Cancer Discovery

Pancreatic ductal adenocarcinoma (PDAC) exhibits profound metabolic rewiring and strong resistance to DNA-damaging therapies, yet how metabolic pathways regulate genome maintenance remains poorly understood. The hexosamine biosynthetic pathway (HBP) integrates nutrient availability with protein glycosylation through production of UDP-GlcNAc, but its role in DNA damage response (DDR) regulation is unclear. Here we show that inhibition of the HBP enzyme phosphoglucomutase-3 (PGM3) reduces DNA repair capacity in pancreatic cancer cells. Transcriptomic and functional analyses reveal that the selective PGM3 inhibitor FR054 amplifies gemcitabine-induced replication stress, disrupts ATR-CHK1 and ATM-CHK2 checkpoint signaling, and selectively impairs homologous recombination. Glycoproteomic profiling identifies the AAA+ ATPase RUVBL2 as a key metabolic-DDR node. Gemcitabine increases RUVBL2 O-GlcNAcylation, with Thr81 identified as a modified residue within the Walker A nucleotide-binding motif. Structural modelling predicts that Thr81 O-GlcNAcylation stabilizes the RUVBL1-RUVBL2 complex without compromising ATP-Mg engagement. PGM3 inhibition and Thr81 mutation similarly reduced ATR and ATM abundance and promoted persistent DNA damage, supporting a role for RUVBL2 Thr81 O-GlcNAcylation in sustaining checkpoint signalling and genome stability. Consequently, PGM3 inhibition induces a BRCAness-like state that sensitizes pancreatic cancer cells to PARP inhibition, both in vitro and in vivo, as well as to ionizing radiation. These findings reveal a nutrient-sensitive mechanism linking protein glycosylation to genome maintenance and identify HBP-dependent DNA repair as a potentially actionable vulnerability in pancreatic cancer.

Solid tumors remain refractory to conventional treatments, yet cell surface proteins, by virtue of their extracellular accessibility and critical roles in tumor signaling, represent an attractive class of targets for precision-targeted therapy. Here, we report that TMEM132A is an essential and previously unrecognized pan-cancer target. TMEM132A interacts directly with EGFR and stabilizes its expression, thereby tethering EGFR at the plasma membrane and sustaining constitutive activation of lipid synthesis. Mechanistically, the TMEM132A-EGFR axis promotes lipogenesis by facilitating SREBP nuclear translocation, which in turn upregulates ACLY and ACSS2 expression to drive acetyl-CoA production and downstream lipid biosynthesis, ultimately disrupting lipid droplet homeostasis. To therapeutically target this axis, we developed a nanobody, LFNanoT132A#3, which effectively blocks the TMEM132A-EGFR interaction, abrogates downstream signaling activation, and potently inhibits proliferation across multiple solid tumor types. Notably, LFNanoT132A also exerts robust antitumor activity against H1975 xenografts, a model resistant to first- and second- generation EGFR inhibitors, underscoring its potential to overcome conventional drug resistance. Our findings establish TMEM132A#3 as a critical node in membrane-tethered oncogenic signaling and metabolic rewiring, and position LFNanoT132A#3 as a promising therapeutic candidate for precision cancer therapy.

KRAS is one of the most frequently mutated oncoproteins in cancer. Its ability to induce malignant transformation relies on metabolic reprogramming that causes cells to become dependent on aerobic glycolysis as a primary source of energy and for generating biological building blocks. Thus far, the signaling mechanism used by oncogenic KRAS to promote these changes in cancer cell metabolism has not been fully elucidated. However, through studies in pancreatic ductal adenocarcinoma (PDAC) cell lines and patient-derived organoids, we now demonstrate how oncogenic KRAS triggers an increase in glycolytic activity and identify Survivin as a newly discovered and critical KRAS-signaling partner essential for promoting these metabolic changes. We show that oncogenic KRAS potently upregulates the expression of Survivin in PDAC cells and patient-derived organoids undergoing increased glycolysis, whereas depleting Survivin expression inhibits their glycolytic activity and growth. Through a combination of cellular, biochemical, and imaging approaches, we further show that Survivin promotes the formation of unique microtubule-based structures that resemble invadosome rosettes, allowing for the recruitment of the glycolytic enzymes triose phosphate isomerase (TPI) and glyceraldehyde-3-phosphate dehydrogenase (GAPDH) to these super-structures which drives the increases in glycolysis. These findings demonstrate that by directing the assembly of a microtubule-based complex of metabolic enzymes, Survivin serves as a vital link in a KRAS signaling pathway responsible for promoting the metabolic changes necessary for the accelerated growth of PDAC cells, and thus potentially highlight new therapeutic strategies for treating KRAS-dependent cancers.

MYCN amplification drives replication stress in high-risk neuroblastoma, yet how MYCN-amplified tumour cells tolerate this stress to sustain proliferation remains poorly understood. Here we show that FBXL12, an SCF ubiquitin ligase substrate receptor that targets the Fanconi anaemia protein FANCD2 for degradation at replication forks, as well as the broader Fanconi anaemia and replication stress transcriptional program are elevated in high-risk and MYCN-amplified neuroblastoma. High FBXL12 expression independently predicts poor survival across neuroblastoma patient cohorts. FBXL12 loss stabilizes FANCD2 on chromatin, elevates ATR-dependent replication stress signalling and DNA damage during S phase, and impairs proliferation of MYCN-amplified neuroblastoma cells in vitro and in vivo. Mechanistically, MYCN directly engages the FBXL12-FANCD2 complex and antagonises FBXL12-mediated degradation of FANCD2 at replication forks, revealing that the oncogenic driver of replication stress also actively preserves the chromatin-bound FANCD2 pool required to tolerate it. Beyond S phase, FBXL12 loss disrupts FANCD2-dependent mitotic DNA synthesis and transmits unresolved replication intermediates into daughter cells. FBXL12-deficient cells consequently show transcriptional activation of MYC target gene, ATR, and mTOR signalling programs, and this pathway-concordant state confers differential sensitivity to ATR, and mTOR-targeting compounds, nominating candidate therapeutic strategies for this disease subset. Together, these findings define a MYCN-FBXL12-FANCD2 axis as a clinically relevant vulnerability in high-risk neuroblastoma.

Lung squamous cell carcinoma (LUSC) is an aggressive malignancy characterized by high cellular plasticity and few targeted treatment options. TNIK overexpression is common in LUSC and promotes tumor growth, with TNIK inhibition sensitizing LUSC to radiotherapy, though the underlying mechanisms are not well defined. Through transcriptomic analyses and functional assays, we identified TNIK as a regulator of a MYC-dependent transcriptional network that coordinates epithelial-mesenchymal plasticity and cell proliferation in LUSC. Depletion of TNIK reprogrammed LUSC cells from a hybrid epithelial/mesenchymal state towards an epithelial, senescent-like state characterized by reduced cell migration, invasion, reduced DNA synthesis, and enhanced {beta}-galactosidase activity. Using a small-molecule screen approach, we found that TNIK inhibitors cooperated with agents suppressing the histone methyltransferase and MYC binding partner EZH2, which further suppressed partial epithelial-to-mesenchymal transition (pEMT). Mechanistically, we identified MYC as a key downstream TNIK effector in LUSC cells: MYC depletion phenocopied the effects of TNIK loss on pEMT and senescence, and restoring MYC expression bypassed the effects of TNIK depletion. Collectively, these results implicate TNIK in the mechanisms linking epithelial-mesenchymal plasticity with proliferation and evasion of senescence and provide insights into future strategies for the clinical deployment of TNIK inhibitors in LUSC and other TNIK-dependent malignancies.

The balance between how much human tumors recapitulate fetal tissue programs versus lose adult tissue identity remains unresolved. I used audited vibe coding, a human-mediated, cross-model critique-and-refinement workflow, to re-analyze a public pan-cancer proteomic atlas. A primary large language model wrote and executed the analysis under scientific direction, while a separate model family audited the code, outputs and claims; findings were returned for correction across seven versioned releases. Among 229 tumor-adjacent pairs in seven organs, tumor-minus-adjacent proteomic change partially aligned with reverse fetal-to-adult maturation (organ-balanced cosine, 0.240; 95% interval, 0.138 to 0.335), with positive alignment in 189 of 229 patients (82.5%). The organ-balanced projection coefficient was 0.195 (95% interval, 0.069 to 0.244), indicating movement along only part of the developmental distance. Although reverse maturation overlapped adult-identity loss, a positive developmental component remained after identity loss entered first (0.203; 95% interval, 0.129 to 0.239). Suppression of adult-high proteins contributed to more positive alignment than reactivation of fetal-high proteins. The vibe coding audits identified substantive defects. A common-mask correction reduced the matched-organ advantage from 0.074 to 0.059; a missing-value correction barely changed aggregate geometry but replaced 5 of the top 40 liver contributors; and coupled resampling repaired uncertainty accounting without changing patient scores. As with any single report, the "vibe reanalysis" biological results are candidate discoveries pending independent replication. The workflow is a single feasibility case, not a reliability benchmark, and shows how conversationally generated analysis can be made more inspectable when model-written code is treated as untrusted, versioned and subject to separate-model critique and executable checks.

Prognostic models in oncology are developed one cancer at a time, from that cancer's own labelled outcomes, and fail where prognostic information is scarcest. Rare cancers account for roughly a fifth of diagnoses and most paediatric malignancies, yet seldom supply enough events for a reliable time-to-event model. We therefore asked whether a representation learned without outcome labels can supply what those cohorts cannot. A Transformer encoder was pretrained by masked field-value modelling on 9425135 tumour records from the SEER 17 registries, diagnosed in 2000 to 2023. Only diagnosis-time fields passing a fail-closed coding-verification gate were admitted, and each record was emitted as an era-specific and a harmonised view, keeping two decades of recoding auditable. The encoder was then frozen and read by a linear Cox head for overall survival. Nine rare cancers were removed from the pretraining corpus entirely, each requiring an independent pretraining run. On a sealed test partition, all nine exceeded an architecture-identical random frozen encoder in Harrell concordance by +0.0034 to +0.0368, every lower confidence limit above zero. At 256 labelled patients, all 67 cancers favoured the pretrained representation over budget-matched Cox regression, median difference +0.0283. The advantage was bounded: given the entire training set, Cox regression was favoured in seven of nine rare cancers. The encoder did not outperform a field-frequency baseline on its own objective, so upstream reconstruction did not predict downstream transfer. Outcome-agnostic registry pretraining carries prognostic signal into cancers it has never seen, and is most useful where labels are fewest, without establishing clinical utility.

Lineage-restricted transcriptional programs establish cell identity and can create selective dependencies in cancer. Here, we identify POU2AF2, encoding the transcriptional co-activator OCA-T1, as a critical lineage-specific dependency in a subset of diffuse large B-cell lymphoma (DLBCL). Pan-cancer dependency analyses and patient cohorts reveal elevated POU2AF2 expression in genetically aggressive DLBCL, where its depletion markedly suppresses tumor growth in vitro and in vivo. Mechanistically, POU2AF2 cooperates with the B-cell lineage-defining transcription factor POU2F2 (OCT2) to activate lymphocyte activation gene programs through direct chromatin engagement, thereby sustaining malignant transcriptional networks. We further identified a key epigenetic regulatory axis composed of the lineage-specific transcription factor TCF3 and the histone methyltransferase SET1A-COMPASS that drives POU2AF2 expression downstream of B-cell receptor signaling. Single-cell transcriptomic analysis reveals that POU2AF2 marks and sustains an innate-like B1 B-cell population in vivo, a candidate cell of origin for lymphoma. Together, these findings define a lineage-restricted POU2AF2/POU2F2 transcriptional module, controlled by a TCF3/SET1A epigenetic network, that sustains both innate-like B-cell identity and malignant fitness in DLBCL. Our study uncovers a previously unrecognized lineage-specific transcriptional dependency and highlights POU2AF2 and its associated regulatory circuitry as potential therapeutic targets in aggressive B-cell malignancies.

Genomic instability is a major driver of tumor evolution, promoting diversification and adaptation while simultaneously increasing the accumulation of deleterious alterations. How tumor populations balance these opposing effects remains poorly understood. Here, we introduce a computational framework that explicitly represents diploid genomes, functional gene classes, point mutations, and chromosome-segregation errors in spatially constrained and well-mixed tumor populations. We identify a viability boundary separating sustained tumor expansion from instability-induced population collapse. Within the viable regime, mutation and selection generate a stable distribution of genomic-instability classes that is accurately captured by an analytical replicator--mutator description. Near the viability boundary, tumor dynamics exhibit prolonged extinction transients and strong sensitivity to stochastic fluctuations, with important differences between solid and liquid architectures. Chromosomal alterations further modify growth by creating transient benefits through increased gene dosage and genetic redundancy, while ultimately increasing genomic fragility. Finally, simulated interventions show that eliminating low-instability subpopulations or increasing the global mutational burden can displace tumors beyond their viability boundary and trigger irreversible collapse. These results identify genome instability as both an evolutionary advantage and an intrinsic vulnerability, providing a quantitative framework for developing therapies that exploit the limits of tumor evolution.

High-grade serous ovarian carcinoma (HGSC) responds poorly to immune checkpoint blockade, partly due to a macrophage-dominated immunosuppressive microenvironment. We integrated single-cell spatial proteomics and spatial transcriptomics across 50 HGSC tumors and applied SPACEstat to resolve higher-order immune communities and their transcriptional programs. We identified six immune community types, with macrophage-dominated Myelonets representing the predominant spatial pattern of immune organisation. In chemotherapy-exposed tumors, Myelonets showed coordinated lipid metabolism-immunosuppression and inflammation-MHC-II macrophage transcriptional programs, with SPP1, C1Q, VEGF, MMPs, and CCL18 linked to immunosuppressive states and fibroblasts emerging as key mediators of macrophage communication. Chemotherapy contracted large Myelonets while increasing CD8+ T-cell organization into Lymphonets. Persistent macrophage dominance within Myelonets was associated with adverse outcomes among patients who achieved a complete response to treatment. Together, we identify Myelonets as clinically relevant, multicellular immunoregulatory niches sustained by spatiotemporally coordinated macrophage programs and stromal crosstalk.

Somatic mutations arise throughout life, with functional consequences tied to the cell populations in which they occur. Genome-wide studies measure somatic variations in bulk tissue, whereas single-cell approaches resolve cell identity but provide limited sensitivity for complex alleles. Here we developed SniffCell, which uses DNA methylation carried on native long reads to assign somatic variant-supporting molecules to methylation-resolvable cell types. SniffCell builds cell-type-discriminatory methylation signatures across eight tissues, assigns long reads to cell types, and provides cell-type-specific variant calling. Across peripheral blood mononuclear cells and brain benchmarks, SniffCell recovered sorted cell identities and validated cell-type-specific variant assignments using purified immune-cell, neuronal, and oligodendrocyte fractions. In blood, SniffCell recovered lineage-restricted antigen receptor rearrangements and localized a somatic tandem-repeat expansion to T cells. In the frontal cortex, SniffCell identified recurrent neuron-specific tandem-repeat expansions in genes including FGF14, LRRC7 and SH3RF3. Across three brain cohorts comprising 172 donors, recurrent neuron-associated expansions were enriched for GAA-rich motifs. In donors with matched blood, and diverged more strongly from the inherited repeat length, whereas oligodendrocyte-associated alleles more often tracked it. SniffCell transforms native bulk long-read genomes into a cell-type-aware resource for somatic variant discovery and reveals recurrent somatic instability in human tissues at cell-type resolution.

Despite the success of chimeric antigen receptor (CAR) T cell therapy for treatment of B cell acute lymphoblastic leukemia (B-ALL), its translation to acute myeloid leukemia (AML) has been hindered by limited efficacy and significant toxicity. Interferon-gamma (IFN{gamma}) blockade with emapalumab has recently emerged as a promising strategy to mitigate CAR T cell-related toxicities in B cell malignancies, based on evidence that IFN{gamma} is largely dispensable for optimal CAR T cell activity in B-ALL. Whether IFN{gamma} signaling is similarly non-essential in the AML context remains unclear. Here, we demonstrate that disruption of the IFN{gamma} axis impedes anti-AML CAR T cell function and prevents upregulation of target antigen CD123, the apoptotic mediator Fas, and the adhesion molecule ICAM-1 on AML cells. Conversely, exogenous IFN{gamma} enhances CAR T cell cytotoxicity and increases CAR T cell avidity for AML targets. These findings identify IFN{gamma} as a critical mediator of CAR T cell efficacy against AML by promoting increased target antigen expression, enhanced cytotoxicity, and stable CAR T/tumor interactions. Our results suggest that therapeutic IFN{gamma} blockade, including with emapalumab, may compromise CAR T cell responses in AML and should be approached with caution in this disease context.

Antral CCK2R+ stem cells are regulated by gastrin, but how endocrine and neural cues integrate under chronic injury remains unclear. Here we show that inducible hypogastrinemia shifts from asymmetric renewal to symmetric expansion of CCK2R+ stem cells. With carcinogenic stress, these cells acquire a cycling, injury responsive progenitor state revealed by single-cell RNA profiling. Acute gastrin loss activates a CCK2R+ nodose DMV vagal reflex that increases acetylcholine release, NGF production, cholinergic innervation, and Chrm3 expression, driving ERK and YAP signaling in CCK2R+ stem cells. Vagotomy, Trk inhibition, or Chrm3 deletion each suppressed stem cell expansion. In H. pylori and MNU injury models, hypogastrinemia amplified inflammation, dysplasia, and CCK2R+ clone expansion, whereas gastrin suppressed these responses. Human scRNA seq and spatial profiling confirmed G cell depletion and progenitor state enrichment. These findings define an endocrine neural epithelial axis in which gastrin loss boosts vagal M3R signaling to initiate antral preneoplasia, highlighting this pathway for early interception.

Accurate, reproducible interpretation of kidney allograft biopsies is critical for diagnosis of graft injury to guide prognosis and management. The international Banff classification is a consensus diagnostic system based on semiquantitative histological lesion scoring on either extent or severity of kidney transplant biopsies. However, pathologist scoring is limited by substantial interobserver variability, constrained scalability, and the inherent nature of the scoring system itself. Here we present BanffNET, a weakly supervised, probabilistic deep learning framework that combines self-supervised feature extraction with a novel Bayesian multiple-instance learning framework to predict (continuously) the full spectrum of Banff lesion scores directly from whole-slide images (WSIs). Using lesion-specific aggregation functions tailored to localized (modeling lesion severity) and diffuse pathologies (modeling lesion extent), BanffNET generates interpretable, patch-level probability maps and calibrated slide-level scores. BanffNET's performance was assessed relative to consensus, biological correlates of rejection and clinical outcome, demonstrating superior consistency, transportability and generalization. Trained on 7,249 WSIs from three cohorts, BanffNET demonstrates consistent performance on 11,028 WSIs across five external test sets, performing on par or exceeding expert consensus across lesions. BanffNET scores align more closely than pathologist Banff scores with molecular profiles of rejection, offering a transparent, biologically grounded framework for computational pathology with relevance beyond transplantation.

Bulk telomere length measured from tumour sequencing is routinely interpreted as a property of the cancer cells. However, a tumour specimen is a mixture, and the patient who supplies it has a telomere length of their own. Here I re-analyse published pan-cancer telomere estimates and ask how much of a tumour's telomere length is patient-specific. A calibration step comes first. Whole-genome and low-pass estimates recover the known cross-sectional attrition of leukocyte telomeres with age, at 26.6 bp per year in blood normals, whereas whole-exome estimates do not. After adjustment for cancer type, sequencing centre and sex, the exome slope is minus 0.6 bp per year. In 684 blood-normal aliquots sequenced by both assays, the whole-genome estimate declines at 38.9 bp per year, whereas the exome estimate from the same DNA shows no detectable decline. The difference between assays is 41.5 bp per year, with P = 3 x 10^-10. Because exome data constitute 78.6% of the original resource, downstream analyses use only whole-genome and low-pass libraries. Within those data, tumour telomere length tracks the patient's matched-normal telomere length. The Spearman correlation is 0.395 in TCGA, with positive associations in 22 of 23 cancer types. This finding replicates in PCAWG using a different telomere estimator, with a correlation of 0.472 and positive associations in all 24 histologies examined. Adjustment for cancer type, sequencing centre and library type leaves a regression coefficient of 0.385. The association is also stable after adjustment for age, sex, tumour purity, leukocyte fraction, ploidy, sequencing coverage and continental ancestry, with coefficients ranging from 0.406 to 0.429. Pure normal-cell admixture is rejected as the sole explanation. Under a two-compartment mixture model, the coefficient for host telomere length is expected to equal 1 and the host-by-purity interaction to equal minus 1. These restrictions are jointly rejected with P = 0.001. Tumour purity, leukocyte fraction and age each explain only about 1 to 3% of within-cohort variance and do not alter the cross-cancer ranking. By contrast, the between-cohort coefficient is not directly interpretable. Its apparent near one-to-one relationship with tissue-associated telomere length depends strongly on which tissue supplies the matched-normal reference and on the statistical spread of that predictor, falling to 0.44 when organ-matched solid tissue is used. Bulk tumour telomere length is therefore a composite phenotype containing a replicated patient-specific component. Telomere biomarker studies should include matched-normal telomere length as a covariate rather than treating tumour telomere length as exclusively tumour-intrinsic.

Disseminated cancer cells (DCC) in non-metastatic carcinoma patient bone marrow (BM) are predictive of metastasis. Those detected by epithelial cytokeratin or EpCAM expression have poorly-characterized transcription profiles due to their extreme rarity: 1~2 cells per two million BM cells in every third non-metastatic patient. We here characterize the transcriptomes of DCCs. Single-cell RNA-sequencing (scRNA-seq) of 864 EpCAM-positive cells (from 1,151 cancer patients) in BM or lymph nodes (LN) revealed plasma, immune, myeloid, erythroid progenitor cells and two candidate DCC populations, termed M0-DCC and M1-DCC. M0-DCC, mostly from non-metastatic M0-stage patients, displayed the highest known adult stemness scores, and were transcriptomically reminiscent of human cleavage-stage, preimplantation embryos. M1-DCC represented cancer cells undergoing the epithelial-mesenchymal transition (EMT), corresponding to later, implanting and gastrulating embryos. Detection of early-embryo-like DCC categorised patients at highest risk for metastatic progression. Furthermore, high M0-DCC scores predicted the metastatic potential of human cell lines from the Cancer Cell Line Encyclopedia. M0-DCC gene expression profiles can be reversibly induced from M1-DCC-like cells in vitro. The close correspondence between gene expression profiles in immediate early embryonic development and metastatic founder cell candidates provides strong evidence that the onset of cancer and metastasis recruits mechanisms employed in fertilization.

Klebsiella pneumoniae (Kp) is a WHO high-priority pathogen for vaccine development, yet previous efforts failed largely because key protective immune mechanisms remain unclear. Here we show that protective immunity conferred by mucosal mRNA vaccines (but not parenteral) require neither serum IgG nor airway secretory IgA, but instead depends on a previously unrecognized lung-resident CD8IL-17 T-cells (Tc17) that rapidly recruits neutrophils/macrophages to eliminate bacteria. To therapeutically harness this paradigm, we developed INSPIRE, a machine learning-engineered exosome platform incorporating donor-screened, miRNA-bioactive backbones (miR-21-mediated airway barrier penetration and miR-155-associated dendritic-cell activation through SOCS1/Inpp5d axis) and computationally designed peptides that boosts 11.6-fold mRNA encapsulation and 3-fold dendritic-cell cross-presentation. Intranasal INSPIRE-mRNA vaccination confers near-complete protection against clinically relevant Kp strains while intramuscular counterparts fail (below ~30% survival). Leveraging pIgR-/- and IL-17-/- mice coupled with T-cell depletions, we demonstrate the protection is Tc17-dependent. This work overturns the antibody-centric dogma and redefines a non-canonical Tc17-correlate for extracellular bacterial pneumonia.

Pathogenic variants in RUBCN, encoding the Run domain Beclin-1 interacting and cysteine-rich domain-containing protein (Rubicon) have been implicated in autosomal recessive spinocerebellar ataxia 15 (SCAR15). However, the molecular mechanisms underlying disease pathogenesis remain poorly understood. Here, we report 18 individuals from 15 unrelated families harbouring biallelic RUBCN variants, who present with an aggressive neurodevelopmental disorder variably characterized by seizures, developmental delay, intellectual disability and movement abnormalities that cause regression, progressive brain atrophy and neurodegenerative features. Through functional characterization, we demonstrate that a subset of disease-associated putative truncating variants disrupt autophagy regulation. In Caenorhabditis elegans models, loss-of-function RUBCN variants result in an increased autophagic flux and impaired neuronal function, recapitulating key features in humans. Correspondingly, cellular assays reveal that nonsense and frameshift RUBCN variants lead to defective autophagy inhibition, underscoring a crucial role for RUBCN as a key negative autophagy regulator. Molecular dynamics simulations rank the eleven missense variants by structural effect, with p.Arg813Trp alone altering the target protein at both the local and the regional level and lying within the RAB7A-binding module that the truncating alleles remove altogether. Our findings establish and expand the RUBCN-related disorders as a clinically and molecularly distinct subset of autophagy-related diseases. By delineating both the genetic landscape and cellular consequences of Rubicon dysfunction, this study enhances our understanding of autophagy-related neurodevelopmental disorders and provides a foundation for future therapeutic investigations.

Bladder carcinoma in situ (CIS) is a multifocal, non-muscle-invasive disease with a high risk of progression to muscle-invasive cancer. Current management strategies are often guided by genomic profiling of single tumor samples, which incompletely capture tumor heterogeneity and may contribute to treatment failure. In particular, the multifocal nature of CIS raises uncertainty regarding the uniformity of genomic, immunologic, and microenvironmental features across anatomically distinct sites within the same patient. To address this, we performed spatial transcriptomic profiling of CIS-containing tissue from four anatomically distinct sites within a single individual. Unsupervised clustering with marker-based annotation, integrated with metabolic inference, identified epithelial tumor populations alongside stromal, immune, and smooth muscle compartments. While key cellular states were conserved, their spatial organization and relative abundance varied by site. Metabolic analysis further revealed region-specific microenvironments shaped by local cellular architecture. These findings indicate that both cellular composition and metabolic activity are spatially structured. Collectively, these results demonstrate that CIS exhibits significant intra-patient heterogeneity not captured by single-site profiling. These findings require validation in larger cohorts but support multi-region sampling could help improve risk stratification, biomarker development, and prediction of response to intravesical therapies, with potential implications for more personalized treatment strategies.

In large data sets discovery is often limited to pre-conceived hypotheses and data fishing. Here we tested whether systematic exploration of AI generated hypotheses could uncover clinically meaningful signals in extensively studied data. We deployed AutoDiscovery, a newly launched large language model (LLM) framework designed to search for hypotheses based on surprisal and systematically interrogate complex datasets, on The Cancer Genome Atlas breast cancer cohort. The system did not identify clinically meaningful novel findings without human input. However, a seeded warm-start run with minimal text input from an oncologist revealed multiple interesting and surprising hypotheses. Among these was that a robust immune signature was present across all subtypes of invasive lobular carcinoma (ILC) that exceeded invasive ductal carcinoma (IDC). This observation was independently validated in independent cohorts and confirmed by high-sensitivity multi-immunofluorescence tumor tissue analyses. These results suggest immunotherapy approaches should be tested in ILC including early stage ER+HER2- ILC; these patients are currently excluded from large neoadjuvant immunotherapy trials. They further demonstrate that surprisal-based hypothesis generation frameworks can extract previously unappreciated patterns from deeply interrogated cancer datasets and imply that disease domain experts working with LLMs can derive more meaningful insights from complex data than either could achieve alone.