What has remained far murkier is how the loss of ATM reshapes the activity of microRNAs, the tiny non-coding RNA molecules that fine-tune gene expression after transcription. MicroRNAs destabilize messenger RNAs or block their translation into proteins, and in the DNA damage response they modulate repair, checkpoint control, apoptosis, senescence and cell-cycle progression. The relationship between ATM and microRNAs runs in both directions. Damage-responsive microRNAs can be transcribed under the control of p53, which itself sits downstream of ATM. ATM can directly activate microRNAs and even promote global microRNA biogenesis: a landmark study showed that ATM phosphorylates the RNA-binding protein KSRP, physically linking ATM signaling to the microRNA processing machinery. Conversely, the microRNA miR-421 can suppress ATM expression by targeting the ATM 3-prime untranslated region, experimentally producing checkpoint defects and increased radiosensitivity. This bidirectional wiring means that a cell lacking functional ATM could suffer microRNA dysregulation through multiple, overlapping mechanisms.

A new study published in Molecular Biology Reports by Muhammad Junaid Iqbal, Michele Menotta and colleagues at the University of Urbino Carlo Bo, together with collaborators in China and Pakistan, set out to characterize this interface in a controlled experimental system. Rather than attempting an unbiased biomarker screen, the team designed a mechanistically motivated assay of four predefined candidate microRNAs chosen to represent complementary arms of the ATM-associated DNA damage response and checkpoint network. miR-34a-5p was selected as the canonical p53-responsive microRNA involved in apoptosis, cell-cycle arrest and senescence, whose mature activity can also be modulated through ATM-dependent 5-prime-end phosphorylation after DNA damage. miR-26a-5p was included for its reported links to genotoxic stress and checkpoint regulation, including regulation of the kinases WEE1 and CHK1 and PTEN-AKT-associated phenotypes. miR-106b-5p and miR-20a-5p were chosen to represent the G1/S and RB/E2F checkpoint-modulatory arm, since miR-106b family members regulate the p21 cell-cycle inhibitor and miR-20a participates in c-Myc/E2F1-associated G1 checkpoint control.

The experimental design was deliberately simple and tightly controlled. The researchers compared one ATM-proficient human fibroblast control line, AG09429, with three patient-derived A-T fibroblast lines, GM00648, GM05849 and GM09607, all obtained from the Coriell Institute for Medical Research. Two of the lines were immortalized in-house by hTERT transduction, while the other two were obtained as transformed lines. To inflict controlled genotoxic stress, the team used bleomycin, a radiomimetic glycopeptide antibiotic that generates oxidative DNA lesions and strand breaks through an iron- and oxygen-dependent mechanism. Because ATM is the central transducer of double-strand break signaling, bleomycin provided a clean way to compare basal and stress-responsive microRNA behavior. Cultures were treated with 8 micrograms per milliliter of bleomycin for three hours, then allowed to recover for 24 hours before RNA extraction. Mature microRNA abundance was quantified by TaqMan reverse transcription quantitative PCR, normalized to the endogenous control miR-16-5p and analyzed with the comparative Ct method across three independent biological replicates per condition.

The basal results were striking. Under untreated conditions, miR-34a-5p showed the most pronounced difference between the control and the A-T panel: all three A-T lines displayed a highly significant reduction in miR-34a-5p compared with the untreated control, each with a p-value below 0.001 after Dunnett’s multiple-comparison testing. miR-26a-5p followed the same pattern, with significantly lower expression in all three A-T lines. miR-106b-5p showed a weaker, line-dependent profile, significantly reduced in two of the three A-T lines but not the third, while miR-20a-5p showed no significant basal differences at all. In other words, the dominant basal phenotype was a reproducible depletion of exactly the two microRNAs most directly tied to p53 signaling and checkpoint regulation, a pattern biologically coherent with impaired ATM-p53 signaling in these cells.

After bleomycin exposure, the between-line differences persisted. miR-34a-5p and miR-26a-5p remained significantly lower in all three bleomycin-treated A-T lines relative to the single bleomycin-treated control, again with p-values below 0.001. Meanwhile, miR-106b-5p and miR-20a-5p behaved in a more condition-dependent manner, with the strongest divergence seen in the AT648 line, which showed significantly higher expression of both microRNAs than the treated control. Crucially, the researchers separated between-line differences from acute treatment responses by performing paired within-line analyses. In the control fibroblasts, bleomycin significantly induced miR-34a-5p and reduced miR-106b-5p, a selective stress response rather than broad induction of the whole panel. AT648 showed the most dramatic behavior: all four candidate microRNAs were significantly induced after treatment, yet the induced miR-34a-5p and miR-26a-5p still failed to reach control-like levels. AT5849 showed only a significant reduction of miR-106b-5p, and AT9607 showed significant induction of miR-34a-5p alone. The response, in short, was heterogeneous and cell-line specific rather than a uniform A-T signature.

To organize the biological context of the candidate panel, the team performed target-based enrichment analysis using DIANA-miRPath v4.0 with the experimentally validated miRTarBase 2022 interaction database, complemented by a sensitivity analysis with TarBase v8.0 and a network-based analysis with MIENTURNET visualized in Cytoscape. Across KEGG, Reactome, Gene Ontology and MSigDB Hallmark gene sets, the documented targets converged on cell-cycle regulation, G1/S transition, RB/E2F signaling, p53 signaling, apoptosis, senescence, PI3K-Akt and mTOR signaling, and DNA damage-associated processes. The network analysis identified ten genes connected to three of the four input microRNAs: CCND1, CCND2, E2F1, E2F3, MYC, PTEN, RB1, SMAD4, VEGFA and WEE1. The authors are careful to note that because the microRNAs were selected a priori for their known involvement in these pathways, the enrichment results annotate the predefined panel rather than independently validating the qPCR findings, and the network remains a hypothesis-generating framework rather than demonstrated regulation.

The mechanistic implications are nonetheless compelling. The persistent depletion of miR-34a-5p across all three A-T backgrounds fits squarely within the ATM-p53 checkpoint axis, since ATM is an upstream regulator of p53 activation and miR-34a is a direct p53-responsive tumor-suppressive microRNA. Importantly, the data support impaired normalization rather than complete absence of inducibility: some A-T lines retained miR-34a-5p inducibility after bleomycin but never reached control-like abundance, a nuance that distinguishes this work from earlier radiation studies in lymphocytes where the miR-34a response was not strongly discriminatory between A-T and control samples. The AT648-specific induction of miR-106b-5p and miR-20a-5p further nominates a p21/RB/E2F checkpoint hypothesis, because miR-106b family members can repress CDKN1A/p21, the major p53 effector of G1/S arrest, and miR-20a modulates E2F1, which intersects functionally with ATM-dependent p53 phosphorylation and apoptosis. Persistent depletion of miR-26a-5p, meanwhile, nominates WEE1, CHK1 and related checkpoint axes for validation.

The authors are equally explicit about the limitations that bound their conclusions. Only one ATM-proficient control line was included, and each A-T line represents a single donor, so donor-specific effects cannot be definitively separated from disease-associated ones. The sample size of three biological replicates per condition limits statistical power, no multiplicity correction was applied across the full set of comparisons, and a single endogenous normalizer, miR-16-5p, is less robust than a geometric mean of multiple validated references. Only one bleomycin dose and one 24-hour recovery time point were tested, mature microRNA abundance was measured without assessing pri-miRNA, AGO loading or target repression, and the database-derived candidate targets were not validated at the mRNA or protein level. Differences in immortalization and transformation status among the cell lines may also have shaped the line-specific responses.

What the study ultimately delivers is a checkpoint-centered model that is genuinely testable. A-T fibroblasts appear to carry a stable, shared depletion of selected ATM, p53 and DNA damage response-associated microRNAs, combined with cell-line-specific stress-induced remodeling of G1/S and RB/E2F-associated microRNA networks. The proposed next steps are concrete: measure p53 activation, p21, WEE1, CHK1, RB1, E2F1, E2F3, the cyclins, CDK6, BCL2, BCL2L11 and PTEN at mRNA and protein levels under the same conditions; prioritize the AT648 line for cell-cycle profiling after bleomycin; and, most decisively, restore miR-34a-5p or miR-26a-5p with mimic experiments to test whether candidate targets and checkpoint phenotypes shift in the expected direction. For a rare disease in which the non-coding RNA literature remains thin, this work provides a clear, mechanistically grounded map of where to look next, and a reminder that in A-T, microRNA biology is model-dependent, stimulus-dependent and time-dependent rather than reducible to any single signature.

Subject of Research: Dysregulation of DNA damage response-associated microRNAs in ataxia-telangiectasia fibroblasts

Article Title: DNA damage response associated miRNA dysregulation in Ataxia-telangiectasia

Article References: Iqbal, M. J., Bucci, C., Ricci, A., Morganti, G., Khan, F. I., Khan, M. J., & Menotta, M. (2026). DNA damage response associated miRNA dysregulation in Ataxia-telangiectasia. Molecular Biology Reports, 53(1), Article 1660. https://doi.org/10.1007/s11033-026-12860-1

Image Credits: AI Generated

DOI: 10.1007/s11033-026-12860-1

Keywords: ataxia-telangiectasia, ATM kinase, microRNA, DNA damage response, miR-34a-5p, miR-26a-5p, p53 signaling, bleomycin, cell-cycle checkpoint, RB/E2F, fibroblasts, genome instability

The central conceptual move in the paper is a reframing of what distinguishes a quiescent cell from a senescent one. Rather than treating the two states as fundamentally different biological categories, Yilmaz and Shay describe them as positions along a spectrum of arrested states, defined by two parameters: the stability of the arrest and the magnitude of the perturbation required to reverse it. A deeply stable senescent cell and a shallowly arrested quiescent cell are, in this view, points on the same landscape of cell-cycle exit decisions, separated not by kind but by degree. The reframing does not erase the distinctive biology of senescence. The authors retain the senescence-associated secretory phenotype, known as SASP, along with metabolic reprogramming and chromatin remodeling, as active, context-dependent components of the senescent program. What changes is the interpretation: senescence is no longer a unique or exceptional state, but one outcome among several within a broader decision-making architecture.

To give this continuum quantitative teeth, the authors introduce what they call the Maintenance Threshold Model, or MTM. The model proposes that the stability of any arrested state emerges from a balance between two competing quantities: the cell’s maintenance capacity, meaning its ability to repair and renew its own components, and the accumulated cellular stress it carries. When maintenance capacity exceeds stress, an arrest remains reversible; when stress overwhelms maintenance, the cell tips into stable senescence. Critically, the model does not pool these inputs into a single average. Instead, it proposes that they are gated in series, a design principle with a striking technical implication: proteostatic and organellar maintenance capacity determines whether an arrested cell can complete division independently of whether the Rb-E2F restriction point is traversed. In other words, having an intact cell-cycle brake is not enough; a cell must also possess the protein-quality and organelle-renewal machinery to survive what comes after reactivation.

This serial gating has deep roots in systems biology. Switch-like, bistable behavior in stress-response networks is what allows cells to integrate signals about cellular stress, metabolic capacity, and quality-control machinery into discrete fate outcomes rather than graded drift. The review argues that such switch-like regulatory behavior, not categorical state identity alone, governs whether an arrested cell stays reversible, progresses toward stable senescent arrest, or resumes proliferation. Bistability, in this framework, means that two stable outcomes can coexist for the same set of underlying conditions, with the direction a cell takes determined by threshold crossings rather than by continuous adjustment. That architecture explains why cells exposed to similar levels of damage can end up in radically different fates, and why intervention at the right moment can, in principle, push an arrested cell back toward a reversible state or forward into irreversible arrest.

One of the model’s most consequential claims concerns why senescent cells persist. The authors propose that senescent-cell persistence is set not by one threshold but by two independent ones, both of which become less reliable with age. The first is cell-autonomous: the intrinsic maintenance capacity of the cell itself, its proteostasis, its macroautophagy, its organellar quality control. The second is non-cell-autonomous: immune surveillance, the ability of the surrounding organism to recognize and clear arrested cells. A senescent cell accumulates when its own maintenance systems fail to restore reactivation competence and when the immune system fails to remove it. Aging degrades both thresholds simultaneously, which, in the MTM framework, explains the age-dependent accumulation of senescent cells far more naturally than any single-mechanism account. The model thereby connects intracellular biochemistry and organism-level immunology in a single quantitative picture of tissue decline.

The review’s synthesis spans three traditionally separate research communities: aging biology, stem cell biology, and cancer research. In aging, the model provides a mechanistic rationale for why senescent cells accumulate in old tissues and why clearing them can restore function. In stem cell biology, it clarifies how tissue reserves maintain the option of re-entry into the cycle, a property essential for homeostasis and regeneration. In cancer, it speaks directly to a long-standing clinical puzzle: therapy-induced senescence. Many cancer treatments, from chemotherapy to radiation, work in part by driving tumor cells into a senescence-like arrest. But therapy-induced senescent cells are not always a victory. Some eventually escape arrest and resume proliferation, often with more aggressive behavior, and their SASP can remodel the tumor microenvironment in ways that support disease progression. The MTM offers a principled way to think about when arrested tumor cells will stay locked down and when they will rebound.

Because fate is plastic under this framework, the authors argue it can be deliberately manipulated, and they lay out a three-part therapeutic framework that exploits that plasticity. The first strategy is state-locking: pushing arrested cells firmly past the point of reversibility so that tumor cells driven into arrest by treatment never escape. The second is metabolic inflexibility targeting: senescent cells typically rely on particular metabolic configurations, and attacking those dependencies selectively eliminates them. The third, and perhaps most conceptually adventurous, is program hijacking: co-opting the senescent program itself, including its secretory machinery, for therapeutic benefit rather than simply suppressing it. Each strategy treats the arrest-state continuum as a control surface with tunable dials rather than a fixed landscape, aiming interventions at the thresholds that determine outcomes.

The same logic runs in reverse for aging interventions. If senescent-cell burden is set by two aging-sensitive thresholds, then interventions could target either side: bolstering cell-autonomous maintenance capacity, through approaches that support proteostasis and macroautophagy, or strengthening non-cell-autonomous immune surveillance, so that the aging immune system regains its ability to find and clear arrested cells. The framework also suggests caution for senolytic approaches: rather than viewing all senescent cells as targets for elimination, a threshold-based view encourages asking whether individual arrested cells sit near the reversible end of the continuum, where restoring maintenance capacity might be preferable to killing. Fate plasticity cuts both ways, offering opportunities to restore youthful function as well as risks of reactivating the wrong cells.

The technical vocabulary of the model is worth pausing on, because it encodes the review’s central bet. The Rb-E2F pathway is the canonical molecular restriction point governing G1-to-S transition, long treated as the decisive gate for cell-cycle re-entry. By insisting that proteostatic and organellar maintenance is gated in series with, and independent of, that restriction point, the MTM makes a falsifiable prediction: cells that pass the cell-cycle gate but lack maintenance capacity should fail to complete division or collapse into stress, while cells with robust maintenance should remain reactivation-competent even after prolonged arrest. Experiments that independently titrate CDK2 activity, autophagic flux, and proteostatic load against one another would directly test this claim. Similarly, the two-threshold account of senescent-cell persistence predicts that impairing immune surveillance should raise senescent-cell burden multiplicatively with intrinsic maintenance decline, a prediction testable in aging animal models.

What emerges from the review is a decision-centric framework in which quiescence, senescence, and proliferation are dynamically stabilized states shaped by tunable regulatory thresholds rather than fixed endpoints. Cellular fate, on this account, is a dynamic balance between damage accumulation and the capacity of quality-control systems to preserve reactivation competence, and the balance can be tipped deliberately. For a field that has spent decades cataloging markers of senescence and debating definitions, the shift from taxonomy to dynamics is significant. If the thresholds the model identifies can be measured, modulated, and eventually targeted in patients, the arrested cell may stop being a biological verdict and become, instead, a clinical choice, one that physicians could push toward clearance, commitment, or recovery depending on what the disease and the patient demand.

Subject of Research: Threshold control of cell fate decisions between quiescence and senescence in aging and cancer

Article Title: Threshold control of growth arrest in quiescence and senescence: implications for cell fate plasticity in aging and cancer

Article References: Threshold control of growth arrest in quiescence and senescence: implications for cell fate plasticity in aging and cancer. (n.d.). https://doi.org/10.1186/s12943-026-02782-8

Image Credits: AI Generated

DOI: 10.1186/s12943-026-02782-8

Keywords: cellular senescence, quiescence, cell fate plasticity, Maintenance Threshold Model, SASP, Rb-E2F, CDK2, proteostasis, macroautophagy, immune surveillance, therapy-induced senescence, aging