Antitumor activity and structure-activity relationship of poly (ADP-ribose) polymerase (PARP)-based dual inhibitors.

Introduction
Despite its stable double-helix structure, DNA within human cells is continually subjected to the risk of damage caused by metabolic by-products and environmental factors1. The DNA damage response (DDR) serves as the primary mechanism for preserving genomic homeostasis, facilitating normal cellular proliferation and maintaining the integrity of organismal functions through the precise repair of genetic material2. When the DDR is compromised, cells are prone to accumulate genetic mutations, leading to a condition of genomic instability. This instability has been extensively documented as a significant contributor to the initiation and progression of cancer3. DNA damage originates from a diverse range of sources, including both exogenous factors and endogenous mutagenic pressures. The primary forms of DNA damage are single-strand breaks (SSBs) and double-strand breaks (DSBs). The repair of SSBs is facilitated by the base excision repair (BER) pathway, the nucleotide excision repair (NER) pathway, and the mismatch repair (MMR) pathway4. Conversely, DSBs are repaired through two principal mechanisms: the high-fidelity homologous recombination (HR) mechanism and the error-prone non-homologous end joining (NHEJ) mechanism. Increasing evidence suggests that deficiencies in these repair pathways can lead to the accumulation of genomic instability, which subsequently acts as a driving force for tumorigenesis and the development of tumour heterogeneity5.
The poly(ADP-ribose) polymerase (PARP) family comprises 17 distinct members, primarily functioning to regulate cellular processes by catalysing the ADP-ribosylation of proteins, a critical post-translational modification mechanism6. Based on their structural characteristics and biological functions, the PARP family can be categorised into four primary groups (Figure 1): (i) DNA-dependent PARPs (PARP1-3), whose activity is stimulated by signals indicative of DNA damage; (ii) the Tankyrase subfamily of telomere-associated polymerases (PARP5a/b), which play a role in telomere maintenance and the Wnt signalling pathway; (iii) CCCH-type PARPs (PARP7, PARP12, and PARP13), distinguished by a characteristic zinc-finger structure; and (iv) large molecular weight PARPs (PARP9, PARP14, and PARP15), which possess multiple functional domains7. Additionally, certain members such as PARP4 and PARP10 remain unclassified at present due to their unique structural features. Significantly, PARP1 and PARP2 are the sole members of this family possessing a fully developed DNA damage recognition module, which comprises zinc-finger and WGR domains8. These proteins are capable of specifically identifying single-strand breaks and initiating the damage response through the BER pathway. Notably, PARP1 is responsible for executing over 90% of DNA repair activities. While PARP2 shares functional similarities with PARP1, it demonstrates complementary characteristics regarding substrate specificity and tissue distribution9,10.
In light of recent advancements in the understanding of DNA damage repair mechanisms and the discovery that PARP inhibition can enhance the cytotoxic effects of DNA-damaging agents, the development of PARP inhibitors has gained significant momentum11. As of the current date, six PARP inhibitors have received global approval for clinical use. Listed in the order of their market introduction, these inhibitors are olaparib, rucaparib, niraparib, talazoparib, fuzuloparib, and pamiparib12. The classical mechanisms of action for this class of drugs primarily involve the following aspects: firstly, the inhibition of PARP enzymatic activity and the process of PARP trapping; secondly, the promotion of PARP-DNA complex formation, which subsequently hinders DNA repair processes and leads to the generation of DSBs; and thirdly, the induction of homologous recombination repair-deficient (HRD) tumour cells to produce irreparable DSBs, thereby achieving a synthetic lethality effect (Figure 2)13. Recent research advancements have led to major breakthroughs in understanding how PARP inhibitors work. The traditional “synthetic lethality” theory has been updated, suggesting that PARP inhibitors primarily cause irreversible DNA damage in HRD tumour cells by disrupting the regulation of transcription-replication conflicts (TRCs), instead of causing replication fork stalling through “PARP trapping.” Experimental findings indicate that the effectiveness of PARP inhibitors is linked to their ability to inhibit PARP enzyme activity, rather than their capacity to trap PARP14. Addressing transcription-replication conflicts presents a promising strategy for the development of safer PARP inhibitors. The significant toxicity associated with conventional inhibitors is primarily due to PARP trapping. Recent insights into the underlying mechanisms suggest that simply inhibiting the enzymatic activity of PARP can effectively target HR-deficient cancer cells, thereby reducing adverse effects on normal cells15.
The continued in-depth investigation of PARP biology has demonstrated that PARP inhibitors possess significant clinical efficacy in the treatment of solid tumours, such as ovarian, breast, prostate, and pancreatic cancers16. However, the broader application of these inhibitors is hindered by the challenge of acquired drug resistance. Research has identified key mechanisms of drug resistance, including restored homologous recombination repair in tumour cells, enhanced DNA replication fork stability, increased activity of drug efflux pumps like P-glycoprotein, and changes in critical pathways such as epigenetic regulation17. To tackle these issues, current research focuses on developing new intervention strategies like PROTAC degraders and PARP subtype-selective inhibitors, creating biomarker systems to predict resistance, and achieving synergistic effects through combination therapies18. Notably, combination drug therapy has proven effective in clinical oncology by improving efficacy and minimising toxicities and side effects19. Extensive preclinical and clinical data show that combining PARP inhibitors with other tumour-targeting inhibitors, such as epigenetic drugs, immunotherapeutics, and RTK inhibitors, creates a synergistic effect20. However, challenges like dose optimisation, pharmacokinetic interactions, and cumulative toxicity hinder clinical use21. Dual-target PARP inhibitors offer a novel approach by modulating PARP and its synergistic targets within one molecule, avoiding the risks of multiple drug administration and balancing pharmacokinetics and safety22. This article conducts a systematic review of recent advancements in the field of dual-target PARP inhibitors, emphasising an analysis of target selection strategies, structure-activity relationships (SAR), and pharmacological activity characteristics. By providing a thorough evaluation of the current state of research, this study seeks to establish a theoretical foundation for further investigations into dual-target PARP inhibitors and to offer guidance for future research trajectories.

PARP as a therapeutic target
Despite the conserved catalytic domains among PARP family members, only PARP1 and PARP2 have distinct DNA-binding domains (DBDs) that enable them to recognise and bind DNA damage sites, activating the BER pathway for DNA repair23. PARP inhibitors target PARP1 and PARP2, hindering their activity and causing DNA damage accumulation, leading to lethal double-strand breaks24. Beyond PARP1 and PARP2, the monoPARP family, including PARP7, plays a crucial role in cellular stress responses related to inflammation, cancer, and neurodegenerative diseases25. PARP7 is notably upregulated in tumours, aiding cancer cell survival, while its inhibition can suppress tumour growth, restore interferon signalling, and counter immune evasion26. At present, PARP1/2 inhibitors have achieved substantial advancements in treating BRCA-mutated tumours by inhibiting enzyme activity and leveraging the “synthetic lethality” effect27. Conversely, highly selective PARP7 inhibitors are designed to minimise off-target effects and improve treatment safety, demonstrating potential benefits in restoring interferon responses and suppressing tumour growth28. Notably, the highly selective PARP7 inhibitor RBN-2397 has progressed to phase II clinical trials. As an emerging target in tumour immunotherapy, the research, development, and clinical translation of PARP7 inhibitors are rapidly advancing29. This chapter presents a comprehensive analysis of the latest clinical advancements concerning PARP1, PARP2, and PARP7, as well as their respective inhibitors. Furthermore, it examines synergistic treatment strategies in conjunction with various antitumor agents. Additionally, the chapter explores the development strategies for dual-target inhibitors that address different members of the PARP family, with the objective of offering guidance for future research initiatives.

PARP1/2/7 structure and functions
The PARP1 protein is a polypeptide consisting of 1,014 amino acids, with an approximate molecular weight of 113 kDa, and is structurally organised into three distinct regions (Figure 1)30: (i) The N-terminal DNA-binding domain (DBD, encompassing amino acids 2–372) comprises three zinc-finger motifs (ZnI-III). ZnI and ZnII are responsible for detecting DNA damage, while ZnIII plays a role in facilitating inter-domain interactions; (ii) The auto-modification domain (AMD, spanning amino acids 373–524) includes the BRCA1 C-terminal (BRCT) motif, which is pivotal for domain interactions and serves as the principal site for auto-poly(ADP-ribosyl)ation; (iii) The C-terminal catalytic domain (CAT, covering amino acids 525–1014) contains a conserved nicotinamide adenine dinucleotide (NAD+)-binding pocket and an ADP-ribosyltransferase (ART) active centre, both of which are crucial for the synthesis of poly(ADP-ribose). Collectively, these regions facilitate poly(ADP-ribosyl)ation signal transduction during the DDR31.
PARP1 plays a crucial role in repairing single-strand DNA breaks via the BER pathway32. Its N-terminal DNA-binding domain, featuring ZnI/ZnII zinc-finger structures, targets these breaks, while the ZnIII zinc-finger aids in homodimer formation to activate PARP133. Once activated, PARP1 splits NAD+ into ADP-ribose units, forming poly(ADP-ribose) chains. These chains modify repair proteins and relax chromatin, exposing DNA damage sites and attracting repair complexes. Proteins with poly(ADP-ribose)-binding motifs, like PBZ and Macro domains, interact with these chains to enhance DNA repair, ensuring timely and precise DNA restoration34.
PARP1 and PARP2, while sharing a conserved catalytic domain with 69% sequence homology, have distinct protein structures. PARP1 (113 kDa) features a complete DNA-binding domain (DBD) with zinc-finger motifs and a BRCT domain for recruiting repair complexes via poly(ADP-ribosyl)ation. In contrast, PARP2 (66 kDa) lacks these domains and instead uses its unique N-terminal intrinsically disordered region, AMD, and CAT for DNA damage binding (Figure 1)35. PARP2 plays a different role in DNA repair, showing a preference for single-strand breaks, while PARP1 can bind to both single-strand and double-strand breaks36. Additionally, PARP2 can inhibit PARP1’s activity upon binding to damaged DNA. In collaboration with BRCA1, PARP2 can also facilitate a DNA end resection repair mechanism that functions independently of ADP-ribosylation activity37.
PARP7 is a member of the CCCH-type PARP subfamily, which also includes PARP12 and PARP13. Its core structure comprises three essential domains: a CCCH-type zinc finger domain characterised by a Cys-Cys-Cys-His motif, a WWE domain (tryptophan-tryptophan-glutamic acid), and a catalytic domain (Figure 1)38. The CCCH domain is pivotal for substrate recognition and binding, particularly in the specific recognition of the androgen receptor (AR) in its agonist-induced conformation, facilitating the mono-ADP-ribosylation modification of AR39. Furthermore, this zinc finger structure is involved in regulating the nuclear localisation of PARP7, enhancing its nuclear import via the nuclear localisation signal to execute its nuclear functions effectively40. The catalytic domain of PARP7, functioning as a mono-ADP-ribosyltransferase, facilitates the transfer of a single ADP-ribose unit from NAD+ to target proteins, resulting in mono-ADP-ribosylation (MARylation), which is essential for the enzyme’s biological activity41. The precise role of the WWE domain in PARP7 remains to be elucidated; however, mutational analyses indicate that it may play a critical role in modulating protein stability and affecting the degradation rate of PARP742.
In contrast to PARP1/2, the expression of PARP7 is induced and upregulated by ligands of the aryl hydrocarbon receptor (AHR), including those found in cigarette smoke. Acting as a negative feedback regulator, PARP7 is capable of inhibiting the expression of the cytochrome P450 genes CYP1A1 and CYP1B143. During viral infections, AHR-induced PARP7 interacts with TBK1, a pivotal kinase in the pattern recognition pathway, which is typically activated to enhance the type I interferon (IFN) response and bolster antiviral immunity. However, PARP7 negatively modulates the type I IFN signalling pathway by catalysing the MARylation of TBK1, thereby inhibiting its kinase activity44. Notably, overexpression or gene amplification of PARP7 has been documented in various cancers, suggesting that this mechanism may facilitate tumour cells in evading natural immune surveillance45.

Research status of PARP inhibitors
As of the current date, six PARP inhibitors have attained regulatory approval and have been commercialised on a global scale. As shown in Table 1, these inhibitors, listed in order of their introduction to the market, include olaparib (1)46, rucaparib (2)47, niraparib (3)48, talazoparib (4)49, fuzuloparib (5)50, and pamiparib (6)51. The approved uses for PARP inhibitors include advanced ovarian cancer, fallopian tube carcinoma, primary peritoneal carcinoma, pancreatic cancer, and prostate cancer (Table 1)52. Current guidelines require molecular testing to identify patients with tumours showing homologous recombination repair deficiency53. Additionally, many new PARP inhibitors are in clinical trials, showing significant anti-tumour efficacy as monotherapy in advanced ovarian cancer, triple-negative breast cancer, and castration-resistant prostate cancer (Table 2)54. It is important to highlight that PARP subtype-selective inhibitors, exemplified by PARP1-selective inhibitors such as AZD5305 (12) and AZD9574 (13), as well as the PARP7-selective inhibitor RBN-2397 (17), demonstrate enhanced target specificity55,56. This increased specificity allows for more precise interactions with target molecules, thereby minimising adverse effects on normal tissues and reducing the incidence of treatment-related toxicity and side effects. Consequently, this class of inhibitors offers a safer and more effective therapeutic strategy for the treatment of tumours57. Additionally, the clinical trials outlined in Table 3 indicate that the combination of PARP inhibitors, including olaparib and niraparib, with agents that target DNA repair mechanisms (e.g., ATR inhibitors), angiogenesis (e.g., VEGFR inhibitors), signalling pathways (e.g., MEK and PI3K inhibitors), or epigenetic factors (e.g., HDAC inhibitors) can significantly reduce the resistance of various cancers, such as ovarian, prostate, and breast cancer, to PARP inhibitors58. Of particular interest are the PARP dual-target inhibitors, exemplified by E7449 (10)59, JPI-547 (11)60, and AMXI-5001 (14)61, which offer distinct advantages. In preclinical studies, these compounds have shown potential in overcoming acquired resistance to conventional PARP inhibitors62. The dual-pathway inhibitory properties of these compounds not only broaden the therapeutic window but also present a novel approach for targeted therapy strategies within the context of precision medicine.
Structurally, PARP inhibitors exhibit common key pharmacophore characteristics. The molecular cores of these inhibitors universally incorporate an amide group that interacts with the catalytic active site of PARP. Typical structural variations include open-chain amides (e.g., niraparib), pyridazinone-modified amides (e.g., olaparib, talazoparib, fuzaparib), and lactam configurations (e.g., rucaparib, pamiparib)63. This amide group, irrespective of its open-chain or cyclic form, is capable of forming a hydrogen-bond network with the conserved amino acid residues Gly863 and Ser904 of PARP. Additionally, the adjacent aromatic ring, through spatial complementarity, facilitates a π-π stacking interaction with tyrosine or histidine residues, collectively ensuring stable binding between the inhibitor and the target protein (Figure 3)64. The catalytic domains of PARP family members, including PARP1 and PARP2, demonstrate significant evolutionary conservation. Specifically, the structure of the NAD+-binding pocket exhibits remarkable similarity across these proteins. Consequently, the PARP inhibitors currently available predominantly function by competitively binding to this conserved site, resulting in a general lack of selectivity among different subtypes. This lack of specificity leads to the concurrent inhibition of both PARP1 and PARP265. Numerous studies have demonstrated a significant correlation between the inhibition of PARP2 and the occurrence of hematological toxicities, such as anaemia and neutropenia, in clinical settings66,67. This relationship is intricately connected to the essential role of PARP2 in maintaining the homeostasis of haematopoietic stem cells68. Therefore, the development of highly selective PARP1 inhibitors is of substantial importance in mitigating the adverse effects associated with PARP2 inhibition. In recent years, the detailed elucidation of the catalytic centre of PARP1 through advanced structural biology techniques, including X-ray crystallography and cryo-electron microscopy, has unveiled significant conformational differences between PARP1 and PARP2. These structural distinctions provide essential targets for the rational design of selective inhibitors. For instance, the selectivity of AZD5305 (12), a highly selective PARP1 inhibitor, arises from the hydrogen-bond network established between the piperazine nitrogen and the His862 residue in PARP1, mediated by a conserved water molecule69. Compared to first-generation PARP inhibitors, AZD5305 demonstrates improved tolerability and clinical efficacy. Significantly, it offers substantial benefits in treating patients with advanced breast, ovarian, prostate, and pancreatic cancers who possess mutations in the BRCA1/2, PALB2, or RAD51C/D genes. Furthermore, AZD9574, a highly selective PARP1 inhibitor developed by AstraZeneca, exhibits significant capability for penetrating the blood-brain barrier70. This compound demonstrates a selectivity for PARP1 over PARP2, PARP3, PARP5a, and PARP6 subtypes that exceeds 8,000-fold, highlighting its exceptional subtype specificity. Preclinical studies show that AZD9574 has strong anti-tumour effects in brain tumour models, highlighting its potential for treating central nervous system cancers like brain metastases and glioblastoma. Presently, the PARP7 selective inhibitor RBN-2397 has been comprehensively identified and assessed in both preclinical and clinical studies. This compound demonstrates highly potent inhibitory activity against PARP7, with an IC50 value of less than 3 nM, and exhibits over 50-fold selectivity for PARP7 relative to other members of the PARP family71. The analysis of the co-crystal structure of RBN-2397 in complex with PARP7 (PDB: 6V3W) demonstrated that the pyridazine moiety of RBN-2397, functioning as an NAD+ analog, occupies the NAD+ binding pocket, a critical interaction for sustaining its inhibitory efficacy against PARP (Figure 3). Concurrently, the 4–(5-(trifluoromethyl)pyrimidin-2-yl)piperazine segment is positioned within the adenosine pocket, which is contiguous to the highly flexible D-loop. The D-loop is known to participate in substrate recognition across all PARP family proteins and exhibits considerable conformational flexibility, thus enabling the accommodation of structurally diverse ligands. The ethoxyethane moiety in RBN-2397 functions as a linking unit, connecting two distinct pharmacophores that interact with the NAD subpocket and the adenosine subpocket, respectively. Subsequent investigations have demonstrated that this compound selectively inhibits PARP7 by modulating the conformation of its flexible ethoxyethane linker and the piperazine component, thereby accommodating the spatial characteristics of the PARP1 binding pocket72.
At present, a central focus of research in PARP-targeted therapy involves the development of inhibitors that selectively target specific PARP subtypes73. These inhibitors not only augment the efficacy of existing cancer therapies by mitigating side effects and enhancing synthetic lethality but also advance the understanding of the functional mechanisms within the PARP family, thereby facilitating biomarker-driven precision treatment strategies. Furthermore, they offer efficient and highly selective tool compounds for the subsequent formulation of precise combination treatment regimens74.

Synergistic effects of PARP inhibitors and other antitumor agents
Recent research on DNA damage repair has expanded the clinical use of PARP inhibitors from maintenance therapy to more comprehensive treatment strategies75. These strategies now include combination therapies with first-line chemotherapeutic agents, molecular targeted therapies, epigenetic drugs and immune checkpoint inhibitors. As shown in Figure 4, this multidimensional strategy not only markedly augments anti-tumour efficacy via the mechanism of synthetic lethality but also extends its therapeutic potential across a range of solid tumours, including ovarian, breast, and prostate cancers76. This is accomplished through various mechanisms, such as epigenetic regulation and the remodelling of the tumour microenvironment, thus offering a novel treatment paradigm for tackling tumour heterogeneity and drug resistance. In this section, we provide a systematic review of the documented mechanisms underlying the synergistic interactions between PARP inhibitors and other anti-tumour agents.

The synergistic effect of PARP inhibitors and chemotherapeutic drugs
Traditional chemotherapeutic agents exert their cytotoxic effects by directly inducing inter-strand or intra-strand DNA cross-links, or by indirectly disrupting DNA replication and repair processes77. PARP inhibitors, as molecularly targeted drugs that act on the DNA damage repair pathway, can achieve a synergistic lethal effect when combined with chemotherapeutic agents through the dual disruption of genomic stability78. Preclinical studies show that combining PARP inhibitors with alkylating agents like temozolomide or platinum-based compounds like carboplatin increases DNA damage by blocking the BER pathway, leading to a synthetic lethal effect79. This research is transitioning from basic to clinical validation, with many Phase I/II trials underway to assess pharmacokinetics and dose-limiting toxicities80.
DNA topoisomerases, which are ubiquitously present nuclear enzymes, play an essential role in modulating DNA entanglement and supercoiling during key cellular processes such as replication, transcription, recombination, and DNA repair81. Based on their functional mechanisms, DNA topoisomerases are classified into two distinct categories: topoisomerase I (Topo I) and topoisomerase II (Topo II). These enzymes facilitate the cleavage and religation of DNA strands during replication, transcription, and repair processes. Inhibitors of topoisomerases disrupt this mechanism, leading to irreparable single-strand or double-strand DNA breaks82. The accumulation of such DNA damage subsequently triggers cell cycle arrest and induces the initiation of programmed cell death, also known as apoptosis, in aberrant cells82. In the combinatorial strategy employing Topo I/II and PARP inhibitors, the latter impede the repair of DNA single-strand breaks induced by Topo I/II inhibitors, resulting in replication fork collapse and the formation of irreparable double-strand breaks. This synergistic interaction, which amplifies replication stress, has demonstrated potential in phase I clinical trials for the treatment of acute myeloid leukaemia (AML)83 and advanced solid tumours84. For instance, the combination of veliparib and topotecan achieved a 32% objective response rate in AML, while veliparib and irinotecan demonstrated manageable safety in advanced colorectal cancer. Ongoing studies (NCT05101551, NCT01012817) are exploring the best dosage for treating acute myeloid leukaemia with talazoparib/topotecan and solid tumours with veliparib/topotecan85,86. Additionally, research shows that doxorubicin works synergistically with PARP inhibitors to cause oxidative DNA damage, providing a basis for new combination therapies87.
Microtubules, composed of α/β-tubulin heterodimers, are key components of the cytoskeleton and play crucial roles in cell shape, intracellular transport, and mitotic spindle formation88. In cancer cells, their dynamics are often altered to aid invasion, metastasis, and proliferation. Microtubule-targeting agents disrupt their polymerisation-depolymerization cycle, activating the spindle assembly checkpoint and causing mitotic catastrophe89. Research shows that microtubule inhibitors increase genomic instability, enhancing the HRD phenotype and boosting the effectiveness of PARP inhibitors against cancer cells. Meanwhile, PARP inhibitors, by blocking DNA repair and worsening replication fork collapse, work synergistically with the chromosomal errors caused by microtubule inhibitors, leading to more effective cancer cell suppression90. Clinical investigations have established that an open-label, multi-centre Phase I/II clinical trial conducted on patients with metastatic TNBC, who had previously undergone anthracycline/taxane treatment, demonstrated a significant improvement in median progression-free survival (mPFS), objective response rate (ORR), and overall survival (OS) with the regimen combining eribulin, a microtubule dynamics inhibitor, and olaparib, compared to eribulin monotherapy. These findings substantiate the efficacy of this combinatorial therapeutic approach91.

The synergistic effect of PARP inhibitors and RTK inhibitors
Receptor tyrosine kinases (RTKs) are essential transmembrane proteins that play a significant role in tumorigenesis and tumour progression92. Under pathological conditions, RTKs can perpetually activate signalling pathways through mechanisms such as activating mutations (e.g., EGFRL858R), gene amplification, aberrant autocrine loops, or persistent downstream signalling93. These processes endow tumour cells with malignant characteristics, including unregulated proliferation, resistance to apoptosis, angiogenesis, and metastatic potential. For example, the amplification or mutation of the epidermal growth factor receptor (EGFR) is a critical driver in the pathogenesis of NSCLC and squamous cell carcinoma of the head and neck94. Simultaneously, the activation of the vascular endothelial growth factor receptor (VEGFR) signalling pathway is closely linked to angiogenesis and metastasis in breast cancer95. Due to the significant pathological association between RTKs and malignant neoplasms, these proteins have become primary molecular targets in modern targeted drug research and development. An in-depth analysis of their mechanisms of action has established a crucial molecular basis for the development of novel and precise anti-tumour therapies.
From a molecular mechanistic standpoint, EGFR inhibitors have the potential to disrupt the DNA damage repair capabilities of tumour cells by specifically targeting and inhibiting the EGFR signalling pathway96. This action may enhance the synthetic lethality of PARP inhibitors in tumours with HRD. Concurrently, the persistent DNA single-strand break damage caused by PARP inhibitors can lead to feedback activation of the EGFR-mediated survival signalling pathway during the process of replication fork collapse. EGFR inhibitors are capable of effectively obstructing this compensatory signalling, thereby creating a cascade effect that results in synergistic inhibition across both pathways97. In ovarian cancer research, preclinical data show that combining the EGFR inhibitor lapatinib with the PARP inhibitor niraparib exhibits strong synergistic anti-tumour effects in vitro and in vivo98. This treatment effectively halts tumour progression in animal models of platinum-resistant ovarian cancer by simultaneously targeting DNA repair and growth factor signalling pathways. Nevertheless, the GOAL study, a phase II trial on advanced lung adenocarcinoma with EGFR mutations, assessed the combination of EGFR inhibitor and the PARP inhibitor olaparib99. Despite theoretical synergy, the clinical results fell short, and treatment-related adverse events, especially severe hematological and gastrointestinal toxicities, were more frequent than with monotherapy. This indicates varied tumour microenvironment responses to the combined treatment100.
At the molecular regulatory level, VEGFR inhibitors have been shown to induce hypoxia within the tumour microenvironment by inhibiting tumour angiogenesis101. This hypoxic condition subsequently results in the downregulation of BRCA1/2 and RAD51 expression levels, which are critical regulatory proteins in the HR repair pathway. This epigenetic modulation can render tumour cells susceptible to “synthetic lethality,” mimicking the phenotype observed in BRCA mutations. As a result, it augments the selective cytotoxic efficacy of PARP inhibitors in tumours exhibiting homologous recombination repair deficiencies102. Pre-clinical data shows that combining the VEGFR inhibitor cediranib with the PARP inhibitor olaparib effectively overcomes resistance to PARP inhibitors in BRCA wild-type ovarian cancer models103. This synergy occurs independently of BRCA mutation status, supporting its potential clinical application. The phase II EVOLVE study (NCT02681237) analysed 34 patients with high-grade serous ovarian cancer who had previously failed PARP inhibitor treatment104. Patients were divided into three groups based on treatment history: post-platinum-sensitive relapse (PS, n = 11), platinum-resistant relapse (PR, n = 10), and advanced-stage re-chemotherapy (PE, n = 13). BRCA mutation rates were 81.8% in the PS group, 80.0% in the PR group, and 53.8% in the PE group. The combination treatment resulted in four partial responses (two in the PR group and two in the PE group) and 18 cases of disease stability. The 12-month overall survival rates were 81.8% for the PS group, 64.8% for the PR group, and 39.1% for the PE group, indicating the combined treatment’s feasibility for patients resistant to PARP inhibitors. A phase II study (NCT02354131) found that combining niraparib with bevacizumab significantly improved mPFS in platinum-sensitive recurrent ovarian cancer compared to niraparib alone (11.9 vs. 5.5 months; Hazard ratio = 0.32)105. A phase I study (NCT01116648) confirmed the safety of combining cediranib and olaparib106. Ongoing large trials, such as phase II/III (NCT02502266), are assessing long-term efficacy in recurrent ovarian cancer107. Another phase II study (NCT03117933) is comparing olaparib alone, olaparib with cediranib, and paclitaxel in platinum-resistant breast cancer108. These studies aim to clarify the significance of combining vascular-targeted therapy with PARP inhibition in treating solid tumours.
MET, an RTK family member, binds HGF to form a dimer, activating its tyrosine kinase domain and triggering RAS-MAPK, PI3K-AKT, and STAT3 pathways. MET dysfunction, including exon 14 skipping mutations, gene amplification, overexpression, or fusion genes, is linked to the malignancy and spread of various solid tumours109. Research has demonstrated that MET contributes to resistance against PARP inhibitors through two principal mechanisms. Firstly, MET phosphorylates PARP1 at the Tyr907 residue, thereby diminishing the binding efficacy of PARP inhibitors and attenuating their therapeutic impact on BRCA1-mutant tumours110. Secondly, MET is integral to the DSB repair network, particularly by modulating the BRCA2/Rad51 complex involved in homologous recombination111. The application of MET kinase inhibitors can mitigate this resistance by disrupting MET’s function in homologous recombination repair, leading to a reduction in Rad51 focus formation and compromising double-strand break repair. This resistance mechanism, which involves both MET signalling and DNA repair pathways, is not confined to breast cancer alone. Clinical studies have indicated that the concurrent inhibition of PARP1 and MET effectively suppresses tumour growth in melanoma and gastric cancer by arresting cell cycle progression and enhancing the response to replication stress112. These findings suggest that this combined therapeutic strategy may be applicable to a broad spectrum of tumour types.

The synergistic effect of PARP inhibitors and PI3K inhibitors
Phosphatidylinositol 3-kinase (PI3K) functions as a pivotal signalling node that regulates a myriad of cellular processes, including cell proliferation, invasion, metastasis, and metabolic reprogramming113. It facilitates the malignant progression of cancers through the PI3K/AKT/mTOR signalling pathway. Foundational research shows that in preclinical glioblastoma models, combining the PARP inhibitor rucaparib with the PI3K inhibitor buparlisib yields significant synergistic anti-tumour effects114. This is due to the interplay between PI3K pathway-induced metabolic stress and PARP-mediated DNA repair disruption. In triple-negative breast cancer (TNBC) treatment, combining buparlisib with olaparib enhances the effectiveness of PARP inhibitors in BRCA wild-type patients by leveraging the interaction between the PI3K pathway and PARP-related DNA repair115. Further investigations have demonstrated that, in both in vitro SCLC cell lines and in vivo xenograft models, the synergistic mechanism of action between buparlisib and talazoparib functions through the dual inhibition of PI3K signalling and PARP enzymatic activity. This dual inhibitory effect significantly suppresses tumour cell proliferation and induces apoptosis, which is associated with the accumulation of DNA damage116. A Phase I dose-escalation trial (NCT01623349) validated clinical research findings by evaluating the maximum tolerated dose and preliminary efficacy of buparlisib and olaparib in recurrent breast or ovarian cancer patients117. Among 59 evaluable patients, the ovarian cancer subgroup had a 29% objective response rate (5/17), and the breast cancer subgroup had a 28% rate (5/18). Both groups had a disease stabilisation rate of about 50%, with an overall disease control rate (DCR) of 76% (45/59). Additionally, a phase III clinical trial (NCT04729387) has begun to assess the efficacy and safety of combining the PARP inhibitor olaparib with the PI3K inhibitor alpelisib in patients with platinum-resistant or refractory high-grade serous ovarian cancer lacking germline BRCA mutations. The primary goal is to measure PFS, with an estimated completion date of June 2025. These results demonstrate the regimen’s feasibility and anti-tumour activity, supporting future pivotal trials.

The synergistic effect of PARP inhibitors and CDK inhibitors
From the perspective of the interaction mechanism between cell cycle regulation and DNA damage repair, the cyclin-dependent kinase 4/6 (CDK4/6) inhibitor palbociclib specifically targets the phosphorylation activity of the CDK4/6-cyclin D complex118. This inhibition results in a blockade at the G1/S phase cell cycle checkpoint, thereby obstructing the progression of the cell cycle in tumour cells and inducing senescence-associated programmed cell death. Based on this mechanism of action, palbociclib has received approval from the FDA for the clinical management of hormone receptor-positive/HER2-negative advanced BC119. Recent studies have shown that in TNBC models resistant to the PARP inhibitor olaparib, the Wnt/β-catenin pathway’s abnormal activation leads to drug resistance by regulating EMT and maintaining stem cell traits. However, combining palbociclib with olaparib can effectively overcome this resistance and improve treatment outcomes by inhibiting cell cycle progression and exploiting HRD-related synthetic lethality120. Similar synergistic effects have been observed in prostate cancer models, involving interactions between CDK4/6-Rb-E2F axis inhibition and PARP-mediated replication stress121. Research has shown that in bladder cancer models with BRCA1/2 mutations, combining the PARP inhibitor talazoparib with palbociclib enhances tumour cell death by inducing G1-phase arrest and accumulating DNA double-strand breaks122. Additionally, studies on TNBC suggest that CDK12 inhibitors can mimic BRCA mutations, making tumour cells more sensitive to PARP inhibitors by reducing the transcription of key NHEJ regulators123. These findings indicate that targeting both cell cycle regulation and DNA repair pathways could be effective for a wide range of tumours.

The synergistic effect of PARP inhibitors and epigenetic drugs
Epigenetic dysregulation is acknowledged as a pivotal factor in the initiation and persistence of tumours, influencing the responsiveness to PARP inhibitors. This understanding suggests a potential strategy for overcoming acquired treatment resistance through the combined application of epigenetic drugs and PARP inhibitors124. BRD4 (bromodomain-containing protein 4), a constituent of the BET (bromodomain and extra-terminal) protein family, is critically involved in the initiation and progression of tumours125. This involvement is mediated through the regulation of key oncogenes such as MYC, CDK, and BCL2. Inhibitors targeting BRD4 demonstrate significant anti-tumour effects by disrupting epigenetic regulatory pathways126. Research has shown that the BET inhibitor JQ-1 increases γH2AX expression and suppresses DNA repair proteins Ku80 and Rad51 through BRD4/BRD2 pathways127. In pancreatic cancer xenografts, JQ-1 and olaparib together enhance anti-tumour effects by jointly inhibiting these repair proteins, a synergy also confirmed in castration-resistant prostate cancer models128. In BRCA wild-type ovarian cancer cell lines, JQ-1 significantly disrupts the regulation of the G2/M checkpoint by downregulating the expression of WEE1 and topoisomerase II binding protein (TOPBP1). When used in conjunction with olaparib, this disruption forces tumour cells with DNA damage to prematurely enter mitosis, leading to catastrophic mitosis and subsequent cell death129. This phenomenon has been substantiated through in vivo experiments. Additionally, studies on SCLC models have demonstrated that the synergistic effect between BET inhibitors and PARP inhibitors may be mediated through the targeted modulation of the MYC/PARP1 signalling axis130.
Histone deacetylases (HDACs) facilitate chromatin condensation by removing acetyl groups from histones, silencing gene transcription. Abnormal expression of HDACs is linked to tumour development. HDAC inhibitors counteract this by promoting histone hyperacetylation, activating tumour suppressor genes, and repressing proto-oncogenes131. They also down-regulate key HR repair pathway genes, inducing a “BRCAness” phenotype to boost the effectiveness of PARP inhibitors on HRD tumours132. Mechanistic investigations have demonstrated that the concomitant use of HDAC inhibitors and PARP inhibitors elicits a synergistic anti-tumour response through the dual modulation of genomic stability and cell cycle checkpoints133. Specifically, HDAC inhibitors induce cell cycle arrest at the G1/S or G2/M phases by disrupting the equilibrium of histone modifications. In contrast, PARP inhibitors lead to replication fork collapse and the accumulation of double-strand breaks by impeding the repair of single-strand DNA breaks. This synergistic interaction effectively circumvents the compensatory mechanisms of cell cycle checkpoints and activates mitochondrial-dependent apoptotic pathways134. Furthermore, this combinatorial approach stimulates the cGAS-STING innate immune signalling pathway by inducing the accumulation of cytoplasmic DNA, thereby promoting the secretion of type I interferons and pro-inflammatory chemokines, such as CXC-chemokine ligand 10 (CXCL10) and chemokine (C-C motif) ligand 5 (CCL5)135. This process reshapes the tumour immune microenvironment and enhances anti-tumour immune responses.
Enhancer of zeste homolog 2 (EZH2), serving as the principal catalytic subunit of the Polycomb repressive complex 2 (PRC2), facilitates the malignant progression of tumours through the mediation of histone H3 lysine 27 trimethylation (H3K27me3), thereby epigenetically silencing tumour suppressor genes, including CDKN2A and PTEN136. CARM1 (coactivator-associated arginine methyltransferase 1) facilitates the establishment of a compact chromatin structure within the promoter region of the Mad2l2 gene, thereby improving the recruitment efficiency of EZH2 through the catalysis of arginine methylation modifications on the BAF155 subunit of the SWI/SNF chromatin remodelling complex137. This process inhibits the expression of critical regulatory factors involved in DNA end resection via an H3K27me3-mediated transcriptional silencing mechanism, ultimately resulting in the aberrant activation of the NHEJ pathway and chromosomal segregation errors during DNA double-strand break repair. Furthermore, this epigenetic regulatory effect significantly increases the sensitivity of PARP inhibitors in ovarian cancer cells characterised by normal HR function but elevated CARM1 expression levels138. Recent research demonstrates that EZH2 plays a dynamic role in modulating the equilibrium between the DDR and the transcriptional regulatory functions of PARP1 through the direct catalysis of PARP1 protein methylation139. This regulation effectively mitigates metabolic stress and genomic instability resulting from excessive PARP1 activation. Subsequent investigations in prostate cancer models reveal that the concurrent administration of EZH2 inhibitors and PARP inhibitors produces a pronounced synergistic anti-tumour effect140.

The synergistic effect of PARP inhibitors and TNKS1/2 inhibitors
Telomeres and tankyrase 1/2 (TNKS1/2) contribute to the stabilisation of β-catenin and the activation of the Wnt/β-catenin signalling pathway through the catalysis of poly(ADP-ribose) modification, which subsequently promotes the ubiquitination and degradation of the Axin scaffold protein. Furthermore, TNKS1/2 plays a critical role in the assembly of poly(ADP-ribose) chain-mediated repair complexes during the DDR141. TNKS1/2 inhibitors can markedly augment the anti-tumour efficacy of PARP inhibitors through two primary mechanisms: first, by downregulating Wnt signalling, they diminish the repair compensation capacity of BRCA wild-type tumours, thereby inducing a “BRCAness” phenotype142; second, by disrupting the alternative lengthening of telomeres mechanism, they further exacerbate genomic instability143. The novel dual-target inhibitor stenoparib (10), in conjunction with JPI-547 (11), concurrently inhibits the enzymatic activities of PARP and TNKS1/2. This dual inhibition synergistically impedes the repair of single-strand DNA breaks and disrupts the maintenance of telomere dynamic equilibrium, ultimately leading to a synergistic lethal effect characterised by replication fork collapse and telomere dysfunction144. This combined therapeutic strategy has successfully undergone initial efficacy and safety evaluations in Phase I/II clinical trials for solid tumours, including ovarian and breast cancer145.

The synergistic effect of PARP inhibitors and NAMPT inhibitors
Nicotinamide phosphoribosyltransferase (NAMPT), recognised as the primary rate-limiting enzyme in the salvage synthesis pathway of NAD+, has garnered significant interest in cancer treatment research, particularly concerning its inhibitors146. Recent studies have revealed a significant synergistic interaction between PARP inhibitors and NAMPT inhibitors. This interaction is largely due to the capacity of NAMPT inhibitors, such as FK866, to markedly reduce intracellular NAD+ levels by inhibiting the rate-limiting step in NAD+ biosynthesis147. NAD+ is a crucial substrate for the DNA repair activities of PARP1/2; consequently, its depletion leads to the suppression of PARP1/2 activity. In conjunction with the direct inhibitory effects of PARP inhibitors on repair mechanisms, this combined approach leads to a persistent accumulation of DNA damage and a further compromise of the repair capacity148. Preclinical studies have indicated that this combinatorial strategy effectively depletes NAD+ precursors, including NMN and NAD+, reduces PARP activity, and increases DNA damage, ultimately resulting in significant inhibition of tumour growth, reduction of tumour volume, delay of disease progression, and a marked extension of survival in tumour-bearing murine models149.

The synergistic effect of PARP inhibitors and proteasome inhibitors
As the principal executor of approximately 80% of protein degradation in eukaryotic cells, the proteasome plays a critical role in regulating essential biological processes, including the cell cycle, signal transduction, and stress responses, through the ubiquitin-proteasome pathway, thus ensuring the maintenance of protein homeostasis within the cell150. Proteasome inhibitors, exemplified by bortezomib, exhibit extensive anti-tumour properties by disrupting the regulatory protein degradation pathway151. These inhibitors demonstrate a dual synergistic interaction with PARP inhibitors. Firstly, the inhibition of proteasome activity results in the accumulation of misfolded proteins, which subsequently impacts the function and expression of proteins involved in DNA damage repair, thereby augmenting the inhibitory effects of PARP inhibitors on the DNA repair pathway152. Secondly, these agents collaboratively enhance the anti-tumour immune response by modulating the tumour microenvironment153. For instance, PARP inhibitors facilitate the generation of tumour neoantigens and upregulate the expression of interferons and PD-L1, while proteasome inhibitors influence immune cell activity and cytokine secretion. Collectively, these mechanisms culminate in a synergistic anti-tumour effect. The combination therapy involving PARP1 inhibitors and proteasome inhibitors has been established as a synergistic treatment for multiple myeloma, with the associated Phase I clinical trial (NCT01495351) having been completed154.

The synergistic effect of PARP inhibitors and DDR inhibitors
The strategic combination of DDR inhibitors with PARP inhibitors represents a promising strategy for addressing resistance to PARP inhibitors and improving therapeutic outcomes155. PARP and ataxia telangiectasia and Rad3-related protein (ATR) serve as pivotal regulatory components within the DNA damage repair pathway. The strategic combination of their inhibitors has emerged as a significant focus in oncological therapeutics156. This synergistic interaction arises from the ability of PARP inhibitors to induce replication fork collapse by impeding SSB repair. Concurrently, ATR, a critical kinase involved in sensing replication stress and activating the S/G2-M checkpoint, is inhibited, thereby disrupting DSB repair and cell cycle arrest mechanisms. The concomitant application of these inhibitors results in the accumulation of lethal DNA damage157. Furthermore, ATR inhibitors have the potential to counteract resistance to PARP inhibitors and re-establish the sensitivity of tumour cells to these agents by inhibiting the activation of compensatory DNA repair pathways in drug-resistant cells158. Current clinical studies have demonstrated that the combination of ATR inhibitors and PARP inhibitors exhibits clinical efficacy in the treatment of various solid tumours159.
WEE1 plays a critical role in suppressing CDKs and activating cell cycle checkpoints, thereby facilitating DNA damage repair160. Consequently, the inhibition of WEE1, which can lead to cell cycle arrest, apoptosis, and transcriptional repression, emerges as a promising combinatorial therapeutic strategy. Recent research has investigated the mechanisms through which WEE1 inhibition overcomes resistance to PARP inhibitors161. Specifically, targeting WEE1 results in the downregulation of ribonucleotide reductase regulatory subunit M2 (RRM2), an essential enzyme in deoxyribonucleoside triphosphate (dNTP) synthesis. This downregulation induces replication fork stalling and triggers a replication stress response, thereby delaying the development of acquired resistance to PARP inhibitor in patient-derived xenograft (PDX) models harbouring BRCA1 mutations162. Furthermore, WEE1 inhibition sensitises BRCA1/2 wild-type TNBC to PARP inhibitors by enhancing DNA damage, promoting apoptosis, inducing replication stress, and activating the stimulator of interferon genes (STING) pathway163.
DNA polymerase θ (Polθ) plays a pivotal role in theta-mediated end joining (TMEJ), a process essential for maintaining genomic stability in tumours deficient in HR by facilitating the repair of DSBs. Consequently, Polθ represents a promising therapeutic target164. The small-molecule inhibitor ART558 allosterically inhibits Polθ, selectively blocking TMEJ without affecting NHEJ165. In HR-deficient tumours, this induces a synthetic lethal effect, enhancing the anti-tumour efficacy of PARP inhibitors. Research shows that when DNA damage occurs, PARP1 is recruited to the site to synthesise poly(ADP-ribose) chains, aiding in the recruitment of repair factors like MRE11 and Polθ, which activate the TMEJ repair pathway166. PARP inhibitors block this process, leading to improper repair of double-strand breaks and cell death. In tumours resistant to PARP inhibitors due to lack of 53BP1, genomic stability depends on Polθ overexpression167. Therefore, inhibiting both PARP and Polθ could overcome resistance, offering a targeted treatment for tumours with HR deficiencies.
Beyond the previously mentioned DDR inhibitors, other DDR inhibitors are promising for combination cancer therapies. For example, the AXL inhibitor bemcentinib increases hepatocellular carcinoma (HCC) sensitivity to olaparib by blocking RPA2/CHK1-mediated HR repair, suggesting AXL as a prognostic biomarker for HCC168. Additionally, RAD51 is crucial for PARP inhibitor sensitivity and is linked to BRCA1/2. Silencing RAD51C enhances olaparib sensitivity, and homozygous methylation of RAD51C is a positive predictive biomarker for PARP inhibitor sensitivity169. Furthermore, ATM, a critical protein in the DDR pathway, when deficient, can heighten cellular sensitivity to the combined treatment of veliparib and irinotecan170. In gastric cancer models characterised by ATM involvement, this combination exhibits a synergistic effect and holds potential for overcoming drug resistance171. Concurrent studies have indicated that ALK promotes HR repair via the phosphorylation of CDK9, thereby contributing to resistance against PARP inhibitors172. Inhibition of ALK can restore the efficacy of PARP inhibitors, suggesting that ALK may serve as a biomarker for predicting resistance to these inhibitors. The concurrent use of ALK inhibitors and PARP inhibitors appears promising as a strategy to improve treatment outcomes.
Studies show that small molecules linked to DDR can predict the effectiveness of PARP inhibitor therapy53. BRCA1 expression is the most researched marker, while the lack of BRCA1 methylation and BRCA1/2 biallelic loss may indicate resistance to these inhibitors173. Other molecules like GLI1174, LIG3175, PPP2R2A176, ERCC1177, and MED12178 might also have prognostic or predictive value for PARP inhibitor treatment.

The synergistic effect of PARP inhibitors and Hsp90 inhibitors
Heat shock protein 90 (HSP90), functioning as an ATP-dependent molecular chaperone, plays a critical role in maintaining the integrity of the HR repair pathway and in the activation of cell cycle checkpoints179. This is achieved through the formation of dynamic complexes with essential DNA repair proteins, including BRCA1, BRCA2, CHK1, RAD51, and MRE11. HSP90 inhibitors promote the degradation of client proteins by obstructing the ATP binding pocket, while concurrently inhibiting the functions of critical components involved in HR repair, such as BRCA1/2 and Rad51, as well as the kinase activity associated with the NHEJ pathway. This dual inhibition results in deficiencies in the repair of DNA double-strand breaks, thereby establishing a synthetic lethal interaction with the single-strand break repair deficiency induced by PARP inhibitors180. Preclinical studies have demonstrated that in non-BRCA mutant ovarian cancer cell models, the HSP90 inhibitor ganetespib significantly enhances the antiproliferative effects of the PARP inhibitor talazoparib through a dose- and time-dependent downregulation of DNA repair proteins, including CHK1 and RAD51141. The mechanism underlying their synergistic effect is directly associated with the extent of inhibition of the compensatory capacity of the DNA repair pathway.

The synergistic effect of PARP inhibitors and immune checkpoint inhibitors
The synergy between PARP inhibitors and immune checkpoint inhibitors is based on interconnected immune regulatory mechanisms181. As shown in Figure 5, this combinatorial therapy involves enhanced immunogenicity due to genomic instability182, activation of the cGAS-STING-type I interferon pathway183, targeting programmed cell death ligand 1 (PD-L1) immune checkpoints184, and remodelling the tumour microenvironment185. Specifically, PARP inhibitors induce DNA damage that increases genomic instability and tumour mutational burden (TMB), a key indicator of neoantigen load that influences the effectiveness of immune checkpoint inhibitors. Higher neoantigen presentation boosts tumour immunogenicity, potentially enhancing sensitivity to immunotherapy. Additionally, PARP inhibitors can lead to cytoplasmic double-stranded DNA accumulation, activating the cGAS-STING pathway and resulting in increased type I interferon production. This promotes various immune-stimulatory effects, including dendritic cell maturation and antigen presentation, improved T cell migration and function, and reduced regulatory T cell activity, all of which enhance the anti-tumour immune response. Preclinical investigations have demonstrated that the combination of PARP inhibitors and STING agonists can surmount the adaptive resistance to olaparib, which is attributed to tumour microenvironment (TME) dependence, by modulating the TME and reprogramming myeloid cells186. Presently, several clinical trials are underway to evaluate STING agonists, including a phase I trial (NCT04144140) assessing the safety of E7766 as a monotherapy in patients with advanced solid tumours or lymphomas187, and a phase I/II trial (NCT04020185) examining the efficacy of IMSA101 as a monotherapy or in conjunction with immune checkpoint inhibitors188. Nevertheless, the combination therapy of STING agonists and PARP inhibitors has not yet progressed to the clinical research phase, and its potential efficacy requires further validation. Notably, PARP inhibitors can significantly alter the tumour immune microenvironment by shifting from chronic to acute inflammation155. This change promotes an anti-tumour environment with increased CD8+ T cell infiltration, reduced myeloid-derived suppressor cells (MDSCs), and altered macrophage phenotypes189. The interplay between DNA damage response and immune regulation underpins the combined use of PARP inhibitors and immune checkpoint inhibitors, potentially overcoming tumour immune escape and reactivating the body’s anti-tumour response.
PD-1 and its ligand, PD-L1, continuously transmit inhibitory signals, leading to the functional exhaustion of T cells. The synergistic anti-tumour effects observed with the combination of niraparib and PD-1 inhibitors are independent of BRCA gene mutation status190. In models of BRCA1-deficient ovarian cancer, olaparib has been shown to induce the secretion of type I interferon through the activation of the cGAS-STING pathway, subsequently leading to the upregulation of PD-L1 expression in tumour cells and the establishment of adaptive immune resistance. Nevertheless, the administration of a PD-1 inhibitor can counteract PD-L1-mediated immune suppression, thereby significantly augmenting the magnitude of anti-tumour immune responses. Experimental data indicate that the group receiving combination therapy exhibited a superior rate of tumour growth inhibition and a significantly extended median survival time in tumour-bearing mice, in comparison to the group receiving single-drug therapy191. The Phase III clinical trials currently in progress, ATHENA (NCT03522246)192 and FIRST (NCT03602859)193, are structured to assess the efficacy of combining PARP inhibitors with immune checkpoint inhibitors in patients newly diagnosed with advanced ovarian cancer. The FIRST and ATHENA trials are Phase III studies evaluating combination therapies for advanced ovarian cancer. FIRST tests rucaparib with nivolumab in 1402 patients, while ATHENA examines niraparib with dostarlimab in 1097 patients. Both trials focus on patients who responded to initial platinum-based chemotherapy, comparing progression-free survival (PFS) of the combination therapies against PARP inhibitors alone or placebo. Results in 2024 may influence first-line treatment strategies and enhance patient outcomes. In a phase III trial (NCT03737643) for newly diagnosed ovarian cancer, researchers evaluated a triple maintenance therapy of bevacizumab, olaparib, and durvalumab. Results showed that, in patients without BRCA mutations, this combination significantly improved PFS compared to bevacizumab alone, regardless of HRD status, with hazard ratios of 0.49 for HRD-positive and 0.68 for HRD-negative patients. However, overall survival data is still immature, and the study design complicates the assessment of immunotherapy’s independent effects193.

Clinical challenges associated with combination therapy involving PARP inhibitors
Research shows that combining PARP inhibitors in cancer treatment offers significant benefits, including overcoming chemotherapy resistance194. However, PARP inhibitors face three main challenges in combination therapies: a limited therapeutic window due to toxicity buildup, complex pharmacokinetic interactions complicating dose optimisation, and unreliable biomarkers with flawed detection systems, all hindering their clinical application195.
Firstly, The toxicity of combination therapies mainly stems from PARP inhibitors impairing DNA repair in normal tissues and the cumulative effects of other drugs. Myelosuppression is the main dose-limiting toxicity, especially when PARP inhibitors are combined with chemotherapy or other DDR inhibitors, occurring in 38% to 96% of cases196. This is largely due to PARP2’s critical role in the survival of haematopoietic stem cells. Furthermore, different combination regimens show tissue-specific toxicity increases159. Especially, combining with radiotherapy worsens skin and mucosal reactions, while pairing with anti-angiogenic agents raises the risk of vascular issues like hypertension and proteinuria197. Nonetheless, these complications occur less frequently than with chemotherapy combinations. Notably, the differences in toxicity among PARP inhibitors are tied to their PARP-trapping abilities198. Talazoparib, for instance, traps PARP 100 times more effectively than veliparib. Greater PARP-trapping increases toxicity risk, especially with DNA-damaging agents, and this risk is not easily mitigated by adjusting doses.
Secondly, the optimisation of dosage regimens for PARP inhibitors is complicated by intricate pharmacokinetic interactions with other drugs. Especially, when rucaparib is administered concurrently with chemotherapy, its dosage is reduced from 600 mg twice daily to 240 mg daily, which may influence plasma drug concentrations199. Due to toxicity concerns, continuous dosing schedules might need to be adjusted to intermittent ones, potentially compromising therapeutic efficacy200. Moreover, the sequence of drug administration is critical; administering carboplatin prior to olaparib can decrease olaparib’s peak concentration and overall exposure, whereas initiating treatment with PARP inhibitors may increase the risk of DNA damage and toxicity201. Additionally, there are significant variations in pharmacokinetic parameters such as bioavailability and half-life among different PARP inhibitors. For instance, the half-life of niraparib extends up to 36 h, while that of olaparib is only 11.9 h202. These differences further complicate the formulation of combination treatment regimens.
Thirdly, the predictive power of biomarkers like BRCA mutations and HRD for evaluating the effectiveness of PARP inhibitors as standalone treatments diminishes when these inhibitors are combined with other therapies7. In such cases, the success of PARP inhibitors depends on the molecular pathways affected by the additional drugs. For example, combining PARP and ATR inhibitors is effective in patients with ATM mutations, which aren’t usually part of standard HRD assessments203. On the other hand, HRD-positive patients may develop resistance if the ATR pathway is activated204. Current biomarker detection systems are inadequate, as genomic biomarkers like HRD scores only indicate past repair status and don’t capture current functional changes. Although functional assays like the HRD foci assay offer a more accurate view of homologous recombination status, they face challenges in clinical use, such as sample accessibility and standardisation issues205.
To improve the clinical utility of functional assays like the HRD foci assay, systematic optimisation is recommended in four areas206: sample considerations, validation design, detection coverage, and reporting specifications. During sample selection, it’s essential to assess tumour cell content and sample type (fresh vs. frozen). The need for internal reference samples should be evaluated, and all related parameters must be thoroughly documented to ensure reliable and comparable analyses; When choosing a validation design, it’s crucial to include enough samples with negative BRCA1/2 mutations but positive or suspected HRD phenotypes. This ensures a thorough evaluation of the assay’s ability to detect homologous recombination deficiencies, enhancing detection accuracy and sensitivity; Detection coverage and interpretation require precise definitions of the genomic regions being analysed (e.g., loss of heterozygosity, large-scale state transitions, telomeric allelic imbalance) and their identification methods. Comprehensive genomic coverage is crucial for accurately representing the HR functional state. Furthermore, establishing clear and reproducible interpretation criteria provides a standardised framework for analysts, ensuring consistent and reliable results across laboratories; At the reporting stage, key details such as the covered HRR genes, detection method limitations, and HRD status thresholds should be clearly and concisely presented. This improves result interpretability, aiding clinicians and researchers in making informed decisions and enhancing the clinical applicability of the HRD foci assay.
Finally, distinct combination strategies require specific biomarkers: immune therapies target PD-L1 and TMB, while anti-angiogenesis therapies focus on VEGFR2 and others. Without standardised phase III trial criteria, patient selection remains arbitrary. Additionally, combination therapies might activate compensatory pathways like ATR-CHK1-WEE1 or cause mutations that reduce the effectiveness of current biomarkers4. Fortunately, recent advancements in clinical translational research, utilising innovative phenotypic screening platforms, have significantly elucidated the molecular mechanisms underlying drug combination synergy207. This progress provides a robust theoretical framework for the rational design of dual-target PARP inhibitors.

Design approach for dual-target PARP inhibitors
In comparison to single-target drugs and combination therapy strategies, dual-target inhibitors exhibit considerable potential for clinical translation. Their primary advantage resides in the ability to facilitate coordinated regulation of multiple pathways through a single molecule208. This approach preserves the biological characteristics of target complementarity and synergistic effects inherent in combination therapies while effectively mitigating the pharmacokinetic heterogeneity often associated with multi-drug regimens. Specifically, dual-target inhibitors significantly diminish the risk of drug-drug interactions, reduce the cumulative off-target toxicities, and enhance the predictability of the in vivo exposure-response relationship209. Furthermore, by concurrently inhibiting the activation of compensatory signalling pathways, dual-target molecules can postpone the emergence of target-dependent drug resistance mechanisms, streamline the dosing regimen to improve patient adherence, and ultimately achieve a more favourable therapeutic window that balances efficacy and safety210.
The systematic screening and validation of appropriate target combinations serve as a prerequisite for the clinical advancement of dual-target PARP inhibitors. This approach can effectively identify critical targets exhibiting synthetic lethal interactions or regulatory network intersections with the PARP signalling pathway, including mechanisms such as the ATM/ATR-mediated DDR, the HDAC/EZH2-involved epigenetic silencing pathway, and the PI3K/AKT/mTOR pro-survival signalling cascade211. These identifications can be achieved through the integration of bioinformatics analysis and functional genomics screening techniques. Furthermore, the CRISPR-Cas9 gene editing system and RNA interference technology can be employed to rigorously validate the synthetic lethal interactions among targets in in vitro cell models212. This methodological framework establishes a foundational biological basis for the rational design of dual-target PARP inhibitors.
The design of dual-target ligands is intended to preserve the activity of the primary target while simultaneously achieving functional activity at a secondary target. The fundamental principle underlying this approach is to ensure that the essential pharmacophores remain compatible with the binding mode of the native target, while also facilitating the effective recognition of the secondary target213. Presently, the predominant strategies for the development of dual-target therapeutics encompass three primary models: drug repurposing, pharmacophore integration, and computer-aided design214. The drug repurposing strategy primarily entails a systematic assessment of the off-target effects of approved PARP inhibitors, such as niraparib, on non-target kinase groups, as well as the structural modification of established secondary target inhibitors, such as the c-Met inhibitor cabozantinib, to incorporate PARP inhibitory activity215. This approach has the potential to significantly reduce the research and development timeline for lead compounds.
The dual-target PARP inhibitors currently being reported are predominantly developed using the pharmacophore integration strategy. This approach involves the spatial optimisation of key pharmacophores from two target-selective inhibitors, achieved through the use of a strategically designed connecting bridge, such as short alkyl chains or aromatic fragments216. The objective is to maintain the molecular weight within a drug-like range (typically less than 500 Da) while enhancing the binding efficiency to both targets. Notable design strategies include the structural integration of characteristic units from PARP inhibitors, such as the phthalazinone ring found in olaparib or the benzimidazole skeleton present in valiparib, with essential binding groups from secondary targets, such as the zinc ion-chelating hydroxamic acid group from HDAC inhibitors or the quinazolinamine structure from EGFR inhibitors217. Similarly, the pharmacophore merging strategy reconstructs the molecular framework by optimising the superimposition of pharmacophores, thereby effectively reducing molecular weight and enhancing physicochemical properties. However, due to the sensitivity of the pharmacophore’s three-dimensional conformation, it is essential to accurately identify and fix the shared binding features of PARP and secondary targets through molecular docking and receptor pharmacophore mapping analysis218. Despite this, current strategies are constrained by the limited scope of existing small molecule databases, posing challenges for the expansion of novel molecular frameworks. In contrast, the utilisation of computational chemistry methodologies, such as molecular docking, pharmacophore matching, and binding cavity similarity analysis, among others, has surmounted this limitation219. Through virtual screening, these techniques have facilitated the discovery of multi-target-adaptable molecules characterised by novel scaffolds. Furthermore, the integration of molecular dynamics simulations has optimised the ligand-receptor interaction energy. Consequently, this approach has successfully identified several innovative molecular entities exhibiting PARP inhibitory activity and synergistic effects in tumour treatment220.

Dual inhibitors targeting PARP and other tumour-associated proteins

Dual PARP-Topo inhibitors
Topo inhibitors cause DNA strand breaks by stabilising the DNA-Topo complex, while PARP1 aids in DNA repair and replication fork restart by destabilising this complex, leading to resistance. Using PARP inhibitors alongside can block this repair, increasing DNA damage and enhancing anti-cancer effects221. Developing a single molecule to inhibit both PARP and Topo could overcome combination therapy challenges and achieve stronger anti-tumour effects.
In 2017, Jiang’s research team employed a pharmacophore combination strategy by integrating the essential pharmacophores of veliparib (8) and Topo I/II inhibitors (18) into a unified molecular framework. This approach facilitated the design and synthesis of 14 novel 4-amidobenzimidazole acridine derivatives222. SAR studies have demonstrated that the position of the benzene ring connection significantly influences biological activity, with the following hierarchy observed: para > meta > ortho. Notably, compound 19 (Figure 6), which features a meta connection with 4-methylacridine, displays the highest inhibitory activity against PARP1 (IC50 = 0.45 μM) and exhibits cytotoxic effects in MCF-7 cells (IC50 = 2.14 μM). Western blot analysis showed that 19 inhibited Topo I and Topo II similarly to the positive control at 100 μM, with notable effects at 2 μM. Mechanistic studies revealed that 19 induces DNA double-strand breaks, degrades Topo I/II proteins, activates the caspase-3/8/9 pathway to promote apoptosis, and causes G0/G1 phase cell cycle arrest. In vivo studies in mice showed that 19 exhibited significant anti-tumour effects and good metabolic stability at a dose of 40 mg/kg.
In 2020, the research group combined the pharmacophore of 18 with olaparib (1) to create a series of phthalazinone-acridine dual-targeting PARP/Topo inhibitors223. Compound 20 (R1 = R2 = H, X = CH) showed the highest antiproliferative activity, with IC50 values of 12.88 μM against HCT116 cells and 7.96 μM against HCC116 cells, outperforming olaparib (Figure 6). All compounds inhibited Topo II at 10 μM, similar to m-AMSA, and 20 also inhibited PARP1 with an IC50 of 11.85 μM. The SAR analysis indicates that both enzyme and cellular activities are optimal when the acridine ring remains unsubstituted. Conversely, the introduction of substituents at either the 2-position or 4-position, or the substitution of the X position with a nitrogen atom (X = N), results in diminished activity (Figure 6).
In 2023, Ananda and colleagues undertook the design and synthesis of a series of molecules exhibiting dual inhibitory activities against Topo1 and PARP1224. This was achieved by employing the phenylphthalazinone structural motif derived from olaparib and the naphthalenedicarboximide moiety characteristic of Topo1 inhibitors as pharmacophores, which were linked via a 1,3,4-oxadiazole bridge. In vitro assessments of cell viability demonstrated that compound 22 (Figure 6), characterised by the presence of a fluorophenyl group, a short alkyl chain, and an oxadiazole thioether structure, exhibited low micromolar antiproliferative activity across various tumour cell lines. Mechanistic investigations indicated that this compound induces DNA double-strand breaks, causes S/G2 phase cell cycle arrest, and promotes apoptosis in tumour cells. This occurs through the synergistic stabilisation of the Topo1-DNA covalent complex (TOP1cc) and the inhibition of the PARP1-mediated repair process involving TOP1cc PARylation. This dual inhibition mechanism presents a novel strategy for addressing resistance to TOPO1 inhibitors. However, the inhibitory activity of compound 22 on the PARP1 enzyme is relatively modest, exhibiting an inhibition rate of only 35% at a concentration of 30 μM. It is hypothesised that this limited efficacy may be attributed to the modification of the phenylphthalazinone structural unit, which potentially impacts its binding affinity to PARP1.
In 2024, Liu et al. utilised the low-toxicity and high water solubility of the natural product matrine (24,
Figure 7) as the foundational structure for their research225. They incorporated fluorene (23) to enhance π-π stacking interactions with DNA and benzimidazole as a critical pharmacophore, thereby synthesising a novel class of benzimidazole matrine derivatives. These derivatives exhibit the potential to dual-target and inhibit both Topo I and PARP1 (Figure 7). Among them, compound 25 (Figure 7) demonstrated potent inhibitory activity against PARP1, exhibiting an IC50 value of 0.09 μM, which is comparable to that of the positive control drug olaparib (IC50 = 0.02 μM). Furthermore, 25 exhibited enhanced inhibitory effects on Topo I in comparison to irinotecan, with a dose-dependent response. This compound also displayed significant anti-proliferative activity across various cancer cell lines, including HGC-27, HepG-2, HeLa, and A549, with IC50 values ranging from 2.49 to 9.48 μM. The underlying anti-cancer mechanisms were found to involve the induction of DNA damage, G0/G1 phase cell cycle arrest, and the promotion of apoptosis. In vivo studies indicated that 25 achieved a tumour inhibition rate of 75.4%, with no significant toxicity observed.

Dual PARP-microtubule inhibitors
In 2020, the Jablons group reported a novel dual-target inhibitor, AMXI-5001 (14,
Table 2), which can simultaneously inhibit PARP1/2 (IC50 values are 5 nM and 50 nM respectively) and microtubule polymerisation (IC50 = 0.99 μM)226. Structurally, AMXI-5001 preserves the essential pharmacophore of olaparib while enhancing its functional capacity through the incorporation of a 2-carbamoyl-1H-benzimidazole moiety. The aromatic ring within its pharmacophore engages in π-π stacking interactions with the Tyr907 and Tyr896 residues located at the PARP1 active site. Concurrently, the carbonyl oxygen and amide nitrogen atoms of the terminal carbamoyl group establish hydrogen bonding networks with the Asp770 and Arg878 residues, facilitating the molecule’s high-affinity occupation of the NAD+ binding pocket. Furthermore, the benzimidazole moiety exhibits specific binding to the colchicine site of the β-tubulin subunit, thereby inhibiting the polymerisation kinetics of tubulin and destabilising microtubules. This unique dual-targeting mechanism allows AMXI-5001 to synergistically inhibit both PARP1/2 and microtubule polymerisation. It is worth noting that the compound has a weak inhibitory effect on PARP5A (IC50 > 4 μM) and shows no significant off-target activity in a screening of 156 kinases. However, it exhibits broad-spectrum and highly efficient anti-proliferative effects on 110 cancer cell lines (with most IC50 < 100 nM, potency 20 to 10,000 times that of clinical PARP inhibitors), and is also effective against BRCA wild-type cancer cells. In rat and dog models, the oral bioavailability of AMXI-5001 is 31% and 64%, respectively. In the BRCA mutant MDA-MB-436 xenograft tumour model, oral monotherapy (50 mg/kg BID) can achieve complete tumour regression. This efficacy not only significantly outperforms olaparib monotherapy (50 mg/kg) and vinblastine monotherapy (1 mg/kg), but also surpasses the combination therapy of olaparib and paclitaxel, and no weight loss or obvious toxicity was observed during the 31-day treatment. Mechanistically, the synergy observed in this dual-target approach arises from two key factors: first, the inhibition of microtubule polymerisation impedes the nuclear transport of DNA repair proteins, such as ATM and Rad50, thereby exacerbating the accumulation of DNA damage induced by PARP inhibitors; second, PARP inhibition leads to the downregulation of checkpoint with forkhead and ring finger domains, which obstructs the activation of the M-phase checkpoint, consequently intensifying the mitotic catastrophe induced by microtubule-targeting agents. In 2024, members of the same research group conducted a study investigating the anti-oesophageal cancer activity of AMXI-5001227. In vitro experiments revealed that the compound exhibited a significantly greater inhibitory effect on the proliferation of nine oesophageal cancer cell lines, which included seven squamous cell carcinomas and two adenocarcinomas, compared to commercially available PARP inhibitors such as olaparib and niraparib. Notably, AMXI-5001 demonstrated an IC50 value of 0.006 μM against the squamous cell carcinoma cell line OE21. In the radiotherapy combination experiments, pre-treatment with 0.008 μM AMXI-5001 for 24 h, followed by administration of 6 Gy radiotherapy, significantly enhanced the growth inhibition of KYSE-70 cells. Notably, the radiosensitization effect of AMXI-5001 was found to be superior to that of olaparib or cisplatin. Furthermore, in vivo studies corroborated these findings, demonstrating that a dosage of 50 mg/kg AMXI-5001 in conjunction with a single 6 Gy local radiotherapy significantly inhibited the growth of KYSE-70 xenografts. Importantly, there was no observed decrease in the body weight of the mice throughout the treatment period, indicating that the systemic toxicity associated with AMXI-5001 was manageable. Currently, AMXI-5001 is in the clinical Phase I/II study stage, aiming to evaluate its therapeutic effect on various malignant solid tumours (including BC and OC). This project is currently in the stage of recruiting participants (Table 2).
In 2021, Niu et al. identified a quinazolinone dual-target inhibitor, designated as compound 26 (Figure 7), from a library comprising 226,019 compounds through the application of structure-based dual pharmacophore virtual screening in conjunction with molecular docking techniques228. This compound demonstrated a synergistic inhibitory effect on tubulin polymerisation, with an IC50 value of 0.94 μM, and on PARP1 enzyme activity, with an IC50 value of 0.48 μM. Notably, its efficacy in inhibiting tubulin polymerisation surpasses that of the positive control drug colchicine (IC50 = 2 μM), while its inhibitory activity against PARP1 is comparable to that of the positive control drug PJ34 (IC50 = 0.51 μM). In vitro cell viability assays revealed that 26 displayed low micromolar anti-proliferative activity across a range of tumour cell lines, with IC50 values generally falling between 1 and 2 μM. Mechanistic investigations have demonstrated that 26 exerts its effects by concurrently targeting tubulin polymerisation and PARP1. This dual action results in the disruption of microtubule architecture and the inhibition of the mitotic process, while simultaneously impairing DNA repair mechanisms. The convergence of these pathways leads to the accumulation of DSBs, thereby promoting enhanced apoptosis in cells. Furthermore, in vivo pharmacodynamic evaluations corroborate these findings, revealing that 26 significantly inhibits tumour growth in the MDA-MB-231 cell xenograft model at a dosage of 5 mg/kg.
In 2023, Sachin et al. conducted a screening of a library of biflavonoids, leading to the identification of lead compound 27 (Figure 7), which exhibited significant inhibitory activity against Ishikawa cells (IC50 = 3 μM)229. Subsequently, they synthesised a series of derivatives through palladium-catalyzed Suzuki arylation to modify the 7-position of the flavonoid scaffold and to modulate the electronic effects of the phenyl group at the 2-position. Among the compounds studied, compound 28 (Figure 7) demonstrated superior anti-proliferative activity against Ishikawa cells, with an IC50 value of 1.0 μM, surpassing the efficacy of 27. The potency of compound 28 significantly exceeded that of the positive controls, olaparib (IC50 = 3.91 μM) and combretastatin (IC50 = 4.21 μM). Further enzyme inhibition assays indicated that 28 exhibited inhibitory activity against PARP1/2, with IC50 values of 74.6 nM and 109 nM, respectively, and against tubulin, with an IC50 value of 1400 nM. Although its PARP inhibition was less potent than that of olaparib (PARP1/2, IC50 = 0.1/0.11 nM), its tubulin inhibition was comparable to that of combretastatin (tubulin, IC50 = 1300 nM). Mechanistic studies have demonstrated that 28 impedes DNA damage repair by inhibiting PARP1, thereby increasing genomic instability. Additionally, it disrupts the cytoskeleton through tubulin inhibition and interferes with mitosis, resulting in a synergistic interplay of replication stress and mitotic catastrophe within tumour cells. This illustrates the synergistic mechanism underlying the efficacy of PARP inhibitors in conjunction with chemotherapeutic agents. However, the current study lacks a comprehensive analysis of the structure-activity relationships at the enzymatic level. Furthermore, the data on cellular activity and enzyme inhibition do not exhibit a consistent correlation, which may be attributed to off-target effects stemming from the multi-target nature of the flavone nucleus230

Dual PARP-EGFR inhibitors/degraders
PARP inhibitors induce cell death in BRCA-mutated cells by obstructing DNA repair mechanisms, a phenomenon known as synthetic lethality. In contrast, EGFR inhibitors attenuate tumour proliferation by inhibiting associated signalling pathways. The concurrent application of these two therapeutic agents can effectively target both DNA repair deficiencies and growth signal dependencies, thereby producing a synergistic anti-tumour effect.
In 2022, Wu et al. successfully synthesised a series of olaparib derivatives that exhibit both PARP inhibition and EGFR phosphorylation inhibition activities231. This was achieved by incorporating a tetrahydropyrido[4,3-d]pyrimidinyl group into the molecular structure of olaparib and substituting the non-essential cyclopropanecarbonyl group (Figure 8). Notably, compound 30, which contains a morpholine ring substituent, demonstrated significantly enhanced proliferation inhibition against TNBC cell lines HCC1937 and MDA-MB-468 compared to olaparib, with IC50 values of 3.23 μM and 15.6 μM, respectively. Furthermore, this compound inhibited PARP enzyme activity by 92.11% at a concentration of 0.1 μM and effectively impeded the phosphorylation of EGFR. Importantly, when compound 30 was administered in conjunction with doxorubicin, it exhibited significant synergistic inhibitory effects in BRCA wild-type MCF-7 cells, as evidenced by a synergy index (CI) of 0.39. Research indicates that the incorporation of a morpholine ring structure significantly enhances the cellular permeability of compounds and facilitates effective dual-target binding.
Utilising the fragment-based drug design (FBDD) methodology, the analysis of the receptor-ligand binding pocket and its interactions with the amino acid side chains of the fragment is a critical step in the development of lead compounds with therapeutic potential. In 2023, Abbas et al. employed structural units such as triazoles, thiazines, and quinoxalines to design a series of dual-target inhibitors aimed at both EGFR and PARP1 through the FBDD approach, ultimately yielding a novel series of dual triazole-thiazine derivatives232. Notably, compound 31 (Figure 8) demonstrated exceptional inhibitory activity against PARP1, with an IC50 value of 1.37 nM, which is superior to that of olaparib (1.49 nM), and also exhibited significant inhibitory effects against EGFR, with an IC50 of 64.65 nM, outperforming erlotinib (80 nM). This compound exhibited the highest cytotoxicity against MDA-MB-231 cells, with an IC50 value of 0.12 μM, significantly surpassing that of the control drug erlotinib, which has an IC50 of 1.02 μM. Molecular docking studies demonstrated that 31 can stably bind to the active sites of PARP1 and the functional regions of EGFR. SAR analysis further indicated that the ortho-substitution pattern, reduced volume of triazole substituents, and shorter alkyl chain structures are critical structural features that enhance anti-breast cancer activity. However, it is important to note that this study lacks a comprehensive investigation into the underlying molecular mechanisms and in vivo efficacy assessments.
In 2020, Jean-Claude et al. proposed a therapeutic approach utilising a single-drug strategy designed to concurrently inhibit multiple overlapping biological effects associated with chemical resistance. Building upon this strategy, the authors developed an innovative linker structure capable of integrating multiple functional modules into a singular molecule, ultimately resulting in the successful synthesis of a trifunctional molecule, designated JS230 (34, Figure 9)233. As the inaugural tri-functional synergistic molecule, JS230 incorporates three distinct functional units—aminoquinazoline (Gefitinib, 32) for EGFR inhibition, alkyl sulphonate (Busulfan, 33) for the induction of DNA damage, and an olaparib analogue for PARP inhibition—through a modular design approach. This compound demonstrates dual inhibition of EGFR phosphorylation (IC50 = 0.326 μM) and PARP enzyme activity, while also inducing sustained DNA damage. Mechanistic studies have demonstrated that in prostate cancer cell lines PC3 and 22Rv1, which exhibit elevated expression of EGFR, compound 34 inhibits EGFR phosphorylation in a dose-dependent manner and downregulates the activity of its downstream signalling pathways, ERK and AKT. Furthermore, this compound displays a seven-fold selective inhibitory effect in NIH-3T3 cells transfected with EGFR. In the context of DNA damage induction, 34 interacts with the N7 position of guanine via its alkyl sulphonate group, resulting in the formation of a stable adduct (with alkylated product identified through LC-MS, Figure 9). This interaction subsequently elevates the phosphorylation level of γ-H2AX and induces DNA double-strand breaks. The results of the alkaline comet assay demonstrated that the DNA damaging capacity of 34 was 3 to 6 times greater than that of busulfan, with the damaging effects persisting for up to 24 h. Moreover, 34 exhibits a PARP inhibition capability comparable to that of olaparib, demonstrating a four-fold selectivity in BRCA-mutated VC8 cells. Importantly, the unique single-molecule structure of 34 confers it with superior subcellular localisation and enhanced stability. In prostate cancer cells characterised by EGFR overexpression and resistance to PARP inhibitors, 34 exhibits significantly greater growth inhibition efficacy compared to the combination therapy of gefitinib, olaparib, and busulfan. Specifically, 34 achieves an IC50 range of 1.9 to 3.2 μM, in contrast to the combination therapy’s IC50 range of 5.6 to 15.5 μM, representing an eight-fold improvement in efficacy. This study introduces an innovative molecular design strategy aimed at overcoming chemotherapy resistance in tumours.
Advancements in precision medicine have made proteolysis targeting chimaera (PROTAC) technology a novel approach to address the limitations of traditional cancer therapies234. Unlike standard small molecule inhibitors, PROTACs selectively degrade target proteins by recruiting E3 ubiquitin ligases235. This method effectively targets drug-resistant sites, modulates non-catalytic protein functions, overcomes drug resistance, prolongs therapeutic efficacy, and improves specificity236. In 2020, Li et al. introduced the “dual-PROTAC” concept, using natural amino acids as linkers to connect EGFR inhibitors, PARP inhibitors, and E3 ligase ligands237. This creates bifunctional PROTAC molecules capable of degrading two target proteins, EGFR and PARP, simultaneously. Examples include DP-C-1 (35, Figure 9) with a CRBN ligand and DP-V-4 (36,
Figure 9) with a VHL ligand. Functional studies showed that compounds 35 and 36 can simultaneously degrade EGFR and PARP in SW1990 and H1299 cells, with compound 35 optimal at 5 μM for 24–36 h and compound 36 having a DC50 of 0.47 μM for PARP. Mechanism studies revealed that the proteasome inhibitor MG132 completely blocked this degradation, confirming reliance on the ubiquitin-proteasome pathway. The experimental results obtained from microScale thermophoresis (MST) demonstrated that the dual-target PROTACs maintained their binding affinity for both EGFR and PARP, with dissociation constants (Kd) of 2.74 μM and 7.89 μM for 35, and 5.47 μM and 12.80 μM for 36, respectively. However, it is noteworthy that these binding affinities were inferior to those of the corresponding monomeric inhibitors. SAR analysis reveals that the linker length substantially influences degradation activity, with the short-chain variant compound 35 in the CRBN group demonstrating superior performance compared to the long-chain variant. Notably, compound 36 in the VHL group exhibits the highest activity. However, an increase in molecular weight adversely affects cell permeability, thereby diminishing its antiproliferative efficacy relative to monomeric drugs. For instance, the IC50 of 36 is 19.92 μM, whereas that of gefitinib is 6.56 μM.

Dual PARP-VEGFR inhibitors
PARP inhibitors are effective for breast cancer with BRCA mutations but less so for BRCA wild-type tumours238. Research shows that VEGFR inhibitors can enhance PARP inhibitor sensitivity in BRCA wild-type cells by creating hypoxic conditions and inhibiting key components of the homologous recombination repair pathway, such as BRCA1/2 and RAD51239,240.
In an effort to address the limitations associated with monotherapy, Ouyang et al. pioneered the application of a pharmacophore hybridisation strategy. This approach involved the integration of essential structural elements from the VEGFR inhibitor pazopanib (specifically, the 2,3-dimethyl-2H-indazole ring and 2-amino-pyrimidine) with the 1H-benzimidazole-4-carboxamide moiety of the PARP inhibitor veliparib, leading to the design and synthesis of dual-target lead compounds 38 and 39 (Figure 10)241. These compounds exhibited significant dual-target inhibitory activity at the enzymatic level, with VEGFR2 IC50 values of 260 nM and 209 nM, and PARP1 IC50 values of 142 nM and 95.7 nM, respectively. However, their antiproliferative efficacy at the cellular level was comparatively modest, as evidenced by IC50 values of 12.1 μM and 7.7 μM for the BRCA wild-type cell line MDA-MB-231, and 10.1 μM and 6.4 μM for MCF-7 cells, respectively. SAR studies revealed that the 2,3-dimethyl-2H-indazole ring is crucial for VEGFR2 binding, and adding a large hydrophobic group could boost cellular activity. The 2-amino-pyrimidine fragment should remain unchanged due to poor modification tolerance. In the phenyl chain, an amide bond is less active than a direct link, and meta-substitution optimises enzyme inhibition and cellular activity. Consequently, compound 40 (Figure 10) emerged as a promising candidate, showing nanomolar dual-target inhibition (VEGFR2, IC50 = 191 nM; PARP1, IC50 = 60.9 nM), strong in vitro antiproliferative activity (MDA-MB-231, IC50 = 4.1 μM; MCF-7, IC50 = 3.5 μM), anti-angiogenic properties, and effective in vivo anti-tumour and anti-metastatic effects (72.1% tumour inhibition at 50 mg/kg orally). Mechanistic investigations have demonstrated that the efficacy of 40 is primarily due to its capacity to induce DNA damage, inhibit the HR repair pathway, and arrest the cell cycle. Additionally, its excellent oral bioavailability (60.1%) and high kinase selectivity present a promising novel strategy for the treatment of BRCA wild-type breast cancer.

Dual PARP-c-Met inhibitors
Resistance to PARP1 inhibitors is linked to c-Met overexpression, which increases PARP1’s enzymatic activity via Tyr907 phosphorylation, reduces its inhibitor binding, and disrupts the BRCA2/RAD51 complex, hindering HR repair238,242. Preclinical studies show that combining PARP and c-Met inhibitors can produce a synergistic anti-tumour effect in various cancers243. Building upon this mechanism, Zhu et al. employed an innovative pharmacophore fusion strategy that combines the PARP1 inhibitory pharmacophore derived from olaparib with the essential structural components of the highly selective c-Met inhibitor compound 41 (Figure 11)244. This approach facilitated the design of a series of novel dual-target inhibitors aimed at both PARP1 and c-Met. SAR studies indicated that the preservation of an unsubstituted piperazine linker is crucial for achieving optimal dual-target inhibitory activity. Conversely, the introduction of a methyl group at the 3-position of the piperazine or its substitution with a piperidine significantly reduced the inhibitory efficacy. Additionally, it was established that an aromatic substituent at the C4 position of the pyridazinone moiety is essential for optimal activity. The selected compound, referred to as compound 42 (Figure 11), demonstrated strong dual-target inhibitory properties, with a PARP1 IC50 of 3.3 nM and a c-Met IC50 of 32.2 nM, in addition to exhibiting significant antiproliferative activity. This compound exhibited significant inhibitory activity against c-Met-amplified BRCA wild-type MDA-MB-231 cells, with an IC50 of 45 nM, and HR function-intact MDA-MB-468 cells, with an IC50 of 170 nM. Furthermore, it successfully circumvented resistance to olaparib, achieving IC50 values of 4.57 μM and 5.33 μM for HCC1937-BRCA1 mutant and HCT116OR-BRCA2 mutant cells, respectively. These results indicate a 4-fold and 10-fold enhancement in efficacy compared to olaparib. Mechanistic studies have demonstrated that compound 42 inhibits c-Met-mediated phosphorylation of PARP1 and down-regulates the expression of BRCA1 and Rad51, thereby impairing HR repair function, inducing DNA damage, and causing G2/M phase cell cycle arrest. In xenograft models of MDA-MB-231 and HCT116OR, tumour growth inhibition rates were observed to be 77% and 69%, respectively. These findings suggest a novel strategy for broadening the clinical applications of PARP inhibitors and addressing issues of drug resistance.

Dual PARP-PI3K inhibitors
In multiple tumour types, the aberrant activation of the PI3K/AKT/mTOR signalling pathway facilitates the repair of DSBs by stabilising and sustaining the interaction with the HR complex245. Consequently, the concurrent targeting of PARP and PI3K is considered a promising therapeutic strategy for BRCA wild-type tumours.
In 2020, Xu et al. successfully synthesised the first lead compound, designated as compound 44 (Figure 12), which exhibited dual inhibitory activity against PARP and PI3K246. This development was based on pharmacophore models derived from olaparib and the PI3K inhibitor BKM120 (43, Figure 12). However, an imbalance was observed in its enzyme inhibitory activities, with PARP1 displaying a pIC50 of 7.81 compared to PI3Kα, which had a pIC50 of 6.49. To address this discrepancy, the researchers employed a scaffold-hopping strategy, substituting the core pyrimidine ring with a 1,3,5-triazine ring, resulting in compound 45 (Figure 12), which demonstrated significantly enhanced inhibitory activity relative to 44 (PARP1, pIC50 = 8.02 vs. PI3Kα pIC50 = 6.77). Subsequently, utilising a ring merging strategy, the researchers modified the piperazine-pyrimidine linker to form a bicyclic system of tetrahydropyrido[3,4-d]pyrimidine, yielding candidate compounds 46 (PARP1, pIC50 = 8.59 vs. PI3Kα, pIC50 = 8.32) and 47 (PARP1, pIC50 = 8.22 vs. PI3Kα, pIC50 = 8.44). In vitro investigations have demonstrated that compounds 46 and 47 possess markedly enhanced anti-proliferative efficacy against both BRCA-deficient cancer cell lines (e.g., HCC1937, HCT116) and BRCA wild-type cancer cell lines (e.g., MDA-MB-468) when compared to the effects of the single-agent treatments olaparib or BKM120, as well as their combination. In vivo studies utilising the MDA-MB-468 xenograft tumour model demonstrated that 47 (administered at a dosage of 50 mg/kg) exhibited a significantly higher tumour growth inhibition rate (TGI = 73.4%) in comparison to 46 (also at 50 mg/kg, TGI = 52.7%) and the combination treatment of olaparib and BKM120 (TGI = 48.4%). Furthermore, no significant toxicity was observed in any of the treatment groups. Mechanistic investigations further elucidate that compounds 46 and 47 effectively inhibit the proliferation of BRCA wild-type tumour cells by inducing HR deficiency, resulting in the accumulation of DNA damage and the promotion of apoptosis through the synergistic inhibition of PARP1 and PI3K. Selectivity assessments indicate that 47 demonstrates potent inhibitory activity against PARP1 (pIC50 = 8.22), PARP-2 (pIC50 = 8.44), PI3Kα (pIC50 = 8.25), and PI3Kδ (pIC50 = 8.13), while exhibiting superior selectivity compared to PI3Kβ and PI3Kγ.
In 2021, researchers from the same research group enhanced the structural configuration of compound 43 by substituting the phthalazinyl moiety with a more compact benzofuran-7-carboxamide group, thereby yielding lead compound 48 (Figure 12), which demonstrated inhibitory activities against both PARP1 and PI3Kα247. At a concentration of 10 nM, compound 48 exhibited inhibition rates of 51% and 80% against PARP1 and PI3Kα, respectively. Building upon the skeleton transition strategy, researchers substituted the pyrimidine skeleton in compound 49 with a 1,3,5-triazine skeleton to improve enzyme inhibition activity. The resulting compound demonstrated IC50 values of 7.3 nM and 39.6 nM against PARP1 and PI3Kα, respectively. However, 49 was characterised by poor solubility, with a solubility of less than 100 μg/mL. To mitigate this limitation, researchers incorporated a hydrophilic N-acetylmorpholine group into the compound’s structure, yielding compound 50 (Figure 12). This modification significantly enhanced the water solubility of its hydrochloride salt to greater than 10 mg/mL; however, it was accompanied by a slight reduction in enzyme inhibition activity, with IC50 values of 13.8 nM for PARP1 and 64.0 nM for PI3Kα. In vitro investigations revealed that 50 demonstrated significantly enhanced antiproliferative activity against the BRCA wild-type TNBC cell line MDA-MB-468, with an IC50 value of 1.4 μM. This activity markedly exceeded that of olaparib (IC50 = 13.72 μM) and BKM120 (IC50 = 3.97 μM). Additionally, 50 exhibited minimal inhibitory effects against 374 kinases, suggesting a targeted mechanism of action primarily involving PARP1/2 and PI3Kα. In vivo studies utilising the MDA-MB-468 xenograft model indicated that administration of compound 50 at a dosage of 50 mg/kg resulted in a tumour growth inhibition rate of 54.6%, surpassing the efficacy of olaparib monotherapy (TGI = 28.5%), BKM120 monotherapy (TGI = 33.4%), and the combined treatment of both agents (TGI = 48.4%).
In 2021, utilising the pharmacophore models of veliparib and the PI3K inhibitor GDC-0980 (46), Xu et al. successfully identified compound 52 (Figure 12), which demonstrated dual inhibitory activity against PARP and PI3K248. This compound exhibited significant enzyme inhibitory activity, with an imbalance favouring PARP1 (IC50 = 0.91 nM) over PI3Kα (IC50 = 1.5 nM). Importantly, 52 displayed high selectivity for PARP1/2 and PI3Kα/δ, with IC50 values below 10 nM, while showing weak inhibition of PI3Kβ/γ and mTOR (IC50 > 12 nM). Additionally, it exhibited no significant hERG toxicity, with an IC50 value exceeding 30 μM. In vitro investigations revealed that 52 demonstrated significant antiproliferative activity against both BRCA wild-type (SW620, MDA-MB-231/468) and BRCA-deficient (HCT-116, HCC-1937) tumour cell lines, surpassing the efficacy of the positive controls olaparib and GDC-0980. Furthermore, in vivo studies indicated that the tumour growth inhibition effect of compound 52 (TGI = 48.77%) was markedly superior to that of olaparib (TGI = 27.91%), GDC-0980 (TGI = 37.98%), and their combination (TGI = 45.93%). SAR studies have demonstrated that the 2-amino-pyrimidine moiety is a critical structural component for the potent activity of PI3Kα. Furthermore, the incorporation of methylthiophene substitutions has been shown to enhance PARP1 inhibition, whereas the introduction of phenyl or heterocyclic substituents markedly diminishes PI3Kα activity. Additionally, reducing molecular weight may enhance drug-like properties; however, this must be carefully balanced with lipophilicity considerations.

Dual PARP-CDK inhibitors
Recent research shows that combining PARP inhibitors with CDK4/6 inhibitors produces a synergistic anti-tumour effect in cancers with high MYC expression and in HR-proficient breast and ovarian cancers120. This effect arises from CDK4/6 inhibitors downregulating key HR proteins like BRCA1/2 and RAD51 during the G2 phase, which increases DNA damage from PARP inhibitors and enhances the sensitivity of HR-proficient cancer cells to treatment249.
In 2023, Wang et al. successfully synthesised the lead compound 54 (Figure 13) by integrating the pyrazinone moiety, known for its PARP binding affinity in olaparib, with the pyrido[3,4-d]pyrimidinone moiety, recognised for its CDK6 binding capability in palbociclib (53, Figure 13), utilising a linker strategy250. Despite demonstrating moderate inhibitory activity against CDK6, the compound exhibited limited antiproliferative efficacy and possessed a molecular weight surpassing the threshold stipulated by Lipinski’s rule of five, necessitating additional optimisation. Initial investigations into the SAR of the connecting group revealed that incorporating a C3 alkyl chain (compound 55, Figure 13) substantially enhanced CDK inhibitory potency and cellular efficacy. Conversely, chains that were excessively long or short diminished activity. Consequently, the pharmacophore of palbociclib was simplified by removing the acetyl group and introducing an alkyl chain connecting group, resulting in compound 56 (Figure 13). This modification led to a notable increase in cellular efficacy (IC50 = 6.46–13.49 μM); however, its CDK6 inhibitory activity (IC50 = 542 nM) was less potent compared to palbociclib (IC50 = 36 nM). To improve enzyme inhibitory activity, a series of molecules were synthesised through the incorporation of rigid linker groups, such as piperidine. Notably, compounds 57 and 59 (Figure 13), which contain piperidine linkers, exhibited exceptional compatibility with the target binding pocket and demonstrated a balanced inhibitory activity against PARP1 and CDK4/6. In particular, compound 59, with IC50 values of 127 nM for PARP1 and 41 nM for CDK6, as well as an IC50 of 1.96 μM against MDA-MB-231 cells, displayed the most favourable balance between inhibitory potency and cellular activity. It is noteworthy that the introduction of a bromine atom at the acetyl group position within the palbociclib pharmacophore (compound 58, Figure 13) selectively enhances inhibitory activity against CDK6. However, this modification does not provide any benefit in terms of PARP1 inhibition or overall cellular efficacy. Conversely, substituting the rigid linker in compound 58 with a flexible linker (compound 60, Figure 13) results in enhanced PARP1 inhibition, albeit at the expense of CDK6 inhibition and overall cellular efficacy. Consequently, compound 59 emerges as the preferred candidate molecule. In vitro studies have demonstrated that 59 significantly down-regulates BRCA1 expression, inhibits RB phosphorylation, and induces DNA damage and apoptosis. It is 10 to 20 times more effective than olaparib against BRCA wild-type TNBC cells and outperforms the drug combination, making it a promising treatment for the 80% of TNBC patients without BRCA mutations.
In 2022, Yang et al. successfully synthesised compound 62 (Figure 14), which exhibits dual inhibitory activities against PARP and CDK4/6. This was achieved by integrating the essential pharmacophores of olaparib and the CDK inhibitor abemaciclib (61, Figure 14) via a linker251. In vitro studies have demonstrated that this compound effectively inhibits the proliferation of breast and ovarian cancer cells (IC50 = 0.435–3.608 μM) through a synergistic mechanism that targets PARP and CDK4/6, leading to G1/S phase arrest and subsequent DNA damage. Furthermore, its inhibitory efficacy is significantly greater than that of olaparib, abemaciclib, or their combination. In the MDA-MB-231 and OVCAR5 mouse xenograft models, the antitumor efficacy of compound 62 administered as a monotherapy at a dosage of 50 mg/kg was found to be significantly superior to that of the combination treatment involving olaparib and abemaciclib, with no discernible toxicity reported. Furthermore, 62 markedly enhanced the cytotoxic effects of cisplatin through a synergistic mechanism that promotes DNA damage and apoptosis. The combination of 62 with cisplatin exhibited enhanced antitumor activity both in vitro and in vivo, thereby presenting a promising therapeutic strategy to address the challenges associated with HR repair status and to effectively treat HR-proficient breast and ovarian cancers.
Recently, Kong et al. synthesised compound 63 (Figure 14) by linking the benzimidazole-4-carboxamide from the PARP inhibitor veliparib with the pyrrolo[2,3-d]pyrimidine-6-carboxamide from the CDK4/6 inhibitor abemaciclib via a benzene ring252. Compound 63 showed moderate enzyme inhibition (PARP1/CDK6, IC50 = 323/467 nM) and good cellular activity (MDA-MB-231/468, IC50 = 0.59/0.25 μM). To augment the enzyme inhibitory activity, researchers undertook a SAR focusing on the linker groups. The findings revealed that compounds with ortho-benzene ring linker arms (compounds 64–66, Figure 14) demonstrated markedly superior PARP1 inhibitory activity compared to the meta-benzene ring linker arm compound (compound 67, Figure 14), which exhibited enhanced CDK6 inhibitory activity (IC50 = 3.2 nM). Furthermore, the position and electronegativity of substituents on the benzene ring of the linker arm were found to significantly influence the activity. Notably, compound 66, featuring a chlorine atom substitution proximal to the CDK6 pharmacophore side, exhibited a balanced enzyme inhibitory profile (PARP1/CDK6, IC50 = 156.8/13.3 nM) and cellular activity (MDA-MB-231/468, IC50 = 1.54/1.69 μM). In contrast, compound 65 (Figure 14), which possesses a chlorine substituent adjacent to the PARP1 pharmacophore, along with compound 64 (Figure 14), which contains methoxy substitutions, demonstrated significantly diminished enzyme inhibitory activity. The subsequent incorporation of either rigid or flexible connecting chains (68–71, Figure 14) based on compound 66 resulted in a reduction of biological activity. Notably, the compounds featuring rigid connecting chains (69–71) exhibited marginally superior activity compared to the compound with a flexible chain (68). Consequently, 66 was identified as the candidate molecule for further investigation. Mechanistic studies have demonstrated that 66 induces excessive DNA damage and apoptosis in cells by synergistically targeting PARP and CDK6. Additionally, it inhibits the proliferation of tumour cells by down-regulating the expression of the oncogene MYC through the inhibition of the Wnt/β-catenin signalling pathway. In vivo investigations have confirmed that administration of 66 at a dosage of 60 mg/kg results in a tumour growth inhibition rate of 97.10% against MDA-MB-231 xenografts, with a TGI of 83.67% observed in the combination treatment group. Furthermore, 66 effectively inhibits lung metastasis and exhibits no significant toxicity, thereby offering a novel therapeutic strategy for BRCA wild-type TNBC.
Recent research has identified CDK12 as a pivotal factor in the transcriptional regulation of genes associated with HR253. The CDK12 inhibitor, dinaciclib, has been shown to reduce the expression levels of RAD51, BRCA1, and BRCA2 in a dose-dependent manner, thereby impairing the restoration of homologous recombination repair functionality and increasing the susceptibility of tumour cells to PARP1 inhibitors254. Currently, a clinical trial investigating the combination of the CDK12 inhibitor dinaciclib and the PARP1 inhibitor veliparib for the treatment of advanced solid tumours with BRCA1/2 mutations is underway in phase I255. The potential synergistic interaction between these two targets, coupled with the structural diversity of their respective inhibitors, suggests a promising avenue for the development of dual-target inhibitors aimed at CDK12 and PARP1.
In 2023, Ouyang and colleagues employed a strategy that integrated pharmacophore fusion with rational drug design256. Utilising the simplified pharmacophore of the CDK12/13 inhibitor THZ531 (72, Figure 15), they fused it with the benzimidazole carboxamide pharmacophore derived from veliparib. This approach facilitated the successful design and synthesis of a series of small molecule compounds exhibiting dual inhibitory activities against PARP1 and CDK12. SAR studies have demonstrated that incorporating electron-withdrawing groups, exemplified by compounds 73 to 75 (Figure 15), at the terminal pyrimidine ring of the CDK12 core structure can markedly enhance inhibitory activity against CDK12 while preserving substantial inhibitory efficacy against PARP1. Conversely, the introduction of electron-donating groups, such as methyl, methoxy, and ethyl, or the inclusion of nitrile, nitro, and pyridine rings, results in a pronounced reduction in CDK12 inhibitory activity. Furthermore, compounds featuring a meta-carbonyl-substituted benzene as the connecting group, such as compounds 77 and 78 (Figure 15), exhibited superior activity, whereas those with a sulphonamide-substituted phenyl group, such as compound 76 (Figure 15), displayed diminished CDK12 inhibitory and cytostatic activities. Additionally, substituting the piperazine ring with a piperidine ring, as seen in compounds 79 and 80 (Figure 15), significantly decreased enzyme inhibitory activity. Based on these optimisation outcomes, compound 78 was identified as a promising candidate molecule, owing to its highly effective and balanced inhibitory activity against both CDK12 (IC50 = 285 nM) and PARP1 (IC50 = 34 nM). Mechanistic studies have demonstrated that 78 exerts anti-tumour effects in TNBC cells by inducing defects in DNA damage repair, causing G2/M phase cell cycle arrest, and promoting apoptosis. Additionally, it has exhibited significant in vivo anti-tumour activity in the MDA-MB-231 xenograft tumour model. However, its oral bioavailability is relatively low (F% = 0.48), indicating that further optimisation is necessary to enhance its pharmacokinetic properties.
Martorana et al. employed a pharmacophore fusion strategy to integrate the key pharmacophore of olaparib with the active fragment 29 (Figure 8), leading to the design and synthesis of a series of novel small molecule compounds exhibiting dual inhibitory activities against PARP1 and CDK12257. The structural modifications of these compounds primarily targeted the terminal substituents of the tetrahydropyrido[4,3-d]pyrimidine fragment. Notably, compound 81 (Figure 8), characterised by a moderate chain length and the presence of a piperidine ring structure, demonstrated the most favourable inhibitory activity. In vitro investigations have demonstrated that this compound effectively inhibits PARP1 enzyme activity at a concentration of 0.1 μM, achieving an inhibition rate of 98.25%. Additionally, it markedly suppresses CDK12 expression. In evaluating its anti-tumour efficacy, compound 81 exhibited an IC50 value of 20.23 μM in HER2-positive SKBR3 cell lines and 1.60 μM in BRCA-mutated MX-1 cell lines, outperforming olaparib. Nonetheless, this study lacks direct experimental data regarding the compound’s inhibitory effects on CDK12 enzyme levels and has not conducted in vivo pharmacodynamic assessments.

Dual PARP-Polθ inhibitors
PARP1 is activated in response to DNA strand breaks and is subsequently recruited to single-strand DNA breaks, facilitating poly(ADP-ribosyl)ation, a critical regulatory event in the Polθ-mediated alternative TMEJ repair pathway258. Notably, the inhibition or knockout of PARP1 results in only a partial reduction of TMEJ activity, with the residual TMEJ process remaining susceptible to Polθ inhibitors259. In tumours characterised by HR deficiency, resistance to PARP inhibitors often arises due to the loss of 53BP1; however, these tumours frequently exhibit overexpression of Polθ and a consequent reliance on this enzyme. Thus, targeting Polθ presents a promising strategy to mitigate PARP inhibitor resistance.
In 2024, Luo and his research team employed a pharmacophore fusion strategy that integrated the phthalazine ring structural domain of olaparib with the biphenyl-sulfodiazole core structure of the Polθ helicase inhibitor IDE95 (82, Figure 16), resulting in the successful design and synthesis of a series of novel dual-target inhibitors for Polθ and PARP. SAR studies have demonstrated that compound 83 (Figure 16), which retains the biphenyl moiety and connects the phthalazine group via a 5-O-thiadiazole linkage, effectively maintains inhibitory activity against Polθ, with an IC50 value of 268.2 nM. Compounds featuring a rigid spirocyclic linker, such as compounds 83 and 84 (Figure 16), exhibit substantial inhibitory activity against PARP1, with inhibition rates exceeding 45% at a concentration of 10 nM, while also retaining some inhibitory capacity against Polθ. Conversely, substitution of the linker with azetidine, as seen in compound 86 (Figure 16), results in a complete loss of inhibitory activity against Polθ, with an IC50 value exceeding 1000 nM. Notably, compound 85 (Figure 16), which incorporates a piperidine linker, achieves the most favourable balance of dual enzyme inhibition, with an IC50 of 117 nM against Polθ and an inhibition rate of 91.6% against PARP1 at a 10 nM concentration. Additional SAR investigations into the biphenyl group’s substitution sites indicated that halogen substitution at the 5-position (88, Figure 16) greatly improves inhibitory activity against Polθ compared to halogen substitution at the 2-position (87, Figure 16), with compound 89 (Figure 16), which has a chlorine atom, demonstrating the highest activity. Following these optimisation findings, the research team substituted the rigid spirocyclic linker with a piperidine linker, resulting in the candidate compound 90 (Figure 16). The IC50 values of this compound for Polθ ATPase and PARP1 are 45.6 nM and 5.4 nM, respectively, demonstrating a significant improvement over those of its lead compound, IDE95 (Polθ, IC50 = 2.4 μM), and olaparib (PARP1 inhibition rate of 94.4% at 10 nM). Selectivity studies indicate that 90 exhibits a high degree of selectivity for PARP1 and PARP2, with IC50 values for PARP3 through PARP10 exceeding 51 nM. Furthermore, no significant inhibitory effects were observed among the 80 kinases tested. Subsequent mechanistic investigations demonstrated that in BRCA1-deficient MDA-MB-436 cells, 90 elicited a more pronounced induction of DNA damage and apoptosis, while effectively inhibiting the TMEJ repair pathway (IC50 = 0.088 μM). These effects were significantly superior to those observed with olaparib or niraparib monotherapy, as well as their combination regimens. Furthermore, 90 was shown to effectively circumvent PARP inhibitor resistance associated with 53BP1 deficiency. In vivo pharmacodynamic assessments revealed that in the MDA-MB-436 xenograft tumour model, 90 administered at a dose of 40 mg/kg achieved a tumour growth inhibition rate of 74.7%, which was significantly greater than the TGI observed with olaparib monotherapy (40.65%), IDE95 monotherapy (45.03%), and the combination of these two agents (62.45%).

Dual PARP-BRD4 inhibitors
Research indicates that BRD4 inhibitors hinder homologous recombination repair by downregulating BRCA1, Rad51, and related genes involved in DNA repair260. In multiple cancer types, combining PARP and BRD4 inhibitors produces a significant synergistic anti-tumour effect, primarily due to the homologous recombination repair deficiency caused by BRD4 inhibition261,262. This potential synergy suggests a promising direction for developing dual-target inhibitors for BRD4 and PARP1.
In 2020, Ouyang et al. identified lead compound 92 (Figure 17), which exhibits inhibitory activity against both BRD4 and PARP1, utilising a fragment-based virtual screening approach263. This compound is chemically characterised as a derivative of the BRD4 inhibitor RVX-208 (91, Figure 17). At a concentration of 10 μM, compound 92 demonstrated inhibition rates of 65% and 30% against BRD4 and PARP1, respectively. SAR studies revealed that the 5,7-dimethoxy substituents on the quinazolinone ring serve as critical pharmacophores for sustaining inhibitory activity against BRD4. Following this finding, researchers substituted the piperidinyl group in compound 92 with various functional groups, including morpholinyl (compound 93, Figure 17), substituted piperazinyl (compound 94, Figure 17), and benzamidyl groups (compound 95, Figure 17). Experimental data demonstrated that these structural modifications significantly enhanced the inhibitory activity of the compounds against PARP1. Notably, compound 95 (Figure 17), characterised by a benzene ring linked to the benzamidyl group via a two-carbon alkyl chain and possessing an unsubstituted benzene ring, exhibited optimal inhibitory activity, with IC50 values of 0.4 μM and 4.6 μM against BRD4 and PARP1, respectively. Subsequent investigations have demonstrated that 95 possesses substantial anti-tumour growth properties both in vitro and in vivo, mediated by the synergistic inhibition of BRD4 and PARP1. In the MDA-MB-468 xenograft tumour model, oral administration of 95 at a dosage of 80 mg/kg resulted in a significant reduction in tumour growth (tumour growth inhibition = 55.3%), with no observed weight loss or other discernible toxic effects. In 2022, researchers from the same research group employed a pharmacophore fusion strategy to amalgamate the key pharmacophores of olaparib and RVX-208 via a phenyl linker, resulting in the successful identification of two lead compounds, designated as 96 and 97 (Figure 17)264. At a concentration of 1 μM, compound 96 exhibited inhibition rates of 47.52% against BRD4 and 72.02% against PARP1, whereas compound 97 demonstrated inhibition rates of 67.82% against BRD4 and 55.04% against PARP1. Subsequently, the researchers undertook three rounds of structural optimisation to achieve a balance in inhibitory activities against the dual targets. Initially, it was established that the presence of dimethoxy substituents at the 5 and 7 positions of the quinazolinone ring is critical for preserving BRD4 activity; substituting these groups with either electron-withdrawing or electron-donating moieties resulted in a marked decline in activity. Furthermore, it was determined that chlorine substitution at the 6 position of the phthalazinone group is essential for PARP1 inhibitory activity, with the absence of this substituent leading to a complete loss of activity. Lastly, it was confirmed that meta-substituted compounds within the phenyl linker exhibited superior enzyme inhibitory activity compared to their para-substituted counterparts, while the introduction of additional substituents on the phenyl ring generally led to a reduction in overall activity. Based on the aforementioned optimisation results, compound 98 (Figure 17) was identified as the candidate molecule due to its highly efficient and balanced inhibitory activities against BRD4 (IC50 = 238 nM) and PARP1 (IC50 = 197 nM). Mechanistic studies indicated that in BRCA wild-type MDA-MB-468 cells, 98 effectively inhibited the DNA damage repair process by down-regulating the expression of key HR repair proteins, including CtIP, Rad51, and Mre11, thereby inducing synthetic lethality. Furthermore, this compound exhibited acceptable oral bioavailability (F% = 23.8%). In the MDA-MB-468 xenograft tumour model, oral administration of the compound at a dose of 80 mg/kg resulted in a tumour growth inhibition rate of 68.7%, significantly surpassing the performance of compound 95 (TGI = 62.5%).
In 2022, Xu et al. implemented a pharmacophore fusion strategy to integrate the critical pharmacophores of veliparib (benzimidazole-4-carboxamide) and BI-2356 (99, tetrahydropteridine), resulting in the successful synthesis of the lead compound 100 (Figure 18), which demonstrated IC50 values of 150 nM for PARP1 and 237 nM for BRD4265. Nonetheless, there remains potential for further optimisation of its inhibitory activity. SAR studies have demonstrated that when the linker group is directly attached to the substituted benzene ring (as observed in compounds 100 and 103, Figure 18), the enzyme inhibitory activity is notably robust. Conversely, elongating the carbon chain of the linker (as seen in compounds 102 and 105, Figure 18) or incorporating benzamide derivatives as the linker (as in compounds 101 and 104, Figure 18) significantly diminishes, or even entirely abolishes, this inhibitory activity. Furthermore, the presence of a cyclopentyl substituent on the tetrahydropteridine ring is essential for maintaining enzyme inhibitory activity; substituting it with an arylmethyl group markedly reduces the inhibitory effect on BRD4. Subsequent optimisation revealed that methylation of the NH group in the linker substantially enhances dual inhibitory activity against PARP1 and BRD4. The resulting candidate molecule 106 (Figure 18) demonstrates exceptional enzyme inhibitory activity (PARP1, IC50 = 13 nM; BRD4, IC50 = 101 nM) and significant antiproliferative efficacy (MDA-MB-231, IC50 = 0.34 μM; SW1990, IC50 = 0.98 μM). Mechanistic studies have demonstrated that this compound inhibits the G1/S phase transition and mitotic processes of the cell cycle, downregulates the expression of HR repair-related genes, specifically BRCA1 and RAD51, and modulates the HEXIM1/c-Myc/FOXM1 signalling pathway to suppress DNA damage repair mechanisms. Concurrently, this compound promotes the expression of LC3B and facilitates the degradation of the P62 protein by activating the AMPKα/ULK1 pathway while inhibiting the Bcl-2/Beclin1 pathway, thereby inducing autophagic cell death. In vitro experiments further corroborated that 106 effectively reverses the accelerated cell cycle and the recovery of DNA repair functions induced by olaparib through the synergistic inhibition of PARP1 and BRD4, thereby overcoming adaptive resistance. Additionally, in vivo studies indicated that administration of 106 at a dosage of 45 mg/kg significantly inhibited the growth of SW1990 transplanted tumours, with no evident organ toxicity observed. In the same year, researchers from our group employed a pharmacophore fusion strategy that involved linking the key pharmacophore of the PARP1-targeting benzimidazole-4-carboxamide with the pharmacophore of the BRD4 inhibitor 107 (Figure 18) via a methylene bridge, ultimately yielding the lead compound 108 (Figure 18)266. This compound exhibited significant inhibitory activity against BRD4 (IC50 = 43 nM), whereas its inhibitory activity against PARP1 was comparatively modest (IC50 = 361 nM). The introduction of a methyl modification at the R1 position of compound 108 yielded compound 109 (Figure 18), which demonstrated a marked increase in inhibitory potency against PARP1 (IC50 = 49 nM), despite a reduction in inhibitory activity against BRD4 (IC50 = 202 nM). Notably, 109 exhibited superior antiproliferative activity in BRCA wild-type cell lines MDA-MB-231 and MDA-MB-468, with IC50 values of 0.14 μM and 0.09 μM, respectively. In contrast, when the linker group was substituted with an amide group in compounds 110, 111, and 113 (Figure 18), a balance in inhibitory activities against both targets was observed (PARP1 IC50 values of 50 nM, 112 nM, and 187 nM; BRD4 IC50 values of 90 nM, 116 nM, and 326 nM). However, this modification led to a significant reduction in cellular activity. Based on the aforementioned structure-activity relationship analysis, 109 was identified as the candidate molecule. Mechanistic studies indicated that 109 effectively mitigated the adaptive resistance conferred by olaparib through a synergistic inhibition of PARP1 and BRD4, as well as a down-regulation of cell cycle-related proteins CDK6 and CCND1, and DNA repair-related proteins BRCA1 and Rad51. Additionally, in the pancreatic cancer SW1990 xenograft model, compound 109 exhibited enhanced in vivo anti-tumour efficacy relative to the single-agent combination treatment regimen, while maintaining a manageable toxicity profile.
In 2024, Hu et al. conducted a study that integrated chemoinformatics, utilising the analysis of amide scaffold networks comprising 3,852 PARP1 inhibitors and 6,182 BRD4 inhibitors, with structural design methodologies, which included molecular docking and the investigation of the co-crystal binding mode of PARPi-BRD4267. The research identified compounds 114 and 115 (Figure 19), both featuring phenanthridin-6(5H)-one as their core. Compound 114, with a non-planar tricyclic structure, showed strong PARP1 inhibition (IC50 = 94 nM) but weak BRD4 inhibition (IC50 = 3560 nM). Conversely, compound 115, with a planar structure, had even weaker BRD4 inhibition (IC50 = 32635 nM), suggesting that planarity reduces BRD4 affinity. To improve dual-target activity, researchers incorporated the side chain substituent of the PARP inhibitor PJ34 (116, Figure 19) onto the core structure of compound 114, resulting in compound 117 (Figure 19). However, this modification did not enhance BRD4 activity, as indicated by an IC50 value of 28805 nM. Consequently, an open-ring strategy was employed based on compound 117 to synthesise compound 118 (Figure 19). Additionally, a large-volume hydrophobic group from the side chain of compound 114 was introduced to yield compound 119 (Figure 19). SAR studies indicated that compounds with polar side chains (118) exhibited a significant reduction in BRD4 activity, whereas compounds with bulky hydrophobic side chains (119) demonstrated strong affinity for BRD4 and greater selectivity for the BRD4-BD1 subtype compared to BD2 (IC50: 204 nM vs. 801 nM). These findings underscore the critical importance of the hydrophobic pocket in BRD4’s functionality. Consequently, 119 was identified as a candidate molecule. Its synergistic mechanism is characterised by the inhibition of the G1/S phase transition, the suppression of DNA damage repair, and the induction of apoptosis. Furthermore, it exhibited significant antiproliferative activity, with an IC50 ranging from 2.2 to 4.9 μM, in both BRCA wild-type and mutant breast cancer cell lines. This study offers valuable insights for the systematic design of dual-target inhibitors related to PARP.

Dual PARP-HDAC inhibitors
Research indicates that the combined use of HDAC inhibitors and PARP inhibitors can effectively circumvent the cell cycle compensation mechanism and activate the mitochondrial apoptotic pathway by synergistically disrupting genomic stability and cell cycle checkpoint regulation268,269. Building on this synergistic mechanism, the development of innovative dual-target inhibitors that concurrently inhibit PARP1 and HDAC through structure-based drug design strategies may substantially enhance cytotoxic efficacy against BRCA wild-type tumour cells, overcome drug resistance, and improve therapeutic outcomes270
.In 2017, Jiang et al. employed a rational drug design strategy that preserved the bis-lactam structure of olaparib as the core component271. They subsequently bridged the key pharmacophore of the HDAC inhibitor panobinostat (120, Figure 20), specifically the hydroxamic acid moiety, with various substituted benzene rings or piperazine rings to develop dual-target inhibitors of PARP1/2 and HDAC1/6 (compounds 121–124, Figure 20). SAR analysis indicates that the piperazine ring is essential for the inhibition of PARP1. Compound 121, which lacks the piperazine moiety, demonstrates the lowest activity with an IC50 value of 68.15 nM. In contrast, compounds 122 to 124, which incorporate the piperazine ring, exhibit enhanced activity, with increases ranging from 3 to 11 times; notably, compound 124 displays an activity comparable to that of olaparib, with an IC50 of 6.16 nM. Furthermore, the selectivity for HDAC inhibition is influenced by the type of linker employed: derivatives featuring a benzene ring (121) or a double bond (122) exhibit superior inhibition of HDAC1 and HDAC6, with IC50 values of 27.26/8.21 nM and 38.25/10.18 nM, respectively. Conversely, compounds 123 and 124, which possess a fatty chain linker, demonstrate a marked reduction in activity against HDAC1 (IC50 > 500 nM), although they retain efficacy against HDAC6, with IC50 values of 129.4 and 16.72 nM, respectively. In vitro investigations have demonstrated that 121 significantly induces DNA damage in cancer cells, promotes hyperacetylation of histones, and triggers caspase-mediated apoptosis. Furthermore, it exhibits enhanced anti-proliferative activity (IC50 = 0.22 μM) in drug-resistant models, such as MDA-MB-231, when compared to the single agents olaparib or SAHA. Notably, 121 displays low cytotoxicity towards normal cells and lacks genotoxic effects.
In 2020, Liao et al. developed a series of molecules exhibiting dual inhibitory activities against HDAC and PARP by strategically fusing the key pharmacophore benzamide from chidamide (125, Figure 20) with the phthalazinone pharmacophore derived from olaparib, utilising multiple connecting linkers272. SAR analysis reveals a significant correlation between the length and flexibility of the connecting chain and the enzyme inhibitory activity. Notably, the derivative of 6-aminocaproic acid (126, Figure 20) exhibits the most pronounced dual inhibitory activity, with IC50 values of 4.2 nM for PARP1 and 340 nM for HDAC1. In contrast, the piperidine connecting chain (131, Figure 20) enhances PARP1 inhibitory activity (IC50 = 1.94 nM) while diminishing HDAC1 activity (IC50 = 810 nM). Furthermore, the rigid cyclic chains (129, 130) preserve HDAC1 inhibitory activity but significantly reduce PARP1 activity. Importantly, the presence of benzamide phenyl ring substituents markedly influences enzyme activity; fluorine-containing substituents (127, 128) exhibit diminished activity, whereas unsubstituted derivatives (126, 129) demonstrate a stronger inhibitory effect on HDAC1. In light of the aforementioned findings, candidate molecule 126 was identified as the preferred compound due to its robust enzyme inhibition and notable anti-proliferative effects in BRCA wild-type cancer cell lines, specifically K562 (GI50 = 5.6 μM) and MDA-MB-231 (GI50 = 4.3 μM). However, it is important to acknowledge that the present study is limited in its exploration of the molecular mechanisms underlying the action of the candidate compounds and lacks in vivo pharmacodynamic validation through animal models.
In 2024, Jiang et al. developed and synthesised a series of dual-target inhibitors that concurrently target PARP and HDAC273. This design preserved the essential pharmacophore of olaparib while incorporating a benzamide moiety as the zinc-binding domain for HDAC. Structural optimisation was achieved through the strategic combination of various linker groups. SAR analysis reveals that the nature of the connecting chain markedly influences the enzyme inhibitory activity of the molecules. Specifically, compounds featuring double bond connecting chains (132, 133) and those with direct connections (134, 135) demonstrate potent inhibitory activity against PARP1/2 and HDAC1 (Figure 20). In contrast, compounds with methylene connecting chains (136, 137) exhibit a significant reduction in enzyme inhibitory efficacy. Furthermore, the introduction of a fluorine atom at the 4-position of the benzamide group in compounds 133 and 135 results in enhanced antiproliferative activity across various tumour cell lines when compared to their non-fluorinated counterparts (132, 134). Consequently, compounds 133 and 135 have been identified as promising candidate molecules. Subsequent mechanistic studies have demonstrated that compounds 133 and 135 restore the synthetic lethality effect by downregulating the expression of HR repair-related proteins, specifically BRCA1 and RAD51. Concurrently, these compounds facilitate the accumulation of cytoplasmic DNA, which activates the cGAS-STING-TBK1-IRF3 signalling pathway, leading to the secretion of IFNβ. This process subsequently enhances the expression of pro-inflammatory chemokines CCL5 and CXCL10 through the JAK-STAT signalling pathway, thereby promoting T cell infiltration. Furthermore, these compounds also upregulate the expression levels of neoantigens (such as CT4541 and SPANX81), antigen presentation-related genes (including HLA-A/B, TAP1/2, LMP2/7, and B2M), as well as PD-L1. In the 4T1 mouse model, which possesses normal immune function, the combination of compounds 133 and 135 with anti-PD-L1 immune checkpoint blockade therapy markedly enhanced the anti-tumour effect, indicating a synergistic interaction. Furthermore, in the MDA-MB-436 and MDA-MB-231 xenograft tumour models, the administration of either 133 (15 mg/kg) or 135 (50 mg/kg) as a single agent significantly inhibited tumour growth and substantially reduced the expression levels of the Ki67 proliferation marker.
RBN-2397 (17, Figure 21) is a selective PARP7 inhibitor with oral bioavailability and is currently in Phase II clinical trials274. The co-crystal structure of RBN-2397 and PARP7 shows that while the trifluoromethyl group interacts with Gly573, this bond isn’t crucial for PARP7 inhibition, indicating it could be optimised structurally275. In 2024, Ye et al. replaced the trifluoromethyl group in RBN-2397 with an isoxazole carboxylic acid moiety derived from vorinostat (138), known for its HDAC inhibitory properties, thus synthesising the lead compound 139 (Figure 21)276. This novel compound not only maintained its potent inhibitory efficacy against PARP7, with an IC50 value of 2.7 nM, but also exhibited substantial inhibitory activity against HDAC enzymes, specifically HDAC1 and HDAC6, with IC50 values of 34.0 nM and 83.7 nM, respectively. Research on the SAR has demonstrated that substituting the pyrimidine ring with either a pyridine ring (compound 140, Figure 21) or a pyrazine ring (compound 141, Figure 21) results in a modest reduction in PARP7 inhibitory activity and a significant decrease in inhibitory activity against HDAC1 (IC50 > 120 nM). This finding underscores the importance of the presence and positioning of the nitrogen atom within the pyrimidine ring for preserving dual enzyme inhibitory activity. Furthermore, replacing the piperazine ring with other rigid aliphatic rings (compounds 142–144, Figure 21) also leads to diminished inhibitory activities against both PARP7 and HDAC1/6. Studies on the structure-activity relationship of the pyrimidine ring and isoxazole acid chain showed that increasing the carbon atoms in the amide chain (compounds 145–147, Figure 21) enhanced enzyme inhibition, with compound 147 having IC50 values of 4.5 nM for PARP7 and 32.0/9.7 nM for HDAC1/6. Adding a pyrazole ring (compound 148) improved HDAC1/6 inhibition (IC50 = 3.5/4.2 nM) but slightly reduced PARP7 inhibition (IC50 = 5.5 nM). Compound 149 (Figure 21), linked by an ether bond, maintained PARP7 inhibition and increased HDAC6 inhibition (IC50 = 6.4 nM) but decreased HDAC1 inhibition (IC50 = 50.2 nM). Compound 150 (Figure 21), with an ether bond between the pyrimidine ring and benzyl group, showed balanced dual-target inhibition (PARP7, IC50 = 3.1 nM; HDAC1/6, IC50 = 35.0/6.7 nM) and effective anti-tumour activity, making it a candidate molecule. Mechanism studies indicate that compound 150 promotes IFN-β and CXCL10 expression by enhancing histone H3 acetylation and STAT1 phosphorylation, though it is less effective than the positive control RBN-2397. Selectivity studies show that compound 150 has a pan-HDAC inhibition profile akin to SAHA, with some selectivity for HDAC6 and PARP7. In the CT26 xenograft model, it achieved a tumour growth inhibition rate of 32.7% at 75 mg/(kg·day), outperforming SAHA (27.8%) and RBN-2397 (28.1%) without causing significant weight loss. However, its oral bioavailability is low (F% = 2.35), necessitating further development to improve its pharmacokinetics.
In preclinical models of Ewing’s sarcoma, the novel bifunctional PARP-HDAC inhibitor KT-3283 (structure undisclosed) exhibited markedly enhanced anti-tumour efficacy compared to the individual agents olaparib or vorinostat, through concurrent targeting of PARP1/2 and HDAC enzymes277. KT-3283 showed high cytotoxicity with an EC50 of 0.0163 μM, caused S/G2-M phase cell cycle arrest at concentrations of 0.2 μM or higher, and induced DNA damage. In three-dimensional tumour spheroid models, KT-3283 demonstrated growth inhibitory potency 30–40 times greater than that of the single agents, with an EC50 as low as 0.053 μM. Furthermore, it effectively inhibited tumour cell colonisation in an ex vivo lung metastasis model at a concentration of 10 nM, while no significant toxicity was observed in murine studies. These findings provide a theoretical foundation for the clinical translation of KT-3283.

Dual PARP-EZH2 inhibitors
Research shows that combining PARP and EZH2 inhibitors creates a synergistic epigenetic effect, reducing tumour cell proliferation, invasion, and metastasis by altering gene expression278. Given the limited effectiveness of PARP inhibitors alone in BRCA wild-type TNBC, researchers are integrating an EZH2 inhibitory fragment into PARP inhibitors to enhance their sensitivity in BRCA wild-type cells, aiming for synthetic lethality279.
In 2021, Kong et al. developed a series of dual-target molecules exhibiting inhibitory activities against both EZH2 and PARP by linking the key pharmacophore of tazemetostat (151) with the pyrazine pharmacophore of olaparib through a pyridine-substituted phenyl linker (Figure 22)280. Their investigation primarily focused on the structure-activity relationship of the substituted phenyl within the linker. Initial findings indicated that the 2-methyl-3-amino substituted phenyl linker demonstrated superior cellular activity, prompting further modifications to the 3-amino group. Subsequent results revealed that compounds (154–156) featuring only acylation of the amino group exhibited significantly enhanced enzyme inhibitory activity compared to those (152, 153) subjected to both acylation and alkylation. Moreover, substituting the phenyl ring with either an indole or indazole ring resulted in a loss of activity. The length of the acyl chain was also found to significantly influence activity, with the acetyl compound (154, Figure 22) displaying the highest efficacy, notably surpassing that of the propionyl (155, Figure 22) and butyryl (156, Figure 22) derivatives. Consequently, 154 was identified as the candidate molecule for further development. In vitro experiments demonstrated that 154 exhibited significantly enhanced inhibitory activity against BRCA wild-type TNBC cell lines MDA-MB-468 (IC50 = 0.41 μM) and MDA-MB-231 (IC50 = 2.63 μM), with potencies approximately 80-fold and 15-fold greater, respectively, than those of olaparib. Furthermore, compound 154 markedly outperformed the combined treatment of olaparib with either tazemetostat or GSK126. Mechanistic investigations revealed that 154 induced autophagic cell death in TNBC cells, as evidenced by the downregulation of the autophagy marker protein P62 and the upregulation of LC3B-II/Beclin-1 expression. Despite these promising results, the compound exhibited low oral bioavailability due to poor permeability in Caco-2 cells, necessitating further optimisation.
In 2024, Yu et al. synthesised a series of dual-target inhibitors exhibiting inhibitory activities against both PARP and EZH2281. This was achieved by retaining the benzimidazole-4-carboxamide moiety from veliparib and the pyridone-benzamide moiety from tazemetostat, employing a fragment-based combination strategy (Figure 22). In a structure-activity relationship study focusing on the linker region, researchers observed that among the initial set of 26 synthesised compounds, those incorporating alkyl chains or N-heterocyclic linkers exhibited weak antiproliferative activity (IC50 > 23 μM) across various tumour cell lines. In contrast, compounds featuring benzyl linkers (compounds 157–140, Figure 22) and phenyl linkers (compounds 161–164, Figure 22) demonstrated significantly enhanced activity. Subsequent modifications, involving the addition of a benzamide group or an indazole ring substituent to the benzyl/phenyl linkers, identified compound 163 as particularly noteworthy. This compound, characterised by a phenyl linker with an indazole ring isopropyl substituent, exhibited superior enzyme inhibition activity (PARP1, IC50 = 6.89 nM; EZH2, IC50 = 27.34 nM) and cellular activity (MDA-MB-231, IC50 = 2.84 μM; BT-549, IC50 = 0.91 μM), thereby establishing it as the lead compound. Mechanistic studies have demonstrated that this compound enhances the sensitivity of tumour cells to PARP1 inhibition through the suppression of EZH2, thereby indirectly inducing synthetic lethality. Concurrently, it promotes excessive autophagy and cell death by down-regulating P62 and up-regulating LC3B-II. In vivo experiments utilising the MDA-MB-231 xenograft model revealed that the tumour growth inhibition rate achieved 75.94% at a dosage of 50 mg/kg, which was significantly superior to the TGI of 57.24% observed with the combination of niraparib and GSK126.

Dual PARP-TNKS1/2 inhibitors
Research shows that TNKS1/2 inhibitors can increase PARP inhibitor sensitivity by inhibiting the Wnt signalling pathway, creating a “BRCAness” phenotype in BRCA wild-type tumours, and by disrupting alternative telomere lengthening, which heightens genomic instability282.
In 2015, Nomoto et al. reported the first small molecule inhibitor stenoparib (10, Figure 23) that co-targets PARP1/2 (IC50 = 1.0–1.2 nM) and TNKS1/2 (IC50 = 50–120 nM)283. Its inhibitory activity against TNKS was significantly stronger than that of olaparib (IC50 > 3000 nM). In vitro and in vivo experiments confirmed that this compound enhanced the cytotoxicity of DNA repair-deficient tumours through PARP trapping and inhibited the Wnt signalling pathway by stabilising axin, thereby demonstrating single-agent anti-tumour activity and chemosensitization effects in the MDA-MB-436 xenograft tumour model (at a dose of 100 mg/kg). Additionally, stenoparib’s unique chemical structure avoided the typical intestinal toxicity of TNKS inhibitors and, when combined with the MEK inhibitor E6201 in Wnt1-driven tumour models, showed a synergistic anti-tumour effect. In a phase II trial (NCT03878849) for platinum-resistant/refractory advanced ovarian cancer, patients with a drug response prediction (DRP) score > 50% and at least two prior treatments received 600 mg of stenoparib daily. After a median follow-up of 21.8 months, the median OS was not reached, but it was estimated to exceed 25 months, surpassing current treatments, which show OS of 16–16.5 months and 11.5–13 months with standard chemotherapy. Stenoparib also outperformed traditional PARP inhibitors, with one BRCA wild-type patient benefiting for 24 months and another heavily pre-treated, primary platinum-resistant patient surviving over two years. Stenoparib was well-tolerated, causing mild bone marrow suppression, mainly anaemia and neutropenia.
JPI-547 (11, Table 2) is a novel dual-target inhibitor of PARP and tankyrase 1/2, showing strong anti-tumour effects in BRCA2-deficient pancreatic ductal adenocarcinoma (PDAC) models. It effectively traps PARP1 on chromatin, disrupting poly-ADP-ribosylation, causing G2/M cell cycle arrest and apoptosis in PDAC cells. Additionally, in RNF43-mutated PDAC cells reliant on Wnt signalling, JPI-547 enhances sensitivity without affecting homologous recombination repair, while also disrupting the Wnt/β-catenin and oncogenic YAP pathways. These findings suggest the potential of JPI-547 as a multi-mechanism therapeutic strategy for PDAC characterised by homologous recombination deficiency or dependence on Wnt signalling284. The first-in-human phase I trial of JPI-547 (11, Table 2) in patients with advanced solid tumours aimed to evaluate safety, tolerability, and determine the recommended phase II dose (RP2D). Secondary goals included assessing pharmacokinetics and preliminary anti-tumour activity. In the dose-expansion phase, 39 evaluable patients showed a confirmed ORR of 28.2% (11 partial responses) and a DCR of 64.1%. The mPFS was 3.5 months, and the median duration of response (mDoR) was 3.4 months. Notably, one ovarian cancer patient, previously treated with olaparib, had a confirmed partial response with a 37% tumour size reduction. The maximum tolerated dose (MTD) was 200 mg, with the RP2D set at 150 mg. Treatment-related adverse events (TRAEs) occurred in 59.1% of patients, primarily anaemia, thrombocytopenia, and neutropenia285.
In 2020, Gajiwala et al. conducted a study examining the selectivity profiles of four clinically utilised PARP inhibitors—talazoparib, olaparib, niraparib, and veliparib—towards PARP1 and TNKS1 (Figure 23)286. Their findings revealed that at clinically relevant concentrations (approximately 17 nM free plasma concentration), talazoparib exhibited the capacity to bind to both PARP1 and TNKS1. Surface plasmon resonance (SPR) experimental data demonstrated that talazoparib possessed a markedly higher affinity for TNKS1 (Kd = 14 nM), which was found to be 2 to 3 orders of magnitude greater than that of the other inhibitors. Mechanistic analysis informed by structural biology reveals that PARP1 possesses a closed binding pocket, which contributes to the enhanced stability of ligands. The D-loop region of PARP1 is characterised by the presence of three proline residues, imparting significant structural rigidity. In contrast, the D-loop of TNKS1 exhibits considerable flexibility and necessitates conformational rearrangement for ligand binding. Furthermore, the Tyr889 residue in PARP1 is capable of forming van der Waals and π-π interactions with talazoparib, whereas the analogous residue in TNKS1, Gly1206, does not facilitate similar interactions. Notably, talazoparib is also able to establish a distinctive hydrogen bond with structural water molecules via its triazole group, a feature absent in other inhibitors. Olaparib and niraparib must assume a high-energy distorted conformation to bind to TNKS1, which consequently diminishes their binding affinity. In contrast, the basic groups present in niraparib and veliparib can establish favourable interactions with acidic residues, such as Glu766, within the helical domain of PARP1. This interaction is not replicated in TNKS1 due to the absence of corresponding acidic residues287.
In 2022, Xu et al. initiated their research with olaparib and the TNKS inhibitor IWR-1, subsequently designing the lead compound 166 (Figure 23) by substituting the cyclopropyl group of olaparib with the 4-amino-(N-quinolin-8-yl)benzamide structural fragment from IWR-1 (165, Figure 23)286. While this compound demonstrated substantial PARP1 inhibitory activity, achieving 96% inhibition at a concentration of 10 nM, it exhibited insufficient inhibitory activity against TNKS1/2, thereby highlighting the necessity for further optimisation. Consequently, the researchers concentrated on a systematic structure-activity relationship study focusing on the R1 group at the terminus of the pharmacophore targeting TNKS1/2 (Figure 23). Research indicates that compound 169 (Figure 23), which features a para-substituted phenyl group, demonstrates superior inhibitory activities against PARP1 and TNKS1/2 when compared to ortho- (167) and meta- (168) substituted phenyl derivatives, exhibiting IC50 values of 0.49 nM for PARP1 and 39.0 nM/14.0 nM for TNKS1/2, respectively. In contrast, the incorporation of electron-withdrawing groups, such as -F or -CF3, results in a marked reduction in TNKS1/2 inhibitory activity, as evidenced by compounds 170 and 171 (Figure 23). Moreover, substituting the R1 group in compound 169 with a 4-methoxyphenyl group (compound 172, Figure 23) yields exceptional inhibitory activities against PARP1 and TNKS1/2, with IC50 values of 0.25 nM for PARP1 and 13.5 nM/4.15 nM for TNKS1/2, thus positioning it as a promising candidate molecule. In the cellular-level assessment, 172 exhibited low micromolar concentration antiproliferative activity in both BRCA mutant (HCT116) and BRCA wild-type (MDA-MB-231/468) cancer cell lines, with its efficacy significantly surpassing that of monotherapy with olaparib or stenoparib. Mechanistic investigations have demonstrated that 172 enhances the sensitivity of BRCA wild-type tumour cells to PARP inhibitors by down-regulating BRCA1 expression via the inhibition of the Wnt/β-catenin signalling pathway. Furthermore, this compound markedly elevates the expression of the DNA double-strand break marker γH2AX while concurrently diminishing the expression of Rad51, a critical component of homologous recombination repair. This results in a defect in homologous recombination repair, which is associated with G2/M phase cell cycle arrest and the induction of apoptosis. In the HCT116 xenograft tumour model, compound 172 demonstrated a tumour growth inhibition rate of 31.73% at a dosage of 25 mg/kg, which was significantly superior to the TGI observed with the positive control agents olaparib (50 mg/kg, TGI = 15.36%) and stenoparib (50 mg/kg, TGI = 20.17%). Furthermore, at a dosage of 100 mg/kg, compound 172 achieved a TGI of 51.74%.

Dual PARP-NAMPT inhibitors
Drawing upon the synergistic anti-tumour mechanisms associated with PARP and NAMPT inhibitors, Zhang et al. developed and synthesised a series of bifunctional inhibitors, specifically acrylamides (175–179, Figure 24) and imidazopyridine carboxamides (180–184, Figure 24)288. This design retained the phthalazinone pharmacophore characteristic of olaparib while substituting the solvent-exposed region with the critical pharmacophore derived from FK866 (173)/GNE-617 (174), namely (E)-3-(pyridin-3-yl)acrylamide or the imidazo[1,2-a]pyridine-6-carboxamide linker. SAR analysis reveals that the length of the connecting chain, comprising 3 to 6 methylene groups, is essential for preserving the inhibitory activity against PARP1, with an IC50 value ranging from 0.6 to 1.8 nM. Notably, the inhibitory activity significantly diminishes or is completely lost when the connecting chain is limited to only 2 methylene groups. Among the compounds studied, the acrylamide derivative 177 (Figure 24), containing 4 methylene groups, demonstrates the highest inhibitory potency against NAMPT, with an IC50 of 10 nM. In contrast, the imidazopyridine carboxamide compound 184 (Figure 24), which has 6 methylene groups, exhibits a comparatively stronger inhibitory effect against NAMPT, with an IC50 of 18 nM. It is noteworthy that 184 demonstrated superior anti-proliferative activity (IC50 = 960/700 nM) in MDA-MB-436 and MDA-MB-231 cell lines, attributable to its enhanced permeability across cell membranes. Mechanistic investigations further revealed that 184 effectively inhibited both the proliferation and migration of BRCA wild-type TNBC cells, while also promoting apoptosis. This was achieved through a significant depletion of intracellular NAD+ levels, down-regulation of BRCA1/Rad51 expression, induction of HR repair deficiency, and a synergistic accumulation of DNA double-strand breaks.

Dual PARP-PD1/PD-L1 inhibitors
PARP7, a gene activated by AHR ligands, suppresses AHR activity through MARylation. Inhibiting PARP7 can restore type I interferon signalling and improve immune responses against tumours289. Furthermore, PARP inhibitors that target DDR proteins can enhance PD-L1 blockade therapy and promote cytotoxic T cell infiltration via the cGAS-STING pathway290. Thus, combining PARP7 inhibition with PD-L1 blockade may synergistically enhance anti-cancer effects291.
In 2024, Ye et al. implemented a “direct connection” strategy to covalently link the PD-L1 inhibitor BMS-1 (185, Figure 25) with the PARP7 inhibitor RBN-2397 through an amide bond, thereby successfully synthesising the inaugural bifunctional conjugate 186 (Figure 25)292. This compound exhibited limited inhibitory activity against the PD-1/PD-L1 interaction, with an inhibition rate of 37.2% at a concentration of 1 μM, while demonstrating potent inhibitory activity against PARP7, with an IC50 value of 29 nM. Molecular docking analysis revealed that fragment 186, which targets the PD-1/PD-L1 interaction, was unable to effectively penetrate the hydrophobic pocket due to the constraints imposed by the rigid amide bond on its conformational flexibility. This limitation prompted the researchers to undertake a systematic SAR study focused on the connecting region. Research findings demonstrate a significant correlation between the length and flexibility of the linker chain and the enzyme inhibitory activity. Specifically, when the linker chain is composed of alkyl or aryl groups (compounds 187–190, Figure 25), the PD-1/PD-L1 inhibitory activity is relatively weak, with an IC50 greater than 1 μM. In contrast, the incorporation of a flexible polyethylene glycol (PEG) chain as the linker arm (compounds 191–194, Figure 25) markedly enhances the PD-L1 inhibitory activity of the molecule. Notably, compound 193, which utilises PEG3 as the linker arm, exhibits a balanced dual-target inhibitory profile, with IC50 values of 426 nM for PD-1/PD-L1 and 2.5 nM for PARP7. Subsequently, the researchers conducted a comprehensive investigation into the structure-activity relationship of the fragments targeting the PD-1/PD-L1 interaction. Their findings revealed that compound 195 (Figure 25), which features a pyridine methyl ether substitution, demonstrated an optimal balance of activity (PD-1/PD-L1, IC50 = 343 nM; PARP7, IC50 = 7.1 nM). In contrast, substitutions with cyano or trifluoromethyl phenyl groups resulted in diminished enzyme inhibition activity. Furthermore, the study confirmed the essentiality of the pyridine-piperazine core structure, as modifications involving open-chain or spiro configurations led to a complete loss of PARP7 inhibitory activity. Building upon compound 195, researchers endeavoured to substitute various linker arms (compounds 196–199, Figure 25). Their findings revealed that the analog 196, which incorporates a piperazine structure, demonstrated remarkable inhibitory activity against the PD-1/PD-L1 interaction (IC50 = 75 nM) while simultaneously preserving potent PARP7 inhibition (IC50 = 1.3 nM). Although the replacement of the piperazine moiety in compound 195 with alkyl chains or a piperidine structure (compounds 196–199) resulted in a significant decrease in activity, these modified compounds still exhibited improved efficacy compared to compound 193. In the hPD-L1 human peripheral blood mononuclear cells (PBMCs) co-culture system, compounds 193 and 195 effectively promoted IFN-γ secretion at 1.0 μM and showed increased effects at 2.0 μM, enhancing secretion by 44.3% and 69.7%, respectively, thus confirming their ability to block the PD-1/PD-L1 pathway and restore T cell function. In contrast, compound 196 showed no effect even at 2.0 μM, indicating a lack of targeted activity. Therefore, compounds 193 and 195 were chosen as candidate molecules. Mechanistic studies further demonstrate that within the co-culture system of MDA-MB-231 and Jurkat T cells, compounds 193 and 195 significantly enhance the cytotoxic efficacy of T cells against tumour cells, as evidenced by an IC50 value of less than 2.5 μM. In the B16-F10 melanoma mouse model, administered at a dosage of 25 mg/kg, the tumour growth inhibition rates of compounds 193 (TGI% = 40.43) and 195 (TGI% = 40.04) were markedly superior to those of the individual agents BMS-1 (TGI% = 3.24) and RBN-2397 (TGI% = 3.94), with no significant toxicity observed.

Dual PARP-proteasome inhibitors
The proteasome maintains protein balance and regulates cellular processes via the ubiquitin-proteasome pathway293. Combining proteasome inhibitors like bortezomib with PARP inhibitors synergistically impacts multiple myeloma by: (i) enhancing DNA repair inhibition through proteasome-mediated damage to DNA repair proteins294, and (ii) jointly modulating the tumour microenvironment to boost anti-tumour immunity295. A supportive phase I clinical trial (NCT01495351) has been completed.
In 2024, Yu et al. successfully synthesised the lead compound 201 (Figure 26), characterised by a PARP1 IC50 of 4.1 nM and a 20S proteasome IC50 of 386 nM, through the fusion of the pharmacophore from the proteasome inhibitor ixazomib (200, Figure 26) with olaparib via a benzamide linker296. Subsequent modification, involving the substitution of the boronic acid moiety with a boronate ester (compound 202, Figure 26), resulted in a marked enhancement of dual-target inhibitory activity, achieving a PARP1 IC50 of 2.1 nM and a 20S proteasome IC50 of 157 nM. Conversely, the replacement with an epoxide ketone led to a loss of activity. Further optimisation of the connecting chain, utilising aryl glycine, yielded compound 203 (Figure 26), which exhibited a 20S proteasome IC50 of 42.6 nM and a PARP1 IC50 of 28 nM, alongside nanomolar-level antiproliferative efficacy across various tumour cell lines (MDA-MB-231, IC50 = 296.8 nM; BT549, IC50 = 216 nM). Subsequent substitutions with isoleucine (compound 204, Figure 26) or methionine (compound 205, Figure 26) resulted in diminished enzyme inhibitory activity; however, these modifications significantly enhanced cellular activity, leading to their selection as candidate molecules for further development. Research on the underlying mechanisms indicates that compounds 204 and 205 diminish the efficacy of DSB repair by downregulating the expression of critical proteins, specifically BRCA1 and RAD51, within the HR repair pathway. Concurrently, these compounds deplete nuclear ubiquitin levels, thereby hindering the recruitment of the BRCA1/RAP80 complex to sites of DNA damage. Notably, compound 205 demonstrates a greater capacity than compound 204 to induce apoptosis and inhibit colony formation, thereby offering a novel strategy to address resistance to PARP inhibitors.

Dual PARP-ATR inhibitors
Preclinical studies show that combining PARP and ATR inhibitors has a synergistic anti-tumour effect in ovarian cancer, particularly in wild-type tumours203. ATR inhibition is crucial for overcoming PARP inhibitor resistance by restoring homologous recombination repair, stabilising replication forks, reversing SLFN11 inactivation, and compensating for PARG deficiency297,298. Molecular evidence includes reduced pATR/pCHK1, decreased RAD51 recruitment, and increased γH2AX accumulation299. Numerous clinical trials are currently evaluating the safety and efficacy of this combination therapy for various tumours, showing promising potential300,301.
Building upon a clinically validated synergistic strategy, Ye et al. modified the pyrrolopyridine core of the ATR inhibitor AZD6738 (206, Figure 27) and the phthalazinone moiety of olaparib302. Specifically, they substituted the iminosulfone group in AZD6738 with an amino group to yield the lead compound 207 (Figure 27), and they replaced the cyclopropyl group with a methylene group to generate compound 208 (Figure 27). Both compounds exhibit moderate inhibitory activities against ATR and PARP1 enzymes, with IC50 values of 215/39.5 nM and 1149/420 nM, respectively; however, their cellular activities are comparatively weak. Subsequently, compounds 209 and 210 (Figure 27) were synthesised by extending the connecting chain, which resulted in a notable enhancement of their PARP1 inhibitory activity; however, their cellular efficacy remained inadequate. The further incorporation of aryl-substituted piperazine or sulfonyl-substituted piperidine moieties led to the development of compounds 211 and 212 (Figure 27). The results indicated a substantial improvement in their PARP1 inhibitory activity, with IC50 values of 0.56 nM and 0.46 nM, respectively. Additionally, compound 211 exhibited enhanced ATR inhibitory activity (IC50 = 43 nM), whereas the ATR activity of compound 212 was comparable to that of compounds 209 and 210. Both compounds 211 and 212 demonstrated significantly improved activity at the cellular level. To enhance the ATR inhibitory activity, aromatic rings were incorporated into the structures of compounds 211 and 212, resulting in the synthesis of compounds 213 and 214 (Figure 27). Among these, 214 demonstrated remarkable ATR and PARP1 inhibitory activities, with IC50 values of 17.3 nM and 0.38 nM, respectively. Additionally, it exhibited significant cellular activity, with IC50 values of 1.89 μM and 0.32 μM against MDA-MB-231 and MDA-MB-468 cells, respectively, thereby qualifying it as a candidate molecule for further investigation. Selective studies have demonstrated that the ATR selectivity of 214 is comparable to that of AZD6738, albeit with slightly reduced activity. However, 214 exhibits superior selectivity for PARP1 compared to PARP2 and PARP7, outperforming olaparib in this regard. Mechanistic investigations have revealed that 214 induces DNA damage, G2/M phase arrest, and apoptosis in MDA-MB-231 and MDA-MB-468 cell lines, while also effectively inhibiting the migration, invasion, and epithelial-mesenchymal transition of tumour cells. In vivo experiments conducted using the MDA-MB-468 xenograft tumour model have shown that the tumour growth inhibition rate (TGI = 68.36%) achieved with 214 at a dosage of 50 mg/kg surpasses that of the combination treatment group (TGI = 54.88%), with no significant weight loss or organ toxicity observed.

Dual inhibitors of PARP and other druggable targets
Dihydroorotate dehydrogenase (DHODH) and PARP play critical roles in regulating pyrimidine synthesis and DNA repair, respectively. Their synergistic interaction in cell proliferation presents a novel approach for cancer therapy303. In 2015, Rahman et al. developed dual-target inhibitors based on a benzimidazole scaffold, revealing that the R1 substituent significantly influenced target selectivity (Figure 28)304. Specifically, compound 215 (Figure 28), which features a carboxyl group, demonstrated potent inhibition of DHODH (IC50 = 0.013 μM) but exhibited diminished PARP1 activity (IC50 = 44 μM). This selectivity was attributed to the formation of hydrogen bonds between the carboxylate anion and the residues Arg136, Gln47, and Thr360 of DHODH. Conversely, compound 216 (Figure 28), which incorporates a formamide group, achieved substantial PARP1 inhibition (IC50 = 0.061 μM) through hydrogen bonding with Ser243 and Gly202 of PARP1, as well as π-π stacking interactions with Tyr246. Furthermore, compound 217 (Figure 28), characterised by a biphenyl moiety with an acetamide group at the para position, exhibited moderate dual-target activity, thereby establishing a foundation for further optimisation. However, it is important to note that this study did not include validation of the proposed synergistic mechanism or in vivo assessments.
Aldose reductase (ALR2), a cytoplasmic NADPH-dependent oxidoreductase, serves as the rate-limiting enzyme in the polyol pathway305. Under physiological conditions, ALR2 primarily catalyses the reduction of aldoses, such as glucose, to sorbitol. However, in the context of diabetes, its excessive activation results in the accumulation of sorbitol, which can induce osmotic damage and contribute to the development of diabetic complications306. Mechanistic studies have demonstrated that hyperglycaemia can synergistically activate the polyol pathway, involving ALR2, alongside the DNA damage pathway mediated by PARP1307. The co-expression of these pathways in retinal cells may lead to a synergistic exacerbation of disease. Consequently, dual-target inhibitors aimed at PARP1 and ALR2 hold promise as a novel therapeutic strategy for diabetic retinopathy. In 2017, Chadha et al. utilised thiazolidinedione as the principal pharmacophore and incorporated an indole-benzyl moiety to effectively engage both the hydrophobic pocket (comprising Trp111, Phe122, and Leu300) of ALR2 and the aromatic residues (Tyr246 and His201) of PARP1308. Additionally, they introduced a 2-chloro substitution on the benzene ring of compound 218 (Figure 28) to enhance hydrophobic interactions, ultimately leading to the identification of a lead compound with low micromolar inhibitory activity against both targets (ALR2/PARP-1 IC50 values of 4.72 and 1.34 μM, respectively). Previous research on ALR2 has primarily examined its link to diabetic complications. Recent findings indicate that ALR2 enhances SLC711 expression by activating STAT3, which promotes glutathione synthesis and contributes to tumour progression. Inhibiting ALR2 can disrupt glutathione synthesis, potentially overcoming resistance to EGFR-targeted therapy in lung cancer309. This study suggests a multi-target approach for diabetic retinopathy and supports the development of new anti-tumour drugs targeting ALR2.
Alzheimer’s disease (AD) has a complex pathology. Current acetylcholinesterase (AChE) inhibitors, like donepezil and rivastigmine, only ease symptoms without halting neuron damage310. New treatments are urgently needed. Research indicates that PARP1 over-activation in neurodegeneration depletes NAD+ and ATP, disrupting energy metabolism, causing mitochondrial dysfunction, and leading to neuron death311. This is linked to early AD events like oxidative stress and mitochondrial damage. Preclinical studies suggest PARP1 inhibitors might protect neurons, making them potential AD treatments312. In advanced AD, AChE levels drop while butyrylcholinesterase (BChE) increases, becoming the main enzyme breaking down acetylcholine (ACh). Thus, BChE inhibitors could be more effective for advanced AD313. Consequently, the development of bifunctional drugs that concurrently inhibit PARP1, thereby mitigating neurodegeneration, and target cholinesterase to elevate acetylcholine levels, is anticipated to synergistically enhance the pathological progression of AD. In 2017, Goodfellow et al. modified the 4-benzylphthalazinone framework of olaparib by substituting the cyclopropyl group with an aryl vinyl group to enhance hydrophobicity and introduce an α,β-unsaturated carbonyl moiety314. Subsequent structure-activity relationship studies revealed that compound 219 (Figure 28), featuring a nitro-substituted benzene ring, exhibited the highest inhibitory activity against PARP1/BChE, with IC50 values of 0.0161 μM and 9.16 μM, respectively. In contrast, the introduction of halogen or bulky substituents, such as trimethoxy groups, resulted in a reduction of inhibitory activity. Molecular docking studies showed that compound 219’s phthalazinone ring formed crucial hydrogen bonds with Gly863 and Ser904 of PARP1, and its acryloyl C = O group interacted with Arg878. Despite a beneficial hydrogen bond between the nitro group and Gly871, compound 219 was less effective than olaparib, likely due to missing interaction with Tyr896. For human butyrylcholinesterase, the phthalazinone ring had hydrophobic interactions with Leu286 and Gly117 but failed to engage key residues Trp82 and Trp231. A hydrogen bond with Tyr128 might have weakened these interactions, explaining the reduced BChE inhibition compared to PARP1. Recent research highlights cholinesterase’s crucial role in regulating the glioma microenvironment, particularly in neuron-tumour interactions315. This study suggests a new therapeutic approach targeting PARP1 and cholinesterase, paving the way for dual-target glioma drugs.
To address the limitations of single-agent efficacy and the issue of drug resistance associated with PARP inhibitors, Bertrand et al. have developed bifunctional molecules that simultaneously target PARP and DNA316. This innovative approach seeks to overcome the efficacy bottleneck by enhancing the synergistic effects of DNA damage and repair inhibition, while circumventing the challenges related to cell penetration and extrusion mechanisms that are prevalent in traditional combination therapies involving PARP inhibitors and DNA-damaging agents317. In this study, a DNA alkylating triazene group, capable of releasing methyl diazonium ions or spontaneously cyclizing to form chloroethyl triazeneium ions, was conjugated to the PARP inhibitor scaffold, 4-amino-1,8-naphthalenedicarboximide, to design the bifunctional molecule 220 (Figure 28). This compound exhibits cooperative inhibition of PARP1 (IC50 = 1.7 μM) and induces DNA breaks, resulting in a significant enhancement of cytotoxicity (2–20 times) in BRCA2-deficient VC8 cells, while preserving a 25-fold selectivity for tumour cells. Fluorescence microscopy imaging further demonstrated that the compound does not accumulate in the nucleus, with its DNA damaging effects being attributed to the presence of the triazene group.
RAD51, a pivotal effector in the HR repair pathway, is situated downstream of the BRCA1 and BRCA2 genes318. Research has demonstrated that the inhibition of RAD51 can increase the sensitivity of breast cancer cells to the PARP inhibitor olaparib169. Furthermore, the concurrent targeting of both PARP and RAD51 may potentially reverse the resistance of tumour cells to PARP inhibitors319. Drawing upon the principle of synthetic lethality, Wiesmuller et al. developed novel bifunctional inhibitors utilising a molecular hybridisation strategy320. This approach involved retaining the phthalazinone pharmacophore of olaparib while covalently linking it to the key structural unit of Rad51 inhibitors through linker chains of varying lengths, thereby generating compounds with dual inhibitory activities against PARP1 and Rad51. Notably, compound 221 (Figure 28), which is directly linked via an amide bond, demonstrated substantial antiproliferative activity in HCC-1937 (BRCA1 mutant TNBC) and MCF-7 (luminal type) cell lines, with IC50 values of 0.31 μM and 0.86 μM, respectively. However, the study’s limitations include an unclear dual-target inhibition mechanism and unverified spatial proximity effects on bifunctional inhibition.
Research has demonstrated that the inhibition of PARP results in the accumulation of single-strand DNA breaks. Concurrently, inhibitors of Hsp90 can downregulate the expression of BRCA, a critical component in the homologous recombination repair pathway321. The combined application of these inhibitors can induce synthetic lethality, which is particularly effective in tumour cells possessing intact homologous recombination repair mechanisms322. In 2020, Wu et al. retained the critical pharmacophoric elements of olaparib and utilised a molecular hybridisation approach to integrate it with the curcumin derivative C0817, recognised for its Hsp90 inhibitory properties323. Concurrently, they identified piperidin-4-one as a pivotal scaffold to replicate the piperazine structure of olaparib, facilitating occupancy of the critical binding site of PARP. This approach culminated in the development of a small molecule inhibitor exhibiting dual inhibitory activity against PARP and Hsp90. Initially, researchers synthesised substituted benzene rings or pyridine at the R1 position (Figure 28), resulting in the identification of 11 active molecules. Following this, they performed cell viability assays across various tumour cell lines. The findings demonstrated that compounds featuring a pyridine substituent at the R1 position (specifically compounds 222–224, Figure 28) exhibited significant anti-proliferative activity. In contrast, compounds with an R1 substituent consisting of a benzene ring containing methoxy or hydroxy groups displayed a loss of cellular activity. Mechanistic investigations indicated that compound 224 was capable of concurrently inhibiting PARP and binding to Hsp90, thereby inducing a “BRCAness” phenotype through the down-regulation of BRCA1 expression. It is important to highlight that the present study does not adequately address the molecular mechanisms underlying the action of the candidate compounds, nor does it provide in vivo pharmacodynamic validation using animal models. Furthermore, there has been no investigation into the toxicity of compound 224 concerning adverse reactions, such as hepatotoxicity and cardiotoxicity, associated with Hsp90 monotherapy.
The RAS gene family, comprising HRAS, KRAS, and NRAS, encodes proteins that function as pivotal molecular switches, regulating downstream signalling pathways through the binding and hydrolysis of GTP/GDP. These proteins are integral to processes of cellular growth, differentiation, and survival. Mutations within RAS genes can result in the constitutive activation of their encoded proteins, leading to the over-activation of downstream pathways such as RAF/MEK/ERK, thereby facilitating tumour initiation and progression324. It is hypothesised that RAS pathway inhibitors and PARP inhibitors may exhibit a synergistic effect, potentially through mechanisms involving the induction of apoptosis, disruption of DNA repair pathways, and attenuation of DNA damage checkpoint activities325. However, the precise mechanisms underlying this synergy require further investigation. In 2024, Huang et al. undertook the design and synthesis of a series of molecules by integrating the pharmacophore of olaparib with the styrylbenzene sulphonamide pharmacophore of rigosertib, known for its RAS signalling pathway interference activity326. These molecules were linked via alkyl chains of varying lengths to confer dual functionality: RAS signalling pathway interference and PARP inhibition. Notably, compound 225 (Figure 28), featuring a four-carbon alkyl chain, demonstrated superior in vitro anti-proliferative efficacy against various breast cancer cell lines. Specifically, its inhibitory potency against BRCA wild-type (MDA-MB-231, IC50 = 0.75 μM) and BRCA-deficient (HCC1937, IC50 = 1.17 μM) TNBC cells was markedly enhanced compared to olaparib (MDA-MB-231, IC50 = 25.46 μM), exhibiting an approximate 34-fold improvement in activity. Mechanistic studies have demonstrated that this compound exerts its anti-tumour effects through the synergistic inhibition of PARP1 and CRAF proteins, leading to the induction of reactive oxygen species (ROS)-mediated DNA damage, collapse of mitochondrial membrane potential, and subsequent apoptosis. In vivo experiments further validated these findings, revealing that in the MDA-MB-231 xenograft tumour model, the tumour growth inhibition rate of compound 225 at a dosage of 38 mg/kg reached 61.3%. This efficacy surpassed that of olaparib alone (TGI 38.5%), rigosertib derivative (TGI 51.8%), and the combination of the two agents (TGI 56.7%), while also demonstrating a lack of significant systemic toxicity.

Conclusion
Given the pivotal regulatory function of PARP1 within the DNA damage repair pathway, the development of PARP inhibitors with dual-target inhibition capabilities represents an innovative strategy to address the limitations associated with monotherapy327. These agents markedly enhance anti-tumour efficacy through the synergistic interaction of multiple targets, simultaneously mitigating the risk of drug resistance and broadening the spectrum of clinical applications. Research into molecular mechanisms has demonstrated that PARP inhibitors can produce substantial synergistic effects when administered in conjunction with various anti-tumour therapeutic agents. These include conventional chemotherapeutic agents, compounds that target cell cycle regulatory factors (e.g., CDK), agents that inhibit epigenetic regulatory factors (such as BRD4, HDAC, and EZH2), and targeted therapies aimed at critical signalling pathways, including EGFR, VEGFR, c-Met, PI3K, and Hsp90. Moreover, the combination therapy regimens in question have demonstrated favourable safety profiles in clinical trials328. Building on the theoretical framework of this combination therapy, the structural optimisation of PARP inhibitors can be achieved through pharmacophore chimerization strategies, which include connection, fusion, and merger modes. This approach seeks to maintain the intrinsic PARP inhibitory activity while integrating biological functions that target an additional pathway, thereby enabling the design of novel chemical entities with dual anti-tumour activities. A multitude of clinical research findings have shown that PARP-based dual-target therapies provide substantial therapeutic benefits and favourable tolerance profiles in the treatment of solid tumours329. This development marks a significant advancement in the clinical application of dual-target drug strategies aimed at synergistic intervention, transitioning from conceptual validation to practical implementation.
However, opportunities and challenges are interconnected and exert mutual influence, collectively shaping the complex landscape of future developments in PARP dual-target inhibitors. This involves three key aspects:The development of dual-target PARP inhibitors, which combine tumour-related targets, is a significant scientific challenge due to the complexity of potential gene combinations. Current efforts focus on clinical research and phenotype-driven screening, but the sheer volume of genetic data makes purely experimental approaches inefficient. To address this, the academic community has created a multi-dimensional computational biology framework that uses statistical modelling, network analysis, and AI to streamline the identification of synthetic lethal gene pairs, providing a theoretical strategy for selecting targets in PARP inhibitor development330. Additionally, the CRISPR-Cas9 high-throughput screening platform is extensively used for large-scale experimental validation. It efficiently identifies new synthetic lethal interactions and aids in developing predictive biomarkers, thus advancing a comprehensive cancer treatment strategy that includes biological validation, therapeutic optimisation, and clinical evaluation331.

The success of dual-target PARP inhibitors in clinical use depends on the balanced interplay of their efficacy, pharmacokinetics, and toxicity. During drug design, it is essential to analyse the spatial configuration of key pharmacophores in the target protein and identify shared motifs with similar coordination patterns. This allows for the structural integration of dual lead ligands while preserving the three-dimensional characteristics and interaction distances of the original pharmacophores through molecular fusion or functional connections. Redundant structural domains are then removed to control molecular volume and mass. Finally, optimising molecular flexibility and lipophilicity is crucial to prevent pharmacokinetic issues like poor drug absorption and distribution imbalances332.

The development of current PARP dual-target inhibitors predominantly employs a structural fusion strategy that integrates core pharmacophores with auxiliary target ligands. While this approach can produce candidate compounds exhibiting dual-target inhibitory activity, significant challenges remain in achieving molecular structural innovation. These challenges impede the advancement of novel chemical entities within the diverse landscape of medicinal chemistry. Through iterative optimisation and innovation in drug discovery models, the paradigm of small molecule drug screening has evolved from traditional approaches, such as screening based on known active molecule libraries and high-throughput screening (HTS), to the incorporation of advanced methodologies. These include structure-based/fragment-based rational drug design (SBDD/FBDD), DNA-encoded compound library (DEL) technology, and PROTAC333,334. This technological evolution has markedly increased the efficiency of new drug research and development by enhancing the discovery rate of lead compounds, reducing the research and development timeline, and optimising associated costs. Notably, the advancement of novel therapeutic agents, exemplified by PARP dual-target inhibitors, is benefiting from this technological integration. Looking ahead, it is anticipated that new molecular entities with improved therapeutic indices, precise targeting capabilities, and optimised toxicity profiles will be developed, offering innovative solutions to address tumour heterogeneity and acquired drug resistance.

This article provides a comprehensive systematic review of the research and development advancements related to PARP dual-target inhibitors. Through a detailed analysis of their mechanisms of action, it elucidates the molecular-level biological functions of the targets and their relevance to disease pathologies. Additionally, the article examines the identification of lead compounds and the optimisation processes aimed at enhancing pharmacological efficacy, while also addressing the design challenges and prospective developmental trajectories for the next generation of dual-target inhibitors.