The ubiquitin (Ub) proteasome system and Ub signalling networks are crucial to cell biology and disease development. Deubiquitylases (DUBs) control cell signalling by removing mono-Ub and polyubiquitin chains from substrates. DUBs take part in almost all processes that regulate cellular life and are frequently dysregulated in disease. We have catalogued 99 currently known DUBs in the human genome and sequence conservation analyses of catalytic residues suggest that 11 lack enzyme activity and are classed as pseudo-DUBs. These pseudoenzymes play important biological roles by allosterically activating catalytically competent DUBs as well as other active enzymes. Additionally, pseudoenzymes act as assembly scaffolds of macromolecular complexes. We discuss how pseudo-DUBs have lost their catalytic activity, their diverse mechanisms of action and their potential as therapeutic targets. Many known pseudo-DUBs play crucial roles in cell biology and it is likely that unstudied and overlooked pseudo-DUB genes will have equally important functions.

Introduction

Ubiquitylation of proteins serves as a post-translational signal to regulate virtually all cellular processes through the precise spatial and temporal control of protein stability, activity and/or localisation [1]. Enzymes involved in the ubiquitin (Ub) system are frequently dysregulated in cancer, neurodegeneration, autoimmunity and other human diseases [25]. Ubiquitylation is a versatile post-translational modification aptly suited for a cellular communication system, with the essential ‘writer, reader and eraser’ core signalling modules (Figure 1).

Overview of Ub signalling.

Figure 1.
Overview of Ub signalling.

Ub signalling can employ the writer, reader and eraser modules as communication tools in a signalling network.

Figure 1.
Overview of Ub signalling.

Ub signalling can employ the writer, reader and eraser modules as communication tools in a signalling network.

Ub processing enzymes (E1, E2 and E3) write the Ub signalling code by adding Ub to substrates (writer module). A single Ub can be conjugated to lysine residues on the surface of substrate proteins (this is known as mono-ubiquitylation), or conjugated further to lysine residues on the surface of Ub itself, thus leading to polyubiquitin (poly-Ub) chains. Seven lysine residues (K6, K11, K27, K29, K33, K48 and K63) and the free amino terminus (M1) on Ub can be targeted for poly-ubiquitylation, and thus eight different Ub chain types can be generated in total, all having different topologies and unique signalling properties [3]. Ub chains can be ‘read out’ by ubiquitin-binding domains (UBDs), which allow signal decoding and transmission (reader module). Ubiquitylation is a reversible process, and Ub is removed by deubiquitylating (DUB) enzymes (eraser module). DUB actions allow the cell to produce monomeric Ub, recycle Ub from chains and reverse signalling events resulting from ubiquitylation [6,7]. To date, 99 DUBs have been identified in the human genome, comprising six major structural families, namely (1) the ubiquitin-specific proteases (USPs), (2) the ubiquitin C-terminal hydrolases (UCHs), (3) the Machado–Joseph domain proteases (Josephins), (4) the ovarian tumour proteases (OTUs), (5) the Jab1/MPN/Mov34 metalloproteases (JAMM/MPNs) and (6) the newly discovered MIU-containing novel DUB family proteases (MINDYs) (Figure 2) [8,9].

Catalogue of known human DUBs and pseudo-DUBs.

Figure 2.
Catalogue of known human DUBs and pseudo-DUBs.

Known DUB families are shown as phylogenetic trees (PhyML) [70] with pseudo-DUBs highlighted in red. USP17 is a highly polymorphic gene with ∼30 copies as a result of a megasatellite repeat in chromosomes four and eight. Since these genes are almost identical, they are represented by USP17-L2. USP, ubiquitin-specific protease; OTU, ovarian tumour protease; JAMM, JAB1/MPN/MOV34 (MPN, Mpr1/Pad1 N-terminal); UCH, ubiquitin C-terminal hydrolase; Josephin (aslo known as Machado–Joseph disease proteases); MINDY, MIU-containing novel DUB family. Gene symbols from the HUGO Gene Nomenclature Committee database (http://www.genenames.org/) are shown where possible. Trees were visualised using the Interactive Tree of Life (iTOL v3) tool accessible online at: https://itol.embl.de.

Figure 2.
Catalogue of known human DUBs and pseudo-DUBs.

Known DUB families are shown as phylogenetic trees (PhyML) [70] with pseudo-DUBs highlighted in red. USP17 is a highly polymorphic gene with ∼30 copies as a result of a megasatellite repeat in chromosomes four and eight. Since these genes are almost identical, they are represented by USP17-L2. USP, ubiquitin-specific protease; OTU, ovarian tumour protease; JAMM, JAB1/MPN/MOV34 (MPN, Mpr1/Pad1 N-terminal); UCH, ubiquitin C-terminal hydrolase; Josephin (aslo known as Machado–Joseph disease proteases); MINDY, MIU-containing novel DUB family. Gene symbols from the HUGO Gene Nomenclature Committee database (http://www.genenames.org/) are shown where possible. Trees were visualised using the Interactive Tree of Life (iTOL v3) tool accessible online at: https://itol.embl.de.

The JAMM/MPN family is unique in using zinc during catalysis, whereas all other DUBs are cysteine proteases. Cysteine-dependent DUBs contain a common catalytic triad with a cysteine, a histidine and an acidic residue, although their structural folds differ markedly [6,10]. Nevertheless, four of the DUB families (USPs, UCHs, Josephins and OTUs) can be assigned to the papain-like structural fold, according to Pfam and SCOP databases. Representative structures for each DUB family have been reported and our understanding of how the families function as proteases has greatly improved in recent years [10]. However, fundamental questions remain as to how DUB function is regulated (i.e. switched on and off) and how substrate specificity is controlled. For example, while some DUBs display exquisite selectivity for specific Ub chain types in vitro and in vivo, others appear highly promiscuous, at least in vitro. Similarly, the catalytic and non-catalytic functions of some DUBs are highly regulated, while others appear constitutively active or permanently repressed (e.g. pseudo-DUBs). This variability of function is underpinned by diverse regulatory molecular mechanisms, even within the same structural family. As such, DUBs represent a fascinating family of enzymes that require a close inspection of mechanism of action as well as biological function.

Overview of pseudoenzymes and pseudo-DUBs

The past 10–15 years have seen a surge in studies showing how pseudoenzymes (enzymes predicted to lack catalytic activity) employ innovative and unexpected mechanisms to regulate biological functions. Pseudoenzymes perform these important roles by allosterically activating catalytically competent enzymes, acting as crucial scaffolds and as molecular switches. The best studied pseudoenzymes are pseudokinases, which have been covered in numerous reviews [1115]. Around 50 proteins containing the kinase fold have been classed as pseudokinases, representing ∼10% of the human kinome [16]. While some predicted pseudokinases have shown vestigial kinase activity in vitro, the majority are classed as true pseudokinases with no catalytic activity, and perform important biological roles as pseudoenzymes. It is now clear that almost every enzyme family contains some inactive ‘rogue’ members. The precise fraction may vary depending on the enzyme class, but in the relatively small number of families studied, ∼5–15% of proteins are inactive pseudoenzymes [1720]. This is true for pseudokinases, pseudoproteases, pseudo-E2 Ub ligases and pseudo-DUBs [2123].

The human genome contains 99 known DUBs (Figure 2) and this number is likely to increase as new members and families are discovered [9]. Of the 99 known DUB proteins, 11 are predicted to be catalytically inactive pseudoenzymes and classed as pseudo-DUBs. Interestingly, the number of pseudo-DUBs differs widely for each family, and two families, namely UCH and Josephins, have no pseudo-DUBs. The JAMM/MPN family contains the largest number (seven pseudo-DUBs), followed by the USP family with two inactive pseudo-DUBs, while the OTU and MINDY families each contain one pseudo-DUB (Figure 2). Despite the lack of catalytic residues, pseudo-DUBs perform regulatory roles in many biological functions, such as gene expression and regulation [2426], DNA damage repair [2729], protein turnover [30,31], cytokine signalling and immune responses [32].

DUB–pseudo-DUB heterodimers

The JAMM/MPN family of DUBs contains 14 members in humans. They share a common metalloprotease domain architecture, denoted the MPN domain, harbouring a zinc chelating active site [3335]. Seven out of the 14 family members are predicted to be catalytically active (denoted MPN+), while the others, which bear substitutions to essential catalytic residues, are predicted to be catalytically dead (denoted MPN). MPN family proteases typically function within larger multi-subunit complexes, which imparts great potential for multi-layer regulation.

The MPN family DUBs regulate diverse aspects of cellular biology, most notably DNA repair through BRCC36 (MPN+) in the BRCA1-A complex (BRCC36–ABRAXAS1–BRCC45–MERIT40–RAP80 complex) [29,3638], interferon receptor signalling through BRCC36 (MPN+) in the BRISC–SHMT2 complex (BRCC36–ABRAXAS2–BRCC45–MERIT40 complex) [32,39] (Figure 3), and protein degradation through Rpn11/PSMD14 (MPN+) in the proteasome lid complex [40]. The CSN5–CSN6 MPN+–MPN pair is part of the Cop9 signallosome (CSN) complex that deneddylates cullin proteins modified with the ubiquitin-like (UBL) small protein NEDD8. Thus, CSN5 is a deneddylase rather than a DUB.

BRCC36 functions as part of two distinct protein complexes.

Figure 3.
BRCC36 functions as part of two distinct protein complexes.

Illustration of the BRCC36 deubiquitylase (DUB) functions. Cytoplasmic BRCC36 regulates interferon receptor signalling as part of the BRISC complex. Nuclear BRCC36 takes part in DNA damage repair as the active DUB in the BRCA1-A complex. Pseudo-DUBs, ABRAXAS1 and ABRAXAS2 are the main determinants for BRCC36 complex formation and subcellular localisation.

Figure 3.
BRCC36 functions as part of two distinct protein complexes.

Illustration of the BRCC36 deubiquitylase (DUB) functions. Cytoplasmic BRCC36 regulates interferon receptor signalling as part of the BRISC complex. Nuclear BRCC36 takes part in DNA damage repair as the active DUB in the BRCA1-A complex. Pseudo-DUBs, ABRAXAS1 and ABRAXAS2 are the main determinants for BRCC36 complex formation and subcellular localisation.

Interestingly, many of the MPN pseudo-DUBs directly interact with active MPN+ domains to form stable heterodimers (Figure 4). In a remarkable analogy to kinase-pseudokinase heterodimers where a pseudokinase allosterically activates a kinase [12,41,42], some pseudo-DUBs also allosterically activate DUBs (Figures 4A and 5A).

Heterodimerisation of DUBs with inactive pseudo-DUBs.

Figure 4.
Heterodimerisation of DUBs with inactive pseudo-DUBs.

Cartoon representation of active DUBs (blue) in complex with inactive pseudo-DUBs (pink). (Left panels) Comparison of MPN+–MPN modes of heterodimerisation between (A) BRCC36–ABRAXAS2(KIAA0157) (PDBid 5CW3), (B) Rpn11–Rpn8 (PDBid 4O8X) and (C) CSN5–CSN6 (PDBid 4D10). (Right panels) Zoom in view of the superimposed MPN+ active site regions with the same region of MPN counterparts. Amino acid side chains are shown as sticks, except glycine residues which are depicted as spheres. Atom colours are as follows: nitrogen = blue, oxygen = red, sulphur = yellow.

Figure 4.
Heterodimerisation of DUBs with inactive pseudo-DUBs.

Cartoon representation of active DUBs (blue) in complex with inactive pseudo-DUBs (pink). (Left panels) Comparison of MPN+–MPN modes of heterodimerisation between (A) BRCC36–ABRAXAS2(KIAA0157) (PDBid 5CW3), (B) Rpn11–Rpn8 (PDBid 4O8X) and (C) CSN5–CSN6 (PDBid 4D10). (Right panels) Zoom in view of the superimposed MPN+ active site regions with the same region of MPN counterparts. Amino acid side chains are shown as sticks, except glycine residues which are depicted as spheres. Atom colours are as follows: nitrogen = blue, oxygen = red, sulphur = yellow.

Mechanism of MPN domain heterodimerisation.

Figure 5.
Mechanism of MPN domain heterodimerisation.

(A) Cartoon representation of BRCC36 allosteric activation by ABRAXAS2. In the absence of ABRAXAS2, BRCC36 exists in a forced homodimer (left panel) with a disordered active site (depicted as dashed lines) and unstructured C-terminus (not shown). Binding of ABRAXAS2 orders the active site regions and ensures catalytic residues are in optimal position. ABRAXAS2 C-terminus interacts with BRCC36 C-terminus leading to a helical bundle which causes the formation of an active super-dimer. (B) Cartoon representation of various MPN domain heterodimerisation. Rpn11–Rpn8, CSN5–CSN6 and eIF3f–eIF3h heterodimers are similar to each other but differ from the BRCC36–ABRAXAS2 heterodimer and are not known to form higher order super-dimers.

Figure 5.
Mechanism of MPN domain heterodimerisation.

(A) Cartoon representation of BRCC36 allosteric activation by ABRAXAS2. In the absence of ABRAXAS2, BRCC36 exists in a forced homodimer (left panel) with a disordered active site (depicted as dashed lines) and unstructured C-terminus (not shown). Binding of ABRAXAS2 orders the active site regions and ensures catalytic residues are in optimal position. ABRAXAS2 C-terminus interacts with BRCC36 C-terminus leading to a helical bundle which causes the formation of an active super-dimer. (B) Cartoon representation of various MPN domain heterodimerisation. Rpn11–Rpn8, CSN5–CSN6 and eIF3f–eIF3h heterodimers are similar to each other but differ from the BRCC36–ABRAXAS2 heterodimer and are not known to form higher order super-dimers.

BRCC36–ABRAXAS1 and BRCC36–ABRAXAS2 (BRCC36–KIAA0157)

The active DUB BRCC36 (MPN+) is found in two macromolecular complexes in the cell (Figure 3), depending on which partnering MPN pseudo-DUB it interacts with (ABRAXAS1 or ABRAXAS2). The BRCC36–ABRAXAS1 complex translocates to the nucleus via the ABRAXAS1 nuclear localisation signal (NLS) and is part of a larger DUB complex called the BRCA1-A complex. This complex localises to sites of DNA damage (e.g. a double-strand break) through RAP80 anchoring to K63-linked poly-Ub chains (Figure 3). Therefore, the BRCA1-A complex plays at least two roles that ultimately lead to DNA damage repair: (1) binding and cleaving K63-linked poly-Ub chains and (2) recruitment of BRCA1 to sites of DNA damage.

In the cytoplasm, BRCC36 interacts with ABRAXAS2 (aslo known as KIAA0157) and is part of a larger complex called BRISC–SHMT2 (Figure 3). This cytoplasmic DUB complex interacts with and deubiquitylates interferon receptors 1 and 2 (IFNAR1/2) and prevents the receptors from being prematurely endocytosed and entering the lysosomal degradation pathway. Thus, the BRISC–SHMT2 complex regulates interferon-dependent immune response by stabilising the interferon receptors and ensuring their availability at the membrane [32].

BRISC deficiency in mice due to genetic deletion of the BRISC specific pseudo-DUB ABRAXAS2 resulted in resistance to bacterial lipopolysaccharide [32]. Interestingly, the mice did not display any adverse phenotypes or deficiency in the DNA damage response, raising the possibility that inhibitors of pseudo-DUB function in the BRISC may have clinical utility against disease processes stemming from elevated inflammatory cytokine signals.

Biochemical studies demonstrated that BRCC36 is inactive in isolation and that heterodimerisation with the pseudo-DUB ABRAXAS2 was required for DUB activity [4345]. The crystal structure and biochemical studies of BRCC36–ABRAXAS2 (aslo known as BRCC36–KIAA0157) provided insights into the mechanism of BRCC36 activation [45]. BRCC36 forms an obligate heterodimer with ABRAXAS2 and interactions are mediated via MPN domains and C-terminal helices from both proteins. The C-terminal helical bundle participates in higher order interactions to form a dimer of heterodimers (super dimer), which constitutes the minimally active assembly capable of cleaving Lys63-linked poly-Ub chain substrates (Figures 4A and 5A). A forced BRCC36–BRCC36 homodimer was also observed by X-ray crystallography; however, the BRCC36 active site residues in this crystal structure were not optimally positioned for catalysis (Figure 5A). Instead, BRCC36 requires direct interactions with ABRAXAS2 to attain an active conformation (Figure 5A) [45].

The JAMM motif is typical of MPN domain-containing proteases and is required for catalysis. It contains five conserved residues from four key structural motifs, namely:

  • (1) E-loop, which houses a glutamic acid residue (E30, Figure 4A). Its role is to co-ordinate a catalytic water molecule.

  • (2) H-strand, which houses the His-X-His motif with two histidine residues (H94 and H96, Figure 4A) involved in Zn2+ binding.

  • (3) S-loop, which houses a serine residue (S104, Figure 4A) involved in stabilising the transition state intermediate.

  • (4) D-helix, which houses an aspartic acid residue (D107, Figure 4A) involved in Zn2+ binding.

ABRAXAS1 and 2 both have several changes in the JAMM motif that preclude function as an active DUB, with changes in the E-loop, H-strand and D-helix (Supplementary Figure S1). The structure of BRCC36–ABRAXAS2 complex revealed that ABRAXAS2 is incapable of interacting with Zn2+ and is therefore a pseudo-DUB [45]. Structural comparison of active BRCC36 with the pseudo-DUB ABRAXAS2 revealed that while there are non-conservative changes in the active site residues, the overall MPN fold and the position of the JAMM motif residues are the same for both active and inactive MPN domains (Figure 4A). As a result, the MPN fold maintains its structural integrity, despite the loss of Zn2+ and several changes to conserved catalytic residues. Interestingly, the positive charge lost by the absence of the Zn2+ ion is compensated by a guanidinium group from Arg103 (ABRAXAS2) which replaces the Zn2+-interacting His96 (BRCC36) residue (Figure 4A).

The biochemical and structural data thus far support the notion that ABRAXAS2 and the related pseudo-DUB ABRAXAS1 serve as allosteric activators of BRCC36 and have a non-catalytic role as molecular scaffolds for the correct assembly of BRCA1-A and BRISC–SHMT2 complexes.

Rpn11–Rpn8 (PSMD14–PSMD7 in humans)

Yeast Rpn11 and Rpn8 (PSMD14 and PSMD7, respectively, in humans) are part of the 19S regulatory particle of the proteasome and play a key role in Ub processing and recycling. The MPN+ domain protein Rpn11 removes poly-Ub chains from ubiquitylated substrates before they translocate to the 20S proteasome. Similar to BRCC36, Rpn11 partners with the pseudo-DUB Rpn8 (Figure 4B), and is part of a larger macromolecular complex. Rpn8 supports the role of Rpn11 through direct interactions and is required for full Rpn11 catalytic activity [31,46].

The yeast pseudo-DUB Rpn8 contains three notable changes in catalytic residues: E-loop (Glu to Arg), H-strand (His-X-His motif changed to His-X-Gly) and S-loop (Ser to Arg). Changes in the H-X-H motif render Rpn8 incapable of binding the catalytic Zn2+ (Figure 4B). Interestingly, the overall fold and integrity of the MPN domain remains unaffected by compensating interactions in the form of a salt bridge between Asp103 (D-loop) and Arg32 (E-loop), and Arg100 (S-loop) hydrogen bonding with a backbone carbonyl group from the E-loop.

Collectively, these active site residue changes abrogate Rpn8 catalytic activity. It is still unclear if Rpn8 allosterically activates Rpn11 or if its main role is to stabilise the Rpn11 fold within the multiprotein complex.

CSN5–CSN6

The COP9 signalosome (CSN complex) has eight subunits amounting to ∼450 kDa [47]. CSN5, the active protease subunit, removes Nedd8 from cullins and thus regulates the activity of cullin RING E3 ligases [48,49]. CSN5 and CSN6 genes are overexpressed in many cancers, and the Cop9 signalosome is implicated in Ub-mediated degradation of several cancer-related proteins: p27, p53, Mdm2, Smad7, Runx3, Id1, Skp2 and HIF1 [50]. Despite the fact that CSN5 is not specific for cleaving Ub, CSN5 (MPN+) and CSN6 (MPN) heterodimers are similar in topology to Rpn11–Rpn8 complexes (Figures 4B,C and 5B) and represent a recurring theme in MPN domain heterodimerisation.

The pseudo-DUB CSN6 contains four notable changes in the E-loop (Glu to Gln), H-strand (His-X-His motif changed to Thr-X-Gly) and S-loop (Ser to Asn). Changes in the H-X-H motif render CSN6 incapable of binding the catalytic Zn2+ (Figure 4C). Interestingly, the position of the D-helix remains unchanged by compensating interactions between Asp135 and Lys153 from a neighbouring β-strand. This ensures minimal disruption to the CSN6 MPN fold in the absence of the Zn2+ ion.

Collectively, these active site residue changes abrogate CSN6 catalytic activity. It is still unclear if CSN6 allosterically activates CSN5 or if its main role is to stabilise the CSN5 fold within the multiprotein complex.

Heterodimerisation of CSN5–CSN6, Rpn11–Rpn8 and eIF3f–eIF3h (discussed below) is topologically similar, but differs from the heterodimer formed by BRCC36–ABRAXAS2 (Figure 4, left panels and Figure 5). This may be because BRCC36–ABRAXAS1/2 require further assembly into a dimer of heterodimers (super dimer), which represents the active form of the enzyme (Figures 4A and 5A) [45]. This diverse mode of activation could be exploited to develop selective BRCC36 inhibitors by targeting the dimerisation and higher order super-dimerisation sites, instead of the common MPN+ active site.

Pseudo-DUBs as scaffolding modules of large complexes

Pseudo-DUBs can also act as molecular scaffolds of many macromolecular complexes, which are central to cellular biology. Examples include the spliceosome, eukaryotic initiation factor 3 (eIF3) complex and the RNA deadenylating complex Pan2–Pan3 (Figure 6).

Pseudo-DUBs as molecular scaffolds.

Figure 6.
Pseudo-DUBs as molecular scaffolds.

(A) S. cerevisiae spliceosome structure (U4/U6.U5 tri-snRNP complex; PDBid 5GAN). Prp8 is represented as ribbons and transparent surface. Prp8 contains four pseudoenzyme domains: MPN domain (pseudo-DUB), restriction endonuclease (RE) homology domain (pseudo-RE); reverse transcriptase (RT) homology domain (pseudo-RT) and RNaseH homology domain (pseudo-RNAseH). (B) Structure of eIF3 complex (PDBid 5A5T). Pseudo-DUBs eIF3f and eIF3h contain an MPN domain and are represented as ribbons and transparent surface. (C) Structure of Pan2–Pan3 deadenylase complex (PDBid 4XR7). Pan2 contains a UCH-like domain as a pseudo-DUB and is represented as ribbons and transparent surface. The pseudokinase Pan3 forms a homodimer.

Figure 6.
Pseudo-DUBs as molecular scaffolds.

(A) S. cerevisiae spliceosome structure (U4/U6.U5 tri-snRNP complex; PDBid 5GAN). Prp8 is represented as ribbons and transparent surface. Prp8 contains four pseudoenzyme domains: MPN domain (pseudo-DUB), restriction endonuclease (RE) homology domain (pseudo-RE); reverse transcriptase (RT) homology domain (pseudo-RT) and RNaseH homology domain (pseudo-RNAseH). (B) Structure of eIF3 complex (PDBid 5A5T). Pseudo-DUBs eIF3f and eIF3h contain an MPN domain and are represented as ribbons and transparent surface. (C) Structure of Pan2–Pan3 deadenylase complex (PDBid 4XR7). Pan2 contains a UCH-like domain as a pseudo-DUB and is represented as ribbons and transparent surface. The pseudokinase Pan3 forms a homodimer.

Prp8 and USP39

Splicing of pre-mRNA is an essential process performed by the spliceosome complex and a crucial step in gene expression and protein translation. Correct processing of introns from pre-mRNAs requires a high degree of accuracy and the spliceosome machinery is highly regulated. The pseudoenzyme Prp8 is central to the assembly of this large multiprotein–nucleic acid complex (Figure 6A). Prp8 is a large protein (human Prp8 has 2335 amino acids) and contains four different inactive pseudoenzyme domains: (1) a restriction endonuclease (RE) homology domain (pseudo-RE), (2) a reverse transcriptase (RT) homology domain (pseudo-RT), (3) an RNaseH homology domain (pseudo-RNAseH) and (4) an MPN domain (pseudo-DUB). Each of these domains has a role in binding to either a protein or nucleic acid component of the spliceosome.

The Prp8 MPN domain contains two active site residue changes, one in the E-loop (Glu changed to Gln) and the other in the H-strand (His-X-His motif changed to His-X-Gln). It was shown that the single His to Gln change in the H-X-H motif of yeast Prp8p is sufficient to completely abolish Zn2+ binding [51,52] and potentially catalysis; however, the Prp8 MPN domain retains the ability to interact with Ub [53]. It is as yet unclear if the Prp8 pseudo-DUB domain has other roles in the spliceosome beyond acting as a macromolecular scaffold.

Interestingly, the spliceosome complex also contains the pseudo-DUB USP39 (Sad1 in yeast) [54], although the role of this pseudoenzyme in the splicing reaction is not clear. Mutations in the active site of key catalytic residues (Cys to Asp and His to Ser) render USP39 inactive (Supplementary Figure S2). Similarly, yeast Sad1 is devoid of catalytic activity, although Sad1 is still capable of interacting with Ub [55]. As with the Prp8 MPN domain, the exact biological role of Sad1 Ub interaction is not clear, although they could serve as readers of Ub signalling by interacting with Ub and/or Ub chains.

eIF3f–eIF3h complex

eIF3f and eIF3h are MPN domain pseudo-DUBs that are part of the eIF3 complex. Binding of initiation factors is required to initiate protein synthesis and recruit transfer and messenger RNAs to the 40S subunit before assembly of active ribosomes. eIF3 is involved in both ribosome biogenesis and protein synthesis [56]. The full eIF3 complex consists of 13 subunits and amounts to ∼800 kDa [57]. Interestingly, eIF3 displays a significant degree of similarity with the CSN complex and both contain an MPN–MPN heterodimeric core. In the eIF3 complex, eIF3f and eIF3h are pseudoenzymes with their main role thought to be of a scaffolding nature (Figure 6B). Both eIF3f and eIF3h are involved in direct binding to other members of the eIF3 complex and although they are devoid of catalytic activity, their scaffolding roles are essential [57,58].

The pseudo-DUB eIF3f contains five notable residue changes in the catalytic motifs: E-loop (Glu to Arg), H-strand (His-X-His motif changed to Ala-X-Gly), S-loop (Ser to Thr) and D-Helix (Asp to Ser). Changes in the H-X-H motif and D-helix render eIF3f incapable of binding the catalytic Zn2+ and therefore inactive.

The pseudo-DUB eIF3h also contains five notable residue changes in the catalytic motifs: E-loop (Glu to Val), H-strand (His-X-His motif changed to Gln-X-Thr), S-loop (Ser changed to Thr) and D-Helix (Asp to Leu). Similar to eIF3f, changes in the H-X-H motif and D-helix render eIF3h incapable of binding the catalytic Zn2+ and inactive.

Taken together, active site amino acid substitutions in eIF3f and eIF3h, which collectively abrogate catalytic function, suggest that their primary function is to serve as molecular scaffolds for assembly of the eIF3 complex.

USP52/Pan2

The yeast Pan2–Pan3 deadenylating complex takes part in shortening of poly(A) tails of mRNA [59]. The Pan2 orthologue in humans is the pseudo-DUB USP52 and many of the functions are evolutionarily conserved. USP52 was recently described as an important regulator of the HIF1A-mediated hypoxic response, by regulating Hif1A mRNA stability [24].

USP52/Pan2 lack the catalytic cysteine (Cys mutated to Ala) and histidine (His mutated to Ser) residues required for DUB activity and are classed as pseudo-DUBs. However, Pan2 contains a catalytically active RNAse domain fused to the pseudo-DUB domain (Figure 6C). Interestingly, Pan3 contains a pseudokinase domain that dimerises, resulting in a Pan2–Pan3 complex with 1 : 2 stoichiometry (Figure 6C). Dimerisation of the Pan3 pseudokinase and the interactions of the Pan2 pseudo-DUB domain are both required for RNAse activity and biological function [60,61]. Furthermore, it is possible that USP52 binds Ub and acts as a UBD and a reader of Ub signalling.

Predicted pseudo-DUBs with unknown functions

FAM105A

FAM105A is an OTU domain containing protein that is closest to OTULIN (aslo known as FAM105B or Gumby) (Figure 2). The latter is an active DUB, which specifically cleaves Met1-linked linear Ub chains involved in cellular signalling [62,63]. FAM105A lacks the catalytic cysteine residue, which has been mutated to an Asp (Supplementary Figure S3). FAM105A is conserved as an inactive pseudo-DUB in vertebrate species, including mammals, birds, frogs and a turtle species (Pelodiscus sinensis), where the active site Cys is replaced by Asp or Ser (data not shown). Therefore, FAM105A is a predicted pseudo-DUB, the function of which remains to be identified. Interestingly, the Drosophila genome contains two OTU domain pseudo-DUBs, although these are related to OTUD4 and are involved in regulation of apoptosis [64]. Similar pseudo-OTUD4 proteins are found in the tunicate Ciona and in the brachiopod Lingula.

MINDY4B

The MINDY DUB family was only recently identified and contains four members, MINDY1–4 [9]. MINDY DUBs are active against K48-linked poly-Ub chains, displaying exquisite specificity [9]. Interestingly, a previously unannotated member (MINDY4B) is predicted to have the same fold, but lacks all three catalytic residues (Supplementary Figure S4). MINDY4B appears conserved as an inactive pseudo-DUB across vertebrates as well as in some invertebrates, for example in the tunicate Ciona intestinalis, in the aphid Acyrthosiphon pisum and in the crab Limulus polyphemus. The function of MINDY4B predicted pseudo-DUBs remains unknown.

Evolution of pseudo-DUBs

One striking observation is that 50% of JAMM/MPN DUBs are classed as pseudo-DUBs, leading to questions as to how these evolved. A possible explanation is that most JAMMs have a propensity to function as dimers with examples of JAMM/MPN functional dimers observed in many eukaryotes (e.g. yeast, ants, rodents and humans), as well as archaea [65]. It is possible that after gene duplication events, accumulation of mutations in active site residues of one protomer is tolerated, since dimerisation ensures that at least 50% enzyme activity is always present. Over time, the supporting role as an allosteric activator or as a scaffolding protein may be more important than catalytic activity, thus leading to the evolution of pseudoenzymes. This would represent a gain in evolutionary terms, since repurposing an established fold may be more advantageous than creating a new one.

Pseudo-DUBs as drug targets

Given that many pseudo-DUBs take part in influencing major cellular functions, including activity of the proteasome (Rpn11–Rpn8), protein synthesis (Prp8 and USP52), Cop9 signallosome (CSN5–CSN6), DNA damage repair (ABRAXAS1) and interferon signalling (ABRAXAS2), they present an opportunity for therapeutic intervention in disease conditions. Developing inhibitors of active DUBs represents challenges, since compounds that target JAMM domain DUBs have so far only targeted the active site Zn2+ moiety [66,67], rendering them relatively non-specific. Similarly, many compounds that target other DUB families react with the active site cysteine [4,68]. Therefore, developing small molecule inhibitors against pseudo-DUBs by exploiting their allosteric mode of action rather than their active site may be a more selective way of inhibiting active DUB complexes such as Rpn11–Rpn8, CSN5–CSN6 and BRCC36–ABRAXAS1/2.

Concluding remarks

The DUB families, like most other enzyme families, contain a fraction of pseudoenzyme members that lack catalytic activity. These pseudo-DUBs have important biological functions involved in cell signalling, protein turnover and Ub recycling. Despite their lack of catalytic activity, they have evolved different ways to influence cell signalling and cell biology. One key observation is the various mechanisms of action that pseudo-DUBs employ, in many cases still influencing Ub signalling systems by partnering with active DUBs. Pseudo-DUBs are often found as part of large macromolecular complexes where they act as allosteric activators of other enzymes, but also as molecular scaffolds required for complex assembly.

While the roles of most pseudo-DUBs are known, many questions still remain. For instance, what is the functional relevance of Prp8 and USP39/Sad1 interaction with Ub? Do other pseudo-DUBs retain Ub-binding ability and Ub chain selectivity, thereby acting as readers of Ub signalling? Pseudo-DUBs that interact with mono-Ub or poly-Ub chains may still have the same effect as active DUBs through a ‘substrate trapping’ mechanism. For instance, recent studies show that alanine substitutions of active site residues can convert DUBs into avid UBDs [69]. By interacting with Ub or poly-Ub, pseudo-DUBs can sterically hinder binding of other proteins that act as readers of Ub signals, thus reversing the signal in the same way that an active DUB can by removing Ub.

Finally, some pseudo-DUBs like FAM105A and MINDY4B remain unstudied with no assigned functional roles, despite their evolutionary conservation. As new DUBs continue to be discovered, the pseudo-DUB field is also likely to see additional new members. These have the potential to reveal new mechanisms of action of how pseudoenzymes work as well as important functional and biological insights.

Abbreviations

     
  • CSN

    Cop9 signallosome

  •  
  • DUBs

    deubiquitylases

  •  
  • JAMM/MPNs

    Jab1/MPN/Mov34 metalloproteases

  •  
  • MINDYs

    MIU-containing novel DUB family proteases

  •  
  • OTUs

    ovarian tumour proteases

  •  
  • poly-Ub

    polyubiquitin

  •  
  • RE

    restriction endonuclease

  •  
  • RT

    reverse transcriptase

  •  
  • Ub

    ubiquitin

  •  
  • UBDs

    ubiquitin-binding domains

  •  
  • UCHs

    ubiquitin C-terminal hydrolases

  •  
  • USPs

    ubiquitin-specific proteases

Funding

Our work on pseudoenzymes is supported by a Sir Henry Dale Fellowship from the Wellcome Trust and the Royal Society [200523/Z/16/Z] to E.Z., and a grant from the Polish National Science Centre [2014/15/B/NZ1/03359] to K.P.

Acknowledgments

We thank Roger Greenberg, Richard Bayliss and Derek Ceccarelli for critically reading of this manuscript and useful discussions.

Competing Interests

The Authors declare that there are no competing interests associated with the manuscript.

References

References
1
Hershko
,
A.
and
Ciechanover
,
A.
(
1992
)
The ubiquitin system for protein degradation
.
Annu. Rev. Biochem.
61
,
761
807
2
Cohen
,
P.
and
Tcherpakov
,
M.
(
2010
)
Will the ubiquitin system furnish as many drug targets as protein kinases?
Cell
143
,
686
693
3
Komander
,
D.
and
Rape
,
M.
(
2012
)
The ubiquitin code
.
Annu. Rev. Biochem.
81
,
203
229
4
Harrigan
,
J.A.
,
Jacq
,
X.
,
Martin
,
N.M.
and
Jackson
,
S.P.
(
2018
)
Deubiquitylating enzymes and drug discovery: emerging opportunities
.
Nat. Rev. Drug Discov.
17
,
57
78
5
Rape
,
M.
(
2018
)
Ubiquitylation at the crossroads of development and disease
.
Nat. Rev. Mol. Cell Biol.
19
,
59
70
6
Komander
,
D.
,
Clague
,
M.J.
and
Urbé
,
S.
(
2009
)
Breaking the chains: structure and function of the deubiquitinases
.
Nat. Rev. Mol. Cell Biol.
10
,
550
563
7
Leznicki
,
P.
and
Kulathu
,
Y.
(
2017
)
Mechanisms of regulation and diversification of deubiquitylating enzyme function
.
J. Cell Sci.
130
,
1997
2006
8
Nijman
,
S.M.B.
,
Luna-Vargas
,
M.P.A.
,
Velds
,
A.
,
Brummelkamp
,
T.R.
,
Dirac
,
A.M.G.
,
Sixma
,
T.K.
et al. 
(
2005
)
A genomic and functional inventory of deubiquitinating enzymes
.
Cell
123
,
773
786
9
Rehman
,
S.A.A.
,
Kristariyanto
,
Y.A.
,
Choi
,
S.-Y.
,
Nkosi
,
P.J.
,
Weidlich
,
S.
,
Labib
,
K.
et al. 
(
2016
)
MINDY-1 is a member of an evolutionarily conserved and structurally distinct new family of deubiquitinating enzymes
.
Mol. Cell
63
,
146
155
10
Mevissen
,
T.E.T.
and
Komander
,
D.
(
2017
)
Mechanisms of deubiquitinase specificity and regulation
.
Annu. Rev. Biochem.
86
,
159
192
11
Boudeau
,
J.
,
Miranda-Saavedra
,
D.
,
Barton
,
G.J.
and
Alessi
,
D.R.
(
2006
)
Emerging roles of pseudokinases
.
Trends Cell Biol.
16
,
443
452
12
Zeqiraj
,
E.
and
van Aalten
,
D.M.F.
(
2010
)
Pseudokinases-remnants of evolution or key allosteric regulators?
Curr. Opin. Struct. Biol.
20
,
772
781
13
Eyers
,
P.A.
and
Murphy
,
J.M.
(
2013
)
Dawn of the dead: protein pseudokinases signal new adventures in cell biology
.
Biochem. Soc. Trans.
41
,
969
974
14
Kornev
,
A.P.
and
Taylor
,
S.S.
(
2009
)
Pseudokinases: functional insights gleaned from structure
.
Structure
17
,
5
7
15
Zhang
,
H.
,
Photiou
,
A.
,
Grothey
,
A.
,
Stebbing
,
J.
and
Giamas
,
G.
(
2012
)
The role of pseudokinases in cancer
.
Cell. Signal.
24
,
1173
1184
16
Manning
,
G.
,
Whyte
,
D.B.
,
Martinez
,
R.
,
Hunter
,
T.
and
Sudarsanam
,
S.
(
2002
)
The protein kinase complement of the human genome
.
Science
298
,
1912
1934
17
Murphy
,
J.M.
,
Mace
,
P.D.
and
Eyers
,
P.A.
(
2017
)
Live and let die: insights into pseudoenzyme mechanisms from structure
.
Curr. Opin. Struct. Biol.
47
,
95
104
18
Murphy
,
J.M.
,
Farhan
,
H.
and
Eyers
,
P.A.
(
2017
)
Bio-Zombie: the rise of pseudoenzymes in biology
.
Biochem. Soc. Trans.
45
,
537
544
19
Pils
,
B.
and
Schultz
,
J.
(
2004
)
Inactive enzyme-homologues find new function in regulatory processes
.
J. Mol. Biol.
340
,
399
404
20
Todd
,
A.E.
,
Orengo
,
C.A.
and
Thornton
,
J.M.
(
2002
)
Sequence and structural differences between enzyme and nonenzyme homologs
.
Structure
10
,
1435
1451
21
Reiterer
,
V.
,
Eyers
,
P.A.
and
Farhan
,
H.
(
2014
)
Day of the dead: pseudokinases and pseudophosphatases in physiology and disease
.
Trends Cell Biol.
24
,
489
505
22
Adrain
,
C.
and
Freeman
,
M.
(
2012
)
New lives for old: evolution of pseudoenzyme function illustrated by iRhoms
.
Nat. Rev. Mol. Cell Biol.
13
,
489
498
23
Jacobsen
,
A.V.
and
Murphy
,
J.M.
(
2017
)
The secret life of kinases: insights into non-catalytic signalling functions from pseudokinases
.
Biochem. Soc. Trans.
45
,
665
681
24
Bett
,
J.S.
,
Ibrahim
,
A.F.M.
,
Garg
,
A.K.
,
Kelly
,
V.
,
Pedrioli
,
P.
,
Rocha
,
S.
et al. 
(
2013
)
The P-body component USP52/PAN2 is a novel regulator of HIF1A mRNA stability
.
Biochem. J.
451
,
185
194
25
Lee
,
M.-H.
,
Zhao
,
R.
,
Phan
,
L.
and
Yeung
,
S.-C.J.
(
2011
)
Roles of COP9 signalosome in cancer
.
Cell Cycle
10
,
3057
3066
26
Shi
,
Y.
(
2017
)
The spliceosome: a protein-directed metalloribozyme
.
J. Mol. Biol.
429
,
2640
2653
27
Coleman
,
K.A.
and
Greenberg
,
R.A.
(
2011
)
The BRCA1-RAP80 complex regulates DNA repair mechanism utilization by restricting end resection
.
J. Biol. Chem.
286
,
13669
13680
28
Shao
,
G.
,
Lilli
,
D.R.
,
Patterson-Fortin
,
J.
,
Coleman
,
K.A.
,
Morrissey
,
D.E.
and
Greenberg
,
R.A.
(
2009
)
The Rap80-BRCC36 de-ubiquitinating enzyme complex antagonizes RNF8-Ubc13-dependent ubiquitination events at DNA double strand breaks
.
Proc. Natl Acad. Sci. U.S.A.
106
,
3166
3171
29
Sobhian
,
B.
,
Shao
,
G.
,
Lilli
,
D.R.
,
Culhane
,
A.C.
,
Moreau
,
L.A.
,
Xia
,
B.
et al. 
(
2007
)
RAP80 targets BRCA1 to specific ubiquitin structures at DNA damage sites
.
Science
316
,
1198
1202
30
de Poot
,
S.A.H.
,
Tian
,
G.
and
Finley
,
D.
(
2017
)
Meddling with fate: the proteasomal deubiquitinating enzymes
.
J. Mol. Biol.
429
,
3525
3545
31
Worden
,
E.J.
,
Padovani
,
C.
and
Martin
,
A.
(
2014
)
Structure of the Rpn11–Rpn8 dimer reveals mechanisms of substrate deubiquitination during proteasomal degradation
.
Nat. Struct. Mol. Biol.
21
,
220
227
32
Zheng
,
H.
,
Gupta
,
V.
,
Patterson-Fortin
,
J.
,
Bhattacharya
,
S.
,
Katlinski
,
K.
,
Wu
,
J.
et al. 
(
2013
)
A BRISC-SHMT complex deubiquitinates IFNAR1 and regulates interferon responses
.
Cell Rep.
5
,
180
193
33
Birol
,
M.
and
Echalier
,
A.
(
2014
)
Structure and function of MPN (Mpr1/Pad1 N-terminal) domain-containing proteins
.
Curr. Protein Pept. Sci.
15
,
504
517
34
Sato
,
Y.
,
Yoshikawa
,
A.
,
Yamagata
,
A.
,
Mimura
,
H.
,
Yamashita
,
M.
,
Ookata
,
K.
et al. 
(
2008
)
Structural basis for specific cleavage of Lys 63-linked polyubiquitin chains
.
Nature
455
,
358
362
35
Ambroggio
,
X.I.
,
Rees
,
D.C.
and
Deshaies
,
R.J.
(
2003
)
JAMM: a metalloprotease-like zinc site in the proteasome and signalosome
.
PLoS Biol.
2
,
e2
36
Dong
,
Y.
,
Hakimi
,
M.-A.
,
Chen
,
X.
,
Kumaraswamy
,
E.
,
Cooch
,
N.S.
,
Godwin
,
A.K.
et al. 
(
2003
)
Regulation of BRCC, a holoenzyme complex containing BRCA1 and BRCA2, by a signalosome-like subunit and its role in DNA repair
.
Mol. Cell
12
,
1087
1099
37
Kyrieleis
,
O.J.P.
,
McIntosh
,
P.B.
,
Webb
,
S.R.
,
Calder
,
L.J.
,
Lloyd
,
J.
,
Patel
,
N.A.
et al. 
(
2016
)
Three-dimensional architecture of the human BRCA1-A histone deubiquitinase core complex
.
Cell Rep.
17
,
3099
3106
38
Wu
,
Q.
,
Paul
,
A.
,
Su
,
D.
,
Mehmood
,
S.
,
Foo
,
T.K.
,
Ochi
,
T.
et al. 
(
2016
)
Structure of BRCA1-BRCT/Abraxas complex reveals phosphorylation-dependent BRCT dimerization at DNA damage sites
.
Mol. Cell
61
,
434
448
39
Cooper
,
E.M.
,
Cutcliffe
,
C.
,
Kristiansen
,
T.Z.
,
Pandey
,
A.
,
Pickart
,
C.M.
, and
Cohen
,
R.E.
(
2009
)
K63-specific deubiquitination by two JAMM/MPN+ complexes: BRISC-associated Brcc36 and proteasomal Poh1
.
EMBO J.
28
,
621
631
40
Yao
,
T.
and
Cohen
,
R.E.
(
2002
)
A cryptic protease couples deubiquitination and degradation by the proteasome
.
Nature
419
,
403
407
41
Zeqiraj
,
E.
,
Filippi
,
B.M.
,
Deak
,
M.
,
Alessi
,
D.R.
and
van Aalten
,
D.M.F.
(
2009
)
Structure of the LKB1-STRAD-MO25 complex reveals an allosteric mechanism of kinase activation
.
Science
326
,
1707
1711
42
Littlefield
,
P.
,
Liu
,
L.
,
Mysore
,
V.
,
Shan
,
Y.
,
Shaw
,
D.E.
and
Jura
,
N.
(
2014
)
Structural analysis of the EGFR/HER3 heterodimer reveals the molecular basis for activating HER3 mutations
.
Sci. Signal.
7
,
ra114
43
Cooper
,
E.M.
,
Boeke
,
J.D.
and
Cohen
,
R.E.
(
2010
)
Specificity of the BRISC deubiquitinating enzyme is not due to selective binding to Lys63-linked polyubiquitin
.
J. Biol. Chem.
285
,
10344
10352
44
Patterson-Fortin
,
J.
,
Shao
,
G.
,
Bretscher
,
H.
,
Messick
,
T.E.
and
Greenberg
,
R.A.
(
2010
)
Differential regulation of JAMM domain deubiquitinating enzyme activity within the RAP80 complex
.
J. Biol. Chem.
285
,
30971
30981
45
Zeqiraj
,
E.
,
Tian
,
L.
,
Piggott
,
C.A.
,
Pillon
,
M.C.
,
Duffy
,
N.M.
,
Ceccarelli
,
D.F.
et al. 
(
2015
)
Higher-order assembly of BRCC36-KIAA0157 is required for DUB activity and biological function
.
Mol. Cell.
59
,
970
983
46
Pathare
,
G.R.
,
Nagy
,
I.
,
Sledz
,
P.
,
Anderson
,
D.J.
,
Zhou
,
H.J.
,
Pardon
,
E.
et al. 
(
2014
)
Crystal structure of the proteasomal deubiquitylation module Rpn8-Rpn11
.
Proc. Natl Acad. Sci. U.S.A.
111
,
2984
2989
47
Lingaraju
,
G.M.
,
Bunker
,
R.D.
,
Cavadini
,
S.
,
Hess
,
D.
,
Hassiepen
,
U.
,
Renatus
,
M.
et al. 
(
2014
)
Crystal structure of the human COP9 signalosome
.
Nature
512
,
161
165
48
Cavadini
,
S.
,
Fischer
,
E.S.
,
Bunker
,
R.D.
,
Potenza
,
A.
,
Lingaraju
,
G.M.
,
Goldie
,
K.N.
et al. 
(
2016
)
Cullin–RING ubiquitin E3 ligase regulation by the COP9 signalosome
.
Nature
531
,
598
603
49
Wei
,
N.
and
Deng
,
X.W.
(
2003
)
The COP9 signalosome
.
Annu. Rev. Cell Dev. Biol.
19
,
261
286
50
Lee
,
M.-H.
,
Zhao
,
R.
,
Phan
,
L.
and
Yeung
,
S.-C.J.
(
2011
)
Roles of COP9 signalosome in cancer
.
Cell Cycle
10
,
3057
3066
PMID:
[PubMed]
51
Pena
,
V.
,
Liu
,
S.
,
Bujnicki
,
J.M.
,
Lührmann
,
R.
and
Wahl
,
M.C.
(
2007
)
Structure of a multipartite protein-protein interaction domain in splicing factor Prp8 and its link to retinitis pigmentosa
.
Mol. Cell
25
,
615
624
52
Zhang
,
L.
,
Shen
,
J.
,
Guarnieri
,
M.T.
,
Heroux
,
A.
,
Yang
,
K.
and
Zhao
,
R.
(
2007
)
Crystal structure of the C-terminal domain of splicing factor Prp8 carrying retinitis pigmentosa mutants
.
Protein Sci.
16
,
1024
1031
53
Bellare
,
P.
,
Kutach
,
A.K.
,
Rines
,
A.
,
Guthrie
,
C.
, and
Sontheimer
,
E.J.
(
2006
)
Ubiquitin binding by a variant Jab1/MPN domain in the essential pre-mRNA splicing factor Prp8p
.
RNA
12
,
292
302
54
Bertram
,
K.
,
Agafonov
,
D.E.
,
Liu
,
W.-T.
,
Dybkov
,
O.
,
Will
,
C.L.
,
Hartmuth
,
K.
et al. 
(
2017
)
Cryo-EM structure of a human spliceosome activated for step 2 of splicing
.
Nature
542
,
318
323
55
Hadjivassiliou
,
H.
,
Rosenberg
,
O.S.
and
Guthrie
,
C.
(
2014
)
The crystal structure of S. cerevisiae Sad1, a catalytically inactive deubiquitinase that is broadly required for pre-mRNA splicing
.
RNA
20
,
656
669
56
Zhou
,
M.
,
Sandercock
,
A.M.
,
Fraser
,
C.S.
,
Ridlova
,
G.
,
Stephens
,
E.
,
Schenauer
,
M.R.
et al. 
(
2008
)
Mass spectrometry reveals modularity and a complete subunit interaction map of the eukaryotic translation factor eIF3
.
Proc. Natl Acad. Sci. U.S.A.
105
,
18139
18144
57
Georges des
,
A.
,
Dhote
,
V.
,
Kuhn
,
L.
,
Hellen
,
C.U.T.
,
Pestova
,
T.V.
,
Frank
,
J.
et al. 
(
2015
)
Structure of mammalian eIF3 in the context of the 43S preinitiation complex
.
Nature
525
,
491
495
58
Masutani
,
M.
,
Machida
,
K.
,
Kobayashi
,
T.
,
Yokoyama
,
S.
and
Imataka
,
H.
(
2013
)
Reconstitution of eukaryotic translation initiation factor 3 by co-expression of the subunits in a human cell-derived in vitro protein synthesis system
.
Protein Expr. Purif.
87
,
5
10
59
Wolf
,
J.
and
Passmore
,
L.A.
(
2014
)
mRNA deadenylation by Pan2–Pan3
.
Biochem. Soc. Trans.
42
,
184
187
60
Wolf
,
J.
,
Valkov
,
E.
,
Allen
,
M.D.
,
Meineke
,
B.
,
Gordiyenko
,
Y.
,
McLaughlin
,
S.H.
et al. 
(
2014
)
Structural basis for Pan3 binding to Pan2 and its function in mRNA recruitment and deadenylation
.
EMBO J.
33
,
1514
1526
61
Schäfer
,
I.B.
,
Rode
,
M.
,
Bonneau
,
F.
,
Schüssler
,
S.
and
Conti
,
E.
(
2014
)
The structure of the Pan2-Pan3 core complex reveals cross-talk between deadenylase and pseudokinase
.
Nat. Struct. Mol. Biol.
21
,
591
598
62
Keusekotten
,
K.
,
Elliott
,
P.R.
,
Glockner
,
L.
,
Fiil
,
B.K.
,
Damgaard
,
R.B.
,
Kulathu
,
Y.
et al. 
(
2013
)
OTULIN antagonizes LUBAC signaling by specifically hydrolyzing Met1-linked polyubiquitin
.
Cell
153
,
1312
1326
63
Rivkin
,
E.
,
Almeida
,
S.M.
,
Ceccarelli
,
D.F.
,
Juang
,
Y.-C.
,
MacLean
,
T.A.
,
Srikumar
,
T.
et al. 
(
2013
)
The linear ubiquitin-specific deubiquitinase gumby regulates angiogenesis
.
Nature
498
,
318
324
64
Louis
,
M.
,
Hofmann
,
K.
and
Broemer
,
M.
(
2015
)
Evolutionary loss of activity in de-ubiquitylating enzymes of the OTU family
.
PLoS ONE
10
,
e0143227
13
65
Cao
,
S.
,
Engilberge
,
S.
,
Girard
,
E.
,
Gabel
,
F.
,
Franzetti
,
B.
and
Maupin-Furlow
,
J.A.
(
2017
)
Structural insight into ubiquitin-like protein recognition and oligomeric states of JAMM/MPN+ proteases
.
Structure
25
,
823
833.e6
66
Li
,
J.
,
Yakushi
,
T.
,
Parlati
,
F.
,
Mackinnon
,
A.L.
,
Perez
,
C.
,
Ma
,
Y.
et al. 
(
2017
)
Capzimin is a potent and specific inhibitor of proteasome isopeptidase Rpn11
.
Nat. Chem. Biol.
13
,
486
493
67
Lauinger
,
L.
,
Li
,
J.
,
Shostak
,
A.
,
Cemel
,
I.A.
,
Ha
,
N.
,
Zhang
,
Y.
et al. 
(
2017
)
Thiolutin is a zinc chelator that inhibits the Rpn11 and other JAMM metalloproteases
.
Nat. Chem. Biol.
13
,
709
714
68
Kemp
,
M.
(
2016
)
Recent Advances in the Discovery of Deubiquitinating Enzyme Inhibitors
.
1st ed.
Elsevier B.V
69
Morrow
,
M.
,
Morgan
,
M.
,
Clerici
,
M.
,
Growkova
,
K.
,
Yan
,
M.
,
Komander
,
D.
et al. 
(
2017
)
Active site alanine substitutions can convert deubiquitinating enzymes into avid ubiquitin-binding domains
. biorxiv.org
70
Dereeper
,
A.
,
Guignon
,
V.
,
Blanc
,
G.
,
Audic
,
S.
,
Buffet
,
S.
,
Chevenet
,
F.
et al. 
(
2008
)
Phylogeny.fr: robust phylogenetic analysis for the non-specialist
.
Nucleic Acids Res.
36
,
W465
W469

Supplementary data