Discovery of potent and selective activity-based probes (ABPs) for the deubiquitinating enzyme USP30

Ubiquitin-specific protease 30 (USP30) is a deubiquitinating enzyme (DUB) localized at the mitochondrial outer membrane and involved in PINK1/Parkin-mediated mitophagy, pexophagy, BAX/BAK-dependent apoptosis, and IKKβ-USP30-ACLY-regulated lipogenesis/tumorigenesis. A USP30 inhibitor, MTX652, has recently entered clinical trials as a potential treatment for mitochondrial dysfunction. Small molecule activity-based probes (ABPs) for DUBs have recently emerged as powerful tools for in-cell inhibitor screening and DUB activity analysis, and here, we report the first small molecule ABPs (IMP-2587 and IMP-2586) which can profile USP30 activity in cells. Target engagement studies demonstrate that IMP-2587 and IMP-2586 engage active USP30 at nanomolar concentration after only 10 min incubation time in intact cells, dependent on the presence of the USP30 catalytic cysteine. Interestingly, proteomics analyses revealed that DESI1 and DESI2, small ubiquitin-related modifier (SUMO) proteases, can also be engaged by these probes, further suggesting a novel approach to develop DESI ABPs.


Introduction
Decoration of proteins with the small protein modifier ubiquitin (Ub) is a reversible post translation modification (PTM) that regulates almost all cellular functions including proteolysis, transcriptional regulation, cellular trafficking, localization, inflammation, and autophagy. 1Hundreds of Ub ligases and deubiquitinases (DUBs) are involved in the addition or removal of ubiquitin, respectively, and modulation of this system has emerged as an important therapeutic strategy across many diseases. 2Ubiquitin specific peptidase 30 (USP30) is a member of the USP DUB family, which typically harbours a Cys-His-Asp catalytic triad. 3USP30 is the only DUB known to be present in the outer mitochondrial membrane due to its unique transmembrane domain. 46][7][8] It has recently been shown that USP30 can also antagonise basal mitophagy mediated by the Parkinson's disease-associated kinase PINK1, even in cells which do not express Parkin. 91][12] USP30 may also play a role in drug resistant lymphoma, leukaemia, multiple myeloma, and BAX/BAKdependent apoptosis. 3USP30 depletion sensitizes cancer cells to BH3-mimetics (e.g., ABT-737), making it a potential target for cancer therapy. 8The USP30 inhibitor MTX652 entered clinical trials recently for Acute Kidney Injury, after exhibiting protective outcomes in various preclinical models. 13Despite being important therapeutic target and ongoing clinical trials of USP30 inhibitors, the endogenous substrates of USP30 have yet to be firmly established in vivo and regulation of its activity remains only partly understood.
Ub-derived activity-based probes (Ub-ABPs) bearing varied electrophilic warheads have been explored to better understand the function and mechanism of DUBs as well as for DUB inhibitor screening. 11While these Ub-ABPs are widely used and have greatly expanded our knowledge of DUB biology, they can be applied only in cell lysate due to their lack of cell permeability. 14In contrast, ABPs based on small molecules offer complementary properties since they are readily cell permeable, and whilst selective probes are challenging to develop they can be powerful tools to profile DUB activity in live cells, which is helpful to uncover their function in health and disease. 15,16Moreover, they have the potential to identify off-targets of structurally-related DUB inhibitors, which is not possible with highly DUB-specific Ub-ABPs. 172][23][24][25] However, no ABP specific for USP30 has been developed to date, limiting opportunities to profile USP30 activity in intact cells.Here we report the first small molecule ABPs which target USP30 in an activity-dependent manner and demonstrate their application to identify active DUBs in cancer cells by activitybased protein profiling (ABPP).

Results
To identify a suitable ABP scaffold for USP30, we considered several reported covalent USP30 inhibitors with low nanomolar potency, prioritising those bearing a CNPy warhead, expected to react with the DUB active cite cysteine residue to form an isothiourea adduct. 16,26We synthesised CNPy 1 (Fig. 1 and Scheme S1, ESI †) which has a reported IC 50 of 1-10 nM against USP30 25 (following 30 minutes incubation), and designed ABP (2, which we named IMP-2587) (Fig. 1 and Scheme 1) bearing a terminal alkyne functionality, based on reported structure-activity relationship (SAR) data suggesting that meta substitution on the aryl ring would retain activity, and compatibility with bioorthogonal click chemistry.IMP-2587 was synthesised through a short synthetic route (Scheme 1a).In parallel, we explored a CNPy with a different scaffold, synthesising the highly potent CNPy-bearing covalent inhibitor 3 (Fig. 1 and Scheme S1, ESI †) with reported IC 50 value = 1.5 nM against USP30 18 (following 30 minutes incubation), and high biochemical selectivity for USP30 among a diverse set of DUB family members, although with some appreciable off-target specificity against USP6 at higher concentration. 18We designed ABP IMP-2586 based on 3.For installation of the affinity handle (azide), the reported SAR supported para substitution at the terminal aryl ring, so we synthesised 4 by a short synthetic route (Scheme 1b).An azide bioorthogonal tag was preferred in this case due to the problematic instability of an alkyne in this relatively electron rich position during pyrrolidine Boc deprotection.
We next determined cellular engagement of endogenous USP30 with IMP-2587 or IMP-2586.HEK293T cells were treated with ABPs for 1 hour followed by western blot analysis of in-lysate competitive activity-based profiling against hemagglutinintagged Ub-vinyl methyl ester probe (HA-Ub-VME) (Fig. 2a).Incell concentration-dependent competition by IMP-2587 and IMP-2586 for USP30 were confirmed below 100 nM probe (Fig. 2b and  c), whilst engagement of an exemplar unrelated DUB, UCHL1, occurred only at a much higher concentration (30 000 nM).We hypothesise that incomplete modification of USP30 by HA-Ub-VME is likely due to limited retention of USP30 activity in cell lysates, which have likely had the USP30 transmembrane domain compromised, highlighting the potential utility of a small molecule USP30 ABP applicable directly in intact cells.Direct USP30 target engagement by IMP-2587 and IMP-2586 was confirmed in HEK293T cells following 1 h probe treatment followed by cell lysis and CuAAC bio-orthogonal ligation to capture reagent azido-TAMRA-biotin (AzTB) for IMP-2587 or alkyne-TAMRA-biotin for IMP-2586 (Fig. 3a). 27In-gel fluorescence revealed one major band at ca. 60 kDa for IMP-2587 starting as low as 3 nM probe, with some labelling of additional bands above 30 nM (Fig. 3b).Similarly, labelling with IMP-2586 revealed a major band at ca. 60 kDa, but in this case in-gel fluorescence showed a degree of unspecific labelling across all concentrations, including the vehicle (DMSO)treated control (Fig. 3d).This unspecific background labelling is presumably due to the previously reported higher off-target CuAAC reactivity of an alkyne-bearing capture reagent, applied here due to incorporation of an azide tag in IMP-2586. 28Moreover, the labelling of USP30 was further confirmed by biotin pulldown and immunoblotting.USP30 was enriched at 10 nM probe and  maximal at 100 nM, resulting in almost complete engagement and enrichment with both IMP-2587 and IMP-2586 (Fig. 3c and e), which was in stark contrast to the B50% USP30 engagement observed with the Ub-based probe, HA-Ub-VME (Fig. 2b and c).
To validate these probes as ABPs, we sought to demonstrate dependence on USP30 catalytic activity for cellular target engagement, and specificity for the active site Cys residue among the 420 Cys residues present in USP30.We expressed HA-tagged wild-type (WT) USP30 and active site cysteine to serine mutant (CS) in HEK293T cells (Fig. 4a).Cells were treated by IMP-2587 or IMP-2586 for 10 min or pre-treated with corresponding parent inhibitors for 1 hour followed by probe treatment.Western blot analysis of affinity enriched USP30 demonstrates that USP30 labelling for both probes requires the presence of the catalytic cysteine, and this enrichment can be readily outcompeted by the parent inhibitors (Fig. 4b and c), confirming the activity-dependence of these ABPs.
To explore the selectivity of IMP-2587 and IMP-2586 across the whole proteome, we performed quantitative activity-based protein profiling (ABPP) in HEK293T cells following treatment with IMP-2587 or IMP-2586 (Fig. 3a).After ABP incubation, cells were lysed, and probe-labelled proteins were ligated to biotinylated capture reagents. 27Incubation of cells with 30 nM probe for 10 min showed highly significant enrichment of USP30 (Fig. 5a and c and Tables S2 and S4, ESI †), consistent with results from in-gel fluorescence (Fig. 3) and pull-down data (Fig. 3 and 4).USP30 was outcompeted by corresponding parent inhibitors for each ABP (Fig. 5b and d and Tables S3 and S5, ESI †), confirming the high potency and rapid engagement of IMP-2587 and IMP-2586 in cells.Interestingly, we also observed less potent but still significant enrichment of two desumoylating isopeptidases (DESI1 and DESI2), which were further validated by pull-down immunoblot analyses (Fig. 5e-h).DESI1 and DESI2 are small ubiquitin-related modifier (SUMO) proteases, 3,[29][30][31][32] and IMP-2587 and IMP-2586 may offer a starting point for future development of DESI ABPs.

Discussion
Small molecule DUB ABPs circumvent the limitations of Ubbased ABPs, which can only be applied in cell lysates.Cell lysis may disrupt native DUB activity resulting in inconsistent and potentially misleading engagement by Ub-ABPs, as seen in the present study in the incomplete modification of USP30 by HA-Ub-VME.In contrast, recently developed pan-DUB small molecule ABPs enable exploration of DUB activity in intact cells, and assessment of cellular target engagement and selectivity of novel inhibitors. 15,33ysregulation of USP30 is implicated in a range of rare genetic mitochondrial diseases, and neurodegenerative diseases including Parkinson's disease. 12Accordingly, there is significant interest in developing USP30 inhibitors for the clinic, and Mission Therapeutics initiated the first clinical trials of a USP30 inhibitor in 2022 for muscular, cardiac and kidney pathologies, 34 reporting encouraging Phase I safety data.Selective and potent small molecule USP30 ABPs would be useful tools to explore the role of USP30 activity in intact cells for these diseases.
In this work, we designed, synthesised and validated two novel USP30 ABPs bearing a CNPy.Cyanoamines, particularly CNPy, have been established as a privileged warhead class for DUBs, which have proven challenging to target selectively with other warhead classes (e.g.chloroacetamides). 3,16,26,33,35,36ecent reports have shown that minimal CNPy probes lacking an extended structure lose activity toward DUBs, suggesting that the scaffold beyond the warhead plays a key role in enabling covalent DUB active site modification, and in selectivity within the DUB family. 26Interestingly, IMP-2586 and IMP-2587 showed very similar selectivity for USP30 in cells despite their divergent scaffolds.Unlike most reported DUB ABPs, IMP-2586 bears an azide tag rather than a terminal alkyne.Given the prevalence and convenience of Boc protecting group chemistry in CNPy synthetic routes, this design choice overcomes problems of acid instability of an electron rich alkyne in this position.Despite reduced CuAAC bioorthogonal ligation specificity using an alkyne capture reagent (Fig. 3), the utility of IMP-2586 as a selective USP30 ABP is preserved in cellular studies.
Our data suggest that both parent inhibitors from which IMP-2587 and IMP-2586 are derived also target DESI1 and DESI2 in cells and these may need to be considered as offtargets where related compounds are used as inhibitors.Like DUBs, DESIs are cysteine proteases, but catalyse hydrolysis of SUMO modifications rather than Ub. 29 No probe or inhibitor has been reported to date for DESIs, although these enzymes

RSC Chemical Biology
Paper are reported to be involved in a wide range of cellular pathways, including PI3K/AKT/mTOR signalling and P53-induced apoptosis. 30,31uture optimisation of IMP-2586 and IMP-2587 to minimise USP30 activity may provide the first ABPs for studying DESIs.In summary, IMP-2586 and IMP-2587 represent the first potent and selective ABPs to explore USP30 activity in intact cells, with the potential to facilitate studies of USP30 biology and target engagement, supporting future development of appropriate therapeutic strategies.

Fig. 1
Fig.1Chemical structure of parent compounds and activity-based probes with their determined IC 50 and K i accordingly.Fluorescent Polarisation assay were performed for IC 50 (30 minutes incubation against USP30), k obs /I, and k inact /K i determination of IMP-2587, and IMP-2586 against USP30 using Ub-Lys(TAMRA)-Gly (Fig.S1, ESI †).

Fig. 2
Fig. 2 (a) Workflow of immunoblot analysis of HA-Ub-VME USP30 labelling following HEK293T treatment with IMP-2587 (b) and IMP-2586 (c) for 1 h.Dose-dependent competition for USP30 labelling occurs for both the ABPs but not with UCHL1 at similar concentration range.

Fig. 3
Fig. 3 USP30 activity profiling using IMP-2587 and IMP-2586 in HEK293 cells.(a) Affinity pulldown and chemical proteomics workflow.Cells were incubated with IMP-2587 and IMP-2586 for 1 hour, separately and labelled proteins ligated via CuAAC to AzRB or AzT for IMP-2587 and YnB or YnT for IMP-2586 for in-gel fluorescence and/or affinity enrichment for immunoblotting and proteomics.Total cell lysates are labelled as TL samples; supernatants following affinity enrichment were collected as SN samples; protein bound to the beads was collected as PD samples.(b) In-gel fluorescence shows dose-dependent labelling by IMP-2587.(c) Probe-labelled (IMP-2587) protein identification by enrichment and immunoblotting for USP30.(TL: Total Lysate, SN: Supernatant, PD: Pull-Down) (d) In-gel fluorescence shows dose-dependent labelling by IMP-2586.(e) Probe-labelled (IMP-2586) protein identification by enrichment and immunoblotting for USP30.

IMP- 2586
and IMP-2587 enable sensitive and rapid detection of USP30 activity in intact cells at probe concentrations as low as 3-10 nM following 1 hour incubation, or 30 nM at 10 min incubation.Parent inhibitors 1 and 3, IMP-2587 and IMP-2586, also showed high inhibitory potency against USP30 with low nM IC 50 and fast USP30 modification kinetics both biochemically and in cells.Low nanomolar cellular target engagement and selectivity were examined with a range of orthogonal approaches, including HA-Ub-VME ABP assays, and analysis of direct target engagement through CuAAC bio-orthogonal ligation, in-gel fluorescence, enrichment, immunoblotting and ABPP proteomics.Both probes are strictly activity-based and selective for the active site cysteine of USP30 for labelling among the 21 Cys residues present in USP30.