such stress is implicated in diseases including cancer, neurodegeneration, diabetes, and other metabolic disorders. During ER stress, cells attempt to restore homeostasis by initiating a multi-pronged unfolded protein response (UPR); when the UPR fails to restore homeostasis, cell death pathways are activated (reviewed in Tsai and Weissman, 2010). ERAD involves convergent pathways that target both lumenal and transmembrane proteins for degradation. For transmembrane proteins, the cytosolic region is believed to be a primary site for ubiquitination, leading to the binding of a cytosolic AAA-ATPase complex consisting of a p97 hexamer and its associated factors. The p97 complex is thought to “ratchet” substrates from the ER membrane, presumably through an aqueous channel. Dislocated substrates are ultimately degraded by the proteasome.

There is increasing appreciation for the importance of intramembrane proteolysis in both cellular homeostasis and signaling (reviewed in Lemberg, 2011; Brown et al., 2000). In the ER, signal peptide peptidase (SPP) cleaves signal peptides characteristic of most secreted and type I transmembrane proteins. In the Golgi, two related enzymes – site-1 and site-2 proteases (S1P and S2P) – are implicated in the cleavage and activation of transcription factor precursors SREBP (S1P and S2P) and ATF6 (S2P), which are transported from the ER to the Golgi in response to low levels of sterol precursors or to ER stress, respectively. Another enzyme involved in such cleavage is γ-secretase. This multi-subunit protease, in its various forms, is characterized by a common enzymatic subunit, presenilin, and cleaves proteins including amyloid protein precursor, releasing Alzheimer’s disease pathogenic Aβ peptides into the extracellular milieu, and Notch, releasing the transcriptional activator NICD.

Fleig and colleagues (2012) now report that the rhomboid protease RHBDL4 resides in the ER and contributes to the proteolysis of several ectopically-expressed ERAD substrates with atypical, positively charged transmembrane domains (Figure 1). Rhomboid proteases represent a family of polytopic intramembrane serine proteases and pseudo-proteases (reviewed in Lemberg, 2011). The first described member of this family was Rhomboid-1, initially identified in Drosophila, which catalyzes the release of growth factors from their transmembrane precursors in the Golgi. Rhomboid proteases are found in many membrane-limited organelles including the mitochondria where, in mammals, presenilin-associated rhomboid-like (PARL) cleaves OPA1, which regulates mitochondrial cristae integrity and apoptosis (Cipolat et al., 2006). Catalytically inactive members of the rhomboid family, iRhoms and Derlin-1, are rhomboid pseudo-proteases implicated in ERAD (Figure 1) (Greenblatt et al., 2011; Zettl et al., 2011). Mouse iRhom2 has also been implicated in the trafficking of TNF-converting enzyme from the ER to the Golgi where it is activated by proteolysis (Adrain et al., 2012).

Unlike the iRhoms and Derlin-1, however, RHBDL4 requires its serine protease activity to enhance protein destruction. Fleig and colleagues (2011) show that RHBDL4 cleaves ERAD substrates, including pre-T cell receptor α (pTα), a type I transmembrane protein, as well as polytopic proteins, resulting in multiple N-terminal proteolytic fragments. The authors find that p97, recruited by interaction with a C-terminal binding motif of RHBDL4, is required for proteasomal degradation of these fragments, likely by facilitating their dislocation from ER membranes. Consistent with a role in ERAD, RHBDL4 is induced by the UPR and knockdown of RHBDL4 activates the UPR, suggesting that it may enhance ERAD of select proteins under conditions of ER stress. Moreover, they show that placement of positively charged amino acids into a normally uncharged transmembrane domain can render a protein a substrate for RHBDL4. It is suggested that RHBDL4 is responsible for proteolytic breakdown beyond the initial cleavage, however, this remains to be determined.

Perhaps most striking about RHBDL4, and what distinguishes it from other known intramembrane proteases, is its ubiquitin-interacting motif (UIM). Fleig and colleagues (2012) find that the UIM is required for RHBDL4 function in protein degradation. Overexpression of an inactive protease mutant of RHBDL4 accumulates ubiquitinated substrates, suggesting that substrate ubiquitination is required for recognition by RHBDL4. This added layer of complexity to ERAD therefore continues to rely on ubiquitination to initiate substrate degradation.

RHBDL4 and it role in generating lumenal degradation intermediates necessarily leads us to question assumptions about ERAD. Particularly, whether ERAD is solely the province of 26S proteasome activity towards fully intact proteins dislocated from the ER and whether lumenal-intramembrane and cytosolic-nuclear proteolysis might function coordinately in protein degradation to a greater extent than generally appreciated. Interestingly, a previous report demonstrated that SPP could play a role in ERAD of ectopically expressed TCR-α, which also has a positively charged transmembrane domain (Loureiro et al., 2006). It is now of interest to assess the range of intramembrane ER proteases and substrates that might be responsible for, or targets of, intramembrane cleavage as part of ERAD.

It is important, however, to note that bona fide endogenous substrates for RHBDL4 have yet to be identified and the proteasome is sufficient to compensate for RHBDL4 knockdown in degradation of substrates. Therefore, the overall significance of this protease in ERAD awaits determination. Beyond this, much of the known intramembrane or juxtamembrane proteolysis serves roles in either signaling or protein processing. Thus, whether RHBDL4 functions primarily in protein degradation vs. ubiquitin-dependent protein processing and signaling pathway(s) becomes of great interest. The provocative findings of Fleig and colleagues (2012) will no doubt spur further studies to decipher the functions of RHBDL4 in physiological contexts.