USP1 Inhibition: A Journey from Target Discovery to Clinical Translation

Abstract

Ubiquitin-specific protease 1 (USP1) is a deubiquitinating enzyme involved in the DNA damage response. Upon DNA damage, USP1 stabilizes replication forks by removing monoubiquitin from PCNA and FANCD2-FANCI, thereby catalyzing critical final steps in translesion synthesis and interstrand crosslink (ICL) repair. This function is particularly crucial in BRCA1 mutant cancers, where the homologous recombination pathway is compromised, leading tumors to rely on USP1 for effective repair. USP1 is also overexpressed in BRCA1 mutant cancers, as well as other tumor types. Preclinical studies have demonstrated that knockout of USP1 is synthetically lethal in tumors with biallelic BRCA1 mutations, and this relationship is enhanced by combination with PARP inhibitors. Newly developed USP1 inhibitors have confirmed this synthetic lethality in BRCA1-deficient tumor cells. Moreover, these drugs have the potential for resensitizing platinum-resistant tumors. Currently, potent and specific USP1 inhibitors are undergoing evaluation in phase I clinical trials. RO7623066 (KSQ-4279) reported an acceptable safety profile during a phase I dose escalation study, with anemia being the most common side effect, and demonstrated robust pharmacokinetic, pharmacodynamic, and clinical activity. Other USP1 inhibitors, including SIM0501, XL309–101, and HSK39775, are currently in early clinical development. In this review, we provide an overview of the molecular function of USP1 and its importance as a therapeutic target in oncology, before focusing on the current state of preclinical and clinical development of USP1 inhibitors.

1. Introduction

The DNA Damage Response (DDR) is a complex network comprising sensors, transducers, and effectors that collaboratively manage cell cycle checkpoints, repair processes and apoptosis. When DNA damage occurs, DDR components detect the impairment, pause the cell cycle, and coordinate the necessary repair mechanisms. If the damage is irreparable, the cell may initiate apoptosis to eliminate the potentially harmful cells. Repair mechanisms include base excision repair, nucleotide excision repair, homologous recombination (HR), and non-homologous end joining (Jackson & Bartek, 2009).

Loss-of-function mutations in DNA damage response proteins can lead to an accumulation of mutations and genomic instability, which may drive normal cells toward oncogenesis (Tian et al., 2015). For example, mutations in HR repair genes such as BRCA1 or BRCA2 (BRCA1/2) impair the cell’s ability to effectively repair DNA double strand breaks and can subsequently lead to malignant transformation (Rempel et al., 2022). However, other repair mechanisms, such as base excision repair and translesion synthesis, can still operate and partially compensate for the deficiencies in HR. These alternative mechanisms are upregulated in BRCA1/2-deficient cells and can help mitigate the impact of accumulated mutations and support cancer cell viability (Voutsadakis & Stravodimou, 2023).

Synthetic lethality occurs when the simultaneous loss of function of two genes leads to cell death, while the loss of either gene alone does not significantly impair cell viability (O’Neil et al., 2017). This concept has been exploited in cancer therapy, where tumors with a specific genetic deficiency are targeted with a drug that inhibits a compensatory pathway, leading to selective cancer cell death. For instance, cancers with deficiencies in HR due to BRCA1/2 mutations are exquisitely sensitive to Poly(ADP-Ribose) polymerase (PARP) inhibitors. PARP is crucial for base excision repair, and its inhibition exacerbates the repair deficits in these cancers, leading to increased cell death (Jurkovicova et al., 2022).

Several proteins involved in the DDR are currently being investigated as potential synthetic lethality targets for new cancer treatments. For instance, ATM, ATR, Wee1, WRN, PARG, POLQ, PMTRT, PKMYT1, and ACL1 are novel synthetic lethal drug candidates currently in preclinical and clinical development aimed at modulating the DDR machinery (Li et al., 2023).

Ubiquitin-specific protease 1 (USP1) is a deubiquitinating enzyme (DUB) that plays a key role in regulating the DNA damage response (Kim et al., 2009; Murai et al., 2011). USP1 deubiquitinates its substrates, mono-ubiquitinated FANCD2-FANCI, and PCNA, to regulate interstrand crosslink (ICL) repair and translesion synthesis (TLS), respectively (Huang et al., 2006; Nijman et al., 2005) (Figure 1). USP1 expression and activity is upregulated in BRCA1-deficient tumors, osteosarcoma, colorectal cancer, and other tumor cells (Lim et al., 2018; Williams et al., 2011; Xu et al., 2019), suggesting that these cancers are hyper-reliant on USP1. Importantly, synthetic lethality between USP1 and BRCA1/2 has been demonstrated experimentally, setting the stage for the development of USP1 inhibitors (Lim et al., 2018). Additionally, the inhibition of both PARP and USP1 has shown strong synergy in BRCA1/2 mutant tumors (Cadzow et al., 2024; Simoneau, Engel, et al., 2023). This research has spurred the discovery of USP1 inhibitors such as RO7623066 (KSQ-4279), SIM0501, XL309–101, and HSK39775. These inhibitors have advanced into phase 1/2 clinical trials, either as monotherapies or in combination with PARP inhibitors, with promising early results emerging (Yap et al., 2024).

Figure 1.

Mechanism of Action of USP1 Inhibitors, their Synthetic Lethality in Cancers with BRCAness, and the enhancement of this activity with PARP inhibitors.

This review provides a comprehensive overview of the fundamental mechanisms of USP1 function and inhibition, as well as the current state of preclinical and clinical development of USP1 inhibitors.

2. The Translational Landscape of USP1

2.1. USP1 function

Ubiquitin is a small regulatory protein that can be covalently attached to lysine residues of substrate proteins. This involves the formation of an isopeptide bond between the carboxyl-terminal glycine (Gly76) of ubiquitin, and the amino group of target lysine residues. Ubiquitination is catalyzed by a three-step enzymatic cascade of ubiquitin-activating enzymes (E1s), ubiquitin-conjugating enzymes (E2s) and ubiquitin ligases (E3s) (Pickart & Eddins, 2004).

Ubiquitination occurs in two distinct forms. Monoubiquitination occurs when a single ubiquitin molecule is attached to a lysine residue on a substrate protein. This modification can alter the protein’s activity, localization, or interaction with other proteins. Proteins can also be polyubiquitinated, where the attached ubiquitin molecule is itself ubiquitinated at one of 7 lysine residues. The successive ubiquitination of conjugated ubiquitin molecules allows for the formation of ubiquitin chains. Chains formed through lysine 48 of ubiquitin typically mark the protein for degradation via the proteasome (Clague et al., 2015).

PCNA is monoubiquitinated at lysine residue 164 by the E2 ubiquitin conjugating-enzyme Rad6, and the E3 ligase Rad18 (Hibbert et al., 2011; Hoege et al., 2002), while FANCD2 and FANCI are monoubiquitinated by UBE2T (E2) and FANCL (E3) within the Fanconi anemia core complex (FANCA, -B, -C, -E, -F, -G, -L, and -M) (Machida et al., 2006; Sims et al., 2007; Smogorzewska et al., 2007). This monoubiquitination is essential for the proper activation of both proteins in their respective DDR pathways: translesion synthesis (TLS) (PCNA) and the repair of DNA ICL lesions (FANCD2) (Lemonidis et al., 2022; Shaheen et al., 2010). Monoubiquitination of PCNA allows for the regulated interaction of PCNA with TLS polymerases (Bienko et al., 2005; Guo et al., 2006; Plosky et al., 2006), while monoubiquitination of FANCD2 allows for the generation of a functional FANCD2/FANCI DNA-binding clamp, a critical step in ICL repair (Alcon et al., 2024; Alcon et al., 2020; Wang et al., 2020) (Figure 2).

Figure 2. USP1 catalyzes the deubiquitination of two mono-ubiquitinated DNA clamps - PCNA and the FANCI-FANCD2 heterodimer.

(A) PCNA mono-ubiquitination triggers a switch from normal replication to DNA synthesis by translesion polymerases. The arrow indicates the direction of DNA synthesis. Models are derived from crystal structures of native and mono-ubiquitinated PCNA (PDBs 6gis and 3tbl) (De March et al., 2017; Zhang et al., 2012). (B) Mono-ubiquitination the FANCI-FANCD2 heterodimer on crosslink-containing DNA, facilitating crosslink repair. Models are derived from cryoEM structures of non-ubiquitinated and mono-ubiquitinated FANCD2-FANCI (PDBs 6vaa and 6vae) (Wang et al., 2020).

The timely deubiquitination of PCNA and FANCD2-FANCI are also critical events for the resumption of DNA synthesis and successful DNA repair. These reactions are catalyzed by the enzyme USP1, which directs the hydrolysis of the ubiquitin-lysine isopeptide bond, and the release of ubiquitin and the non-modified substrate. Both functions of USP1 are dependent on its interaction with its close binding partner, USP1-associated factor 1 (UAF1; also known as WD repeat-containing protein 48 [WDR48]). In addition, deubiquitination of PCNA requires an additional complex member, ATPase family AAA domain-containing protein 5 (ATAD5) (Cohn et al., 2009; Lee et al., 2010). Failure to deubiquitinate PCNA and FANCD2-FANCI, resulting from knockdown of USP1 or from inhibition of its enzymatic activity, causes the toxic accumulation of these monoubiquitinated substrates and disruption of their respective pathways. USP1 inhibition leads to destabilization of replication forks, which is especially toxic in cells with BRCA1 deficiency, accounting for the synthetic lethality of USP1 disruption in these cells (Lim et al., 2018). Inhibition of USP1 can also lead to a subsequent accumulation of polyubiquitinated PCNA and FANCD2, promoting the proteasomal degradation of these proteins and thereby further impairing TLS and ICL repair (Simoneau, Engel, et al., 2023).

In homologous recombination-deficient (HRD) tumors, the inability to perform homologous recombination DNA repair leads to a reliance on alternative repair pathways, such us the TLS and DNA ICL repair (Taglialatela et al., 2021). TLS is a DNA damage tolerance mechanism that employs specialized translesion polymerases to continue replication past sites of DNA damage. Many of these enzymes are “Y-family” polymerases, including polymerases η (eta), κ (kappa), ι (iota), and Rev1). Unlike high-fidelity replicative polymerases of the “B-family” (e.g. Pol delta (δ) and Pol epsilon (ε)), whose tight active sites only accommodate undamaged templates and correctly paired nucleotides, the large and permissive active sites of TLS polymerases allow them to synthesize new DNA using templates containing a range of DNA adducts (McCulloch & Kunkel, 2008). This increased permissivity however occurs at the expense of reduced accuracy; TLS is therefore considered to be an error-prone mechanism of DNA synthesis (Maiorano et al., 2021). The activity of most DNA polymerases is dependent on their interaction with PCNA, which forms a trimeric ring around DNA and acts as a processivity factor for DNA polymerases during replication (Maga & Hubscher, 2003). Under normal conditions, TLS polymerases are unable to engage in replication, as they cannot compete with the replicative polymerases for access to PCNA. The DNA damage-induced ubiquitination of PCNA lysine 164 however creates an additional binding site for Y-family polymerases, which they can interact with via their ubiquitin-binding domains (Bienko et al., 2005; Guo et al., 2006) (Figure 2A). Accordingly, when USP1 deubiquitinates PCNA, the TLS polymerase is released and an error-free replicative polymerase, such as Polδ/ε, can bind PCNA and resume DNA replication. This switch-back to the replicative polymerase is essential for fork stability and efficient bypass of DNA adducts.

In addition to their roles in the “classical TLS” described above, TLS polymerases are also thought to synthesize new DNA at sites of single strand DNA gaps caused by replication stress that does not necessarily cause DNA adducts, e.g. ssDNA gaps induced by oncogene activation (Nayak et al., 2021; Nayak et al., 2020; Neelsen et al., 2013). This role for TLS polymerases is also dependent on PCNA mono-ubiquitination by RAD6/RAD18, as well as efficient deubiquitination by USP1. Indeed, inhibiting USP1 in BRCA1-deficient cells results in the accumulation of ssDNA gaps, which correlates with drug sensitivity (da Costa et al., 2024). The accumulation of ssDNA gaps may therefore be a useful biomarker for USP1 drug sensitivity in ongoing clinical trials. This likely results from the prolonged deployment of the TLS polymerase, fork instability, and enhanced mutagenesis. Interestingly, one prediction of this model is that an increase in TLS-mediated mutagenesis by USP1 inhibitors might promote the development of PARP resistance.

The FANCD2-FANCI heterodimer, the other main substrate of USP1, is a central component of the Fanconi Anemia (FA) pathway and is crucial for ICLs and coordinating other DNA repair processes. ICLs are covalent bonds formed between the two strands of the DNA helix, preventing them from separating. This crosslinking blocks DNA replication and transcription, leading to stalled replication forks and potential genomic instability if unrepaired (Kais et al., 2016). The timely monoubiquitination and deubiquitination of FANCD2 and FANCI is required for the regulation of the FA pathway and ICL repair (Lemonidis et al., 2022). A prediction of this model is that a USP1 inhibitor will disrupt the FA pathway and lead to a decrease in ICL repair and, consequently, cisplatin sensitivity. As USP1 inhibitors enter the clinic, this impact on ICL repair must be considered as an additional DNA repair alteration.

The FA pathway involves a multiprotein complex called the FA core complex, which is responsible for the initial recognition and processing of ICLs. This complex includes several FA proteins that work together to address DNA crosslinks. One of the key steps in the FA pathway is monoubiquitination of the FANCD2-FANCI heterodimer. These modifications FANCD2-FANCI from an open dimer into a closed DNA-binding clamp (Figure 2B) (Alcon et al., 2020; Rennie et al., 2020; Tan et al., 2020; Wang et al., 2020). Mono-ubiquitination thereby allows FANCD2-FANCI to localize to chromatin near the crosslink and to coordinate downstream FA proteins, including FANCD1 (BRCA2), FANCP (SLX4), FANCQ (ERCC1) and FANCR (RAD51). The crosslink is usually resolved through a series of incisions, leading to the removal of the damaged DNA segment. The excised DNA is then repaired using the sister chromatid as a template, a process that relies heavily on HR (Nijman et al., 2005). The deubiquitination of FANCD2-FANCI by USP1 is crucial for the timely termination of the DNA repair process once the damage is repaired. Indeed, while USP1 is not considered an FA protein per se, genetic deletion of murine USP1 resulted in a Fanconi anemia phenotype, demonstrating that USP1 is an important regulator of this pathway (Kim et al., 2009).

When USP1 is inhibited, FANCD2-FANCI remains monoubiquitinated because USP1 can no longer remove the ubiquitin, resulting in dysregulation of interstrand crosslink repair process. This persistent activation can disrupt the balance of DNA repair mechanisms, particularly in tumors, where precise timing is essential. Improper regulation of FANCD2, in the absence of effective HR function, may lead to dysregulated DNA repair activities, causing the repair machinery to engage in error-prone or inappropriate repair. (Kim et al., 2009)

2.2. Structure of USP1

Human USP1 is a 785 amino acid cysteine protease and member of the USP family of deubiquitinating enzymes (Fujiwara et al., 1998). This family includes 57 members, each of which possess a structurally similar catalytic domain, characterized by a “thumb, palm, fingers” architecture (Komander et al., 2009; Reyes-Turcu et al., 2009) (Figure 3).

Figure 3. The catalytic domain of USP1 interacts with ubiquitin and UAF1.

(A) A schematic illustrating the domain structure of USP1. The numbers below the schematic refer to amino acid number. CD = catalytic domain, NLS = nuclear localization signal, PIP = PCNA-interacting peptide. (B) A cryo-EM model of the USP1 catalytic domain in complex with ubiquitin (red) and UAF1 (green) (PDB: 8a9k) (Rennie et al., 2022).

The fingers of USPs are responsible for binding ubiquitin and directing its substrate-bound C-terminus to the catalytic cleft formed between the palm and thumb (Villamil et al., 2012). Catalysis is mediated by catalytic cysteine (Cys90), histidine (His593), and aspartate (Asp752) residues referred to as the catalytic triad. The catalytic cysteine acts as the nucleophile in hydrolysis of the ubiquitin-substrate isopeptide bond, which is activated and deprotonated by the catalytic histidine (Villamil et al., 2012). The catalytic aspartate forms a hydrogen bond with the catalytic histidine, aligning its side chain, and allowing it to stabilize the catalytic cysteine. Interestingly, while Asp751 was originally considered a component of the catalytic triad, based on homology with other USPs, recent findings have since demonstrated that this role is instead played by the adjacent residue, Asp752 (Keijzer et al., 2024).

The catalytic residues are positioned within a three-dimensional structure that creates a tunnel-like environment. This tunnel accommodates the ubiquitin molecule and its attached substrate, guiding them towards the catalytic residues. The tunnel’s shape and the positioning of additional residues can influence the enzyme’s specificity for different substrates. This means that while the catalytic triad is essential for catalysis, the surrounding regions help determine which substrates can access the tunnel and be deubiquitinated (Huang et al., 2006; Rennie et al., 2022).

USP1 is most closely related to two other USPs, USP12 and USP46 (Mevissen & Komander, 2017). USP1 however contains three additional inserts within the coding sequence of its catalytic domain. These inserts contain important regulatory motifs essential for USP1 function (Dharadhar et al., 2021). In addition, inserts 1 and 2 are required for an interaction between USP1 and fork-like DNA, suggesting that these inserts house DNA-binding motifs (Lim et al., 2018).

Apart from binding ubiquitin, the fingers of the USP1, USP12, and USP46 catalytic domains, also interacts directly with the β-propeller domain of UAF1 (Dharadhar et al., 2016; Li et al., 2016; Rennie et al., 2021; Yin et al., 2015). UAF1 functions as a cofactor for these USPs, by interacting directly with their substrates. UAF1 binds the FANCD2-FANCI heterodimer through an extensive interaction with FANCI, involving the UAF1 β-propeller and ancillary domains, as well as SUMO-like domain 2 (SLD2) (Rennie et al., 2021). USP1 and UAF1 initially interact in the cytoplasm, but are translocated to the nucleus due to two nuclear localization signals found within USP1 (Garcia-Santisteban et al., 2012). Although USP1, USP12, and USP46 all interact with UAF1, only USP1-UAF1 can deubiquitinate FANCD2-FANCI (Cohn et al., 2009). One explanation for this may relate to the observation that the ubiquitin groups on mono-ubiquitinated FANCD2-FANCI are buried within the structure of the closed clamp (Rennie et al., 2021). While USP1 can effectively detect and remove these buried ubiquitin molecules, they may not efficiently be detected by the USP12 and USP46 catalytic domains. USP1 also contains a unique N-terminus that interacts with FANCD2, and is required for FANCD2 deubiquitination (Arkinson et al., 2018). This may provide another explanation for the specific role of USP1 in deubiquitinating FANCD2-FANCI.

UAF1 is also important for the USP1-mediated deubiquitination of PCNA. The β-propellor and SLM domains of UAF1 interact with the protein ATAD5, a PCNA-interacting protein that functions as a binding platform for deubiquitinating enzymes that target PCNA - USP1-UAF1, USP7, and USP11 (Lee et al., 2010; Ryu et al., 2024). In addition to this indirect interaction, Insert L1 of USP1 also contains a PCNA-interaction protein (PIP) motif, which is known to be important for PCNA deubiquitination by USP1 in vitro (Dharadhar et al., 2021).

Importantly, USP1 also has an autocleavage site at an internal ubiquitin-like amino sequence embedded in the primary protein sequence (Huang et al., 2006). Autocleavage by the active site cysteine results in a catalytically inactive form of USP1 lacking part of the catalytic domain including catalytic residue, Asp752. Autocleavage subsequently leads to degradation and clearance of the enzyme from the replication fork (Piatkov et al., 2012). USP1 turnover is mediated by a degron sequence with insert L1, that signals USP1 for proteasomal degradation. The degron of USP1 binds to a ubiquitin E3 ligase complex called Anaphase Promoting Complex/CyclosomeCdh1(APC/CCdh1). This complex mediates the K48-linked polyubiquitination of USP1, targeting the protein for degradation by the proteasome (Cotto-Rios et al., 2011; Gallastegui & Groll, 2010).

Mutation of the autocleavage site, or exposure of a cell to a USP1 inhibitor, results in the accumulation of uncleaved USP1 protein at the replication fork. A recent study indicates that this accumulation of uncleaved USP1 can further compromise replication fork progression and stability (Coleman et al., 2022). Interestingly, the accumulation of uncleaved USP1 protein is an additional pharmacodynamic biomarker that can be used to follow the on-target activity of a USP1 inhibitor in anti-cancer clinical trials.

2.3. USP1 expression in cancer

USP1 overexpression is more prominent in breast cancers than in other cancer types, with particularly elevated levels in hormone receptor-deficient breast cancers compared to other subtypes. Tumors deficient in BRCA1 that exhibit high USP1 expression also show a greater number of genomic alterations. In addition, among breast tumors with the highest USP1 overexpression, BRCA1 was the second most frequently mutated gene (Cancer Genome Atlas, 2012; Lim et al., 2018).

USP1 expression is elevated in platinum-resistant ovarian cancers with BRCA1 mutations and in BRCA1-deficient ovarian tumors. Comparative analysis of USP1 expression in platinum-resistant versus platinum-sensitive ovarian cancers revealed that higher USP1 levels are linked to platinum resistance (Lim et al., 2018).

Additionally, USP1 overexpression has been described in osteosarcoma, colorectal cancer, non-small cell lung cancer, and gastric cancers. Osteosarcoma overexpression is linked to the maintenance of a mesenchymal stem cell program (Williams et al., 2011).

In colorectal cancer, the levels of USP1 are associated with shorter overall survival in patients and correlate with more advanced cancer stages (Xu et al., 2019). A study demonstrated that USP1 is upregulated in gastric cancer cells compared to paired normal tissues. This overexpression may serve as a valuable diagnostic marker for gastric cancer and was associated with poorer overall survival in patients (Meng & Li, 2022). Increased translation of USP1 mRNA has been identified as a major mechanism of platinum chemotherapy resistance in non-small cell lung cancer (NSCLC) cells (Sourisseau et al., 2016).

2.4. Synthetic lethality and USP1

The synthetic lethal relationship of USP1 and BRCA1 mutations has been demonstrated in several studies. Silencing USP1 through siRNA transfection resulted in reduced growth in BRCA1 mutant ovarian cancer cell lines, while cells reconstituted with BRCA1 (functionally wild-type) showed no such effects (Lim et al., 2018). Furthermore, in CRISPR-Cas9 screens, depletion of USP1 preferentially impaired the proliferation of cell lines with biallelic loss-of-function mutations in BRCA1 (Simoneau, Engel, et al., 2023). These synthetic lethality effects were confirmed in an in vivo mouse xenograft model. BRCA1-deficient breast cancer cells with CRISPR-mediated knockout of USP1 also exhibited significantly reduced tumor growth compared to those with BRCA1 reconstitution (functionally BRCA wild-type) (Lim et al., 2018). Notably, this mechanism of synthetic lethality was not observed in BRCA2 deficient cells in two different studies where BRCA2 mutated cell lines showed no sensitivity to USP1 depletion (Lim et al., 2018; Simoneau, Engel, et al., 2023).

2.5. Discovery of USP1 inhibitors

The first selective USP1 inhibitor, ML323, was developed through quantitative high-throughput screening (qHTS) and subsequent medicinal chemistry optimization of the hit compound 1, N-(thiophen-2-ylmethyl)-2-(2-(trifluoromethyl)phenyl)quinazolin-4-amine (Dexheimer et al., 2010). ML323 is a small-molecule inhibitor that targets an allosteric pocket on the catalytic domain of USP1 and causes subtle changes in the positioning of a loop containing catalytic residue Asp751 (Figure 4). Consequently, PCNA mono-ubiquitination serves as a pharmacodynamic marker for the activity of ML323 and other USP1 inhibitors. ML323 demonstrated half-maximal inhibitory concentration (IC₅₀) values of 76 nM in a ubiquitin-rhodamine (Ub-Rho) assay and 820 nM in orthogonal gel-based assays using K63-linked Ub-PCNA as substrate (Liang et al., 2014). Although ML323 was designed to bind to USP1, it also partially inhibits the closely related USP enzymes USP12 and USP46 at concentrations ~100 x greater than the IC₅₀ value for USP1 (Rennie et al., 2024).

Figure 4. Allosteric USP1 inhibitors target a pocket adjacent to the active site.

A model derived from a cryo-EM structure of the USP1 catalytic domain in complex with ubiquitin (red) and ML323 (green) (PDB: 8a9k) (Rennie et al., 2022). ML323, TNG348 and RO7623066 each bind the same pocket of USP1.

RO7623066 is a USP1 inhibitor currently being evaluated in early phase clinical trials. The methoxyl substituent of RO7623066 provides greater selectivity for USP1 vs other USPs, even at inhibitor concentrations >10,000 times the IC₅₀ value. Like ML323, RO7623066 binds to an allosteric pocket on the USP1 catalytic domain through an induced-fit mechanism that subtly disrupts the catalytic aspartate residue (Rennie et al., 2024). Both ML323 and RO7623066 result in near complete inhibition of USP1-UAF1.

TNG348 is another selective, noncompetitive, potent, reversible, allosteric inhibitor of USP1, which induces dose-dependent pharmacodynamic (PD) modulation of the USP1 substrate, Ub-PCNA. In the ub-Rho110 biochemical assay, TNG348 inhibited USP1-UAF1 activity with an IC50 of 82 nM, indicating its strong ability to block the deubiquitinating action of USP1. In a cellular PD assay conducted using BRCA1 mutant cells, TNG348 exhibited an IC50 of 95 nM for the reduction of Ub-PCNA. TNG348 also demonstrated selectivity across a panel of deubiquitinating enzymes (DUBs), confirming its specific targeting of USP1 over other DUBs (Simoneau, Pratt, et al., 2023). TNG348 has been reported to bind the same allosteric pocket of USP1 as ML323, based on as yet unpublished but publicly reported Cryo-EM data (Throner et al., 2024). Another related USP1 inhibitor, I-138, is also presumed to target the allosteric pocket of USP1, as its interaction with USP1 was mutually exclusive with ML323 (Simoneau, Engel, et al., 2023).

Most of the other USP1 inhibitors that are currently in development bind close to the catalytic site and include XL309–101, SIM0501 and HSK39775. The molecular structures of these compounds have not however been publicly disclosed.

2.6. In-vitro and in-vivo efficacy of USP1 inhibitors

The first in-vitro and in-vivo experiments were conducted with ML323. Inhibition of USP1 activity with ML323 results in reduced growth of BRCA1 mutation ovarian cancer cell lines compared to BRCA1 wild-type cells (Lim et al., 2018). ML323 also sensitized colorectal cancer (CRC) cells to DNA-targeting chemotherapeutics, including doxorubicin, TOPI/II inhibitors and PARP inhibitor, but not to flurouracil (5-FU) chemotherapy (Xu et al., 2019). ML323 also re-sensitized resistant cells to cisplatin in NSCLC cell cultures (Sourisseau et al., 2016).

RO7623066 has shown selective inhibitory activity in cells with HR repair deficiencies, including BRCA1/2 mutations. RO7623066 has demonstrated dose-dependent anti-tumor effects across various triple-negative breast cancer (TNBC) and ovarian cancer models both as monotherapy and in combination with olaparib. Moreover, CRISPR screens using RO7623066, cisplatin, and olaparib revealed a distinct resistance profile for RO7623066 that was non-overlapping with olaparib, highlighting the rationale for combination therapy using these agents in the clinic (Cadzow et al., 2024; Shenker et al., 2021).

TNG348 has been shown to effectively suppress BRCA1/2 mutation cancer cell lines, with its antitumor activity further enhanced when combined with PARP inhibitors. One prediction is that a USP1 inhibitor will increase TLS activity and thereby increase point mutagenesis in a treated tumor cell mutagenesis. Accordingly, co-treatment of a tumor with a combination of a USP1 inhibitor and a PARPi inhibitor could, in principle, result in an increase in mutagenesis and in acquired PARP inhibitor resistance. In a patient-derived xenograft (PDX) model of acquired resistance to PARP inhibitors, TNG348 exhibited strong synergy with PARP inhibition, demonstrating the ability of the combination to restore PARP inhibitor sensitivity in resistant tumors. This synergy may result, at least in part, from the high level of USP1 mRNA and protein expression resulting from PARP inhibitor exposure. Notably, resistance to TNG348 was uniquely conferred by knocking out genes involved in PCNA ubiquitination and translesion synthesis (Simoneau, Pratt, et al., 2023).

Importantly, early studies suggest that PARP inhibitors will increase the expression of USP1 mRNA and protein, leading to an enhancement of USP1 inhibitor responsiveness. Indeed, PARP inhibitor can lead to an increase in single strand DNA gaps (ssGAPs) (Cong et al., 2021) in BRCA1-deficient cells. This increase in ssGAPs will subsequently increase the expression of USP1 mRNA (da Costa et al., 2024), leading to increased refilling of the ssGAPs and to PARPi resistance. Taken together, this model provides a rationale for combining a PARP inhibitor with a USP1 inhibitor.

SIM0501 is a selective USP1 inhibitor that has been evaluated both as monotherapy and in combination with olaparib in cancer cell lines and in vivo tumor models. It exhibited synergistic effects with olaparib in breast, ovarian, and prostate cancer cell lines, most of which harbored HR repair pathway mutations. In in vivo efficacy studies, SIM0501 inhibited tumor growth in a dose-dependent manner when combined with olaparib, including BRCA1 mutated breast cancer, BRCA-wildtype olaparib-resistant ovarian cancer and olaparib-resistant ovarian PDX models. Additionally, biomarker studies showed a dose-dependent increase in Ub-PCNA in these in vivo models (Yang et al., 2024).

Preclinical data on other USP1 inhibitors that are currently in clinical development, e.g. XL309–101 and HSK39775, are not currently available.

2.7. Clinical Development of USP1 inhibitors

The first-in-human phase I trial of RO7623066, the first-in-class oral USP1 inhibitor, comprised a two-part study (Table 1). Results from Part 1, which involved dose escalation, were recently presented at the 2024 American Society of Clinical Oncology (ASCO) Annual Meeting (Yap et al., 2024). The presentation included data from dose escalation arms: RO7623066 as monotherapy (Arm 1), RO7623066 in combination with olaparib at 200 mg BID (Arm 2), and RO7623066 in combination with AUC 4 carboplatin (Arm 3). The maximum tolerated dose (MTD) was not reached in all 3 arms. The most common treatment-emergent adverse event (TEAE) was anemia, occurring in 36% of patients receiving RO7623066 alone, 87% in combination with olaparib, and 71% with carboplatin. At 16 weeks, the disease control rates were 28% in Arm 1, 40% in Arm 2, and 29% in Arm 3. Additionally, induction of PCNA mono-ubiquitination was observed in paired tumor biopsies from patients treated with RO7623066, providing evidence of pharmacodynamic activity. Part 2 of the study (NCT05240898) will focus on assessing the efficacy and safety of the combination with olaparib in expansion cohorts (Yap et al., 2024).

Table 1.

Clinical Development of USP1 inhibitors.

Investigational Agent	Company	NCT number	Treatment Schedule	Trial Status	Expected Enrollment	Results
KSQ-4279/RO7623066 (MedChem Express)	Hoffmann-La Roche	NCT05240898	Monotherapy or combination with Olaparib or Carboplatin	Recruiting	140	Part 1 (dose escalation) proved an acceptable safety profile. MTD was not reached. The most common TEAEs was anemia. The disease control rate at 16 weeks was 28% in monotherapy, 40% in combination with olaparib, and 29% for the combination with carboplatin (Yap et al., 2024).
TNG348 (MedChem Express)	Tango Therapeutics	NCT06065059	Monotherapy or in combination with Olaparib/Gemcitabine + carboplatin/mFOLFOX	Terminated	N/A	The study was terminated due to Grade 3/4 liver function abnormalities observed in the initial study cohorts (Therapeutics, 2023).
XL309–101	Exelixis	NCT05932862	Monotherapy or in combination with olaparib	Recruiting	66	N/A
SIM0501	Simcere	NCT06331559	Monotherapy or in combination with olaparib	Recruiting	176	N/A
HSK39775	Xizang Haisco Pharmaceutical Co., Ltd	NCT06314373	Monotherapy	Recruiting	243	N/A

Abbreviations: TEAES: Treatment Adverse Events

TNG348, an allosteric USP1 inhibitor, was evaluated in a Phase 1/2 multi-center study (NCT06065059). However, on May 23, 2024, the trial was terminated due to the observation of Grade 3/4 liver function abnormalities in patients who were on TNG348 monotherapy (Therapeutics, 2023). This liver function toxicity is believed to be a compound-specific to TNG348, rather than a class-specific adverse event. Similar liver function abnormalities were not observed on the phase I trial with RO7623066.

The Phase 1 clinical trial evaluating XL309 both as a monotherapy and in combination with olaparib is currently ongoing and recruiting patients. The monotherapy arm includes a dose-escalation phase with patients with advanced solid tumors, such as breast, ovarian, pancreatic, and prostate cancers. Once the recommended dose is established, the safety and efficacy of XL309 as a single agent will be further assessed in specific cohorts. A separate dose-escalation phase will determine the recommended dose for the combination of XL309 and olaparib, followed by planned expansion cohorts. (NCT05932862).

SIM0501 is currently being evaluated in a Phase 1 clinical trial at a single site in China, both as a monotherapy and in combination with olaparib (NCT06331559). Additionally, a Phase I/II study is actively recruiting patients to assess HSK39775 as a monotherapy for advanced solid tumors, also at a single institution in China (NCT06314373).

3. Conclusions

Targeting the DNA damage response with PARP inhibitors has provided the initial proof-of-concept for a synthetic lethal approach in oncology. Preclinical data have demonstrated a synthetic lethal relationship between USP1 inhibitors and BRCA1 mutant tumors, which is further enhanced when combined with PARP inhibitors. This targeted approach with USP1 inhibitors represents a promising novel anti-tumor strategy that may address key clinical areas of unmet need, such as PARP inhibitor resistance. Multiple potent and specific USP1 inhibitors are undergoing clinical assessment in early phase trials, including RO7623066, SIM0501, XL309–101, and HSK39775.

Abbreviations

APC/CCdh1

Anaphase Promoting Complex/CyclosomeCdh1

ASCO

American Society of Clinical Oncology

CD

Catalytic domain

CRC

Colorectal cancer

DDR

DNA Damage Response

DUB

Deubiquitinating enzyme

FA

Fanconi Anemia

HR

Homologous recombination

ICL

Interstrand crosslink repair

NLS

Nuclear localization signal

NSCLC

Non-small cell lung cancer

PARP

Poly(ADP-Ribose) polymerase

PDX

patient-derived xenograft

PIP

PCNA-interacting peptide

SLD2

SUMO-like domain 2

ssGAPs

Single strand DNA gaps

TLS

Translesion synthesis

TNBC

Triple-negative breast cancer

UAF1

USP1-associated factor 1

USP1

Ubiquitin-specific protease 1