The biological rationale behind the combination rests on how cancer cells cope with DNA damage. ATR, or ataxia telangiectasia and Rad3-related protein kinase, is a central regulator of the DNA damage response. When replication stress causes DNA replication forks to stall, ATR stabilises those forks, delays the firing of late replication origins and activates S-phase and G2-M checkpoints, buying the cell time to repair damage before it divides. Blocking ATR therefore allows double-strand DNA breaks to accumulate, driving genomic instability and ultimately cell death. PARP1, meanwhile, is involved in repairing single-strand breaks during base excision repair, and PARP inhibitors such as olaparib work partly by trapping PARP1 on DNA, causing stalled replication forks to collapse into double-strand breaks.

Those double-strand breaks are normally repaired by the homologous recombination repair pathway, in which the tumour suppressor proteins BRCA1 and BRCA2 play key roles. In tumours carrying BRCA1/2 mutations, homologous recombination is defective, so PARP inhibitors kill the cells through synthetic lethality: two independent weaknesses that are individually survivable but lethal together. PARP inhibition also increases replication stress, making tumour cells increasingly dependent on ATR for survival. Preclinical work had shown selective synergy between ceralasertib and olaparib in BRCA-deficient and ATM-deficient models, and in most models of acquired PARP inhibitor resistance, raising hopes that the pairing could both deepen responses and overcome resistance in the clinic.

The trial, registered as NCT02264678, was a modular, open-label, multicentre study. In the dose-escalation phase, Part A1 tested ceralasertib alone at 80 mg twice daily or 160 mg once daily in 13 patients, while Part A2 escalated ceralasertib in combination with olaparib across 67 patients with advanced solid tumours. Dose expansion then followed in two breast cancer cohorts: Part B3 enrolled 37 patients with BRCA1/2-mutated, HER2-negative breast cancer who were PARP inhibitor-naive, and Part B4 enrolled 25 patients with BRCA1/2 wild-type triple-negative breast cancer. All expansion patients received the recommended Phase 2 dose until disease progression or discontinuation criteria were met.

The safety findings shaped the final dosing regimen. No dose-limiting toxicities occurred with ceralasertib monotherapy, but haematological toxicity emerged as the limiting factor when the two drugs were combined. Dose-limiting toxicities appeared in several escalation cohorts, predominantly involving neutropenia and thrombocytopenia, and attempts to push the ceralasertib dose higher or extend the number of dosing days proved intolerable. The recommended Phase 2 dose was set at ceralasertib 160 mg once daily on days 1 to 7 of each 28-day cycle, plus olaparib 300 mg twice daily given continuously. At this schedule, tolerability was illustrated by limited dose reductions and discontinuations, with a median relative dose intensity of 100 percent for ceralasertib in the expansion cohorts.

Across the expansion cohorts, treatment-emergent adverse events were nearly universal, affecting 97.3 percent of patients in Part B3 and 96.0 percent in Part B4, but grade 3 or worse events were less frequent, at 43.2 percent and 36.0 percent respectively. The most common severe toxicities were haematological: anaemia and neutropenia predominated in the BRCA-mutant cohort, while anaemia, thrombocytopenia and neutropenia appeared in the triple-negative cohort. Adverse events led to discontinuation in only a small minority of patients, two of 37 in Part B3 and one of 25 in Part B4, and no deaths from adverse events occurred in Part B3. The investigators concluded that no new safety signals were observed beyond those already known for each drug class.

The efficacy results revealed a striking split between the two expansion cohorts. In Part B3, 14 of 37 patients, or 37.8 percent, achieved a confirmed objective response, comprising one complete response and 13 partial responses, with a median duration of response of 9.8 months. Among the 30 patients whose BRCA1/2 mutation was centrally confirmed by next-generation sequencing, the response rate rose to 40.0 percent, the median duration of response reached 11.0 months, and median progression-free survival was 9.0 months. Median overall survival in this centrally confirmed group was 26.2 months. In sharp contrast, Part B4 produced no objective responses at all among patients with BRCA wild-type triple-negative breast cancer lacking homologous recombination repair mutations, with median progression-free survival of just 3.1 months.

Pharmacokinetic and pharmacodynamic analyses added mechanistic texture to the clinical findings. Ceralasertib was rapidly absorbed, reaching peak plasma concentrations within about an hour and a half, and exposures at both 160 mg once daily and 80 mg twice daily provided 24-hour coverage over the concentration required for 90 percent ATR inhibition. When the drugs were combined, single-dose ceralasertib exposure was modestly reduced compared with monotherapy, and ceralasertib did not accumulate at steady state. In paired tumour biopsies, levels of phosphorylated RAD50, a pharmacodynamic biomarker of DNA damage response activation, increased on treatment in all three evaluable samples, alongside rises in the DNA damage markers gamma-H2AX and phosphorylated CHK1, confirming target engagement, although the small sample size limited interpretation.

Exploratory biomarker work examined whether ATM status or expression of Schlafen-11, an emerging predictor of sensitivity to DNA-damaging therapy, could identify responders. Responders showed a higher proportion of ATM-low tumours than non-responders, and median SLFN11 H-scores were lower in patients achieving complete or partial responses than in those with stable or progressive disease, with 85.7 percent of responders classified as SLFN11-low. However, neither marker was strongly associated with clinical outcome, and the authors note that predictive biomarkers for ceralasertib remain to be elucidated, partly because replication stress, the key hypothesised predictor of ATR inhibitor sensitivity, still cannot be measured accurately in clinical tissue samples.

The authors place their findings in context by comparing them with landmark PARP inhibitor monotherapy trials in BRCA-mutated metastatic breast cancer, where median progression-free survival has ranged from 4.0 to 8.6 months and response rates have varied widely depending on prior platinum exposure. Without a monotherapy comparator arm, the contribution of ceralasertib to the observed activity cannot be conclusively determined, and the parallel randomised VIOLETTE study found no significant progression-free survival benefit from adding ceralasertib to olaparib in triple-negative breast cancer, possibly because haematological toxicity forced a relatively low ceralasertib dose. Nevertheless, the combination demonstrated preliminary activity with manageable safety in PARP inhibitor-naive patients with BRCA1/2-mutated HER2-negative breast cancer, and the authors argue that further work should identify which patient groups, defined by prior PARP exposure, specific mutations or biomarkers, are most likely to benefit from adding ATR inhibition to PARP therapy.

Subject of Research: Phase 1 trial of the ATR inhibitor ceralasertib combined with the PARP inhibitor olaparib in advanced solid tumours and breast cancer

Article Title: Safety and efficacy of ceralasertib, an ATR kinase inhibitor, combined with olaparib: dose escalation in advanced solid tumours and dose expansion in advanced breast cancer

Article References: Lopez, J. S., Im, S.-A., Postel-Vinay, S., Campone, M., El-Khoueiry, A. B., Abida, W., Arkenau, T., Lee, K.-W., Rha, S. Y., Jodrell, D. I., Roylance, R., Irurzun-Arana, I., Jones, G. N., Lukashchuk, N., Stephens, C., Norris, C., Loembé, A.-B., Dean, E., & Krebs, M. G. (2026). Safety and efficacy of ceralasertib, an ATR kinase inhibitor, combined with olaparib: dose escalation in advanced solid tumours and dose expansion in advanced breast cancer. British Journal of Cancer. https://doi.org/10.1038/s41416-026-03601-z

Image Credits: AI Generated

DOI: 10.1038/s41416-026-03601-z

Keywords: ceralasertib, olaparib, ATR inhibitor, PARP inhibitor, BRCA1/2, breast cancer, triple-negative breast cancer, DNA damage response, replication stress, Phase 1 trial, synthetic lethality, targeted therapy

The research, led by Chenxi Li, Yingying Qin and Yajuan Yang of Shandong University and collaborating institutions in China, focused on what happens when mouse primordial germ cells lose RAD54 like, or RAD54L. The team found that without this factor, the cells proliferate poorly, leaving an insufficient pool of germ cells and ultimately impairing fertility. The defect traces back to a specific molecular hazard: the accumulation of topoisomerase I cleavage complexes, abbreviated TOP1ccs, which are among the most common and most dangerous lesions that arise during DNA replication.

To appreciate why TOP1ccs matter, it helps to understand what topoisomerase I normally does. As the two strands of the double helix are prised apart for copying, the DNA ahead of the replication machinery becomes overwound, like a rope twisted too tightly. Topoisomerase I relieves this torsional stress by nicking one strand of the DNA, allowing it to swivel freely, and then sealing the break again. The enzyme normally completes this cut-and-paste cycle in a fraction of a second. Occasionally, however, the enzyme becomes trapped mid-cycle, covalently bonded to the DNA at the nick it just created. That trapped enzyme-DNA adduct is the TOP1 cleavage complex, and if it is not removed promptly, the approaching replication fork collides with it, converting a transient intermediate into a double-strand break and stalling DNA synthesis.

Cells possess dedicated machinery to clear these protein-DNA adducts, and the new study adds a surprising player to that list. The researchers showed that RAD54L promotes the degradation of TOP1ccs through the ubiquitin-proteasome pathway, the cellular recycling system that tags unwanted proteins with chains of ubiquitin and delivers them to the proteasome for destruction. The key to this activity lies in a partnership between RAD54L and tripartite motif-containing protein 21, or TRIM21, which the authors identify as a newly recognized E3 ubiquitin ligase responsible for clearing TOP1ccs. E3 ligases are the enzymes that confer specificity on the ubiquitin system, deciding which targets get marked for degradation. In this case, TRIM21 appears to be the enzyme that flags trapped topoisomerase I for removal.

The mechanistic twist is that RAD54L does not itself ligate ubiquitin. Instead, it interacts with TRIM21 and stabilizes the protein, keeping sufficient TRIM21 available to do its clearing work. When RAD54L is absent, TRIM21 is not maintained properly, TOP1ccs accumulate, and the resulting DNA damage escalates. In RAD54L-deficient primordial germ cells, this cascade of events exacerbates DNA damage, undermines proliferation and erodes the reproductive reserve. The finding reframes RAD54L, historically viewed as a homologous recombination factor that helps search for and invade homologous DNA templates during double-strand break repair, as also acting upstream of repair, at the stage of removing the lesions that would otherwise create breaks in the first place.

Primordial germ cells are an especially revealing setting for this kind of analysis. These cells undergo rapid mitotic divisions during embryonic development, and their genome must be transmitted faithfully across generations. High levels of replication, combined with the metabolic demands of a growing embryo, make TOP1ccs a constant threat. The DNA damage response mechanisms that protect somatic cells have been studied extensively, but the regulatory networks that safeguard the germline during this early mitotic phase have remained largely unexplored. By showing that a defect in a single DNA damage response factor can compromise the founding population of the germline, the study connects genome maintenance at the molecular level to fertility at the organismal level.

The experimental logic of the work follows a path familiar to genome stability researchers but with distinctive results. Mice lacking RAD54L in their germ cells show proliferation defects in the primordial germ cell population. The DNA damage observed in these cells is not random; it is specifically tied to TOP1cc accumulation, which means the damage is preventable if the trapped complexes are cleared. The demonstration that RAD54L stabilizes TRIM21 provides a coherent causal chain: loss of RAD54L destabilizes the E3 ligase, the ligase can no longer ubiquitinate trapped topoisomerase I efficiently, the adducts persist, replication forks collide with them, and DNA damage mounts until cells falter or die.

The discovery also resonates with a broader and clinically important theme in cancer medicine. Drugs called topoisomerase I poisons, including camptothecin and its clinical derivatives irinotecan and topotecan, work precisely by stabilizing the TOP1 cleavage complex, trapping the enzyme on DNA and forcing tumor cells into lethal collisions between replication forks and the drug-stabilized adducts. Understanding how healthy cells, and particularly vulnerable populations like germ cells, clear TOP1ccs has implications for how such treatments affect fertility and for how resistance to these drugs emerges. A pathway involving RAD54L and TRIM21 that governs the lifespan of TOP1ccs could, in principle, influence both the toxicity and the efficacy of these widely used chemotherapeutics, although the new study addresses physiology rather than treatment outcomes.

TRIM21 itself carries an interesting scientific history. Long studied in immunology as an antibody receptor inside cells, it has more recently been appreciated as a versatile quality-control factor that recognizes and ubiquitinates a range of intracellular targets. Its identification here as the E3 ligase for TOP1cc clearance in germ cells extends that repertoire into genome maintenance and adds a new dimension to how the ubiquitin system participates in the DNA damage response. The RAD54L-TRIM21 axis suggests that repair factors may do more than mend breaks after they occur; some may actively manage the burden of endogenous lesions so that breaks are less likely to arise at all.

For the field of reproductive biology, the study fills in a piece of a larger puzzle: how the mitotic phase of gametogenesis, before meiosis begins, is protected against the wear and tear of rapid proliferation. The authors frame their results as extending understanding of the regulatory mechanisms that safeguard genome integrity during this process, and the practical significance is clear. A diminished primordial germ cell pool is a diminished reproductive reserve, and the work points to TOP1cc accumulation as one preventable driver of that loss. As with any mouse study, translating the findings to human fertility will require further work, but the core machinery of topoisomerase I, the ubiquitin-proteasome system and the DNA damage response is conserved across mammals, making the pathway a plausible target for future investigations into reproductive health and the side effects of topoisomerase-targeting therapies.

Subject of Research: The role of RAD54L and TRIM21 in clearing topoisomerase I cleavage complexes to maintain genome stability in primordial germ cells

Article Title: RAD54L counteracts topoisomerase I cleavage complexes by stabilizing E3 ligase TRIM21 to maintain genome stability

Article References: Li, C., Wang, S., Xu, W., Kong, Z., Wen, C., Zhao, S., Cao, L., Chen, Z.-J., Zhao, S., Qin, Y., & Yang, Y. (2026). RAD54L counteracts topoisomerase I cleavage complexes by stabilizing E3 ligase TRIM21 to maintain genome stability. Cellular and Molecular Life Sciences. https://doi.org/10.1007/s00018-026-06458-w

Image Credits: AI Generated

DOI: 10.1007/s00018-026-06458-w

Keywords: RAD54L, TRIM21, topoisomerase I cleavage complexes, primordial germ cells, genome stability, DNA damage response, ubiquitin-proteasome pathway, replication stress, fertility, gametogenesis, E3 ubiquitin ligase, DNA repair

Replication stress arises from many sources, and the review methodically maps them. Oncogene activation is a principal culprit. When growth-promoting signals such as MYC or cyclin E are unleashed, they push cells through the cell cycle faster than replication machinery can be properly licensed, forcing replication to begin in inappropriate places. Shortened G1 phases leave origin licensing incomplete, and transcription programs fire at full blast while forks attempt to copy the same sequences. The resulting collisions between replication and transcription complexes are among the most potent sources of chromosome breakage, particularly at long, highly transcribed genes known as common fragile sites, which have been recognized as breakage hotspots since experiments with the polymerase inhibitor aphidicolin in the 1980s.

Transcription-replication conflicts come in mechanistically distinct flavors. Head-on conflicts, where the replication fork meets an oncoming RNA polymerase, and co-directional conflicts produce different R-loop levels and activate different DNA damage responses. R-loops, three-stranded structures containing an RNA-DNA hybrid and a displaced single strand, have emerged as pivotal intermediates that can either protect or poison genomes depending on how they are resolved. Studies of BRCA1 and BRCA2 have shown that these breast and ovarian cancer tumor suppressors do more than repair double-strand breaks; they actively suppress R-loop accumulation by coordinating transcription elongation and RNA processing. Similarly, chromatin remodelers such as the SWI/SNF complex and spliceosome factors help clear conflicts, explaining why mutations in these complexes, common in myelodysplastic syndromes and Ewing sarcoma, produce cells exquisitely sensitive to ATR inhibition.

Beyond collisions, the review highlights metabolic and structural pressures. Nucleotide pools are a striking example: early work demonstrated that deoxyribonucleotide deficiency distorts origin choice and spacing, while later studies showed that oncogene-induced senescence is partly a consequence of nucleotide starvation. Cancer cells also accumulate unprocessed uracil and ribonucleotides in DNA, both of which impede fork progression. DNA secondary structures add another layer. Unstable inverted repeats fold into hairpins and cruciforms, and expanded TA-dinucleotide repeats in microsatellite-unstable colorectal cancers form structures that grind forks to a halt. These repetitive obstacles, rather than mismatch repair failure alone, are now understood to be why such tumours become fatally dependent on the WRN helicase, which unwinds these structures and prevents catastrophic chromosome shattering.

The consequences of unresolved replication stress reverberate through the cancer genome. Stalled forks are vulnerable to nuclease degradation unless BRCA2 protects the nascent DNA, and collapse generates breaks that drive chromosomal instability, linking structural and numerical chromosome chaos. Under-replicated regions persist into mitosis, producing anaphase bridges, micronuclei and copy-number changes that mirror the polymorphic and pathogenic variants seen in human populations. Several avenues of rescue have been characterized, including fork reversal, repriming by the PRIMPOL primase-polymerase and specialized translesion polymerases that replicate past damaged bases, but these repair routes are intrinsically error-prone. Translesion synthesis underlies acquired chemotherapy resistance, and repriming leaves single-stranded DNA gaps that, recent work shows, are a key determinant of PARP inhibitor sensitivity in BRCA-deficient cells.

The ATR-CHK1 checkpoint pathway is the master guardian that keeps replication stress survivable. ATR is recruited to RPA-coated single-stranded DNA through its partner ATRIP and activated by TopBP1 or ETAA1, with the MRN complex and the 9-1-1 clamp sharpening its activation. Once active, ATR suppresses excess origin firing, stabilizes stalled forks, enforces an intrinsic S/G2 checkpoint that prevents cells from entering mitosis with unfinished replication, and prevents global exhaustion of RPA, a failure mode that culminates in replication catastrophe and genome fragmentation. Without ATR, forks collapse and single-stranded DNA accumulates beyond rescue, a finding that transformed the enzyme from a basic research curiosity into a top-priority drug target.

That targeting has now reached the clinic. ATR inhibitors such as ceralasertib and elimusertib have produced durable responses in tumours with high replication stress, including small cell lung cancer, and synthetic lethal combinations with ATM loss, ERCC1 deficiency, CCNE1 amplification and APOBEC3 activity are being pursued. WEE1 inhibitors override the G2 checkpoint and force stressed cells into lethal mitosis, with combination strategies including sequential PARP and WEE1 dosing designed to minimize toxicity. The clearest success story remains PARP inhibition, introduced in landmark 2005 studies demonstrating synthetic lethality in BRCA-mutant cells. Mechanistic work has since refined the model: PARP inhibitors trap PARP1 and PARP2 on DNA and act through persistent single-stranded gaps that BRCA-deficient cells cannot fill, and next-generation PARP1-selective agents such as AZD5305 promise better tolerability. WRN inhibition in microsatellite-unstable cancers represents the newest synthetic lethal frontier, with resistance mechanisms beginning to be mapped by CRISPR screens.

The review also reframes replication stress as a chronic, sublethal condition that sculpts tumour evolution over years rather than an acute lethal event. This slow burn generates mutational signatures, including those of APOBEC3 cytidine deaminases, that fuel heterogeneity, therapy resistance and metastasis, and therapy itself can amplify APOBEC3A-driven evolution in persistent cancer cells. Importantly, replication stress intersects with the immune system. Under-replicated DNA in micronuclei activates the cGAS-STING pathway, and chromosomal instability promotes metastasis through cytosolic DNA sensing, while SAMHD1 and MYC normally suppress inflammatory signaling from replication-derived nucleic acids. ATR inhibition can potentiate the inflammatory tumour microenvironment, opening rational combinations with immune checkpoint blockade, an approach already validated in microsatellite-unstable and BRCA-mutated tumour settings.

The authors position replication stress not as a single weakness but as a dynamic, context-dependent vulnerability landscape shaped by each tumour’s oncogenic, metabolic and chromatin state. Biomarkers such as replication stress signatures, gap formation and fork protection status may soon guide which patients receive ATR, WEE1, PARP or WRN inhibitors, and in which combinations. The therapeutic opportunity, they argue, lies in recognizing that tumours pay a permanent price for their genomic chaos, and that the rescue pathways they depend upon are now precisely druggable targets.

Subject of Research: Replication stress in cancer: origins, consequences and therapeutic opportunities

Article Title: Replication stress in cancer: origins, consequences and therapeutic opportunities

Article References: Chen, J., & Zou, L. (2026). Replication stress in cancer: origins, consequences and therapeutic opportunities. Nature Reviews Cancer. https://doi.org/10.1038/s41568-026-00979-z

Image Credits: AI Generated

DOI: 10.1038/s41568-026-00979-z

Keywords: replication stress, cancer, ATR, CHK1, PARP inhibitors, synthetic lethality, R-loops, genomic instability, WRN helicase, WEE1 inhibitors, precision oncology, DNA replication

ATR, short for ataxia-telangiectasia and Rad3-related, belongs to the PIKK family of giant kinases that also includes ATM, DNA-PKcs, mTOR and SMG1. Like its siblings, ATR is an enormous HEAT-repeat scaffold carrying a kinase domain buried deep within, and it operates as a dimer paired with its obligate partner ATRIP. Since the first cryo-EM structures of human ATR–ATRIP appeared in 2018, researchers have known the complex’s overall architecture but not the physical choreography of its activation. The central puzzle has been phosphorylation: ATR autophosphorylates itself, and this modification is considered a hallmark of the active kinase, yet nobody had seen structurally what phosphorylation does to the machine or how the activator protein TopBP1 drives the process. The new study addresses that gap directly by trapping the complex at successive stages of its catalytic cycle.

The team began with precatalytic snapshots. In two structures determined with the non-hydrolysable ATP analogue ATPγS wedged in the active site, they found that nucleotide binding alone accomplishes something remarkable: it primes the catalytic center for chemistry without wholesale rearrangement of the kinase domain. When the researchers then added a fragment of Chk1, the key downstream effector of the ATR pathway, the substrate docked neatly onto the primed enzyme. In other words, ATP binding and substrate positioning can proceed as relatively independent, modest steps rather than requiring a dramatic conformational detonation. This finding revises the textbook expectation that PIKK kinases need massive global movements to become catalytically competent, and it suggests that the rate-limiting event in ATR activation lies elsewhere — specifically, in the regulatory rearrangements triggered by TopBP1.

That TopBP1-dependent rearrangement is the structural centerpiece of the paper. In the activated structure, two molecules of TopBP1 bind simultaneously to the ATR–ATRIP dimer, one engaging each protomer from opposite faces. TopBP1 latches onto the HEAT-repeat scaffold far from the active site, and in doing so it locks the N-HEAT and M-HEAT domains of ATR together in both protomers. In the precatalytic and substrate-bound states, these scaffold elements make intimate contact in one protomer while hovering apart in the other — an inherent asymmetry. TopBP1 binding abolishes that asymmetry, snapping the dimer into a near-symmetric conformation that promotes ATR–ATRIP autophosphorylation and drives efficient phosphorylation of Chk1. Detailed atomic contacts reveal how this is achieved: an intensive network of hydrogen bonds welds a TopBP1 beta strand to the M-HEAT domain, the tryptophan residue W1145 of TopBP1 buries into a hydrophobic pocket at the N-HEAT/M-HEAT junction, and salt-bridge interactions fasten a TopBP1 helix onto the N-HEAT superhelix.

Perhaps the most provocative observation concerns what happens after activation. The researchers purified ATR–ATRIP complexes that had been fully autophosphorylated in the presence of TopBP1, then stripped the activator away and determined the structure of the phosphorylated kinase on its own. Remarkably, the phosphorylated complex remained fully competent to phosphorylate Chk1, and it did so regardless of how much TopBP1 was added back. Structurally, the released enzyme adopted an asymmetric conformation, yet its kinase domain was essentially indistinguishable from that of the TopBP1-bound, near-symmetric activated state. The authors conclude that catalytic activity is compatible with both near-symmetric and asymmetric architectures, meaning that phosphorylation — not a rigid, TopBP1-imposed symmetry — is the true durable switch. Once ATR has phosphorylated itself, the checkpoint engine keeps running even if its ignition key is removed, a property with clear implications for how sustained checkpoint signaling is maintained at damaged replication forks.

The autophosphorylation story was reinforced biochemically. Mass spectrometry mapped dozens of newly phosphorylated sites on ATR and ATRIP, and phosphatase treatment confirmed that the mobility shifts seen in gels were genuinely phosphorylation-dependent. Activity assays showed a clean dose-response: non-phosphorylated ATR–ATRIP phosphorylated Chk1 only when TopBP1 was titrated in, whereas the pre-phosphorylated complex catalyzed Chk1 modification at TopBP1 concentrations approaching zero. Interestingly, an isolated ATR head domain displayed only basal activity with or without TopBP1, underscoring that the full dimeric scaffold — and its TopBP1-driven reorganization — is required to reach full catalytic power. Together these experiments elevate ATR autophosphorylation from a correlative biomarker to a mechanistically demonstrated molecular switch for checkpoint activation.

The paper’s second act concerns therapy. ATR inhibitors such as berzosertib (VX-970/M6620), ceralasertib (AZD6738), gartisertib (VX-803/M4344) and elimusertib (BAY 1895344) are in clinical development as sensitizers that push tumor cells with pre-existing DNA repair defects over the edge. The team determined four separate inhibitor-bound structures of ATR–ATRIP, one for each compound, providing the first side-by-side atomic views of how this pharmacological class occupies the ATR active site. The nucleotide-binding pocket turns out to be divisible into four sub-compartments: an adenine-binding pocket lined by residues including M2325, Y2365, I2377, W2379 and V2380; an inner hydrophobic pocket; a ribose-binding site; and a triphosphate-binding site. All four inhibitors anchor themselves in the adenine pocket, but each then exploits the pocket’s flexibility differently, extending flanking chemical groups into the hydrophobic cavity, the ribose site, or the triphosphate region.

These structures explain both potency and selectivity. ATR’s adenine and hydrophobic sub-pockets are shifted toward the C-lobe relative to the corresponding sites in DNA-PKcs, mTOR and ATM, a displacement caused by the bulky side chains of M2325 and V2493 in ATR. Inhibitors can therefore nestle into an ATR-specific geometry that the closely related PIKK kinases cannot accommodate. Superimposition with structures of DNA-PKcs, mTOR and ATM further showed that a key hydrogen bond formed by berzosertib with glycine 2385 of ATR would be sterically blocked by the larger residues occupying the equivalent positions in the sibling kinases. In effect, the drugs exploit the conformational plasticity of the ATR active site — the very same flexibility that allows ATPγS to prime the kinase — to achieve multipocket engagement that favors ATR over its family members. For medicinal chemists, the four structures amount to a comparative map of which sub-pockets each clinical candidate touches, a starting grid for designing next-generation inhibitors with improved selectivity and resistance profiles.

Taken together, the study delivers a structural framework in which ATR regulation can finally be read as a sequence of physically defined events. ATP binding primes the catalytic cleft and positions Chk1; TopBP1 binding to the HEAT scaffold reorganizes the dimer and licenses autophosphorylation; phosphorylation converts the enzyme into a durably active state that survives TopBP1 dissociation; and small molecules shut the whole system down by wedging into the plastic ATP pocket. The work also resonates with parallel findings on ATM, where asymmetric activation of the dimeric kinase has recently been visualized, hinting at shared regulatory logic across the PIKK family. For cancer biology, the implications are tangible: tumors rely obsessively on ATR to survive their own replication chaos, and a structure-level understanding of both the activation switch and the drug-binding pocket should accelerate the design of inhibitors that hit ATR harder and more selectively. As ATR inhibitors advance through clinical trials, the molecular movie assembled by Wang, Cheng, Qiao and colleagues provides the kind of mechanistic blueprint on which the next generation of checkpoint-targeted therapies can be drawn.

Subject of Research: Cryo-EM structural mechanisms of phosphorylation-driven activation and inhibitor binding of the human ATR–ATRIP checkpoint kinase complex

Article Title: Mechanistic insights into phosphorylation-driven activation and therapeutic inhibition of human ATR–ATRIP

Article References: Wang, L., Wang, M., Zhao, L., Rao, Q., Wu, H., Ma, B., Wang, J., Zheng, J., Li, Y., Xu, Y., Guo, J., Cheng, J., & Qiao, S. (2026). Mechanistic insights into phosphorylation-driven activation and therapeutic inhibition of human ATR–ATRIP. Nature Structural & Molecular Biology. https://doi.org/10.1038/s41594-026-01887-4

Image Credits: AI Generated

DOI: 10.1038/s41594-026-01887-4

Keywords: ATR, ATRIP, TopBP1, Chk1, cryo-electron microscopy, DNA damage response, replication stress, autophosphorylation, kinase activation, ATR inhibitors, berzosertib, cancer therapy