Send Orders of Reprints at [email protected] Current Pharmaceutical Design, 2013, 19, 000-000 1

Emerging Role of the -proteasome System as Drug Targets

Gozde Kar1, Ozlem Keskin1, Franca Fraternali2 and Attila Gursoy1,*

1Center for Computational Biology and Bioinformatics and College of Engineering, Koc University, Rumelifeneri Yolu, 34450 Sari- yer Istanbul, Turkey; 2Randall Division of Cell and Molecular Biophysics, New Hunt's House, King's College London, Guy's Campus, SE1 1UL London UK

Abstract: The ubiquitin-proteosome system (UPS) regulates a wide range of cellular processes including degradation, DNA re- pair, transcription regulation, and . Alterations and mutations in UPS components give rise to various human diseases, most prominently cancer and neurodegenerative disorders. Here, we review recent advances in UPS-based drug discovery highlighting the emerging relationships between the UPS and disease and discuss potential future therapeutic interventions. In particular, we focus on re- cent structural approaches in UPS and explain how the knowledge of protein structural details can guide the design of specifically tar- geted inhibitors. Keywords: Ubiquitin-proteosome system, disease, drug design.

1. INTRODUCTION particularly on targeting protein-protein interactions and interfaces The ubiquitin-proteosome system (UPS) is essential for the of UPS components in disease with the ultimate purpose of design- protein turnover and has a wide variety of functions ranging from ing more efficient and selective inhibitors. cell-cycle control and transcription to development [1]. Ubiquitin is 2. TARGETING UPS IN DISEASE a small 76-aminoacid protein that can reversibly bind to and label other in a highly specific manner. Ubiquitin attachment Ubiquitin attachment to proteins (ubiquitination) is a major takes place via a sequential enzymatic route involving ubiquitin post-translational modification that can have profound effects on activation (by E1 enzymes), ubiquitin conjugation (by E2 en- protein stability, localization or interaction pattern [8]. Several pro- zymes), and ubiquitin substrate tagging (by E3 enzymes). In the teins regulated by ubiquitination control cell signaling events, and first step, ubiquitin is activated by an ATP-dependent E1 and E1 hence are crucial for cell-cycle progression, proliferation and apop- delivers the activated ubiquitin to E2 through a transthiolation reac- tosis [8]. Such signaling processes are frequently altered in cancer tion. E3s recognize specific target substrates and catalyze the ubiq- with several tumor suppressors and oncogenes representing the uitin transfer from the E2 to the substrate. Ubiquitin can be conju- enzymes of the ubiquitin conjugation and deconjugation pathways. gated to the substrate as a monomer on one (monoubiquitination) or The first inhibitor targeting the proteosomal degradation of proteins more substrate (multiubiquitination) or as a polymer in cancer is bortezomib, which has illustrated a clinical success and (polyubiquitination) by forming ubiquitin chains [2]. Seven stimulated further research in the design of specific UPS-based residues of ubiquitin can participate in ubiquitin conjugation, lead- inhibitors. Subsequently, other UPS enzymes such as E1, E2 and E3 ing to ubiquitin chains of different lengths, topologies and func- have started to be exploited as targets in disease. Below, we start by tions. Substrates that are tagged by ubiquitin chains conjugated describing the role of each enzyme classes (E1, E2, E3, DUBs and through the lysine 48 of ubiquitin are targeted by proteosome and proteasome) in the UPS pathway and in disease and by listing all undergo degradation process, whereas chains linked at lysine 63 are the so far available inhibitors and therapeutic approaches. The UPS associated with signaling and trafficking events [3]. Importantly, pathway and the various possibilities for therapeutic intervention ubiquitination is a reversible process; the deubiquitinating enzymes are displayed in (Fig. 1). (DUBs) can in fact remove from modified substrates, thus rescuing the substrate from degradation. Targeting E1 Activating Enzymes Considering the crucial multiple biological roles of UPS en- The first step of the UPS pathway is ubiquitin activation, per- zymes, it is expected that malfunctioning of any component partici- formed by E1 enzymes. Ubiquitin (Ub) or ubiquitin-like proteins pating in the regulation of these enzymes would result in disease. (Ubl) are adenylated at their C-terminal glycine residue and react There is in fact accumulating evidence of altered functions of UPS with the E1 active site thiol to become charged as a thiol ester. enzymes in numerous cancer types [4], cardiovascular and viral Then, Ub or Ubl are transferred to E2 conjugating enzymes also diseases [5, 6] and neurodegenerative disorders [7]. In treating such through a thiol ester bond. There are eight structurally and func- diseases, developing efficient small-molecule inhibitors targeting tionally similar E1s that can activate Ub and Ubls. Although Ub- the UPS pathway rather than the single molecules is of crucial im- activating E1s can charge a variety of E2s, Ubl-activating E1s ap- portance and has elicited significant interest in recent drug design pear to charge only a single E2 [1]. studies. In this review, we first summarize role of UPS in human For inhibiting ubiquitin activation, there are a couple of points diseases along with the current approaches in the development of to target (Fig. 1). One of them is to prevent ATP from binding to E1 small-molecule inhibitors targeting UPS components. We then and hence to inhibit Ub-adenylate formation using small-molecule discuss the potential future therapeutic interventions, focusing inhibitors. Another possibility is to target the E1 active site thiol for blocking the formation of E1-Ub complex. One example is PYR41 which is a pyrazone derivative designed for inhibiting HDM2- *Address correspondence to this author at the Center for Computational dependent ubiquitination. The nitrogen dioxide group of Biology and Bioinformatics and College of Engineering, Koc University, PYR41 covalently modifies the active site cysteine of E1 while, Rumelifeneri Yolu, 34450 Sariyer Istanbul, Turkey; notably, it does not inhibit other thiol-dependent enzymes including Tel:/Fax: ???????????????????????; E-mail: [email protected] many E2s [9]. The downstream effects of PYR41 treatment are NF-

1381-6128/13 $58.00+.00 © 2013 Bentham Science Publishers 2 Current Pharmaceutical Design, 2013, Vol. 19, No. 00 Kar et al.

Fig. (1). UPS mechanism and several potential UPS-based drug targets. Ubiquitin attachment takes place via a sequential enzymatic route. In the first step, ubiquitin is activated by an ATP-dependent E1 and E1 delivers the activated ubiquitin to E2 through a transthiolation reaction. E3s recognize specific target substrates and catalyze the ubiquitin transfer from the E2 to the substrate. Substrates that are tagged by ubiquitin chains conjugated through the lysine 48 of ubiquitin are targeted by proteosome and undergo degradation process, whereas monoubiquitination is associated with signaling and localization events [3]. The deubiquitinating enzymes (DUBs) can remove ubiquitins from modified substrates. There are several potential drug targets in UPS in disease: the activity of E1, E2, E3 enzymes, DUB and proteasome can be inhibited. Additionally, inhibition by blocking the interface regions of E1-E2, E2-E3 and E3-susbtrate interactions would provide an efficient means in treating UPS-associated diseases.

B activation, stabilization of p53 and induction of p53-dependent (~500 or more have been proposed to exist in human) [18], which transcription [9]. Besides PYR41, the nitric oxide (NO)-producing suggests that one E2 can recognize several different E3s [19]. The drug JS-K also inhibits E1-ubiquitin thioester formation through an principles determining the E2-E3 selectivity are unclear and in interaction between the E1 active-site thiol and NO [10]. Similarly, many cases it is not known which E3s interact with which E2s [20]. this compound results in an increased p53 expression and decreased Since the knowledge about the functions of E2s is also limited, it is levels of total ubiquitinated proteins. E1s can also be inhibited difficult to design selective E2 inhibitors. Despite these challenges, through adduct formation. One example of such compounds is there are emerging approaches for inhibiting E2s (Fig. 1). One ex- Nedd8-activating enzyme (NAE-E1) inhibitor; MLN4924, which is ample for blocking E1-E2 interaction is a synthetic peptide called an adenosine sulphamate analogue [11, 12] that binds to Nedd8 UBC12N26, which was shown to compete for UBC12-NAE bind- thioster form of NAE-E1 and then forms a covalent adduct with ing, hence blocking transthiolation of Nedd8 to UBC12 and pre- Nedd8. Blocking NAE by MLN4924 results in the inhibition of venting the transfer of Nedd8 to the substrate targets [21]. For some CRL (-Ring Finger Ligase) activity and thus stabilization of E2s such as UBE2G2, E3 binding can allosterically activate ubiq- CRL substrates (such as CDT1), leading to deregulation of DNA uitination activity, then another potential mode of inhibition may be synthesis during the of the cell division cycle and apoptosis the blocking E2-dependent ubiquitin transfer using allosteric effec- in proliferating cells [11]. MLN4924 has entered Phase II clinical tors. An additional approach is to inhibit catalytic activity of E2s by trials for treating multiple myeloma and non-Hodgkin’s lymphoma. targeting the conserved asparagine residue of E2s.

Targeting E2 Conjugating Enzymes Targeting E3 Ubiquitin Ligases E2s are involved in ubiquitin conjugation and ubiquitin transfer E3s recognize specific target substrates and catalyze the ubiq- to the substrates. E2s mediate ubiquitin chain assembly and ubiq- uitin transfer from the E2 to the substrate. E3s are represented by uitin linkage topology, therefore together with E3s, leading to de- three main classes: Homologs of E6AP Carboxy Terminus (HECT), termination of the fate of the substrates. E2s have been implicated Really Interesting New (RING), and the UFD2 homology (U- in many diseases including cancer. For example, UBCH10 was box) family proteins. HECT is a domain of ~350 amino acids; its observed in chromosomal instability [13], UBE2T in lung cancer N-terminal lobe contains the E2-binding site, and the C-terminal [14], UBE2Q2 in head and neck squamous cell carcinoma [15] and lobe confers catalytic activity. The conserved Cys residue in the C- the SUMO E2 UBC9 in ovarian carcinoma and breast cancer [16, terminal lobe forms thioester complexes with ubiquitin [22]. The 17]. RING finger domain contains a short motif rich in histidine and Currently, there is a limited number of known E2s (~40 in hu- cysteine residues that coordinate zinc atoms in a cross-brace struc- man); however the number of known E3s is increasing rapidly ture, characterized by a central -helix and variable-length loops separated by small beta strands [23]. The U-box domain is similar Emerging Role of the Ubiquitin-proteasome System Current Pharmaceutical Design, 2013, Vol. 19, No. 00 3 to the structure of the RING domain with the exception that it lacks age repair and check point control processes. It is mutated the conserved histidine and cysteine residues [24]. The ubiquitina- in more than 50% of inherited breast cancers [36] and its mutation tion mechanisms for these three E3 classes differ. For HECT do- leads to early-onset breast and ovarian cancers and to the develop- main E3s, the ubiquitin is first transferred from E2 to the active site ment of a malignant phenotype [37]. BRCA1 heterodimerizes with residue of HECT E3, with subsequent transfer from E3 to the sub- another RING finger protein, BRCA1 Associated Ring Domain 1 strate protein. For RING and U-box type E3s, ubiquitin is directly (BARD1) increasing its activity [38]. transferred from E2 to the substrate without an E3 intermediate The gene encodes a RING finger E3 that can ubiquiti- linkage. nate itself and the -synuclein interacting protein, Synphilin 1. Fa- Although the encodes numerous E3 ligases [25] milial associated mutations in Parkin leads to a deficiency in E3 and there are several efforts in identifying functions of E3s and activity and thus to Parkinson’s disease [39]. Histopathological their association to human disease, progresses in this field are still hallmark of Parkinson’s disease are the so-called Lewy bodies in its infancy. In the UPS, E3 ligases are probably the most attrac- which are abnormal protein aggregates developing inside nerve tive drug targets since the substrate selectivity relies mainly on the cells. Lew bodies contain ubiquitin and filamentous aggregates of specificity of E3s. Potential points to target are: E2-E3 binding denatured proteins including -synuclein [40]. interface, E3-substrate recognition and the activity of the E3 itself Another E3 ubiquitin ligase is VHL (von Hippel-Lindau) that is (Fig. 1). Below, we present E3s, their alterations in diseases, the mutated in a familial cancer susceptibility syndrome, von Hippel- available inhibitors targeting these E3s; a summary is provided in Lindau syndrome. This disease is characterized by the development Table 1. of various highly angiogenic tumors such as renal cell carcinoma, RING-finger Domain Type E3s pancreatic cysts and tumors [41]. VHL ligase functions as a sub- is a RING-finger E3 ligase that mediates ubiquitination strate recognition subunit in the VCB-Cullin 2-VHL complex and of tumor suppressor protein p53. Mdm2 is an oncogenic protein mediates ubiquitination of HIF-1 (Hypoxia-inducible factor-1) whose gene amplification and overexpression are prominently pre- protein [42]. HIF-1 functions in oxygen homeostasis and is a regu- sent in human cancers including breast cancer, lung cancers and lator of energy metabolism and angiogenesis processes [42]. Under many other tumor types [26, 27]. There have been so far a numer- normal oxygen levels, HIF-1 is hydroxylated, recognized by VHL ous approaches that have proposed compounds disrupting the inter- and ubiquitinated by VCB-Cullin 2-VHL complex leading to its action between E3 and its substrates. One of such inhibitors is Nut- proteasomal degradation. However, hypoxia decreases the capacity lin /R7112, which has entered clinical trials. Nutlin /R7112 targets of prolyl hydroxylases to hydroxylate HIF-1, and in turn HIF-1 the interface of E3 ligase MDM2 (HDM2) and its substrate p53 to escapes from VHL-induced degradation, which promotes expres- prevent p53 degradation and thus to elevate p53 expression [28]. sion of hypoxia inducible such as Glucose Transporter 1 and Similarly, the compound RITA is used to promote p53 stabilization Vascular Endothelial Growth Factor (VEGF) [4]. Mutations in [29]. Known disease associations include various cancers; breast VHL are observed to prevent degradation of HIF-1 which leads to and lung, blood cancers and multiple myeloma [30]. upregulation of HIF-1-induced genes and to high levels of tumor vascularization [43]. Consequently, reactivating the VHL ligase Another ligase class that can be targeted in multiple cancers and would be a possible strategy to treat VHL-associated tumors. In one disorders linked to the NF-B pathway is cullin-RING E3 ligases Skp2 such effort, a bioengineered VHL complex is exploited to increase (CRLs). SCF is a CRL that ubiquitinates p27 targeting it for HIF-1 degradation [44]. proteosomal degradation [31]. Elevated levels of SCFSkp2 and de- creased levels of p27 are observed in some human cancers. Chen et HECT Domain Type E3s al. [32] have identified a compound that is active against leukemia E6-Associated Protein (E6-AP) is a member of HECT E3 ubiq- cells and prevent the formation of the functional SCF complex, uitin ligase family that catalyzes ubiquitination and subsequent which stabilizes p27, induce G1/S cell cycle arrest, and thereby degradation of p53 upon infection by Highrisk human papillomavi- triggers cell death [32]. SCFTRCP is a CRL that mediates the degra- ruses (HPV) [45]. HPV 16 and 18 are associated prominently with dation of IB; the inhibitory component of the proinflammatory cervical cancer together with genital and head and neck cancers transcription factor NF-B [33]. An inhibitor of SCFTRCP was [46]. Additionally, mutations in E6-AP are associated with Angel- found as well, which prevents the polyubiquitination and subse- man’s syndrome [47]. quent degradation of IB [33]. Another example is the inhibitor of ARF-BP1 (HUWE1) is another HECT-type E3 ligase that can Cdc4, the yeast ortholog of the mammalian Cullin RING E3 ligase induce p53 ubiquitination and degradation [48]. ARF-BP1 is highly Fbw7, that was identified in a biochemical screen and disrupts bind- expressed in breast cancer cell lines and therefore appears to have a ing of Cdc4 to its substrate SIC1 [34]. Inhibition is achieved potential role in breast cancer tumorigenesis [49]. through an allosteric mechanism: the inhibitor binds to Cdc4 at a site distant from SIC1 binding, which induces a large conforma- Another HECT type E3 is Nedd4-1, belonging to Nedd4-like tional change resulting in a distortion of the substrate binding subgroup, negatively regulates tumor suppressor Phosphatase and pocket and impediment of the substrate recognition [34]. Although Tensin Homolog protein (PTEN) by ubiquitinating it for degrada- these inhibitors of CRL have not entered into clinical trials yet, they tion [50]. Overexpression of Nedd4-1 was detected in multiple represent a strong new avenue for treating several human diseases. human cancer samples and the depletion of Nedd4-1 inhibited xenograft tumor growth in a PTEN-dependent manner [50]. On the The Inhibitors of Apoptosis Proteins (IAPs) are also actively contrary, one study reported no dysregulation of PTEN protein targeted E3 ligases in disease. They mediate degradation of several level upon Nedd4-1 knockout [51]. PTEN is frequently mutated in substrates involved in apoptosis and signaling and are linked to human tumors [52]. A more exact explanation of the role of Nedd4- several cancer types such as liver and lung, ovarian carcinoma, 1 in regulating PTEN is therefore necessary to elucidate mecha- solid and lymphoid tumors [30]. There are a few antagonists of nisms of tumor formation and progression. IAPS such as GDC-0152, LCL161, YM155 and TL32711 that have entered clinical trials. These compounds are mimetics of SMAC Smurf2 is also a HECT E3 and has a key function in mediating proteins that antagonize IAPs and promote IAP auto-ubiquitination TGF signaling pathway [53, 54]. It mediates ubiquitination of and degradation, leading to the death of cancerous cells through Smad1, Smad2 and TGF receptor 1. High-level expression of TNF pathway [35]. Smurf2 has been observed to correlate with tumor development, lymph node metastasis and a poor survival rate in esophageal BRCA1 (Breast cancer type 1 susceptibility protein) is an E3 squamous cell cancer [55]. ligase with tumor suppressor activity that is involved in DNA dam- 4 Current Pharmaceutical Design, 2013, Vol. 19, No. 00 Kar et al.

Table 1. E3 ligases and Disease

E3 Substrate Associated disease Alterations in disease Inhibitor/Treatment

Brca1 NPM1 [170], Breast and ovarian cancer [37] Promoter hypermethylation [174], - (RING-finger) RBBP8 [171], genomic deletions and duplica- RNA polymerase tions [175] [172], ESR1 [173]

E6-AP p53 [45], Cervical, genital, head and neck Mutations associated with Angel- mRNA decay factor TTP promotes (HECT) p27 [176] carcinomas [45, 177], Angelman’s man’s syndrome [47], infection by rapid decay of E6-AP and stabi- syndrome [47] high-risk HPV in cancer [45, 177] lizes p53 in cervical cancer [178]

IAP Several substrates Various cancers including liver, Overexpression of IAPs in various Many inhibitors targeting IAPs (RING-finger) lung, ovarian carcinoma, lymphoma cancers [179] such as GDC-0152, AEG35156, YM155, LCL161, AT-406

Mdm2 p53 and p27 [180] Various cancers including breast and Overexpression and amplification Nutlins, RITA, Parthenolide (RING-finger) lung [27], point mutations and nucleo- tide insertions in human tumors [181]

Nedd4-1 PTEN [50] Multiple cancers, including bladder, Overexpression and mutations in - (HECT) prostate, colorectal and gastric can- cancer [50, 182] cer [50, 182]

Parkin PARKIN, Synphilin Parkinson’s disease [39] Familial associated mutations [39] Nitric oxide to inhibit Parkin’s (RING-finger) 1 [39] ligase activity [183]

SCFTRCP IB [184] Multiple cancers and NF-B related Overexpression Inhibitor to prevent SCFTRCP activ- (RING-finger) disorders ity [33]

SCFFbw7 Notch [185], c-Myc Colon, stomach cancer, T-cell acute Mutations in substrate recognition Allosteric inhibitor of its yeast (RING-finger) [186], c-Jun [187], lymphocytic leukemia [190] domain [190] ortholog, Cdc4 impedes recogni- cyclin E [188] and tion of the substrate [34] mTOR [189]

SCFSkp2 p27 [191] Multiple cancers including breast Gene amplification and overex- Inhibitor to prevent SCFSkp2 activ- (RING-finger) and colon [192], gastric, lung, pros- pression in cancer [193] ity and stabilize p27 [32] tate, ovarian cancer [191]

Smurf2 Smad1 [53], Smad2 Esophageal squamous cell cancer Overexpression [55] - (HECT) [54], TGF receptor [55] 1 [194]

VHL HIF-1 [42] Renal cell carcinoma, pancreatic Germline deletion, mutations [43] Bioengineered VHL complex to (RING-finger) cancer, VHL syndrome [195] increase HIF-1 degradation [44]

Wwp1 p53 [57], p63 [196], Breast and prostate cancer [56] Overexpression [56] - (HECT) KLF5 [197], ERBB4 [198]

Xiap MEKK2, AIF, Breast cancer and acute myeloid Overexpression [202] Inhibitor compound AEG35156 to (RING-finger) TGF-activated leukaemia [199, 201] inhibit XIAP [203] kinase [199, 200]

E3 ligases together with their substrates, associated disease, alterations and available inhibitors/treatment are listed.

Wwp1 is a HECT E3 that is aberrantly regulated in cancer. lent thioester intermediates with ubiquitin before transferring ubiq- Overexpression of Wwp1 is observed in breast and prostate cancer uitin to the substrate. Since this step requires a large conformational samples [56]. Wwp1 has been shown to regulate p53 ubiquitination change in the HECT domain, it may be possible to block this transi- tanslocating p53 to the cytoplasm and diminishing its transcrip- tion by small-molecule inhibitors [1]. tional activity [57]. Inactivation of Wwp1, therefore, could activate p53, which may provide a means for treating cancer. Targeting Deubiquitinating Enzymes The HECT E3 ligases may be more amenable to small- Deubiquitinating enzymes regulate various ubiquitin-mediated molecule intervention than other classes of E3s as they form cova- biological processes such as proteosome-dependent degradation, Emerging Role of the Ubiquitin-proteasome System Current Pharmaceutical Design, 2013, Vol. 19, No. 00 5 protein localization and endocytosis by reversing the action of the Another member of USP family of DUBs is USP8 (UBPY), ubiquitin conjugation mechanism [58]. Deubiquitinases belong to which functions in receptor endocytosis and trafficking processes five different gene families; four are cysteine peptidases (the ubiq- and interact with several proteins including the epidermal growth uitin-specific proteases (USPs), the ubiquitin C-terminal hydrolases factor receptor (EGFR) [66]. The depletion of USP8 leads to accu- (UCHs), ovarian tumor domain proteins (OTU), and Machado- mulation of ubiquitinated EGFR in endosomes [67]. Although an Joseph disease protein (MJD), and the fifth family consists of met- inhibitor of USP8; HBX 90397 is identified, the resulting effects alloproteinases from the JAMM/MPN domain family [58]. are questionable since knockout of USP8 was observed to cause a The dysregulation of DUBs is highly observed in human dis- liver failure in adult mice [67, 68]. eases, especially cancer. In recent years, there is a growing interest USP2a deubiquitinates and stabilizes FAS, a metabolic onco- in developing inhibitors against DUBs although there has been a gene overexpressed in many cancers [69]. The depletion of USP2a limited success in turning these into drugs [59]. The DUBs have was observed to destabilize FAS and leads to apoptosis [69]. USP2a been extensively reviewed in [59]. Here, we classify the DUBs deubiquitinates Mdm2 as well, leading to stabilization of Mdm2 according to their domain type, overview advances in targeting levels and induction of p53 degradation [70]. USP2a deubiquiti- these DUBs and provide a summary in Table 2. nates also Mdmx, a Mdm2-like p53 repressor [71]. Interestingly, USP Family of Deubiquitinating Enzymes USP2a has been observed to have dual functions in disease. While overexpression of USP protects prostate cancer cells from apoptosis USP7 (also known as Herpes virus associated USP (HAUSP)) by reducing p53 stability, on the contrary, another study has illus- belongs to USP family of DUBs that has a critical role in cell pro- trated that it is downregulated in breast carcinomas suggesting that liferation and differentiation through regulating the activity of USP2a may have an antitumor property [72]. phosphatase and tensin homolog and localization of FOXO [60]. USP7 participates in p53 regulation and catalyzes deubiquitination USP28 is another cancer-associated DUB; it is overexpressed in of both p53 and Mdm2 [61, 62]. Therefore, USP7 can act as a tu- breast and colon carcinomas, and is essential regulating the stability mor suppressor or an oncogene depending on whether it predomi- of c-Myc, a central modulator of cell growth, proliferation and nantly deubiquitinates p53 or Mdm2, respectively [63]. USP7 also apoptosis [73]. Since c-Myc is dysregulated in many human cancers deubiquitinates a tumor suppressor PTEN leading to cytoplasmic and cancer cells require c-Myc activity for growth [74], inhibiting accumulation of this protein [60]. USP7 has been observed to be USP28 may be useful in targeting such malignancies by accelerat- overexpressed in prostate cancer and associated with tumor aggres- ing degradation of c-Myc [59]. siveness [60], which makes this DUB is an important target for USP14 is a DUB that blocks the degradation of ubiquitinated pharmaceutical intervention [59]. A high-throughput screen identi- substrates. Loss of USP14 is related to neurodegenerative diseases fied a small molecule USP7 inhibitor; HBX 41108 that can activate resulting in an ataxic neurological phenotype in mice [75]. A small- p53 resulting in p53-dependent apoptosis [64]. A fluorescence- molecule inhibitor of USP14, called IU1, has been identified [76]. based multiple assay has recently characterized another inhibitor IU1 did not inhibit eight other tested deubiquitinating enzymes (P022077) for USP7 [65]. making it a relatively specific inhibitor of USP14 [76]. Treating the

Table 2. DUBs and Disease

DUB Substrate Associated disease Alterations in disease Inhibitor

A20 RIP, RIP2, TRAF6, NEMO and Several lymphomas including Hodgkin’s Promoter methylation, mutations - MALT1 [93-98] lymphoma and B cell lymphoma [204] [204]

CYLD Bcl-3, TRAF2, TRAF6, Kidney, colon, B-cell malignancies, multi- Mutations, decreased expression - TRAF7, RIP1, TAK1 [86-89] ple myeloma, uterine cervix carcinoma and in cancer [84] melanoma [84]

USP2a MDM2, FAS and MDMX [69- Prostate cancer [205] Overexpression [205] - 71]

USP28 c-Myc [73], 53BP1 [206] Breast and colon cancers [73] Overexpression, somatic muta- - tions in cancer [73]

USP7 p53, MDM2, PTEN, FOXO4 Prostate, colon, non-small cell lung cancer Overexpression [60], USP7 inhibitor, [60-62, 207, 208] [60, 82] Downregulation [82] HBX41108 activates p53 [64] and P022077 [65]

USP8 Several receptor tyrosine Multiple abnormalities in RTK-dependent The depletion of USP8 leads to USP8 inhibitor, HBX kinases including MET, EGFR pathways [209] accumulation of ubiquitinated 90397[67] [66] EGFR in endosomes [67]

USP9x Several proteins including - Colon, breast, small cell lung carcinoma, Overexpression [79, 210] - catenin, Smad4 and MCL1 [77- lymphoma [79, 210] 79]

USP14 Tau, ataxin-3 [76] Neurodegenerative diseases, ataxia [75] Loss of USP14 [75] USP14 inhibitor, IU1 [76]

DUBs together with their substrates, associated disease, alterations and available inhibitors are listed. 6 Current Pharmaceutical Design, 2013, Vol. 19, No. 00 Kar et al. cells with IU1 increases the degradation of several substrates in- Targeting Proteasome cluding Tau and ataxin-3, both of which have been implicated in The proteasome is a large, 26S, multicatalytic protease in which neurodegenerative diseases [76]. Such inhibitors targeting USP14 proteolysis of short-lived or regulatory proteins from the cytoplasm, could, therefore, potentially be used to reduce or eliminate mis- the nucleus or the endoplasmic reticulum is carried out predomi- folded and aggregated proteins that accumulate in neurodegenera- nantly [101]. The 26S proteasome (proteasome holoenzyme) con- tive diseases and improve the prognosis in such disorders [25]. sists of two major subcomplexes: the 20S core particle (CP) that USP9x is another DUB from USP family that deubiquitinates contains the protease subunits and the 19S regulatory particle (RP) several proteins related to cancer disease such as -catenin, Smad4 that regulates the function of 20S CP [101]. The 19S regulatory and MCL1 [77-79]. -catenin is known to promote colon cancer particle contains two mutisubunit structures, a lid and a base. It has cell growth and it is stabilized by USP9x [77]. By an siRNA-based multiple roles in regulating proteasomal activity: selecting sub- screen, USP9x has been found to be important for TGF signaling strates, preparing them for degradation and translocating them into [78]. An essential component of TGF signaling pathway, Smad4, the CP. The proteolytic activity of the proteasome occurs in the 20S becomes inactive by undergoing monoubiquitination. USP9x was CP that contains two inner -rings that have the proteolytic active observed to deubiquitinate Smad4, hence, it positively regulates sites [101]. Active sites of 20S proteasome are the caspase-like TGF signaling pathway, which is related to tumor initiation and (1), trypsin-like (2) and chymotrypsin-like (5) that use an amino progression [78]. USP9x also deubiquitinates MCL1, that is an anti- terminal threonine as the catalytic aminoacid residue. There are one apoptotic BCL-2 homolog leading to blocking of MCL1 degrada- or two regulatory particles attached to the surface of the outer rings tion and stabilizing MCL1 levels, thus leading to cell survival [79]. of the 20S CP to form 26S proteaome. After degradation of the However, having a dual function, it can also promote apoptosis by substrate, short peptides derived from the substrate and reusable stabilizing apoptosis signal-regulating kinase 1 [80]. ubiquitin are released. Other DUBs belonging to USP family that are linked to disease Alterations in the proteasome are associated with many human are: USP1 related to Fanconi anaemia DNA repair pathway [81]; diseases including neurodegenerative disorders, cardiac dysfunc- USP6 to aneurismal bone cysts [58]; and USP7 to non-small cell tion, cataract formation, cachexia and rheumatoid diseases [102]. lung adenocarcinoma [82]. Several antiapoptotic and proliferative signaling pathways such as UCH Family of Deubiquitinating Enzymes pro-oncogenic NF- pathway require proteasomal activity [4]. NF-  pathway is activated in many tumor types [103]. The inhibitors CYLD is a DUB belonging to UCH family and is mutated in of NF-s (IBs) keep NF- factors in an inactive state. Upon Familial Cylindromatosis disease, in which numerous skin tumors phoshorylation and polyubiquitination of IBs and subsequent deg- develop [83]. Mutation or decreased expression levels of CYLD radation by the proteaosome, NF- factors become activated have been observed in kidney cancer, colon cancer, B-cell malig- [103]. The first clinically validated drug to target UPS is borte- nancies, multiple myeloma, uterine cervix carcinoma and mela- zomib, which reversibly inhibits the active sites in the 20S protea- noma [84]. CYLD deubiquitinates the Bcl-3 protein, which is a some. Bortezomib-mediated proteasome inhibition leads to down- transcriptional co-activator of NF-B, blocks its nuclear transloca- regulation of NF- signaling through blocking IB degradation tion and hence inhibits NF-B activation [85]. CYLD also deubiq- [104] and to endoplasmic reticulum stress, which in turn promotes uitinates TRAF2, TRAF6, TRAF7 (TNF Receptor Associated Fac- cell death [105]. The effects of NF- inhibition are the upregula- tors), RIP1, TAK1 leading to inactivation of NF-B signaling [86- tion of the cyclin-dependent kinase inhibitors p21 and p27 and 89]. CYLD is also associated with Wnt signaling pathway that is downregulation of pro-inflammatory response genes, consequently essential for embryonic development and is frequently activated in leading to increased apoptosis in tumor cells [106]. Bortezomib is cancer [90]. It negatively regulates Wnt signaling through deubiq- currently used as an anticancer drug for the treatment of multiple uitinating Dishevelled protein and activating -catenin [91]. Having myeloma, relapsed mantle cell lymphoma. It is also in many other a tumor suppressor role, increasing the activity of CYLD that is lost clinical trials including Phase III clinical trials for follicular non- during tumor progression may be an efficient treatment in cancer. Hodgkin’s lymphoma and Phase II trials for diffuse large B cell OTU Family of Deubiquitinating Enzymes lymphoma [25]. A20 contains an OTU domain at the amino-terminus that pos- Owing to success of Bortezomib, several new generation of sesses DUB activity. Additionally, it also functions as an E3 ligase proteasome inhibitors such as Carfilzomib (PR-171), MLN9708, through its carboxyl-terminal consisting of seven zinc fingers. A20 CEP18770, NPI-0052, ONX0912, PR957 and IPSI-001 are cur- can act as tumor suppressor by downregulating NF-B signalling rently in development [107-114]. These proteasome inhibitors tar- [92]. It removes K63-linked ubiquitin chains from RIP, RIP2, get different active sites of the 20S proteasome and differ in binding TRAF6, NEMO and MALT1 [93-98]. With its E3 ligase activity, it kinetics. For example, while Bortezomib inhibits the caspase-like conjugates K48-linked polyubiquitin chains to RIP1 and TRAF2 (1) and chymotrypsin-like (5) activities of the proteasome in a targeting them for degradation [99]. Inactivation of A20 protein is reversible manner, NPI-0052 irreversibly blocks the trypsin-like associated with several chronic inflammatory and autoimmune (2) and chymotrypsin-like (5) sites. NPI-0052 is in Phase I stage diseases. and when combined with Bortezomib in low doses has been ob- JAMM Family of Deubiquitinating Enzymes served to show synergistic effects on cancer cells in vitro [115]. Another proteasome inhibitor, Carfilzomib, which has entered The JAMM domain proteins are zinc metalloisopeptidases [58]. Phase III trials targets mainly the 5 site of the proteasome in an Members of this class include AMSH, BRCC36, CSN5 and POH1. irreversible manner and has been developed for the treatment of The activity of AMSH regulates trafficking of G protein-coupled relapsed multiple myeloma [107]. Other proteasome inhibitors are receptors and receptor tyrosine kinases [100]. BRCC36 has a role in detailed in Table 3. the DNA damage response. CSN5 mediates NEDD8 attachment to , regulating the cullin-RING ligase activity and stability . Despite the success of Bortezomib and efforts for developing Finally, POH1 functions as a part of 19S cap of the 26S proteasome other inhibitors, targeting proteasome is somewhat problematic to trim polyubiquitin chains from the substrates. Though no drugs since this final common step of the degradative pathways are used are yet available, these DUBs are potential drug targets. by hundreds of different processes in the cell. Therefore, inhibition of the proteasome is likely to interrupt several diverse pathways.

Emerging Role of the Ubiquitin-proteasome System Current Pharmaceutical Design, 2013, Vol. 19, No. 00 7

Table 3. Proteasome Inhibitors, Development Stage, Binding Kinetics and Targeted Disease

Inhibitor Development Stage Disease Binding Kinetics

Bortezomib Approved Multiple myeloma and mantle cell lymphoma Slowly reversible

Carfilzomib Phase III Multiple myeloma and other cancers Irreversible

CEP18770 Phase I Multiple myeloma and other cancers Slowly reversible

IPSI-001 Preclinical Multiple myeloma and other hematologic malignancies Not reported

MLN9708 Phase I Multiple myeloma and other cancers Reversible

NPI-0052 Phase I Multiple myeloma and leukemia Irreversible

ONX0912 Phase I Multiple myeloma and other cancers Irreversible

PR957 Preclinical Autoimmune disorders Irreversible

Developing proteasome inhibitors with distinct substrate selectivity interactions and to designation of novel compounds. A potent in- and lower toxicity would reduce such inconveniences in treating hibitor mimicking such key residue interactions can achieve favor- associated diseases. able interaction energy; thus, can compete with the original partner for binding [121]. One such competitive inhibitor in treating im- 3. IMPORTANCE OF STRUCTURES IN TARGETING UPS mune disorders is SP-4206 that binds to hot spot residues of cyto- INTERACTIONS kine IL-2 with high affinity and block the formation of the natural Protein-protein interactions have a central role in regulating complex of IL-2 and its receptor IL-2R [125]. The most widely many biological processes, cellular and signaling pathways. Altera- known example of competitive inhibitors of the UPS is nutlins that tions in protein-protein interactions may lead to several diseases prevent the Mdm2-p53 binding, thus enhance the tumor suppressor such as cancer and neurological disorders. Therefore, protein- activity of p53. Alanine-scanning of the Mdm2-p53 binding region protein interactions are recurrently considered as drug targets in illustrated that there are three hot spot residues, Phe19, Trp23 and disease states [116-122]. Drugs targeting protein-protein interac- Leu26, on p53 critical for binding to Mdm2 [126]. Nutlin family of tions, ultimately head protein interfaces where two proteins come compounds such as Nutlin-2 and Nutlin-3 mimics p53 and binds to into contact. Understanding the details and principles of protein Mdm2 even with a greater affinity than p53 [127]. Mdm2-p53 and interfaces is, therefore, immensely essential to develop efficient Mdm2-Nutlin-2 binding are visualized in (Fig. 2). strategies in drug design. At protein interfaces there are some criti- In the UPS, there are five major steps: activation, conjugation, cal residues that account for the majority of the binding energy ubiquitin-conjugate recognition, ubiquitin removal, and substrate called hot spots [123]. A hot spot is defined as a residue that, when degradation by the proteasome [1]. Since the greatest amount of mutated to alanine, leads to a dramatic decrease in binding free specificity is the conjugation step that is mainly mediated by E3s, energy (Gbinding > 2 kcal/mol) [123, 124]. Since these residues the interactions of E3s represent important drug targets. The E3 and are more critical than others to the stability of the complex, target- E2 proteins function in a concerted manner; however, the principles ing the hot spots may lead to improved inhibition of protein-protein determining E2-E3 selectivity are still not entirely understood. In

Fig. (2). Nutlin-2 targeting Mdm2-p53 interaction interface. (A) Mdm2 is an E3 ligase mediating ubiquitination and subsequent degradation of p53. Alanine-scanning of the Mdm2-p53 binding region illustrated that there are three hot spot residues, Phe19, Trp23 and Leu26, on p53 critical for binding to Mdm2 [126]. (B) Nutlin-2 prevents the interaction between Mdm2 and p53 by mimicking the conformation of Phe19, Trp23 and Leu26 of p53 to block Mdm2 p53-binding region [127]. 8 Current Pharmaceutical Design, 2013, Vol. 19, No. 00 Kar et al. addition, in many cases it is not known which E3s interact with complex (pdb code: 1fbv:AC), HECT(NEDD4L)/ UBE2D2 com- which E2s and how they interact with each other. Identifying all E2 plex (pdb codes: 3jw0:AC and 3jvz:AC), TNF receptor-associated partners of a specific E3 is essential for inferring the principles factor 6 (TRAF6)/UBE2N complex (pdb code: 3hct:AB), STIP1 guiding E2 selection by an E3. In one experimental approach, all E2 homology and U-Box containing protein 1 (CHIP)/UBE2D1 com- partners of Brca1 ubiquitin ligase were identified by two-hybrid plex (pdb code: 2oxq:AC), Ubiquitin-protein ligase Ube3a (E6AP)/ experiments [128, 129]. Brca1 was found to interact with multiple UBE2L3 complex (pdb code: 1c4z:AD) and U-box domain of hu- different E2s and to possess E2-specific ubiquitin-transfer proper- man E4B ubiquitin ligase/UBE2D3 complex (pdb code: 3l1z:AB). ties. Gardner et al. identified the interactions of HRD1 (HMG-CoA Comparison of UBE2L3 binding to RING type c-Cbl and HECT reductase degradation) ubiquitin ligase by in vivo cross-linking type E6-AP revealed that there are common structural elements of methods [130]. Ballar et al. illustrated how HRD1 can further act as this E2 recognized by both E3s: most extensive contacts are pro- an E3 for another E3, gp78 involved in ER-associated degradation vided by F63, P97 and A98 residues of UBE2L3 binding to both [131]. Two large-scale studies have addressed this E2-E3 identifica- RING type c-Cbl and HECT type E6-AP [134]. Comparison of all tion problem. In a global yeast-two hybrid screen, over 300 high known E2-E3 structures through identifying the structurally con- quality interactions are uncovered [132]. They generated mutants of served contacts implied that loop L1 second position residue con- UBE2N to mimic the interaction pattern of another E2, UBE2D2, tacts are only employed when interacting with HECT E3s [135]. which resulted in new E3 interactions and hence illustrated how the Combining such structural details with protein interactions net- E3 interactions of the E2 specific for K63 linkages UBE2N can be works of UPS, obviously, improves drug design concept as outlined altered towards the K48-specific UBE2D2 [132]. Markson et al. in the following section. combined yeast two-hybrid screens with homology modeling meth- ods yielding a map of human E2-E3 RING interactions [133]. Their Structural Interaction Network of UPS Proteins data includes 535 experimentally defined novel E2-RING E3 inter- Network-based approaches have gained importance in drug actions and more than 1300 E2-RING E3 pairs with favorable pre- discovery with the realization that diseases are complex as they are dicted free energy values. They extended this network by including often consequences of multiple molecular abnormalities rather than known human E2-RING E3 pairs, hence, assembled a network of being the result of a single defect or phenomena [136]. To go a step 2644 proteins and 5087 interactions [133]. further in the mechanistic understanding one has to combine inter- Including structural details of the interactions can assist in un- action networks information with molecular details of protein inter- derstanding principles beyond E3 selectivity. The first structural faces in the effort of identifying all target proteins that are influ- clues of E2-E3 interaction specificity were discovered from the enced by a drug, either positively or negatively. crystal structure of the E3 ligase c-Cbl RING and the E2, UBE2L3 The number of distinct binding motifs is limited in nature [137] [134] (Fig. 3A). The -helix and zinc-chelating loops of the RING implying that structurally different proteins can share similar inter- domain are in contact with loop L1 and L2 of UBE2L3. In particu- face architectures [138]. Following this notion, an algorithm called lar, F63 residue in loop L1, P97 and A98 in loop L2 of UBE2L3 Prism has been developed that predicts potential protein associa- mediate its binding to c-Cbl. The c-Cbl linker region interacts with tions based on known protein interface motifs [139, 140]. Exploit- -helix 1 of UBE2L3 [134]. Currently, there are 9 E2-E3 complex ing this algorithm and the available structural proteome, recently, structures deposited in PDB; UBE2D2/CNOT4 complex (pdb code: we have constructed a human E2-E3 structural interaction network 1ur6:AB), Baculoviral IAP repeat-containing protein 3/ UBE2D2 [135]. Considering the all human genome, protein structures for 24 (pdb code: 3eb6:AB), Signal transduction protein C-/UBE2L3

Fig. (3). Interactions of c-Cbl E3 ubiquitin ligase with two E2 proteins, UBE2L3 and UBE2D1. (A) c-Cbl interacts with UBE2L3 (pdb code: 1fbv:AC). - helix 1, loop L1 and loop L2 regions of UBE2L3 are displayed in yellow, green and orange color, respectively. Two loop regions, L1 and L2 of UBE2L3 are in contact with the -helix and zinc-chelating loops of the RING domain. The central F63 residue in loop L1, P97 and A98 in loop L2 of UBE2L3 mediate its binding to c-Cbl. The c-Cbl linker region interacts with -helix 1 of UBE2L3 [134]. (B) Modeled c-Cbl-UBE2D1 complex. Although the interaction was re- ported earlier [157], binding details and the structure of the complex was not available. c-Cbl residues labeled in red color was observed to be critical: C384, C404, R420 were mutated in patients with acute myelogenous leukemia [158]. Additionally, Ile383 and Trp408 of c-Cbl were shown to have a critical role in E2 binding [134]. These residues correspond to computational hotspots predicted by Hotpoint server indicating the usefulness of structural modeling of E2-E3 associations. Emerging Role of the Ubiquitin-proteasome System Current Pharmaceutical Design, 2013, Vol. 19, No. 00 9

E2s are available (out of ~40) while the structural coverage is rela- able to demonstrate, by analyzing the dynamic behavior of ubiq- tively low for E3 proteins; 32 of them (out of ~500) have resolved uitin from the the picosecond to microsecond time scale, that for structures in PDB. Among these, we predict 107 interactions be- each bound ubiquitin structure there is a member of the unbound tween 22 E2s and 16 E3s. In our network, the number of all possi- sampled ensemble that is structurally similar to it [151]. The work ble E2-E3 pairs is 352 (i.e. 22x16) and 51 of these were already strongly supported an underlying conformational selection mecha- reported in earlier studies. We recover 76% (39/51) of the known nism for these multibinding promiscuous proteins. Lately, by Mo- pairs through 107 predicted interactions verifying 36% (39/107) of lecular Dynamics investigations of the same data by Wlodarski and the network. On the basis of known E2-E3 interaction data, the Zagrovic suggested that during the binding event the conformation accuracy of the predictions is 76% indicating that the predicted E2- structurally most similar to the bound structure is the dominant one E3 interactions are in agreement with validated functional E2-E3 via the conformational selection model. After this selection, re- pairs. Our method depends on the coverage of known interface optimization of the binding interface occurs via a ‘residual induced architectures and there are only 9 E2-E3 known complex structures fit’ mechanism, therefore calling in cause both mechanisms of bind- in the PDB (as aforementioned above) which are used as a basis for ing for these enzymes [152]. Long and Brüschweiler performed modeling. Because the size and coverage of the PDB increases microsecond time scale all-atom MD simulations of ubiquitin and exponentially, we expect the prediction efficiency of our method to its binding partners using the latest generation of molecular me- increase. chanics force field [153]. In this study, the effect of ligand mole- We observed that E3 proteins could interact with multiple E2s cules on the flexibility and plasticity of ubiquitin binding properties and likewise E2 proteins could interact with multiple E3s, as ex- is analyzed. The study focuses on the ubiquitin interacting motif pected. To reveal which residues are conserved among E2s interact- (UIM), a short alpha-helical conserved motif for ubiquitin recogni- ing with a common E3 and which residues differ, hence, to provide tion. The energy landscape and dynamics of ubiquitin in response to clues to specificity, we analyzed the E2-E3 interfaces and predicted the approach of the UIM ligand is analyzed in details. By the use of hotspot residues at interfaces using the Hotpoint server [141]. The Boltzmann sampling of molecular dynamics snapshots, the authors Hotpoint server determines computational hot spot residues based show a statistical reweighing of the populations of ubiquitin con- on conservation, solvent accessibility and statistical pairwise resi- formers in the presence of its ligand molecule at intermediate dis- due potentials of the interface residues. The predicted hot spots are tance range (1–2 nm) and are able to examine the population redis- observed to match with the experimental hot spots with an accuracy tribution mechanisms. of 70% [141]. Considering E2-E3 interface regions, some residues Nussinov and collaborators focused on cullin-RING E3 ubiq- are structurally conserved among E2 proteins and appear to be es- uitin ligases and their flexible role in binding to many partners sential for all E2-E3 interactions, whereas others, particularly in [145, 146]. They performed structural analysis and MD simulations loop L1, appear to play important roles in E3 selectivity. The E2 starting from Cul1, Cul4A, and Cul5 crystal structures. The MD proteins which have been characterized so far are known to recog- work clearly shows that at least three cullins do not act as merely nize E3s through the L1 and L2 loops and the N-terminal -helix 1 rigid scaffolds but are flexible with conserved hinges in the N- on the E2 surface [142]. In particular, loop L1 has been shown to be terminal domain. In another study, dynamical properties of the E2 critical in E3 binding. In our network, E2 interface residues also enzymes of family 3R are characterized, showing a structural role mostly correspond to the -helix 1, loop L1 and L2 regions. The of Loop 7 and Loop 8 in stabilizing closed conformations [154]. residue in the sixth position in loop L1 is widely utilized as an inter- Taken together, all these evidences highlight the importance of face hot spot and appears indispensible for E2 interactions. Other flexibility in the recognition and functional activity of these E3 loop L1 residues also confer specificity on the E2-E3 interactions: ligase enzymes and pave the way for more studies in which Mo- HECT E3s are in contact with the residue in the second position in lecular Dynamics techniques are used for the conformational char- loop L1 of E2s; but this is not the case for the RING finger type E3s acterization and identification of flexible regions to be targeted in [135]. drug design. Knowledge of such structural details and all interacting partners of E3s at a proteome-level are crucial in drug design. This effort 4. CASE STUDIES will aid in explaining: i) which E3s interact with which E2s and Here we present two case studies of UPS proteins relevant to which substrates; ii) the molecular determinants of these interac- drug design. The first example illustrates how structural modeling tions; iii) what the corresponding functions and outcomes are. In the of E2-E3 associations may be beneficial to drug discovery. In the drug design field, to date, researchers have focused primarily on second example, an interesting approach of designing a small- blocking E3-substrate interfaces (as aforementioned above). How- molecule antagonist to target an E3 protein in cancer is explained. ever, targeting and disrupting E2-E3 interactions may be relatively easy since these interactions are usually relatively weak [142]. Such Modeling E2-E3 Complexes: UBE2D1-c-Cbl Ligase Example inhibitors can block the interaction by binding to the E2, the E3 or c-Cbl is an E3 ubiquitin ligase that attenuates signaling through the E2-E3 interface. Structural examples showing the binding of the poly- or monoubiquitination of several activated receptor tyrosine inhibitors are provided as case studies in the next section. kinases (such as Flt-3, c-kit, and M-CSF) and other tyrosine kinases of the Scr family [155]. The RING domain has a central role in c- Flexibility and Molecular Dynamics of UPS Proteins Cbl function because its deletion or disruption abolishes the func- It is more and more recognized that increment in protein flexi- tion of c-Cbl [156]. From its RING domain, c-Cbl binds to an E2 bility is paralleled by an increment in binding promiscuity of the and mediates ubiquitin transfer from E2 to the target substrates. flexible region. This is particularly true for those region that un- Structure of c-Cbl interacting with one of its E2 partners, UBE2L3, dergo post-translational modifications such as phosphorylation is deposited in (PDB) (pdb code: 1fbv) [134]. [143, 144] or ubiquitination [145, 146] and for their partners [147, Two loop regions, L1 and L2 of UBE2L3 are in contact with the - 148]. These flexible changes are mostly studied by NMR and Mo- helix and zinc-chelating loops of the RING domain. The central lecular Dynamics (MD) [149]. Studies using NMR residual dipolar F63 residue in loop L1, P97 and A98 in loop L2 of UBE2L3 medi- couplings has shown neatly that free ubiquitin samples conforma- ate its binding to c-Cbl. The c-Cbl linker region interacts with - tions globally similar to those in the bound state [150, 151]. Lange helix 1 of UBE2L3 [134]. c-Cbl ligase-UBE2L3 complex showing and colleagues compared a number of X-ray structures of ubiquitin, the critical binding regions (-helix 1, loop L1 and L2) of UBE2L3 bound to different partners, with NMR structures of the ubiquitin is visualized in (Fig. 3A). protein free in solution. In this pioneering study the authors were 10 Current Pharmaceutical Design, 2013, Vol. 19, No. 00 Kar et al.

c-Cbl is also reported to interact with another E2, UBE2D1, and molecules that mimic the binding of IAP antagonist, second mito- ubiquitinate epidermal growth factor receptor [157], although the chondria-derived activator of caspases/direct IAP-binding protein structure of the complex is not available as yet. Using Prism algo- with low pI (SMAC/DIABLO) [166, 167], to baculoviral IAP re- rithm, we modeled c-Cbl-UBE2D1 complex based on a known peat (BIR) domains [168]. The IAP BIR domains are necessary for interface of another E2-E3 complex; UBE2D2-CNOT4 (pdb code: the antiapoptotic function of the IAP proteins [165]. One such an- 1ur6:AB) [135]. We observe that UBE2D1 (pdb code: 2c4p:A) tagonist is GDC-0152, which has entered Phase I clinical trials, can binds to c-Cbl (pdb code: 1fbv:A) mainly through its -helix 1, loop bind to BIR3 domain of cIAP1 [165]. Binding of GDC-0152 to L1 and L2 regions. Earlier work suggested that mutations in the c- cIAP1 promotes degradation of cIAP1, induces activation of Cbl RING domain are associated with acute myelogenous leukemia caspase-3/7, and leads to decreased viability of breast cancer cells and myeloproliferative neoplasms and the impairing the degrada- without affecting normal mammary epithelial cells [165]. (Fig. 4A) tion of tyrosine kinases is an important mechanism in cancer [158]. illustrates N-terminal peptide from SMAC/DIABLO binding to In particular, on the RING domain of c-Cbl, the substitution of BIR3 domain of cIAP1 (pdb code: 3d9u). GDC-0152 antagonist cysteine or arginine residues at position 384 (C384Y), 404 (C404S), (pdb code: 3uw4), mimicking SMAC/DIABLO binding to cIAP1, and 420 (R420Q) are observed in patients with acute myelogenous is visualized in (Fig. 4B). With the complete knowledge of struc- leukemia [158]. Additionally, Ile383 and Trp408 of c-Cbl were tural and binding details of UPS proteins, it will become undoubt- shown to have a critical role in E2 binding [134] and the mutation edly easier to design such small-molecule compounds. of Trp408 to alanine reduces c-Cbl’s affinity for the E2 and elimi- nates its ubiquitin-ligase activity in vitro [159]. Interface analysis of 5. CONCLUSIONS our c-Cbl-UBE2D1 model indicates that these critical residues; Since the approval of the proteasome inhibitor Bortezomib in I383, C384, C404 and W408, correspond to computational hotspots 2003, UPS components have emerged as crucial drug targets in predicted by Hotpoint server [141], implying their importance in treatment of cancer, neurodegenenerative diseases, immunological binding and function of c-Cbl. Such structural modeling of E2-E3 disorders and microbial infection. Small-molecule inhibitors of associations could help in understanding E2-E3 selectivity and ubiquitin-conjugating and –deconjugating enzymes are continu- discovering drug candidates targeting E3s. (Fig. 3B) illustrates the ously being developed and evaluated for their potential use as modeled c-Cbl-UBE2D1 complex and the critical residues in bind- therapeutics. The major difficulty is, however, to design target- ing. specific inhibitors since UPS has diverse roles affecting hundreds of different processes in the cell. Although attracted little attention so A Potent Small Molecule Antagonist of IAP Proteins: GDC- far, developing selective compounds targeting E2-E3 interactions 0152 would be a crucial potential therapeutic intervention: inhibitors can Inhibitors of apoptosis (IAP) family of proteins function in the be designed to bind to E3 ligases proximal to E2 binding region, regulation of signal transduction pathways associated with malig- thus blocking E2-E3 interface or disrupt E2-E3 interactions allos- nancy [160, 161] and are frequently overexpressed in cancer cells terically by inducing long-range conformational changes. At this [162, 163]. Cellular IAP (cIAP) proteins have been observed to point, knowledge of structural details of the interactions in UPS, regulate TNF-mediated NF-B activation via its RING domain especially at a proteome-scale, appears to be essential. and ubiquitinate receptor interacting protein RIP-1 and NF-B in- In addition to targeting UPS with a single small-molecule com- ducing kinase, NIK [164]. pound, combinatorial drug therapy in UPS can be also efficient Viability of cancer cells depend on the aberrations in the apop- since drug combinations currently provide a route to overcome the tosis signaling pathways. Therefore, drugs that can restore apopto- complexity of human cancers [169]. Novel approaches including sis in tumor cells might be efficient in treating cancer [165]. In large-scale genomics, systems and structural biology will set the cancer, one approach to target the IAP proteins is to design of small basis for future potential therapeutic interventions in UPS-

Fig. (4). A potent small molecule antagonist targeting an E3, cIAP1 in cancer treatment. (A) cIAP1 interacting with N-terminal peptide from SMAC/DIABLO (pdb code: 3d9u). (B) GDC-0152 antagonist mimicks SMAC/DIABLO binding to cIAP1 (pdb code: 3uw4). Binding of GDC-0152 to cIAP1 promotes degradation of cIAP1, induces activation of caspase-3/7, and leads to decreased viability of breast cancer cells without affecting normal mammary epithelial cells [165]. Emerging Role of the Ubiquitin-proteasome System Current Pharmaceutical Design, 2013, Vol. 19, No. 00 11 associated diseases and guide the identification of best possible [25] Cohen P, Tcherpakov M. Will the ubiquitin system furnish as many drug combinations targeting UPS. drug targets as protein kinases? Cell 2010; 143: 686-93. [26] Quesnel B, Preudhomme C, Fournier J, Fenaux P, Peyrat JP. CONFLICT OF INTEREST MDM2 gene amplification in human breast cancer. Eur J Cancer 1994; 30A: 982-4. The authors confirm that this article content has no conflicts of [27] Momand J, Jung D, Wilczynski S, Niland J. The MDM2 gene interest. amplification database. Nucleic Acids Res 1998; 26: 3453-9. [28] Vassilev LT, Vu BT, Graves B, et al. In vivo activation of the p53 ACKNOWLEDGEMENTS pathway by small-molecule antagonists of MDM2. Science 2004; Declared none. 303: 844-8. [29] Issaeva N, Bozko P, Enge M, et al. Small molecule RITA binds to REFERENCES p53, blocks p53-HDM-2 interaction and activates p53 function in tumors. Nat Med 2004; 10: 1321-8. [1] Nalepa G, Rolfe M, Harper JW. Drug discovery in the ubiquitin- [30] Bedford L, Lowe J, Dick LR, Mayer RJ, Brownell JE. Ubiquitin- proteasome system. Nat Rev Drug Discov 2006; 5: 596-613. like protein conjugation and the ubiquitin-proteasome system as [2] Pickart CM. Mechanisms underlying ubiquitination. Annu Rev drug targets. Nat Rev Drug Discov 2011; 10: 29-46. Biochem 2001; 70: 503-33. [31] Merlet J, Burger J, Gomes JE, Pintard L. Regulation of cullin- [3] Bennett EJ, Harper JW. DNA damage: ubiquitin marks the spot. RING E3 ubiquitin-ligases by neddylation and dimerization. Cell Nat Struct Mol Biol 2008; 15: 20-2. Mol Life Sci 2009; 66: 1924-38. [4] Hoeller D, Dikic I. Targeting the ubiquitin system in cancer [32] Chen Q, Xie W, Kuhn DJ, et al. Targeting the p27 E3 ligase therapy. Nature 2009; 458: 438-44. SCF(Skp2) results in p27- and Skp2-mediated cell-cycle arrest and [5] Herrmann J, Ciechanover A, Lerman LO, Lerman A. The activation of autophagy. Blood 2008; 111: 4690-9. ubiquitin-proteasome system in cardiovascular diseases-a [33] Nakajima H, Fujiwara H, Furuichi Y, Tanaka K, Shimbara N. A hypothesis extended. Cardiovasc Res 2004; 61: 11-21. novel small-molecule inhibitor of NF-kappaB signaling. Biochem [6] Petroski MD. The ubiquitin system, disease, and drug discovery. Biophys Res Commun 2008; 368: 1007-13. BMC Biochem 2008; 9 Suppl 1: S7. [34] Orlicky S, Tang X, Neduva V, et al. An allosteric inhibitor of [7] Lehman NL. The ubiquitin proteasome system in neuropathology. substrate recognition by the SCF(Cdc4) ubiquitin ligase. Nat Acta Neuropathol 2009; 118: 329-47. Biotechnol 2010; 28: 733-7. [8] Hershko A, Ciechanover A. The ubiquitin system. Annu Rev [35] Wu H, Tschopp J, Lin SC. Smac mimetics and TNFalpha: a Biochem 1998; 67: 425-79. dangerous liaison? Cell 2007; 131: 655-8. [9] Yang Y, Kitagaki J, Dai RM, et al. Inhibitors of ubiquitin- [36] Ford D, Easton DF, Stratton M, et al. Genetic heterogeneity and activating enzyme (E1), a new class of potential cancer penetrance analysis of the BRCA1 and BRCA2 genes in breast therapeutics. Cancer Res 2007; 67: 9472-81. cancer families. The Breast Cancer Linkage Consortium. Am J [10] Xu GW, Ali M, Wood TE, et al. The ubiquitin-activating enzyme Hum Genet 1998; 62: 676-89. E1 as a therapeutic target for the treatment of leukemia and [37] Scully R. Role of BRCA gene dysfunction in breast and ovarian multiple myeloma. Blood 2010; 115: 2251-9. cancer predisposition. Breast Cancer Res 2000; 2: 324-30. [11] Soucy TA, Smith PG, Milhollen MA, et al. An inhibitor of [38] Ohta T, Fukuda M. Ubiquitin and breast cancer. Oncogene 2004; NEDD8-activating enzyme as a new approach to treat cancer. 23: 2079-88. Nature 2009; 458: 732-6. [39] Dawson TM, Dawson VL. The role of parkin in familial and [12] Soucy TA, Smith PG, Rolfe M. Targeting NEDD8-activated cullin- sporadic Parkinson's disease. Mov Disord 2010; 25 Suppl 1: S32-9. RING ligases for the treatment of cancer. Clin Cancer Res 2009; [40] Bedford L, Hay D, Devoy A, et al. Depletion of 26S proteasomes 15: 3912-6. in mouse brain neurons causes neurodegeneration and Lewy-like [13] van Ree JH, Jeganathan KB, Malureanu L, van Deursen JM. inclusions resembling human pale bodies. J Neurosci 2008; 28: Overexpression of the E2 ubiquitin-conjugating enzyme UbcH10 8189-98. causes missegregation and tumor formation. J Cell [41] Lonser RR, Glenn GM, Walther M, et al. von Hippel-Lindau Biol 2010; 188: 83-100. disease. Lancet 2003; 361: 2059-67. [14] Hao J, Xu A, Xie X, et al. Elevated expression of UBE2T in lung [42] Maxwell PH, Wiesener MS, Chang GW, et al. The tumour cancer tumors and cell lines. Tumour Biol 2008; 29: 195-203. suppressor protein VHL targets hypoxia-inducible factors for [15] Maeda H, Miyajima N, Kano S, et al. Ubiquitin-conjugating oxygen-dependent proteolysis. Nature 1999; 399: 271-5. enzyme UBE2Q2 suppresses cell proliferation and is down- [43] Kim WY, Kaelin WG. Role of VHL gene mutation in human regulated in recurrent head and neck cancer. Mol Cancer Res 2009; cancer. J Clin Oncol 2004; 22: 4991-5004. 7: 1553-62. [44] Sufan RI, Moriyama EH, Mariampillai A, et al. Oxygen- [16] Zhu S, Sachdeva M, Wu F, Lu Z, Mo YY. Ubc9 promotes breast independent degradation of HIF-alpha via bioengineered VHL cell invasion and metastasis in a sumoylation-independent manner. tumour suppressor complex. EMBO Mol Med 2009; 1: 66-78. Oncogene 2010; 29: 1763-72. [45] Scheffner M, Huibregtse JM, Vierstra RD, Howley PM. The HPV- [17] Duan X, Trent JO, Ye H. Targeting the SUMO E2 conjugating 16 E6 and E6-AP complex functions as a ubiquitin-protein ligase in enzyme Ubc9 interaction for anti-cancer drug design. Anticancer the ubiquitination of p53. Cell 1993; 75: 495-505. Agents Med Chem 2009; 9: 51-4. [46] zur Hausen H. Papillomaviruses and cancer: from basic studies to [18] Wong BR, Parlati F, Qu K, et al. Drug discovery in the ubiquitin clinical application. Nat Rev Cancer 2002; 2: 342-50. regulatory pathway. Drug Discov Today 2003; 8: 746-54. [47] Matsuura T, Sutcliffe JS, Fang P, et al. De novo truncating [19] Dominguez C, Bonvin AM, Winkler GS, et al. Structural model of mutations in E6-AP ubiquitin-protein ligase gene (UBE3A) in the UbcH5B/CNOT4 complex revealed by combining NMR, Angelman syndrome. Nat Genet 1997; 15: 74-7. mutagenesis, and docking approaches. Structure 2004; 12: 633-44. [48] Chen D, Kon N, Li M, et al. ARF-BP1/Mule is a critical mediator [20] Weissman AM. Themes and variations on ubiquitylation. Nat Rev of the ARF tumor suppressor. Cell 2005; 121: 1071-83. Mol Cell Biol 2001; 2: 169-78. [49] Chen D, Brooks CL, Gu W. ARF-BP1 as a potential therapeutic [21] Huang DT, Miller DW, Mathew R, et al. A unique E1-E2 target. Br J Cancer 2006; 94: 1555-8. interaction required for optimal conjugation of the ubiquitin-like [50] Wang X, Trotman LC, Koppie T, et al. NEDD4-1 is a proto- protein NEDD8. Nat Struct Mol Biol 2004; 11: 927-35. oncogenic ubiquitin ligase for PTEN. Cell 2007; 128: 129-39. [22] Rotin D, Kumar S. Physiological functions of the HECT family of [51] Fouladkou F, Landry T, Kawabe H, et al. The ubiquitin ligase ubiquitin ligases. Nat Rev Mol Cell Biol 2009; 10: 398-409. Nedd4-1 is dispensable for the regulation of PTEN stability and [23] Borden KL. RING domains: master builders of molecular localization. Proc Natl Acad Sci U S A 2008; 105: 8585-90. scaffolds? J Mol Biol 2000; 295: 1103-12. [52] Baker SJ. PTEN enters the nuclear age. Cell 2007; 128: 25-8. [24] Ohi MD, Vander Kooi CW, Rosenberg JA, Chazin WJ, Gould KL. [53] Zhang Y, Chang C, Gehling DJ, Hemmati-Brivanlou A, Derynck Structural insights into the U-box, a domain associated with multi- R. Regulation of Smad degradation and activity by Smurf2, an E3 ubiquitination. Nat Struct Biol 2003; 10: 250-5. ubiquitin ligase. Proc Natl Acad Sci U S A 2001; 98: 974-9. 12 Current Pharmaceutical Design, 2013, Vol. 19, No. 00 Kar et al.

[54] Lin X, Liang M, Feng XH. Smurf2 is a ubiquitin E3 ligase [78] Dupont S, Mamidi A, Cordenonsi M, et al. FAM/USP9x, a mediating proteasome-dependent degradation of Smad2 in essential for TGFbeta signaling, controls transforming growth factor-beta signaling. J Biol Chem 2000; 275: Smad4 monoubiquitination. Cell 2009; 136: 123-35. 36818-22. [79] Schwickart M, Huang X, Lill JR, et al. Deubiquitinase USP9X [55] Fukuchi M, Fukai Y, Masuda N, et al. High-level expression of the stabilizes MCL1 and promotes tumour cell survival. Nature 2010; Smad ubiquitin ligase Smurf2 correlates with poor prognosis in 463: 103-7. patients with esophageal squamous cell carcinoma. Cancer Res [80] Nagai H, Noguchi T, Homma K, et al. Ubiquitin-like sequence in 2002; 62: 7162-5. ASK1 plays critical roles in the recognition and stabilization by [56] Chen C, Zhou Z, Sheehan CE, et al. Overexpression of WWP1 is USP9X and oxidative stress-induced cell death. Mol Cell 2009; 36: associated with the estrogen receptor and insulin-like growth factor 805-18. receptor 1 in breast carcinoma. Int J Cancer 2009; 124: 2829-36. [81] Cohn MA, Kowal P, Yang K, et al. A UAF1-containing [57] Laine A, Ronai Z. Regulation of p53 localization and transcription multisubunit protein complex regulates the Fanconi anemia by the HECT domain E3 ligase WWP1. Oncogene 2007; 26: 1477- pathway. Mol Cell 2007; 28: 786-97. 83. [82] Masuya D, Huang C, Liu D, et al. The HAUSP gene plays an [58] Singhal S, Taylor MC, Baker RT. Deubiquitylating enzymes and important role in non-small cell lung carcinogenesis through p53- disease. BMC Biochem 2008; 9 Suppl 1: S3. dependent pathways. J Pathol 2006; 208: 724-32. [59] Shi D, Grossman SR. Ubiquitin becomes ubiquitous in cancer: [83] Bignell GR, Warren W, Seal S, et al. Identification of the familial emerging roles of ubiquitin ligases and deubiquitinases in cylindromatosis tumour-suppressor gene. Nat Genet 2000; 25: 160- tumorigenesis and as therapeutic targets. Cancer Biol Ther 2010; 5. 10: 737-47. [84] Sun SC. CYLD: a tumor suppressor deubiquitinase regulating NF- [60] Song MS, Salmena L, Carracedo A, et al. The deubiquitinylation kappaB activation and diverse biological processes. Cell Death and localization of PTEN are regulated by a HAUSP-PML Differ 2010; 17: 25-34. network. Nature 2008; 455: 813-7. [85] Massoumi R, Chmielarska K, Hennecke K, Pfeifer A, Fassler R. [61] Li M, Brooks CL, Kon N, Gu W. A dynamic role of HAUSP in the Cyld inhibits tumor cell proliferation by blocking Bcl-3-dependent p53-Mdm2 pathway. Mol Cell 2004; 13: 879-86. NF-kappaB signaling. Cell 2006; 125: 665-77. [62] Li M, Chen D, Shiloh A, et al. Deubiquitination of p53 by HAUSP [86] Brummelkamp TR, Nijman SM, Dirac AM, Bernards R. Loss of is an important pathway for p53 stabilization. Nature 2002; 416: the cylindromatosis tumour suppressor inhibits apoptosis by 648-53. activating NF-kappaB. Nature 2003; 424: 797-801. [63] Fraile JM, Quesada V, Rodriguez D, Freije JM, Lopez-Otin C. [87] Kovalenko A, Chable-Bessia C, Cantarella G, et al. The tumour Deubiquitinases in cancer: new functions and therapeutic options. suppressor CYLD negatively regulates NF-kappaB signalling by Oncogene 2012; 31: 2373-88. deubiquitination. Nature 2003; 424: 801-5. [64] Colland F, Formstecher E, Jacq X, et al. Small-molecule inhibitor [88] Reiley WW, Jin W, Lee AJ, et al. Deubiquitinating enzyme CYLD of USP7/HAUSP ubiquitin protease stabilizes and activates p53 in negatively regulates the ubiquitin-dependent kinase Tak1 and cells. Mol Cancer Ther 2009; 8: 2286-95. prevents abnormal T cell responses. J Exp Med 2007; 204: 1475- [65] Tian X, Isamiddinova NS, Peroutka RJ, et al. Characterization of 85. selective ubiquitin and ubiquitin-like protease inhibitors using a [89] Yoshida H, Jono H, Kai H, Li JD. The tumor suppressor fluorescence-based multiplex assay format. Assay Drug Dev cylindromatosis (CYLD) acts as a negative regulator for toll-like Technol 2011; 9: 165-73. receptor 2 signaling via negative cross-talk with TRAF6 AND [66] Mizuno E, Kobayashi K, Yamamoto A, Kitamura N, Komada M. A TRAF7. J Biol Chem 2005; 280: 41111-21. deubiquitinating enzyme UBPY regulates the level of protein [90] Klaus A, Birchmeier W. Wnt signalling and its impact on ubiquitination on endosomes. Traffic 2006; 7: 1017-31. development and cancer. Nat Rev Cancer 2008; 8: 387-98. [67] Colombo M, Vallese S, Peretto I, et al. Synthesis and biological [91] Tauriello DV, Haegebarth A, Kuper I, et al. Loss of the tumor evaluation of 9-oxo-9H-indeno[1,2-b]pyrazine-2,3-dicarbonitrile suppressor CYLD enhances Wnt/beta-catenin signaling through analogues as potential inhibitors of deubiquitinating enzymes. K63-linked ubiquitination of Dvl. Mol Cell 2010; 37: 607-19. ChemMedChem 2010; 5: 552-8. [92] Wertz IE, O'Rourke KM, Zhou H, et al. De-ubiquitination and [68] Niendorf S, Oksche A, Kisser A, et al. Essential role of ubiquitin- ubiquitin ligase domains of A20 downregulate NF-kappaB specific protease 8 for receptor tyrosine kinase stability and signalling. Nature 2004; 430: 694-9. endocytic trafficking in vivo. Mol Cell Biol 2007; 27: 5029-39. [93] Hitotsumatsu O, Ahmad RC, Tavares R, et al. The ubiquitin- [69] Graner E, Tang D, Rossi S, et al. The isopeptidase USP2a regulates editing enzyme A20 restricts nucleotide-binding oligomerization the stability of fatty acid synthase in prostate cancer. Cancer Cell domain containing 2-triggered signals. Immunity 2008; 28: 381-90. 2004; 5: 253-61. [94] Turer EE, Tavares RM, Mortier E, et al. Homeostatic MyD88- [70] Stevenson LF, Sparks A, Allende-Vega N, et al. The dependent signals cause lethal inflamMation in the absence of A20. deubiquitinating enzyme USP2a regulates the p53 pathway by J Exp Med 2008; 205: 451-64. targeting Mdm2. EMBO J 2007; 26: 976-86. [95] Duwel M, Welteke V, Oeckinghaus A, et al. A20 negatively [71] Allende-Vega N, Sparks A, Lane DP, Saville MK. MdmX is a regulates T cell receptor signaling to NF-kappaB by cleaving Malt1 substrate for the deubiquitinating enzyme USP2a. Oncogene 2010; ubiquitin chains. J Immunol 2009; 182: 7718-28. 29: 432-41. [96] Boone DL, Turer EE, Lee EG, et al. The ubiquitin-modifying [72] Metzig M, Nickles D, Falschlehner C, et al. An RNAi screen enzyme A20 is required for termination of Toll-like receptor identifies USP2 as a factor required for TNF-alpha-induced NF- responses. Nat Immunol 2004; 5: 1052-60. kappaB signaling. Int J Cancer 2011; 129: 607-18. [97] Mauro C, Pacifico F, Lavorgna A, et al. ABIN-1 binds to [73] Popov N, Wanzel M, Madiredjo M, et al. The ubiquitin-specific NEMO/IKKgamma and co-operates with A20 in inhibiting NF- protease USP28 is required for MYC stability. Nat Cell Biol 2007; kappaB. J Biol Chem 2006; 281: 18482-8. 9: 765-74. [98] Lee EG, Boone DL, Chai S, et al. Failure to regulate TNF-induced [74] Albihn A, Johnsen JI, Henriksson MA. MYC in oncogenesis and as NF-kappaB and cell death responses in A20-deficient mice. a target for cancer therapies. Adv Cancer Res 2010; 107: 163-224. Science 2000; 289: 2350-4. [75] Anderson C, Crimmins S, Wilson JA, et al. Loss of Usp14 results [99] Song HY, Rothe M, Goeddel DV. The tumor necrosis factor- in reduced levels of ubiquitin in ataxia mice. J Neurochem 2005; inducible zinc finger protein A20 interacts with TRAF1/TRAF2 95: 724-31. and inhibits NF-kappaB activation. Proc Natl Acad Sci U S A [76] Lee BH, Lee MJ, Park S, et al. Enhancement of proteasome 1996; 93: 6721-5. activity by a small-molecule inhibitor of USP14. Nature 2010; 467: [100] Sierra MI, Wright MH, Nash PD. AMSH interacts with ESCRT-0 179-84. to regulate the stability and trafficking of CXCR4. J Biol Chem [77] Murray RZ, Jolly LA, Wood SA. The FAM deubiquitylating 2010; 285: 13990-4004. enzyme localizes to multiple points of protein trafficking in [101] Glickman MH, Ciechanover A. The ubiquitin-proteasome epithelia, where it associates with E-cadherin and beta-catenin. Mol proteolytic pathway: destruction for the sake of construction. Biol Cell 2004; 15: 1591-9. Physiol Rev 2002; 82: 373-428. Emerging Role of the Ubiquitin-proteasome System Current Pharmaceutical Design, 2013, Vol. 19, No. 00 13

[102] Dahlmann B. Role of proteasomes in disease. BMC Biochem 2007; [127] Wells JA, McClendon CL. Reaching for high-hanging fruit in drug 8 Suppl 1: S3. discovery at protein-protein interfaces. Nature 2007; 450: 1001-9. [103] Karin M. Nuclear factor-kappaB in cancer development and [128] Christensen DE, Brzovic PS, Klevit RE. E2-BRCA1 RING progression. Nature 2006; 441: 431-6. interactions dictate synthesis of mono- or specific polyubiquitin [104] Adams J. The development of proteasome inhibitors as anticancer chain linkages. Nat Struct Mol Biol 2007; 14: 941-8. drugs. Cancer Cell 2004; 5: 417-21. [129] Christensen DE, Klevit RE. Dynamic interactions of proteins in [105] Gu H, Chen X, Gao G, Dong H. Caspase-2 functions upstream of complex networks: identifying the complete set of interacting E2s mitochondria in endoplasmic reticulum stress-induced apoptosis by for functional investigation of E3-dependent protein ubiquitination. bortezomib in human myeloma cells. Mol Cancer Ther 2008; 7: FEBS J 2009; 276: 5381-9. 2298-307. [130] Gardner RG, Shearer AG, Hampton RY. In vivo action of the HRD [106] Ling YH, Liebes L, Jiang JD, et al. Mechanisms of proteasome ubiquitin ligase complex: mechanisms of endoplasmic reticulum inhibitor PS-341-induced G(2)-M-phase arrest and apoptosis in quality control and sterol regulation. Mol Cell Biol 2001; 21: 4276- human non-small cell lung cancer cell lines. Clin Cancer Res 2003; 91. 9: 1145-54. [131] Ballar P, Ors AU, Yang H, Fang S. Differential regulation of [107] Demo SD, Kirk CJ, Aujay MA, et al. Antitumor activity of PR- CFTRDeltaF508 degradation by ubiquitin ligases gp78 and Hrd1. 171, a novel irreversible inhibitor of the proteasome. Cancer Res Int J Biochem Cell Biol 2010; 42: 167-73. 2007; 67: 6383-91. [132] van Wijk SJ, de Vries SJ, Kemmeren P, et al. A comprehensive [108] Zhou HJ, Aujay MA, Bennett MK, et al. Design and synthesis of framework of E2-RING E3 interactions of the human ubiquitin- an orally bioavailable and selective peptide epoxyketone proteasome system. Mol Syst Biol 2009; 5: 295. proteasome inhibitor (PR-047). J Med Chem 2009; 52: 3028-38. [133] Markson G, Kiel C, Hyde R, et al. Analysis of the human E2 [109] Kupperman E, Lee EC, Cao Y, et al. Evaluation of the proteasome ubiquitin conjugating enzyme protein interaction network. Genome inhibitor MLN9708 in preclinical models of human cancer. Cancer Res 2009; 19: 1905-11. Res 2010; 70: 1970-80. [134] Zheng N, Wang P, Jeffrey PD, Pavletich NP. Structure of a c-Cbl- [110] Piva R, Ruggeri B, Williams M, et al. CEP-18770: A novel, orally UbcH7 complex: RING domain function in ubiquitin-protein active proteasome inhibitor with a tumor-selective pharmacologic ligases. Cell 2000; 102: 533-9. profile competitive with bortezomib. Blood 2008; 111: 2765-75. [135] Kar G, Keskin O, Nussinov R, Gursoy A. Human proteome-scale [111] Dorsey BD, Iqbal M, Chatterjee S, et al. Discovery of a potent, structural modeling of E2-E3 interactions exploiting interface selective, and orally active proteasome inhibitor for the treatment motifs. J Proteome Res 2012; 11: 1196-207. of cancer. J Med Chem 2008; 51: 1068-72. [136] Barabasi AL, Gulbahce N, Loscalzo J. Network medicine: a [112] Muchamuel T, Basler M, Aujay MA, et al. A selective inhibitor of network-based approach to human disease. Nat Rev Genet 2011; the immunoproteasome subunit LMP7 blocks cytokine production 12: 56-68. and attenuates progression of experimental arthritis. Nat Med 2009; [137] Aloy P, Bottcher B, Ceulemans H, et al. Structure-based assembly 15: 781-7. of protein complexes in yeast. Science 2004; 303: 2026-9. [113] Kuhn DJ, Hunsucker SA, Chen Q, et al. Targeted inhibition of the [138] Keskin O, Tsai CJ, Wolfson H, Nussinov R. A new, structurally immunoproteasome is a potent strategy against models of multiple nonredundant, diverse data set of protein-protein interfaces and its myeloma that overcomes resistance to conventional drugs and implications. Protein Sci 2004; 13: 1043-55. nonspecific proteasome inhibitors. Blood 2009; 113: 4667-76. [139] Tuncbag N, Gursoy A, Nussinov R, Keskin O. Predicting protein- [114] Feling RH, Buchanan GO, Mincer TJ, et al. Salinosporamide A: a protein interactions on a proteome scale by matching evolutionary highly cytotoxic proteasome inhibitor from a novel microbial and structural similarities at interfaces using PRISM. Nat Protoc source, a marine bacterium of the new genus salinospora. Angew 2011; 6: 1341-54. Chem Int Ed Engl 2003; 42: 355-7. [140] Aytuna AS, Gursoy A, Keskin O. Prediction of protein-protein [115] Chauhan D, Singh A, Brahmandam M, et al. Combination of interactions by combining structure and sequence conservation in proteasome inhibitors bortezomib and NPI-0052 trigger in vivo protein interfaces. Bioinformatics 2005; 21: 2850-5. synergistic cytotoxicity in multiple myeloma. Blood 2008; 111: [141] Tuncbag N, Keskin O, Gursoy A. HotPoint: hot spot prediction 1654-64. server for protein interfaces. Nucleic Acids Res 2010; 38: W402-6. [116] Arkin MR, Wells JA. Small-molecule inhibitors of protein-protein [142] Ye Y, Rape M. Building ubiquitin chains: E2 enzymes at work. Nat interactions: progressing towards the dream. Nat Rev Drug Discov Rev Mol Cell Biol 2009; 10: 755-64. 2004; 3: 301-17. [143] Papaleo E, Ranzani V, Tripodi F, et al. An acidic loop and cognate [117] Chene P. Drugs targeting protein-protein interactions. phosphorylation sites define a molecular switch that modulates ChemMedChem 2006; 1: 400-11. ubiquitin charging activity in Cdc34-like enzymes. PLoS Comput [118] Fry DC. Protein-protein interactions as targets for small molecule Biol 2011; 7: e1002056. drug discovery. Biopolymers 2006; 84: 535-52. [144] Autore F, Pagano B, Fornili A, Rittinger K, Fraternali F. In silico [119] Gonzalez-Ruiz D, Gohlke H. Targeting protein-protein interactions phosphorylation of the autoinhibited form of p47(phox): insights with small molecules: challenges and perspectives for into the mechanism of activation. Biophys J 2010; 99: 3716-25. computational binding epitope detection and ligand finding. Curr [145] Liu J, Nussinov R. Flexible cullins in cullin-RING E3 ligases Med Chem 2006; 13: 2607-25. allosterically regulate ubiquitination. J Biol Chem 2011; 286: [120] Kar G, Kuzu G, Keskin O, Gursoy A. Protein-protein interfaces 40934-42. integrated into interaction networks: implications on drug design. [146] Liu J, Nussinov R. Molecular dynamics reveal the essential role of Curr Pharm Des 2012; linker motions in the function of cullin-RING E3 ligases. J Mol [121] Keskin O, Gursoy A, Ma B, Nussinov R. Towards drugs targeting Biol 2010; 396: 1508-23. multiple proteins in a systems biology approach. Curr Top Med [147] Patel S, George R, Autore F, et al. Molecular interactions of Chem 2007; 7: 943-51. ASPP1 and ASPP2 with the p53 protein family and the apoptotic [122] Zhao L, Chmielewski J. Inhibiting protein-protein interactions promoters PUMA and Bax. Nucleic Acids Res 2008; 36: 5139-51. using designed molecules. Curr Opin Struct Biol 2005; 15: 31-4. [148] Patel S, Bui TT, Drake AF, Fraternali F, Nikolova PV. The p73 [123] Bogan AA, Thorn KS. Anatomy of hot spots in protein interfaces. J DNA binding domain displays enhanced stability relative to its Mol Biol 1998; 280: 1-9. homologue, the tumor suppressor p53, and exhibits cooperative [124] Thorn KS, Bogan AA. ASEdb: a database of alanine mutations and DNA binding. Biochemistry 2008; 47: 3235-44. their effects on the free energy of binding in protein interactions. [149] Bruschweiler R. Protein dynamics: whispering within. Nat Chem Bioinformatics 2001; 17: 284-5. 2011; 3: 665-6. [125] Thanos CD, DeLano WL, Wells JA. Hot-spot mimicry of a [150] Lakomek NA, Lange OF, Walter KF, et al. Residual dipolar cytokine receptor by a small molecule. Proc Natl Acad Sci U S A couplings as a tool to study molecular recognition of ubiquitin. 2006; 103: 15422-7. Biochem Soc Trans 2008; 36: 1433-7. [126] Picksley SM, Vojtesek B, Sparks A, Lane DP. Immunochemical [151] Lange OF, Lakomek NA, Fares C, et al. Recognition dynamics up analysis of the interaction of p53 with MDM2;--fine mapping of to microseconds revealed from an RDC-derived ubiquitin ensemble the MDM2 binding site on p53 using synthetic peptides. Oncogene in solution. Science 2008; 320: 1471-5. 1994; 9: 2523-9. 14 Current Pharmaceutical Design, 2013, Vol. 19, No. 00 Kar et al.

[152] Wlodarski T, Zagrovic B. Conformational selection and induced fit [175] Hogervorst FB, Nederlof PM, Gille JJ, et al. Large genomic mechanism underlie specificity in noncovalent interactions with deletions and duplications in the BRCA1 gene identified by a novel ubiquitin. Proc Natl Acad Sci U S A 2009; 106: 19346-51. quantitative method. Cancer Res 2003; 63: 1449-53. [153] Long D, Bruschweiler R. In silico elucidation of the recognition [176] Mishra A, Godavarthi SK, Jana NR. UBE3A/E6-AP regulates cell dynamics of ubiquitin. PLoS Comput Biol 2011; 7: e1002035. proliferation by promoting proteasomal degradation of p27. [154] Papaleo E, Casiraghi N, Arrigoni A, et al. Loop 7 of e2 enzymes: Neurobiol Dis 2009; 36: 26-34. an ancestral conserved functional motif involved in the e2- [177] Scheffner M. Ubiquitin, E6-AP, and their role in p53 inactivation. mediated steps of the ubiquitination cascade. PLoS One 2012; 7: Pharmacol Ther 1998; 78: 129-39. e40786. [178] Sanduja S, Kaza V, Dixon DA. The mRNA decay factor [155] Swaminathan G, Tsygankov AY. The Cbl family proteins: ring tristetraprolin (TTP) induces senescence in human papillomavirus- leaders in regulation of cell signaling. J Cell Physiol 2006; 209: 21- transformed cervical cancer cells by targeting E6-AP ubiquitin 43. ligase. Aging (Albany NY) 2009; 1: 803-17. [156] Blake TJ, Shapiro M, Morse HC, 3rd, Langdon WY. The [179] Nachmias B, Ashhab Y, Ben-Yehuda D. The inhibitor of apoptosis sequences of the human and mouse c-cbl proto-oncogenes show v- protein family (IAPs): an emerging therapeutic target in cancer. cbl was generated by a large truncation encompassing a proline- Semin Cancer Biol 2004; 14: 231-43. rich domain and a leucine zipper-like motif. Oncogene 1991; 6: [180] Zhang Z, Wang H, Li M, et al. MDM2 is a negative regulator of 653-7. p21WAF1/CIP1, independent of p53. J Biol Chem 2004; 279: [157] Umebayashi K, Stenmark H, Yoshimori T. Ubc4/5 and c-Cbl 16000-6. continue to ubiquitinate EGF receptor after internalization to [181] Schlott T, Reimer S, Jahns A, et al. Point mutations and nucleotide facilitate polyubiquitination and degradation. Mol Biol Cell 2008; insertions in the MDM2 zinc finger structure of human tumours. J 19: 3454-62. Pathol 1997; 182: 54-61. [158] Makishima H, Cazzolli H, Szpurka H, et al. Mutations of e3 [182] Kim SS, Yoo NJ, Jeong EG, Kim MS, Lee SH. Expression of ubiquitin ligase cbl family members constitute a novel common NEDD4-1, a PTEN regulator, in gastric and colorectal carcinomas. pathogenic lesion in myeloid malignancies. J Clin Oncol 2009; 27: APMIS 2008; 116: 779-84. 6109-16. [183] Dawson TM. Parkin and defective ubiquitination in Parkinson's [159] Joazeiro CA, Wing SS, Huang H, et al. The tyrosine kinase disease. J Neural Transm Suppl 2006; 209-13. negative regulator c-Cbl as a RING-type, E2-dependent ubiquitin- [184] Read MA, Brownell JE, Gladysheva TB, et al. Nedd8 modification protein ligase. Science 1999; 286: 309-12. of cul-1 activates SCF(beta(TrCP))-dependent ubiquitination of [160] Bertrand MJ, Milutinovic S, Dickson KM, et al. cIAP1 and cIAP2 IkappaBalpha. Mol Cell Biol 2000; 20: 2326-33. facilitate cancer cell survival by functioning as E3 ligases that [185] Oberg C, Li J, Pauley A, et al. The Notch intracellular domain is promote RIP1 ubiquitination. Mol Cell 2008; 30: 689-700. ubiquitinated and negatively regulated by the mammalian Sel-10 [161] Varfolomeev E, Blankenship JW, Wayson SM, et al. IAP homolog. J Biol Chem 2001; 276: 35847-53. antagonists induce autoubiquitination of c-IAPs, NF-kappaB [186] Yada M, Hatakeyama S, Kamura T, et al. Phosphorylation- activation, and TNFalpha-dependent apoptosis. Cell 2007; 131: dependent degradation of c-Myc is mediated by the F-box protein 669-81. Fbw7. EMBO J 2004; 23: 2116-25. [162] Yang L, Cao Z, Yan H, Wood WC. Coexistence of high levels of [187] Nateri AS, Riera-Sans L, Da Costa C, Behrens A. The ubiquitin apoptotic signaling and inhibitor of apoptosis proteins in human ligase SCFFbw7 antagonizes apoptotic JNK signaling. Science tumor cells: implication for cancer specific therapy. Cancer Res 2004; 303: 1374-8. 2003; 63: 6815-24. [188] Koepp DM, Schaefer LK, Ye X, et al. Phosphorylation-dependent [163] Imoto I, Tsuda H, Hirasawa A, et al. Expression of cIAP1, a target ubiquitination of cyclin E by the SCFFbw7 ubiquitin ligase. for 11q22 amplification, correlates with resistance of cervical Science 2001; 294: 173-7. cancers to radiotherapy. Cancer Res 2002; 62: 4860-6. [189] Mao JH, Kim IJ, Wu D, et al. FBXW7 targets mTOR for [164] Varfolomeev E, Vucic D. (Un)expected roles of c-IAPs in degradation and cooperates with PTEN in tumor suppression. apoptotic and NFkappaB signaling pathways. Cell Cycle 2008; 7: Science 2008; 321: 1499-502. 1511-21. [190] Akhoondi S, Sun D, von der Lehr N, et al. FBXW7/hCDC4 is a [165] Flygare JA, Beresini M, Budha N, et al. Discovery of a potent general tumor suppressor in human cancer. Cancer Res 2007; 67: small-molecule antagonist of inhibitor of apoptosis (IAP) proteins 9006-12. and clinical candidate for the treatment of cancer (GDC-0152). J [191] Frescas D, Pagano M. Deregulated proteolysis by the F-box Med Chem 2012; 55: 4101-13. proteins SKP2 and beta-TrCP: tipping the scales of cancer. Nat Rev [166] Du C, Fang M, Li Y, Li L, Wang X. Smac, a mitochondrial protein Cancer 2008; 8: 438-49. that promotes cytochrome c-dependent caspase activation by [192] Shapira M, Ben-Izhak O, Linn S, et al. The prognostic impact of eliminating IAP inhibition. Cell 2000; 102: 33-42. the ubiquitin ligase subunits Skp2 and Cks1 in colorectal [167] Verhagen AM, Ekert PG, Pakusch M, et al. Identification of carcinoma. Cancer 2005; 103: 1336-46. DIABLO, a mammalian protein that promotes apoptosis by binding [193] Hershko DD. Oncogenic properties and prognostic implications of to and antagonizing IAP proteins. Cell 2000; 102: 43-53. the ubiquitin ligase Skp2 in cancer. Cancer 2008; 112: 1415-24. [168] Flygare JA, Fairbrother WJ. Small-molecule pan-IAP antagonists: [194] Kavsak P, Rasmussen RK, Causing CG, et al. Smad7 binds to a patent review. Expert Opin Ther Pat 2010; 20: 251-67. Smurf2 to form an E3 ubiquitin ligase that targets the TGF beta [169] Al-Lazikani B, Banerji U, Workman P. Combinatorial drug therapy receptor for degradation. Mol Cell 2000; 6: 1365-75. for cancer in the post-genomic era. Nat Biotechnol 2012; 30: 679- [195] Rathmell WK, Chen S. VHL inactivation in renal cell carcinoma: 92. implications for diagnosis, prognosis and treatment. Expert Rev [170] Sato K, Hayami R, Wu W, et al. Nucleophosmin/B23 is a Anticancer Ther 2008; 8: 63-73. candidate substrate for the BRCA1-BARD1 ubiquitin ligase. J Biol [196] Li Y, Zhou Z, Chen C. WW domain-containing E3 ubiquitin Chem 2004; 279: 30919-22. protein ligase 1 targets p63 transcription factor for ubiquitin- [171] Yu X, Fu S, Lai M, Baer R, Chen J. BRCA1 ubiquitinates its mediated proteasomal degradation and regulates apoptosis. Cell phosphorylation-dependent binding partner CtIP. Genes Dev 2006; Death Differ 2008; 15: 1941-51. 20: 1721-6. [197] Chen C, Zhou Z, Guo P, Dong JT. Proteasomal degradation of the [172] Starita LM, Horwitz AA, Keogh MC, et al. BRCA1/BARD1 KLF5 transcription factor through a ubiquitin-independent ubiquitinate phosphorylated RNA polymerase II. J Biol Chem pathway. FEBS Lett 2007; 581: 1124-30. 2005; 280: 24498-505. [198] Li Y, Zhou Z, Alimandi M, Chen C. WW domain containing E3 [173] Ma Y, Fan S, Hu C, et al. BRCA1 regulates acetylation and ubiquitin protein ligase 1 targets the full-length ErbB4 for ubiquitination of estrogen receptor-alpha. Mol Endocrinol 2010; ubiquitin-mediated degradation in breast cancer. Oncogene 2009; 24: 76-90. 28: 2948-58. [174] Esteller M, Silva JM, Dominguez G, et al. Promoter [199] Neil JR, Tian M, Schiemann WP. X-linked inhibitor of apoptosis hypermethylation and BRCA1 inactivation in sporadic breast and protein and its E3 ligase activity promote transforming growth ovarian tumors. J Natl Cancer Inst 2000; 92: 564-9. factor-{beta}-mediated nuclear factor-{kappa}B activation during breast cancer progression. J Biol Chem 2009; 284: 21209-17. Emerging Role of the Ubiquitin-proteasome System Current Pharmaceutical Design, 2013, Vol. 19, No. 00 15

[200] Galban S, Duckett CS. XIAP as a ubiquitin ligase in cellular [206] Zhang D, Zaugg K, Mak TW, Elledge SJ. A role for the signaling. Cell Death Differ 2010; 17: 54-60. deubiquitinating enzyme USP28 in control of the DNA-damage [201] Carter BZ, Mak DH, Schober WD, et al. Simultaneous activation response. Cell 2006; 126: 529-42. of p53 and inhibition of XIAP enhance the activation of apoptosis [207] Faustrup H, Bekker-Jensen S, Bartek J, Lukas J, Mailand N. USP7 signaling pathways in AML. Blood 2010; 115: 306-14. counteracts SCFbetaTrCP- but not APCCdh1-mediated proteolysis [202] Carter BZ, Mak DH, Schober WD, et al. Triptolide sensitizes AML of Claspin. J Cell Biol 2009; 184: 13-9. cells to TRAIL-induced apoptosis via decrease of XIAP and p53- [208] van der Horst A, de Vries-Smits AM, Brenkman AB, et al. FOXO4 mediated increase of DR5. Blood 2008; 111: 3742-50. transcriptional activity is regulated by monoubiquitination and [203] Cummings J, Ranson M, Lacasse E, et al. Method validation and USP7/HAUSP. Nat Cell Biol 2006; 8: 1064-73. preliminary qualification of pharmacodynamic biomarkers [209] Mosesson Y, Mills GB, Yarden Y. Derailed endocytosis: an employed to evaluate the clinical efficacy of an antisense emerging feature of cancer. Nat Rev Cancer 2008; 8: 835-50. compound (AEG35156) targeted to the X-linked inhibitor of [210] Deng S, Zhou H, Xiong R, et al. Over-expression of genes and apoptosis protein XIAP. Br J Cancer 2006; 95: 42-8. proteins of ubiquitin specific peptidases (USPs) and proteasome [204] Schmitz R, Hansmann ML, Bohle V, et al. TNFAIP3 (A20) is a subunits (PSs) in breast cancer tissue observed by the methods of tumor suppressor gene in Hodgkin lymphoma and primary RFDD-PCR and proteomics. Breast Cancer Res Treat 2007; 104: mediastinal B cell lymphoma. J Exp Med 2009; 206: 981-9. 21-30. [205] Priolo C, Tang D, Brahamandan M, et al. The isopeptidase USP2a protects human prostate cancer from apoptosis. Cancer Res 2006; 66: 8625-32.

Received: October 2, 2012 Accepted: November 1, 2012