Báo cáo khoa học: Trafficking and proteolytic processing of RNF13, a model PA-TM-RING family endosomal membrane ubiquitin ligase pot

MINIREVIEW Trafficking and proteolytic processing of RNF13, a model PA-TM-RING family endosomal membrane ubiquitin ligase Jeffrey P. Bocock 1 , Stephanie Carmicle 2 , Mayukh Sircar 1 and Ann H. Erickson 1 1 Department of Biochemistry and Biophysics, University of North Carolina, Chapel Hill, NC, USA 2 Department of Biological Sciences, Mississippi College, Clinton, MS, USA Physiological function of RNF13 RING finger protein 13 (RNF13) has been linked to a variety of physiological conditions through its isolation in multiple screens for functional genes. The C-terminal half of the protein contains a RING domain that func- tions as an ubiquitin ligase in vitro [1,2]. Presumably, the ability of RNF13 to ubiquitinate and so determine the half-life and ⁄ or targeting of other proteins is central to its physiological role, but substrates of this ubiquitin ligase have not yet been identified. Expression levels of the protein are higher in adult tissues than in the corresponding embryonic tissues [2,3], suggesting roles in development, consistent with the absence of homologs in yeast. Expression of RNF13 is upregulat- ed when mouse brain neurons are treated with tenas- cin C [4], in basilar papilla when chickens are exposed to acoustic trauma [5], in pancreatic and other tumors [1,6], when neurons are stimulated to extend neurites on fibronectin [2] and on treatment of chicken fetal myoblasts with myostatin [3]. RNF13 has been reported to promote neurite extension when over- expressed ectopically in PC12 rat adrenal medulla pheochromocytoma cells cultured on collagen [7]. Keywords cellular targeting; endosomes; GRAIL; inner nuclear membrane; protease-associated domain; proteasomes; proteolysis; RING domain; RMR; ubiquitin ligase Correspondence A. H. Erickson, The Department of Biochemistry and Biophysics, CB 7260 GM, University of North Carolina, Chapel Hill, NC 27599, USA Fax: +1 929 966 2852 Tel: +1 919 966 4694 E-mail: ann_erickson@med.unc.edu (Received 17 April 2010, revised 2 July 2010, accepted 15 July 2010) doi:10.1111/j.1742-4658.2010.07924.x RING finger protein 13 (RNF13) is a ubiquitously expressed, highly regu- lated ubiquitin ligase anchored in endosome membranes. A RING domain located in the cytoplasmic half of this type 1 membrane protein mediates ubiquitination in vitro but physiological substrates have not yet been identi- fied. The protein localized in endosomal membranes undergoes extensive proteolysis in a proteasome-dependent manner, but the mRNA level can be increased and the encoded protein stabilized under specific physiological conditions. The cytoplasmic half of RNF13 is released from the membrane by regulatory proteases and therefore has the potential to mediate ubiquiti- nation at distant sites independent of the full-length protein. In response to protein kinase C activation, the full-length protein is stabilized and moves to recycling endosomes and to the inner nuclear membrane, which exposes the RING domain to the nucleoplasm. Thus RNF13 is a ubiquitin ligase that can potentially mediate ubiquitination in endosomes, on the plasma membrane, in the cytoplasm, in the nucleoplasm or on the inner nuclear membrane, with the site(s) regulated by signaling events that modulate protein targeting and proteolysis. Abbreviations APP, amyloid precursor protein; CTF, C-terminal fragment; EGFR, epidermal growth factor receptor; GRAIL, gene related to anergy in lymphocytes; HA, hemagglutinin; ICD, intracellular C-terminal domain; INM, inner nuclear membrane; MVB, multivesicular body; NLS, nuclear localization signal; PA, protease-associated; PKC, protein kinase C; PM, plasma membrane; PMA, 4b-phorbol 12-myristate 13-acetate; RMR, receptor homology region-transmembrane domain-RingH2 motif protein; RNF13, RING finger protein 13; TM, transmembrane. FEBS Journal 278 (2011) 69–77 ª 2010 The Authors Journal compilation ª 2010 FEBS 69 Overexpression of RNF13 suppresses cell proliferation [3] but drives Matrigel invasion [1]. It remains unclear how these seemingly disperate effects are mediated on a biochemical level, but involvement of the extracellu- lar matrix appears to be a recurring theme. Cellular localization of RNF13 Substrate choice of a ubiquitin ligase is regulated by its cellular localization. Using immunofluorescence staining, we observe RNF13 at a steady state in multivesicular endosomes, where it shows partial colocalization with CD63, LAMP2 and the mannose phosphate receptor [2]. Ectopically expressed protein has also been detected in the endoplasmic reticulum [1]. RNF13 is often observed in ring-shaped structures, consistent with localization in the membrane of vesicles (Fig. 1). Two PA-TM-RING family members, RNF128 ⁄ gene related to anergy in lymphocytes (GRAIL) [8,9] and receptor homology region-transmembrane domain- RingH2 motif protein (RMR) [10,11] are also localized in endosomes. Proteolytic processing of RNF13 RNF13 undergoes extensive post-translational proteol- ysis (Fig. 2A). Unless expression levels are artificially high because of transient ectopic expression, we are generally unable to detect the expressed protein in cell extracts by PAGE (Fig. 2B, lanes 1 and 4) or in intact cells by immunofluorescence [2]. Inhibition of lyso- somal proteolysis by modulation of pH does not pre- vent generation of the proteolytic fragments or significantly stabilize them [2]. Inhibiting the protea- some does not prevent proteolysis of RNF13, but unlike inhibition of lysosomal proteolysis, does retard turnover of fragments. Thus the RNF13 fragments must be generated by specific regulatory proteases and are not merely by-products of protein turnover. Inhibition of proteasome proteolysis by treatment of cells with MG132 (Fig. 2, lanes 2 and 5) or epoxomi- cin (Fig. 2, lanes 3 and 6) for  8 h allows sufficient accumulation of protein to render RNF13 detectable Fig. 1. RNF13 is localized in ring-shaped structures. Chinese ham- ster ovary cells stably expressing FLAG-tagged RNF13 were stained with mouse anti-FLAG IgG followed by donkey AlexaFluor 488 anti-mouse IgG (green) and goat anti-(lamin B) IgG followed by donkey AlexaFluor 568 anti-goat serum (red). The inset shows an enlargement of the ring-shaped vesicles. Dimethylsulfoxide MG132 Epoxomicin Dimethylsulfoxide MG132 1 2 3 4 5 6 72 kDa- 36 kDa- 55 kDa- Full-Length + Chondroitin sulfate CTF-43 ICD CTF-39 CTF-37 Epoxomicin Anti-HA Anti-FLAG B A SP PA RING PEST TM Ser-rich CTF ICD Fig. 2. RNF13 undergoes extensive post-translational proteolysis. (A) RNF13 is predicted to possess a transient N-terminal signal peptide (SP), a protease-associated (PA) domain, a hydrophobic transmembrane sequence (TM), and a RING, PEST and Ser-rich domain. The membrane-bound C-terminal fragment(s) (CTF) and the soluble intracellular C-terminal domain (ICD) are indicated. (B) Chinese hamster ovary cells stably expressing RNF13-HAF were treated for 10 h with vehicle (dimethylsulfoxide) (lanes 1 and 4) or with a proteasome inhibitor, either 5 l M MG132 (lanes 2 and 5) or 0.5 l M epoxomicin (lanes 3 and 6). Cellular proteins were resolved on a 12% NuPage Novex bis-Tris polyacrylamide gel (Invitrogen, Carlsbad, CA, USA) run in Mes buffer and RNF13 was visualized on a western blot with horseradish peroxidase-conjugated anti-HA or anti-FLAG IgG, as indicated. Prestained molecular mass markers are indicated on the left. The identity of the forms, as labeled on the right, was established by blotting microsomal membranes with antisera specific for the epitope tags [2]. RNF13 trafficking and proteolysis J. P. Bocock et al. 70 FEBS Journal 278 (2011) 69–77 ª 2010 The Authors Journal compilation ª 2010 FEBS in cells that stably express the protein, suggesting a significant role for proteasomes in the turnover of this endosomal protein. MG132 has been used similarly to stabilize and so aid visualization of biosynthetic forms of other integral membrane proteins including Notch [12], EGF receptor [13], growth hormone receptor [14] and the amyloid precursor protein (APP) [15]. Epox- omicin is a specific inhibitor of the chymotryptic-like activity of the proteasome [16]. MG132, however, not only blocks proteosome proteolysis, but also inhibits calpain [17] and lysosomal cysteine proteases [18]. Thus MG132 potentially achieves its greater stabiliza- tion of RNF13 relative to epoxomicin (Fig. 2, lane 5 versus lane 6) by inhibiting more than one protease that cleaves the protein. To determine the relationship of the biosynthetic forms stabilized by proteasome inhibitors, we added epitope tags to RNF13, marking the N-terminal ecto- domain with a hemagglutinin (HA) epitope at position 38, after the signal peptide cleavage site predicted by the algorithm signalp v.3.0 [19], and tagging the cyto- plasmic half of the protein with a 3· FLAG epitope inserted at the C-terminus after amino acid 381 [2]. The resulting protein was designated RNF13–HAF. Antiserum specific for the HA epitope recognizes only full-length RNF13 and full-length protein modified with chondroitin sulfate (Fig. 2, lanes 2 and 3). Antise- rum specific for the C-terminal FLAG epitope recog- nizes full-length RNF13 as well as C-terminal fragments (CTFs) and the intracellular C-terminal domain (ICD) (Fig. 2, lanes 5 and 6). Proteolytic cleavage of the N-terminal protease-asso- ciated (PA) luminal domain or ectodomain must occur because 43 and 39 kDa membrane-bound CTFs lack- ing the HA tag are detected when microsomal mem- branes isolated from B35 rat neuroblastoma cells treated with MG132 are stripped of peripheral proteins (Fig. 3, lanes 3 versus 5). A third CTF (37 kDa) is detected if cells are treated with epoxomicin (Fig. 2, lane 6). MG132 stabilizes CTF-39 efficiently, but this fragment is barely detectable in epoxomicin-treated cells; epoxomicin stabilizes CTF-37, but this form is hard to detect in MG132-treated cells (Fig. 2, lanes 5 versus 6). The fact that the CTFs differ slightly in size is consistent with the possibility that they are gener- ated by different ectodomain-cleaving proteases. Additional fragments of RNF13 are variably detected, such as an HA-tagged form lacking the C-terminal FLAG epitope (Fig. 3, lane 5, asterisk) [2], suggesting that additional cleavages can occur under specific conditions. The regulation of other PA-TM- RING proteins by proteolysis has not been described, although multiple forms of the homolog h-Goliath have been detected by in vitro translation in the pres- 1 2 3 4 5 Sol Mb Cells Mb Anti-HA Anti-FLAG 54- 38- * ICD Full-length CTFs Fig. 3. Proteolysis releases the cytoplasmic half of RNF13 from the membrane. Microsomal membranes were prepared [2] from MG132-treated B35 rat neuroblastoma cells [61] stably expressing RNF13–HAF. Proteins in the soluble cytoplasmic fraction (lane 1, Sol), in the soluble fraction generated when microsomal mem- branes (Mb) were stripped of peripheral proteins (lane 2), in the stripped microsomal membranes (lanes 3 and 5), and in whole cells (lane 4) were resolved on a 12% Laemmli polyacrylamide gel. Bio- synthetic forms of RNF13 were visualized on a western blot with anti-(FLAG-horseradish peroxidase) or anti-(HA- horseradish peroxi- dase) IgG, as indicated. Prestained molecular mass markers are indicated on the left. C D E B +PMA RNF13 Lamin B Merge A 5 µm 5 µm 5 µm 5 µm 5 µm +Dimethylsulfoxide Fig. 4. Phorbol ester-stabilized RNF13 can target to the inner nuclear membrane. Chinese hamster ovary cells (A,B) stably expressing RNF13–HAF were treated with vehicle (dimethylsulfoxide) (A) or with 1 l M 4b-phorbol 12-myristate 13-acetate (PMA) (B) for 6 h. RNF13– HAF expression was detected with mouse anti-HA IgG followed by donkey Alexa Fluor 568 anti-mouse serum. HeLa cells (C–E) tran- siently expressing RNF13–HAF were serum-starved for 2 h and trea- ted with 1 l M PMA for 4 h. RNF13 expression was detected with mouse anti-FLAG IgG followed by donkey AlexaFluor 568 anti-mouse serum. The inner nuclear membrane protein lamin B was stained with goat anti-(lamin B) IgG, followed by donkey AlexaFluor 488 anti- goat serum. J. P. Bocock et al. RNF13 trafficking and proteolysis FEBS Journal 278 (2011) 69–77 ª 2010 The Authors Journal compilation ª 2010 FEBS 71 ence of microsomal membranes [20], suggesting that proteolysis initiates in the endoplasmic reticulum or cis-Golgi. The extensive post-translational proteolysis of RNF13 occurs constitutively in cultured cells but is potentially regulated by physiological conditions in vivo. Proteolytic processing adds the potential for temporal control of ubiquitination activity, but it could also alter the cellular site of the ligase activity and thus control the potential substrate population. Post-translational modification of RNF13 RNF13 in cell extracts migrates as a 60 kDa [1] to 65 kDa [2] protein on PAGE. Pulse–chase analysis established that this is the initial biosynthetic form of RNF13 [2], but this is larger than predicted by the amino acid sequence, suggesting extensive post-trans- lational modification occurs. As expected, one to two asparagines in the luminal domain acquire high- mannose carbohydrate that becomes modified with complex sugars [1,2]. RNF13 is also modified by addi- tion of heterogeneous chondroitin sulfate glycosamino- glycan chains [2], producing a smear of protein bands migrating above full-length protein (Fig. 2). Proteogly- can is similarly added to integral membrane proteins which localize to endosomes and the plasma membrane, such as APP [21] and the immunoglobulin invariant chain [22,23]. Because most ubiquitin ligases self- ubiquitinate, it is also possible that some of the high- mass forms detected when proteasomal proteolysis is inhibited are heterogeneously ubiquitinated full-length RNF13. Expression in bacteria of a D1–205 RNF13 con- struct, which lacks the ectodomain and the transmem- brane (TM), generates a protein that migrates at 28 kDa, not at the expected 20 kDa, despite the absence of post-translational modification in prokary- otic cells [2]. When a similar construct is expressed in eukaryotic cells, it comigrates with the 38 kDa ICD [2]. Modifications that might contribute to the molecu- lar mass of the C-terminal tail in eukaryotic cells, but have not yet been identified on RNF13, include phos- phorylation and tyrosine sulfation, observed on APP [24], which is also a type 1 endosomal integral mem- brane protein, ubiquitination of the PEST sequence [2] or other lysines in the C-terminal half of the protein, and methylation, acetylation and sumoylation. PA domain cleavage The fate of the PA domain released from the mem- brane by proteolysis is unknown. If cleavage occurs on the endosomal membrane, the domain could be rapidly degraded in lysosomes. If cleavage occurs on the plasma membrane or if vesicles containing cleaved RNF13 fuse with the plasma membrane, the soluble PA domain could be released outside the cell. Under steady-state conditions, most of the protein resides on endosomes so the percentage released to the cell exterior is expected to be small, but physiological sig- nals could change the amount of proteolysis or the protein location relative to proteases. The PA domain has been predicted to be a protein interaction domain [25,26], but proteins which interact with the PA domain, either as part of the full-length RNF13 protein or as a solubilized domain, have not yet been identified. In analogy to epidermal growth factor receptor (EGFR) [27], the particular ligand bound could ultimately determine the cellular targeting and fate of the RNF13 molecules. RNF13 upregula- tion initiates in response to extracellular signals such as tenascin C and myostatin, but there is no evidence the ubiquitin ligase directly binds these molecules. In analogy to GRAIL [28], the RNF13 PA domain may bind the lumenal domain of integral membrane proteins which it subsequently ubiquitinates on the cytoplasmic side of the membrane. In analogy to plant RMR [11], RNF13 PA domain could bind proteases in the Golgi and mediate their targeting to multivesicular endosomes. Because ligands are commonly released by the acidic pH of endosomal vesicles, these two roles need not be mutually exclusive. The shed RNF13 PA ectodomain could, in analogy to EGFR superfamily members, serve extracellularly as a ligand that either acts as a juxtacrine factor or mediates transactivation of a distant unknown recep- tor. Ectodomain cleavage would terminate RING-med- iated ubiquitination if the PA domain selects targets. Activity would also be terminated if the PA domain mediates dimerization critical for function, this protein domain does in tomato subtilase [29,30]. Alternatively, it is possible that in vivo the cleaved PA domain remains associated with a CTF, as observed for the NOTCH heterodimer [31]. Liberation of the C-terminal tail The C-terminal cytoplasmic domain of RNF13 is also shed from the membrane and can be detected as a soluble protein, the ICD, in the cytoplasmic fraction when microsomal membranes are prepared from MG132-treated cells (Fig. 3, lane 1). Full-length RNF13, including protein modified with chondroitin sulfate, and the CTF forms remain in membranes when they are stripped of peripheral proteins (Fig. 3, lanes 3 RNF13 trafficking and proteolysis J. P. Bocock et al. 72 FEBS Journal 278 (2011) 69–77 ª 2010 The Authors Journal compilation ª 2010 FEBS and 5), indicating that these are all integral membrane proteins. Cleavage presumably occurs very near the membrane surface because the ICD comigrates on PAGE with the ectopically expressed cytoplasmic frag- ment RNF13d1–206, generated by insertion of a Met before Val207 [2]. The N-terminal residue of the endogenous ICD is not known because the precise cleavage site has not been determined. To escape rapid degradation, a polypeptide must begin with Met or Val, according to the N-end rule [32,33]. If the ICD initiates with Met202, localized at the TM–cytoplasmic junction, the ICD may evade N-terminal ubiquitination [33] and thus be sufficiently stable to mediate function. This proteolytic processing resembles that of ErbB4 [34]. For this EGFR family member, release from the membrane by proteolysis constitutes a switch from activation of one pathway by signaling as a TM protein to initiation of new functions mediated by the soluble ICD in a different cellular compartment [33]. Two different algorithms, PredictNLS [35] and pSORT II [36], predict that the mouse RNF13 ICD contains a nuclear localization signal (NLS) N-termi- nal to the RING domain. PredictNLS designates the sequence RRNRLRKD as an NLS and predicts the protein binds DNA. pSORT II predicts the NLS is PVHKFKK. These two sequences, separated by five amino acids, might function alone or as part of a tripartite NLS similar to that recently described for EGFR family members [37]. The cytoplasmic tails of proteins released from the membrane by regulated intramembrane proteolysis frequently undergo NLS- mediated import into the nucleus, where they modu- late transcription [33,38]. Thus it is possible that the RNF13 ICD that possesses the RING domain partic- ipates in modulation of transcription, perhaps con- trolling the half-life of transcription factors. RNF13 was initially postulated to regulate gene expression in the nucleus based on detection in the sequence of a leucine zipper, a motif that often mediates protein dimerization, a putative NLS, and the presence of a region rich in acidic amino acids following the RING domain [4]. The presence of a PEST sequence charac- teristic of rapidly turned over proteins is consistent with a role in regulation of transcription. Ectopically expressed RNF13d1–206 does not target to the nucleus [2], however, but interaction with other pro- teins or post-translational modification such as phos- phorylation may be required for protein stabilization and targeting to the nucleoplasm under specific physi- ological conditions. For example, the ICD of APP must associate with a histone acetyltransferase, Tip60, in order to be transported to the nucleus [39,40]. Targeting to the inner nuclear membrane Treatment of cells with phorbol esters such as 4b-phor- bol 12-myristate 13-acetate (PMA) activates protein kinase C (PKC), an enzyme that regulates cell prolifer- ation, differentiation, angiogenesis and apoptosis through the ability of its isoforms to initiate key sig- naling cascades at the plasma membrane (PM) [41]. Upon stimulation, PKCa and PKCbII and plasma membrane receptors, such as EGFR and the transfer- rin receptor, move to the pericentrion, a PKC-depen- dent subset of Rab11-positive recycling endosomes concentrated around the microtubule-organizing cen- ter ⁄ centrosome [42–45]. PMA similarly induces RNF13 to accumulate in perinuclear recycling endosomes (Fig. 4B,C), where it colocalizes with the transferrin receptor [46]. In HeLa cells, a spherical unstained area, characteristic of the centrosome, is often detected when cells are stimulated with PMA and stained for RNF13 (Fig. 4C). RNF13 could reach recycling endosomes via the PM or could be transported directly to recycling endosomes from the trans-Golgi network. There is increasing evidence supporting a role for recycling endosomes in biosynthetic pathways [47–49]. Surpris- ingly, on PMA treatment of cells,  20% of the RNF13 additionally moves to the inner nuclear mem- brane (INM), where it colocalizes with lamin B (Fig. 4E) [46]. Trafficking to recycling endosomes is required for subsequent transport to the INM, as expression of dominant-negative Rab11 blocks nuclear localization of RNF13 [46]. Full-length RNF13 possessing both epitope tags and RNF13 CTFs are found in purified INM fractions [46]. This signal-induced movement to the INM places the RING domain in the nucleoplasm and the PA domain between the two nuclear membranes. The PA domain could bind substrates at this site or possibly a protein bound to the PA domain earlier in the biosyn- thetic pathway might be transported to the membrane space as a PA domain ligand. This unusual targeting pathway has only been described for two PM-localized epidermal growth factor family members, heparin- binding epidermal growth factor (HB-EGF) [50] and proamphiregulin [51]. Both regulate transcriptional activity following localization in the INM. However, they possess short cytoplasmic tails of < 25 residues that are not removed by proteolysis. Selective seques- tration of receptors in the pericentrion is thought to protect them from agonist-induced degradation. Consistent with this, this altered cellular location of RNF13 coincides with an increase in protein stability [46]. These changes in cellular location prolong the J. P. Bocock et al. RNF13 trafficking and proteolysis FEBS Journal 278 (2011) 69–77 ª 2010 The Authors Journal compilation ª 2010 FEBS 73 ubiquitin ligase activity of RNF13 and expose the ubiquitin ligase to different substrates. Trafficking of a PA-TM-RING protein The complex cellular trafficking and regulation of RNF13 by proteolysis is diagrammed in Fig. 5. Transport of the newly synthesized protein across the endoplasmic reticulum membrane is mediated by the transient N-terminal signal peptide. High-mannose carbohydrate added co-translationally is modified with complex sugars in the Golgi and proteoglycan chains are added. On exiting the trans-Golgi network, RNF13 can enter a constitutive pathway (Fig. 5A) or a signal- induced pathway (Fig. 5B). In analogy to APP [24], RNF13 may reach multivesicular bodies (MVBs) fol- lowing transport to the PM, followed by endocytosis into endosomes. The presence of a dileucine motif in the cytoplasmic half of the mouse protein (amino acids 307–312) suggests that the protein is capable of under- going endocytosis. Alternatively, RNF13 could travel directly to MVBs in analogy to the plant homolog RMR, which serves as a targeting receptor delivering enzymes to protein storage bodies [11,52]. Once local- ized in an MVB, RNF13 could mediate ubiquitination of substrates on the endosomal membrane, binding substrates with the lumenal PA domain. In addition, it is well-established that in response to specific stimuli, such as calcium ionophores, MVBs can fuse with the plasma membrane [53,54]. This could position RNF13 on the PM. Endocytosis of PM-localized RNF13 might expose RNF13 to endosome-localized proteases, resulting in solubilization of essentially the entire cytoplasmic tail. The ICD could be turned over by proteasomes or possibly, as a result of post-transla- tional modification and ⁄ or association with interacting proteins, the NLS could mediate import into the nucle- oplasm. Here RNF13 could potentially mediate ubiq- uitination, perhaps utilizing the Ser-rich C-terminal domain to bind soluble substrates in the absence of the PA domain. In response to extracellular signals that activate PKC, RNF13 can enter a regulated trafficking path- way that ultimately delivers the protein to the INM. Our studies indicate that it is newly synthesized RNF13, not protein stored in MVBs, which enters this pathway [46]. Following PKC treatment, the majority of RNF13 localizes to recycling endosomes. Both full- length and RNF13 CTFs ultimately reach the INM [46], where they colocalize with lamin B, a component of the inner nuclear membrane. Thus RNF13 can potentially ubiquitinate substrates in organelles of the biosynthetic pathway, such as the endoplasmic reticulum, Golgi, PM or endosomes. In addition, RNF13 has the potential to ubiquitinate two distinct sets of nuclear proteins. Full-length RNF13 positioned in the INM could capture integral mem- brane protein substrates via its N-terminal PA domain. ICD soluble in the cytoplasm could capture nucleo- plasm substrates via its C-terminal Ser-rich domain. The two pathways targeting RNF13 to the nucleus presumably lead to ubiquitination of distinct sets of substrates. Thus a single ubiquitin ligase may ubiquiti- nate different substrates under different physiological conditions that alter its cellular localization. This complex regulation by cellular targeting and proteolysis is unique for ubiquitin ligases, which are commonly soluble proteins, but similar to that described for such physiologically important proteins as Notch [31], members of the EGFR superfamily of tyrosine kinases [55,56] and APP [24,57]. The growing appreciation of the role of both the nuclear membrane [58,59] and endosomes [60] in the regulation of tran- scription suggests PA-TM-RING ubiquitin ligases are well-positioned to impact key regulatory events of the cell. C N 2 Cleavage EE C N Golgi C N PM Endocytosis Secretion Nucleus C ONM INM C C N MVB C N C N C N Golgi C N RE C N EE C N ONM INM Nucleus AB Constitutive pathway Signal-induced pathway PM Fig. 5. RNF13 targeting pathways. RNF13 trafficking and proteolysis J. P. Bocock et al. 74 FEBS Journal 278 (2011) 69–77 ª 2010 The Authors Journal compilation ª 2010 FEBS Acknowledgements This work was supported by research grants MCB-0235680 and MCB-0938796 (to AHE) from the National Science Foundation. Microscopy was performed at the University of North Carolina in the Microscopy Services Laboratory, Department of Pathology and Laboratory Medicine, under the direction of C. Robert Bagnell, Jr. Cell sorting was per- formed at the UNC Flow Cytometry Facility that is under the direction of L. Arnold. References 1 Zhang Q, Meng Y, Zhang L, Chen J & Zhu D (2008) RNF13: a novel RING-type ubiquitin ligase over- expressed in pancreatic cancer. Cell Res 19, 348–357. 2 Bocock JP, Carmicle S, Chhotani S, Ruffolo MR & Erickson AH (2009) The PA-TM-RING protein RNF13 is an endosomal integral membrane E3 ubiquitin ligase whose RING finger domain is released to the cytoplasm by proteolysis. FEBS J 276, 1860–1877. 3 Zhang Q, Wang K, Zhang Y, Meng J, Yu F, Chen Y & Zhu D (2009) The myostatin-induced E3 ubiquitin ligase RNF13 negatively regulates the proliferation of chicken myoblasts. FEBS J 277, 466–476. 4 Tranque P, Crossin KL, Cirelli C, Edelman GM & Mauro VP (1996) Identification and characterization of a RING zinc finger gene (C-RZF) expressed in chicken embryo cells. Proc Natl Acad Sci USA 93, 3105–3109. 5 Lomax MI, Warner SJ, Bersirli CG & Gong T-WL (1998) The gene for a RING zinc finger protein is expressed in the inner ear after noise exposure. Prim Sens Neuron 2, 305–316. 6 Jin X, Cheng H, Chen J & Zhu D (2010) RNF13: an emerging RING finger E3 ubiquitin ligase important in cell proliferation. FEBS J 278, 78–84. 7 Saito S, Honma K, Kita-Matsuo H, Ochiya T & Kato K (2005) Gene expression profiling of cerebellar development with high-throughput functional analysis. Physiol Genomics 22, 8–13. 8 Anandasabapathy N, Ford GS, Bloom D, Holness C, Paragas V, Seroogy C, Skrenta H, Hollenhorst M, Fathman CG & Soares L (2003) GRAIL: an E3 ubiquitin ligase that inhibits cytokine gene transcription is expressed in anergic CD4+ T cells. Immunity 18, 535–547. 9 Whiting CC, Su LL, Lin JT & Fathman CG (2010) GRAIL: a unique mediator of CD4 T-lymphocyte unresponsiveness. FEBS J 278, 47–58. 10 Jiang L, Phillips TE, Rogers SW & Rogers JC (2000) Biogenesis of the plant storage vacuole crystalloid. J Cell Biol 150, 755–770. 11 Wang H, Rogers JC & Jiang L (2010) Plant RMR proteins: unique vacuolar sorting receptors that couple ligand sorting with membrane internalization. FEBS J 278, 59–68. 12 Kopan R, Schroeter EH, Weintraub H & Nye JS (1996) Signal transduction by activated mNotch: importance of proteolytic processing and its regulation by the extracellular domain. Proc Natl Acad Sci USA 93, 1683–1688. 13 Longva KP, Blystad FD, Stang E, Larsen AM, Johann- essen LE & Madshus IH (2002) Ubiquitination and proteosomal activity is required for transport of the EGF receptor to inner membranes of multivesicular bodies. J Cell Biol 156, 843–854. 14 Cowan JW, Wang X, Guan R, He K, Jiang J, Baumann G, Black RA, Wolfe MS & Frank SJ (2005) Growth hormone receptor is a target for presenilin- dependent gamma-secretase cleavage. J Biol Chem 280, 19331–19342. 15 Steinhilb ML, Turner RS & Gaut JR (2001) The prote- ase inhibitor, MG132, blocks maturation of the amyloid precursor protein Swedish mutant preventing cleavage by beta-secretase. J Biol Chem 276, 4476–4484. 16 Meng L, Mohan R, Kwok BHB, Elofsson M, Sin N & Crews CM (1999) Epoxomicin, a potent and selective proteasome inhibitor, exhibits in vivo anti-inflammatory activity. Proc Natl Acad Sci USA 96, 10403–10408. 17 Tsubuki S, Saito Y, Tomioka M, Ito H & Kawashima S (1996) Differential inhibition of calpain and proeasome activities by peptidyl aldehydes of di-leucine and tri-leucine. J Biochem 119, 572–576. 18 Lee DH & Goldberg AL (1998) Proteasome inhibitors: valuable new tools for cell biologists. Trends Cell Biol 8 , 397–403. 19 Bendtsen JD, Neilsen H, Von Heijne G & Brunak S (2004) Improved prediction of signal peptides: SignalP 3.0. J Mol Biol 340, 783–795. 20 Guais A, Siegrist S, Solhonne B, Jouault H, Guelaen G & Bulle F (2006) h-Goliath, paralog of GRAIL, is a new E3 ligase protein, expressed in human leukocytes. Gene 374, 112–120. 21 Shioi J, Anderson JP, Ripellino JA & Robakis NK (1992) Chondroitin sulfate proteoglycan form of the Alzheimer’s beta-amyloid precursor. J Biol Chem 267, 13819–13822. 22 Sant AJ, Cullen SE, Giacoletto KS & Schwartz BD (1985) Invariant chain is the core protein of the Ia-associated chondroitin sulfate proteoglycan. J Exp Med 162, 1916–1934. 23 Arneson LS & Miller J (2007) The chondroitin sulfate form of invariant chain trimerizes with conventional invariant chain and these complexes are rapidly transported from the trans-Golgi network to the cell surface. Biochem J 406, 97–103. 24 Thinakaran G & Koo EH (2008) Amyloid precursor protein trafficking, processing, and function. J Biol Chem 283, 29615–29619. J. P. Bocock et al. RNF13 trafficking and proteolysis FEBS Journal 278 (2011) 69–77 ª 2010 The Authors Journal compilation ª 2010 FEBS 75 25 Mahon P & Bateman A (2000) The PA domain: a protease-associated domain. Protein Sci 9, 1930–1934. 26 Luo X & Hofmann K (2001) The protease-associated domain: a homology domain associated with multiple classes of proteases. Trends Biol Sci 26, 147–148. 27 Roepstorff K, Grandal MV, Henriksen L, Knudsen SLJ, Lerdrup M, Grovdal L, Willumsen BM & van Deurs B (2009) Differential effects of EGFR ligands on endocytic sorting of the receptor. Traffic 10, 1115–1127. 28 Lineberry N, Su L, Soares L & Fathman CG (2008) The single-subunit transmembrane E3 ligase gene related to anergy in lymphocytes (GRAIL) captures and then ubiquitinates transmembrane proteins across the cell membrane. J Biol Chem 283, 28497–28505. 29 Ottmann C, Rose R, Huttenlocher F, Cedzich A, Huaske P, Kaiser M, Huber R & Schaller A (2009) Structural basis for Ca 2+ -independence and activation by homodimerization of tomato subtilase 3. Proc Natl Acad Sci USA 106, 17223–17228. 30 Rose R, Schaller A & Ottmann C (2010) Structural features of plant subtilases. Plant Signal Behav 5, 180– 183. 31 Fortini ME (2009) Notch signaling: the core pathway and its posttranslational regulation. Dev Cell 16, 633–647. 32 Ravid T & Hochstrasser M (2008) Diversity of degradation signals in the ubiquitin–proteasome system. Nat Rev Mol Cell Biol 9, 679–689. 33 Hass MR, Sato C, Kopan R & Zhao G (2009) Presenilin: RIP and beyond. Semin Cell Dev Biol 20, 201–210. 34 Ni CY, Murphy MP, Golde TE & Carpenter G (2001) Gamma-secretase cleavage and nuclear localization of ErbB-4 receptor tyrosine kinase. Science 294 , 2179–2181. 35 Cokol M, Nair R & Rost B (2000) Finding nuclear localization signals. EMBO Rep 1, 411–415. 36 Nakai K & Horton P (1999) PSORT: a program for detecting the sorting signals of proteins and predicting their subcellular localization. Trends Biochem Sci 24, 34–35. 37 Hsu S-C & Hung M-C (2007) Characterization of a novel tripartite nuclear localization sequence in the EGFR family. J Biol Chem 282, 10432–10440. 38 Wolfe MS (2009) Intramembrane proteolysis. Chem Rev 109, 1599–1612. 39 Cao X & Sudhof TC (2001) A transcriptionally active complex of APP with FE65 and histone acetyltransfer- ase Tip60. Science 293, 115–120. 40 Walsh DM, Fadeeva JV, LaVoie MJ, Paliga K, Eggert S, Kimberly WT, Wasco W & Selkoe DJ (2003) Gamma-secretase cleavage and binding to FE65 regulate the nuclear translocation of the intracellular C-terminal domain (ICD) of the APP family of proteins. Biochemistry 42, 6664–6673. 41 Mellor H & Parker PJ (1998) The extended protein kinase C superfamily. Biochem J 332, 281–292. 42 Becker KP & Hannun YA (2003) cPKC-dependent sequestration of membrane-recycling components in a subset of recycling endosomes. J Biol Chem 278, 52747– 52754. 43 Idkowiak-Baldys J, Becker KP, Kitatani K & Hannun YA (2006) Dynamic sequestration of the recycling compartment by classical protein kinase C. J Biol Chem 281, 22321–22331. 44 Alvi F, Idkowiak-Baldys J, Baldys A, Raymond JR & Hannun YA (2007) Regulation of membrane trafficking and endocytosis by protein kinase C: emerging role of the pericentrion, a novel protein kinase C-dependent subset of recycling endosomes. Cell Mol Life Sci 64, 263–270. 45 Idkowiak-Baldys J, Baldys A, Raymond JR & Hannun YA (2009) Sustained receptor stimulation leads to sequestration of recycling endosomes in a classical protein kinase C- and phospholipase D-dependent manner. J Biol Chem 284, 22322–22331. 46 Bocock JP, Carmicle S, Madamba E & Erickson AH (2010) Nuclear targeting of an endosomal E3 ubiquitin ligase. Traffic 11, 756–766. 47 Ang AL, Taguchi T, Francis S, Folsch H, Murrells LJ, Pypaert M, Warren G & Mellman I (2004) Recycling endosomes can serve as intermediates during transport from the Golgi to the plasma membrane of MDCK cells. J Cell Biol 167 , 531–543. 48 Schmidt MR & Haucke V (2007) Recycling endo- somes in neuronal membrane traffic. Biol Cell 99, 333– 342. 49 Gonzalez A & Rodriguez-Boulan E (2009) Clathrin and AP1B: key roles in basolateral trafficking through trans- endosomal route. FEBS Lett 583, 3784–3795. 50 Hieda M, Isokane M, Koizumi M, Higashi C, Tachibana T, Shudou M, Taguchi T, Hieda Y & Higashiyama S (2008) Membrane-anchored growth factor, HB-EGF, on the cell surface targeted to the inner nuclear membrane. J Cell Biol 180, 763–769. 51 Isokane M, Hieda M, Hirakawa S, Shudou M, Nakashiro K, Hashimoto K, Hamakawa H & Higashiyama S (2008) Plasma-membrane-anchored growth factor pro-amphiregulin binds A-type lamin and regulates global transcription. J Cell Sci 121, 3608– 3618. 52 Park JH, Oufattole M & Rogers JC (2007) Golgi-mediated vacuolar sorting in plant cells: RMR proteins are sorting receptors for the protein aggregation ⁄ membrane internalization pathway. Plant Sci 172, 728–745. 53 Thery C, Zitvogel L & Amigorena S (2002) Exosomes: composition, biogenesis and function. Nat Rev Immunol 2, 569–579. RNF13 trafficking and proteolysis J. P. Bocock et al. 76 FEBS Journal 278 (2011) 69–77 ª 2010 The Authors Journal compilation ª 2010 FEBS 54 Raiborg C, Rusten TE & Stenmark H (2003) Protein sorting into multivesicular endosomes. Curr Opin Cell Biol 15, 446–455. 55 Blobel CP, Carpenter G & Freeman M (2009) The role of protease activity in ErbB biology. Exp Cell Res 315, 671–682. 56 Sorkin A & Goh LK (2008) Endocytosis and intracellu- lar trafficking of ErbBs. Exp Cell Res 314, 3093–3106. 57 De Strooper B (2010) Proteases and proteolysis in Alzheimer disease: a multifactorial view on the disease process. Physiol Rev 90, 465–494. 58 Hetzer MW & Wente SR (2009) Border control at the nucleus: biogenesis and organization of the nuclear membrane and pore complexes. Dev Cell 17, 606–616. 59 Mekhail K & Moazed D (2010) The nuclear envelope in genome organization, expression and stability. Nat Rev Mol Cell Biol 11, 317–328. 60 Pyrzynska B, Pilecka I & Miaczynska M (2009) Endocytic proteins in the regulation of nuclear signaling, transcription and tumorigenesis. Mol Oncol 3, 321–338. 61 Otey CA, Boukhelifa M & Maness P (2003) B35 neuro- blastoma cells: an easily transfected, cultured cell model of central neurvous system neurons. Methods Cell Biol 71, 287–305. J. P. Bocock et al. RNF13 trafficking and proteolysis FEBS Journal 278 (2011) 69–77 ª 2010 The Authors Journal compilation ª 2010 FEBS 77 . MINIREVIEW Trafficking and proteolytic processing of RNF13, a model PA-TM-RING family endosomal membrane ubiquitin ligase Jeffrey P. Bocock 1 , Stephanie Carmicle 2 ,. Isokane M, Hieda M, Hirakawa S, Shudou M, Nakashiro K, Hashimoto K, Hamakawa H & Higashiyama S (2008) Plasma -membrane- anchored growth factor pro-amphiregulin