Human CtIP: A ’double agent’ in DNA repair and tumorigenesis

: Human CtIP was originally identified as an interactor of the retinoblastoma protein and BRCA1, two bona fide tumour suppressors frequently mutated in cancer. CtIP is renowned for its role in the resection of DNA double-strand breaks (DSBs) during homologous recombination, a largely error-free DNA repair pathway crucial in maintaining genome integrity. However, CtIP-dependent DNA end resection is equally accountable for alternative end-joining, a mutagenic DSB repair mechanism implicated in oncogenic chromosomal translocations. In addition, CtIP contributes to transcriptional regulation of G1/S transition, DNA damage checkpoint signalling, and replication fork protection pathways. In this review, we present a perspective on the current state of knowledge regarding the tumour-suppressive and oncogenic properties of CtIP and provide an overview of their relevance for cancer development, progression, and therapy. Human CtIP was originally identified as an interactor of the retinoblastoma protein and BRCA1, two bona fide tumour suppressors frequently mutated in cancer. CtIP is renowned for its role in the resection of DNA double-strand breaks (DSBs) during homologous recombination, a largely error-free DNA repair pathway crucial in maintaining genome integrity. However, CtIP-dependent DNA end resection is equally accountable for alternative end-joining, a mutagenic DSB repair mechanism implicated in oncogenic chromosomal translocations. In addition, CtIP contributes to transcriptional regulation of G1/S transition, DNA damage checkpoint signalling, and replication fork protection pathways. In this review, we present a perspective on the current state of knowledge regarding the tumour-suppressive and oncogenic properties of CtIP and provide an overview of their relevance for cancer development, progression, and therapy.


Introduction
Human CtIP is an 897-amino acid nuclear adaptor protein with multiple roles in DNA metabolism and genome stability [1,2]. Initially thought to function exclusively as a regulator of gene expression, continued research has revealed that CtIP plays a significant role in DNA double-strand break (DSB) repair pathway choice, with opposing consequences for genome stability and tumorigenesis. Herein, we review the current literature on CtIP's molecular and biological activities with a particular focus on their clinical relevance for cancer progression and treatment outcomes.

CtIP: History at a glance
Human CtIP was first isolated in 1998 using yeast two-hybrid screening experiments performed in four independent laboratories. Schaeper et al. identified CtIP as an interacting protein of the oncogenic transcriptional corepressor CtBP. Of note, CtIP binds CtBP through a PLDLS motif conserved among E1A proteins of human adenoviruses [3]. Fusco et al. called the same polypeptide RIM (encoded by RBBP8 located at 18q11.2) by virtue of its ability to interact with the Rb tumour suppressor protein via a conserved LxCxE Rb-binding motif [4]. Shortly thereafter, two studies reported CtIP to specifically interact with the BRCT domain of BRCA1 [5,6]. Together, these early findings implicated CtIP in suppressing tumorigenesis, possibly by regulating the transcription of cell cycle-related genes [7].
Typical for multivalent adaptor proteins, CtIP contains various short linear sequence motifs facilitating interactions between protein-binding partners, some of which are governed by CtIP phosphorylation (Fig. 1) [2,24]. For instance, CDK-mediated phosphorylation of CtIP at S327 is a prerequisite for the binding of BRCA1 BRCT domains and has been proposed to speed up DSB resection in S/G2-cells [25,26]. In contrast, CDK-dependent phosphorylation of CtIP at T847 is indispensable for DNA end resection, most likely by fostering MRN-CtIP complex formation [23,27]. Two additional CDK phosphorylation sites in CtIP, S276 and T315, are required for the binding of PIN1, a unique peptidyl-prolyl cis-trans isomerase found to be overexpressed in many cancers and to promote uncontrolled cell proliferation [28][29][30]. CtIP isomerisation has been proposed to facilitate CUL3-KLHL15-mediated CtIP ubiquitination and subsequent proteasomal degradation to limit CtIP's resection activity [28,31]. APC/C-Cdh1 recognizes CtIP through a conserved KEN box motif and targets CtIP for ubiquitin-proteasome-mediated degradation in G2, presumably to prevent inappropriate DNA resection in mitosis [32]. Following genotoxic insults, CtIP phosphorylation at T859 by the apical DNA damage response (DDR) kinases ATM and ATR is important for stable chromatin association of CtIP and efficient DSB processing [33,34]. Remarkably, a recent biochemical study reported that CtIP phosphorylation on T847 and T859 is critical for MRN endonuclease-mediated removal of DNA-PK from DNA ends to allow resection and repair of DSBs by HR [35]. CtIP also directly associates with FANCD2, the central mediator of the Fanconi Anemia tumour suppressor pathway responsible for the repair of DNA interstrand crosslinks [36,37]. Finally, CtIP structural interaction with the breast cancer oncogene LMO4 has been hypothesised to sequester CtIP away from DNA repair complexes [38].
CtIP expression levels are tightly regulated throughout the cell cycle, remaining low in G1 and increasing in S and G2 phases, when sister chromatids are available as a template for HR repair [1,8]. Besides the aforementioned mechanisms controlling CtIP protein turnover by ubiquitin-proteasomal processing, there is accumulating evidence that CtIP abundance is also regulated at the level of transcription. Liu et al. showed that CtIP can bind to its own promoter, upregulating its expression during G1/S progression (Fig. 2) [39]. Moreover, several microRNAs have been shown to silence CtIP expression. Martin et al. reported that miR-335-mediated suppression of CtIP is released following ionising radiation (IR)-induced ATM activation, classifying miR-335 as a potential radiosensitiser for clinical use [40,41]. Similarly, CtIP transcripts are negatively regulated by miR-19, a member of the miR-17~92 cluster whose expression is upregulated in many cancer types [42,43]. More recently, downregulation of CtIP by miR223-3p was identified as a protective mechanism against chromosomal rearrangements in developing haematopoietic cells [44].

Versatile Roles of CtIP in Genome Maintenance
CtIP's ability to specifically interact with multiple binding partners ( Fig. 1) correlates well with its function in diverse genome maintenance mechanisms, including cell cycle control, DNA damage repair, and DNA replication (Fig. 2) [1,2,8].

Transcriptional regulation and cell cycle checkpoints
Early studies connecting the CtIP-CtBP corepressor complex to Rb family and BRCA1 proteins suggested a prominent role of CtIP in transcriptional regulation of cell cycle events (Fig. 2) [2]. For instance, the interaction between CtIP and Rb has been shown to alleviate Rb-mediated transcriptional repression of certain E2F target genes (e.g. CCND1), thus facilitating G1/S transition [39,45]. Conversely, in response to DNA damage, CtIP was reported to bind to the promoter and induce the expression of the CDK inhibitor p21, hence blocking G1/S transition [46]. Moreover, binding of CtIP-CtBP to BRCA1 has been implicated in preventing p21 promoter transactivation by BRCA1, while DNA damage-induced disruption of the BRCA1-CtIP interaction could relieve p21 and GADD45A transcriptional repression [47,48]. Similarly, association of CtIP with BRCA1 and the BRCA1-binding transcriptional repressor ZBRK1 negatively regulates the expression of pro-tumorigenic factors such as ANG1 and HMGA2, and loss of this heterotrimeric complex has been shown to enhance mammary tumorigenesis [49,50]. In addition, binding of the CtIP-LMO4 complex to BRCA1 inhibits BRCA1-mediated transcriptional activation [51]. Consequently, CtIP appears to exert dual control over gene expression and promoter activities depending on its binding partners.
In contrast, Yu and Chen first demonstrated that formation of the CtIP-BRCA1 complex in G2 is required for the IR-induced phosphorylation of Chk1 and activation of the G2/M transition checkpoint [25]. In agreement with these findings, it was recently shown that CtIP promotes G2/M arrest through ATR-Chk1 activation rather than by induction of the p53-p21/GADD45a pathway [52]. ATR-Chk1 signalling in response to DNA damage is initiated by the recruitment of ATR to RPA-coated ssDNA tracks, key intermediates formed during DSB resection ( Fig. 2) [53]. Therefore, while CtIP regulates G1/S transition mostly at the level of transcription, its G2/M checkpoint function is primarily mediated by DNA end resection.

DNA end resection and DSB repair pathway choice
In a process commonly referred to as 'DNA end resection', CtIP associates with the MRN complex to promote limited (5'-3') nucleolytic processing of DSBs, which commits cells to homology-directed repair mechanisms (HR, SSA, and a-EJ) and prevents repair by c-NHEJ (Fig. 2)

Fig. 1. Human CtIP, a polyvalent adaptor protein.
CtIP is composed of a tetramerisation motif (Tet) and an extended coiled-coil region at the N-terminus, an unstructured central region and an Sae2-like domain at the C-terminus. Two non-canonical nuclease motifs harbouring residues implicated in catalytic activity are highlighted by asterisks. Short linear sequence motifs promoting specific protein interactions and DNA binding are indicated by dashed lines. CtIP interactors classified as tumour suppressors and oncogenes are highlighted in green and red letters, respectively. Selected serine/threonine (S/T) phosphorylation sites are indicated. CtIP C-terminal truncation variants identified in Jawad and Seckel (SCKL2) syndrome patients [83] are indicated by jagged lines. Three RBBP8 mutations (Q643P, E804Δ and R805G) associated with early-onset breast cancer (BC) are indicated by straight lines [78]. [12,14,54,55]. However, while HR faithfully restores the original DNA sequence at the break site in most cases, SSA and a-EJ frequently result in deletions and are associated with pathological chromosome rearrangements [14]. Importantly, BRCA1, in complex with its constitutive partner BARD1, counteracts 53BP1 chromatin accumulation to facilitate CtIP-mediated resection, thereby tipping the balance towards the exclusion of c-NHEJ and induction of HR, especially for breaks encountered during DNA replication [56,57]. Extended resection is carried out by dedicated DNA nucleases and helicases, most notably EXO1, DNA2, and BLM [13]. Interestingly, CtIP restrains EXO1 exonuclease activity but accelerates BLM/DNA2-mediated DNA cleavage, presumably via direct protein-protein interactions, thus coordinating both short-and long-range resection processes [60][61][62]. Resulting 3'-ssDNA overhangs are rapidly coated by the heterotrimeric RPA complex, creating a platform for ATR-Chk1 signalling (Fig. 2) [63]. The choice between error-free HR and mutagenic a-EJ and SSA pathways mainly happens downstream of DSB resection [14,64,65]. In the presence of a homologous DNA template, preferentially the intact sister chromatid, BRCA2 mediates RAD51 filament assembly on ssDNA to promote subsequent strand invasion and exchange steps of HR. When HR is inappropriate, exposure of microhomologies between resected DNA arms can drive RPA removal and gap-filling synthesis by Polθ associated with a-EJ. Alternatively, RAD52-mediated annealing of complementary ssDNA ends between repeated sequences occurs during SSA. Remarkably, DSB resection and repair activities of CtIP have also been reported outside S/G2 phases of the cell cycle. For instance, Löbrich and colleagues could demonstrate that resection-dependent c-NHEJ of IR-induced DSBs in G1-cells contributes to chromosomal translocations and is induced by polo-like kinase 3 (PLK3)-dependent CtIP-S327 phosphorylation and interaction between CtIP and BRCA1 [66,67].
In summary, CtIP-mediated DNA end resection is mostly required for accurate repair of DSBs by HR, but also allows SSA, a-EJ and c-NHEJ to create mutagenic repair products, ranging from simple deletions to complex chromosomal rearrangements.

DNA replication stress
There is substantial evidence that besides its fundamental role in DSB resection, CtIP is engaged in different processes safeguarding genome

Fig. 2. CtIP at the crossroads of multiple genome maintenance mechanisms.
Upper left, CtIP is implicated in the transcriptional regulation of several genes during G1/S transition. Upon interaction with CtIP, Rb is released from E2F-responsive promoters, where CtIP binds and activates gene expression. Under normal conditions, p21 is repressed by BRCA1-CtIP-CtBP. DNA damage disrupts this complex and induces the recruitment of CtIP to p21 promoter elements, enhancing p21 expression. Middle, In response to DSBs, CtIP associates with the MRN complex and stimulates MRE11 endonuclease activity to initiate bidirectional DNA end resection, using EXO1 or BLM/DNA2 in the 5'-3' direction away from the DSB, and MRE11's exonuclease in the 3'-5' direction towards the DSB end (not shown). In addition, CtIP/MRN-mediated endonucleolytic cleavage displaces DNA-PK (composed of the Ku70-80 heterodimer and DNA-PKcs) from broken DNA ends. These processes effectively suppress c-NHEJ. Ensuing resection of the 5' strand leads to the exposure of 3'-ssDNA overhangs, which commits DSB repair to homology-directed pathways (HR, a-EJ or SSA) and activates ATR-Chk1 signalling. HR relies on the sequential actions of BRCA1-BARD1 and BRCA2 to direct RAD51-dependent homologous DNA template search and ensure accurate DSB repair. In the absence of HR or when HR repair is inappropriate due to the lack of a sister chromatid template, mutagenic repair pathways such as Polθ-dependent a-EJ or RAD52mediated SSA are activated. Upper right, Following DNA replication stress, CtIP is recruited to stalled replication forks, where it inhibits DNA2-dependent overresection of nascent DNA, presumably via endonucleolytic incision of the reversed 5'-strand. Therein, CtIP acts synergistically with BRCA1, which protects reversed forks from MRE11-mediated degradation.
integrity in response to DNA replication stress [68,69]. Cortez and colleagues repeatedly found CtIP to be enriched at active replication forks [70,71], presumably through its direct interaction with PCNA ( Fig. 1) [72]. Yeo et al. reported that CtIP binding to FANCD2 facilitates CtIP relocation to damaged replication forks, thereby promoting fork restart and preventing new origin firing during fork recovery [73]. The Hendrickson laboratory showed that CtIP maintains telomere integrity during DNA replication via a similar mechanism [74]. The role of CtIP at stalled forks following hydroxyurea (HU)-induced replication stress is subject to controversy. Initially, it was shown that CtIP promotes the degradation of stalled forks by stimulating MRE11 and EXO1 nuclease activities [75,76]. Our group later demonstrated that CtIP protects reversed forks from nucleolytic degradation by DNA2 (Fig. 2) [77], a finding that was independently confirmed by a study from the Sørensen group [78]. However, while our data implicated CtIP's endonuclease activity in fork protection, Zarrizi et al. proposed that CtIP suppresses fork degradation by counteracting FBH1-mediated removal of RAD51 from stalled forks [77,78]. In addition, we found that CtIP-dependent stimulation of MRE11 endonuclease activity (via pT847) and CtIP-BRCA1 interaction (via pS327) are largely dispensable for fork protection [77]. Of note, recent data suggest that CtIP-BRCA1 interaction is critical for the initation of transcription-associated HR repair of DSBs arising at stalled replication sites upon formation of DNA-RNA hybrids, also termed R-loops [79]. Remarkably, efficient resolution of R-loops was shown to rely on CtIP's endonuclease activity [80]. Recent work from the Morris laboratory established that PIN1-induced conformational change of the BRCA1-BARD1 complex is required for fork protection, but not for HR repair. On this basis, it would be interesting to determine whether a similar separation-of-function mechanism applies to CtIP, an established PIN1 target [28,81]. Interestingly, we could further reveal that CtIP acts in a synergistic or additive manner with BRCA1 in protecting nascent DNA strands from nucleolytic degradation, which potentially contributes to the observed synthetic sick/lethal relationship between CtIP and BRCA1 [77]. Together, these findings highlight an important HR-independent function of CtIP in maintaining genome stability.

CtIP's dual role in tumorigenesis and cancer progression
Deregulation of the cell cycle and high levels of genomic instability are not only characteristics of cancer but also a driving force of tumorigenesis [82]. Therefore, CtIP-ascribed molecular activities have mostly been implicated in tumour suppression. This general assumption has gained further support by an early genetic study in mice showing that homozygous deletion of Ctip causes early embryonic lethality, while Ctip +/-heterozygosity renders mice prone to malignancy, most likely due to CtIP haploid insufficiency [45]. This paradigm has recently been challenged by several studies indicating that CtIP has oncogenic potential facilitating tumorigenesis.

CtIP mutations and genomic alterations in human disease
Given that Ctip is indispensable for viability in mice [45,58], it is not surprising that complete loss-of-function mutations in CtIP/RBBP8 have not been reported in man. Rare, biallelic hypomorphic CtIP variants translating into C-terminal truncated proteins cause Jawad and Seckel syndromes (Fig. 1), genetically heterogeneous conditions characterised by severe microcephaly and mental retardation, but no overt cancer predisposition [83]. Interestingly, Seckel patient-derived cell lines commonly display defects in ATR-dependent DNA damage signalling, certainly a well-established phenotype of CtIP-deficient cells [84,85].
Germline mutations in BRCA1 or BRCA2 greatly increase breast and ovarian cancer risk in patients [86]. Early studies demonstrated that tumour-associated mutations within the BRCA1 BRCT domains can disrupt its interaction with CtIP [2,6]. However, loss of BRCA1-CtIP interaction in mouse models expressing S327A mutant CtIP neither compromises genomic stability nor enhances tumour susceptibility [58,59]. Therefore, it remains an open question whether and how the BRCA1-CtIP interaction contributes to cancer development. Rebbeck et al. revealed that inherited clusters of SNPs at the CtIP locus ('CtIP haplotypes') are associated with modified breast, but not ovarian, cancer risk in women with BRCA1 mutations, supporting a role for CtIP in tumour suppression [87]. Moreover, two different, deleterious RBBP8 frameshift variants were recently detected in two early-onset breast cancer patients without BRCA1 mutation, although their relative contribution to pathogenicity was not determined [88]. A recent study identified twelve rare germline CtIP mutations in Danish cohorts of early-onset breast cancer patients negative for high risk BRCA1/2 mutations [78]. Remarkably, the CtIP-E804Δ variant, lacking a single amino acid within the C-terminal region (Fig. 1), caused severe defects in response to HU-induced replication stress, including deleterious nucleolytic degradation of reversed forks, but were proficient in DNA end resection and HR [78]. Owing to an exonic (A)9 mononucleotide repeat tract prone to mutation in mismatch repair-deficient tumours, RBBP8 has been described as a prominent target for microsatellite instability (MSI)-induced mutations in colorectal, endometrial and myeloid malignancies [89][90][91].
Genomic alteration analyses in patient samples across 32 different cancer types reveal that CtIP/RBBP8 is mainly amplified, but rarely mutated or deleted, which is in marked contrast to the situation of BRCA1 (Fig. 3A). CtIP amplification is most prevalent in pancreatic cancers, with 16% of profiled tumours from two publicly available datasets showing genomic alterations in RBBP8 (Fig. 3A and B). Although CtIP is predominantly amplified in pancreatic cancers, a few missense CtIP mutations of unknown significance have been reported that could potentially drive pancreatic tumorigenesis (Fig. 3B) [5]. Furthermore, Kaplan-Meier survival analysis in a study of 178 patients reveals that high expression of CtIP is significantly linked to poor prognosis in pancreatic adenocarcinoma (PAAD), suggesting an oncogenic activity of CtIP in these tumours (Fig. 3C).
Interestingly, RBBP8 amplification was previously shown to occur mutually exclusive with BRCA1 and RB1 inactivation in high grade serous ovarian cancer, further supporting the idea of RBBP8 acting as a candidate oncogene in certain tumour types [92]. Rb is a master negative regulator of cell cycle progression and numerous cancers are characterised by Rb inactivation [93]. It is therefore conceivable that CtIP upregulation could drive tumorigenesis in a similar manner to RB1 loss-of-function. Presumably, CtIP-mediated Rb inhibition and subsequent de-repression of E2F-responsive genes such as CCND1 could promote sustained proliferation by facilitating G1/S transition [39]. A similar relationship between CtIP and the Rb pathway has recently been proposed to contribute to the development of plasma cell myeloma, wherein high RBBP8 expression correlates with poor survival and tumour recurrence in patients [94]. Accordingly, CtIP downregulation by promoter methylation was found to correlate with a more favourable prognosis in bladder cancer [95]. Yu et al. reported that CtIP expression was upregulated in gastric cancer and adjacent neoplastic tissues as compared with normal tissue [96]. CtIP overexpression in gastric cancer was accompanied by increased CtIP-CtBP-BRCA1-mediated p21 suppression, which could promote G1/S transition and sustained proliferation [96]. Although CtIP expression levels did not significantly affect overall survival in those patients, CtIP-mediated proliferative activity, presumably through p21 repression and CCND1 upregulation (Fig. 2), was associated with worse prognosis at early stages of gastric cancer development [96]. Consistently, CtIP is frequently amplified in oesophageal carcinoma (ESCA) and stomach adenocarcinoma (STAD) (Fig. 3A). Other studies, however, have found that CtIP expression is frequently downregulated in breast cancer patients and that low expression is combined with unfavourable prognosis and metastasis formation [97][98][99], which is markedly different from what is observed in pancreatic cancer patients (Fig. 3C). Thus, the relationship between CtIP expression and disease progression appears to be intricately linked with tumour type, possibly reflecting the relative importance of distinct CtIP functions in the pathogenesis of different cancers.

Anti-and pro-tumorigenic properties of CtIP
Owing to its fundamental role in HR, CtIP may, albeit to a limited extent, be involved in tumour suppression through preventing the accumulation of cancer-enabling mutations and genome instability arising from erroneous repair of DSBs [1,2,82]. Originally, CtIP has been postulated as a haploinsufficient tumour suppressor based on the shortened lifespan of Ctip +/-heterozygous mice, which succumbed to multiple tumours, predominantly large B-cell lymphomas [45]. However, a recent independent study could not recapitulate this finding, showing instead that tumour-free survival is indistinguishable between wild-type mice and mice harbouring a single null allele of the Ctip gene [100]. Moreover, and considerably different from conditional Brca1-null females, mammary-specific biallelic inactivation of Ctip did not induce breast tumours [100]. On the contrary, Reczek et al. discovered that Ctip loss-of-function slowed down mammary tumorigenesis in p53-deficient mice, suggesting an oncogenic activity in CtIP [100]. A conceivable mechanistic explanation for CtIP-driven malignant transformation is that CtIP-dependent resection promotes chromosomal translocations and telomere fusions through a-EJ (Fig. 2) [101][102][103]. In line with this hypothesis, Srinivasan et al. recently reported that miR223-3p-mediated CtIP downregulation prevents mutagenic a-EJ and, thus, chromosomal translocations in normal cells, whilst reduced expression of miR223-3p is frequently observed in BRCA1-deficient cancer cells [44].
Maintenance of telomere length and integrity is essential for preventing genomic instability and oncogenic transformation [104]. Due to the formation of secondary structures and the repetitive nature of telomeric DNA sequences, telomere replication is a complex process that can lead to replication fork stalling and collapse [105]. CtIP was recently shown to protect replicating telomeres from aberrant shortening and fusions, by rescuing stalled replication forks and promoting HR-mediated DSB repair of collapsed forks [74]. In addition, the function of CtIP in protecting nascent DNA at stalled forks from nucleolytic degradation (Fig. 2) may contribute to the protection of replicating telomeres [77]. Furthermore, work from the Scully laboratory revealed that CtIP acts independently from BRCA1 in preventing the formation of microhomology-mediated small tandem duplications, a genome-wide DNA rearrangement signature produced by aberrant processing of stalled replication forks and typically found in BRCA1-deficient cancers [106,107]. Therefore, it is tempting to speculate that CtIP, at least in parts, counteracts tumorigenesis via its role in alleviating replication stress-induced genomic instability.

Clinical relevance of targeting CtIP for cancer therapy
Despite its ambiguous role in tumorigenesis, CtIP has prognostic value in several types of cancer and could serve as a predictive biomarker for treatment response in selected patients [94,95,98,99]. As previously mentioned, low RBBP8 expression correlates with breast  [126,127]. (C) Kaplan-Meier plots using mRNA expression data from primary tumour samples of the TCGA Pan-Cancer Atlas PAAD study [126] were generated in UCSC Xena [128]. High CtIP expression correlates with unfavourable prognosis in PAAD, both for overall (p-value = 0.0032) and progression-free (p-value = 0.0092) survival. For this analysis, patient samples were divided into two RBBP8 expression groups, separated by the median: "High" (expression values ≥ 9.350) and "Low" (expression values cancer aggressiveness and poor response to hormonal therapy in breast cancer [97,99]. However, reduced CtIP protein levels in luminal breast cancers are associated with better response to DNA-damaging chemotherapy [99], presumably due to a decreased DSB repair capacity. In bladder cancer, CtIP promoter methylation, leading to gene silencing, is linked to better prognosis and can be readily detected in non-invasive urine tests [95]. Platinum-based drugs are commonly used for treating bladder cancer [108] and have demonstrated improved efficacy in BRCA1/2-mutated, HR-deficient tumours [109,110]. Therefore, it is plausible that reduced CtIP expression and ensuing HR defects could underlie improved response to cisplatin treatment in bladder cancer. In contrast, selective CtIP inhibition with yet-to-be-discovered targeting molecules may synergise with DNA-damaging drugs, particularly in tumours displaying CtIP amplification such as highly aggressive pancreatic cancer (Fig. 3).
Genomic instability and acquired DNA repair defects are common features of many cancers and can be therapeutically exploited based on the concept of synthetic lethality [82,[110][111][112][113]. A paradigm example is the use of PARP inhibitors for treating BRCA1/2-mutated breast and ovarian cancer [114][115][116]. A major caveat of PARPi therapy is the development of treatment resistance due to secondary mutations leading to the restoration of BRCA1/2 or re-establishment of DNA end resection [117,118]. Therefore, selectively targeting additional DNA damage response components, including CtIP, could potentially re-sensitise these cancers to PARPi [119]. CtIP depletion sensitises breast cancer cells to PARPi, presumably by suppressing HR, suggesting that CtIP could be an interesting biomarker for predicting disease progression and therapy response in breast cancer [98]. Furthermore, impaired recruitment of CtIP to DSBs upon CHD1 loss was shown to sensitise prostate cancer cells to PARPi [120]. Ceccaldi et al. reported that HR-deficient ovarian cancers strongly rely on Polθ-mediated a-EJ to repair DSBs arising at collapsed forks, thereby contributing to tumour development [112,121,122]. Thus, blocking a-EJ via Polθ or CtIP inhibition could be a valid synthetic lethal strategy for selective killing of BRCA1-deficient tumours. We have recently reported that CtIP and BRCA1 operate in separate fork protection pathways, thus providing an alternative mechanism explaining the potential synthetic lethal relationship between CtIP and BRCA1 [77]. A recent synthetic lethality screen for candidate PARPi sensitivity genes in HR-proficient cells identified BRD4, a chromatin-binding protein frequently overexpressed in several types of cancers [123,124]. Remarkably, BRD4 enrichment at the RBBP8 promoter and enhancer, and, thus, CtIP expression, were significantly decreased upon BRD4 inhibition, which ultimately caused HR deficiency and PARPi hypersensitivity [123]. In our opinion, these findings provide a clear rationale for the development of CtIP targeting approaches to potentially improve existing cancer therapies and counteract acquired PARPi resistance in selected cancers.

Conclusions
CtIP is a multifaceted player in DNA repair and genomic stability and acts as a "double-edged sword" in cancer. Its diverse functions in transcriptional regulation, DSB repair and replication stress response have been implicated in different aspects of tumorigenesis, leading to positive or unfavourable outcomes depending on cancer type and biological context. Thus, CtIP (RBBP8) can be classified as 'double agent', a term coined for genes harbouring both tumour suppressive and oncogenic functions, depending on the respective alterations and cellular context, and often functioning as central hubs in protein interaction networks [125]. Further research will be required to more clearly define the role of CtIP in maintaining the balance between oncogenesis and tumour suppression, with potential implications for cancer prognosis and therapy.

Declaration of Competing Interest
The authors report no declarations of interest.