搜索结果：Cancer genomics & proteomics

Breast cancer is a heterogeneous disease characterized by substantial molecular diversity and variable treatment outcomes across patients. Despite advances in targeted and systemic therapies, anticipating individual benefit remains a major clinical challenge. In this context, Artificial Intelligence (AI) can support precision oncology by integrating high-dimensional molecular profiles with clinical and pharmacological information. Here, we present B.R.E.A.S.T. (Breast canceR Enhanced AI-Supported Therapy), an interpretable machine learning framework designed to predict therapy outcome from tumor proteomic profiles integrated with clinical and treatment annotations. Proteomic data from The Cancer Genome Atlas (TCGA) and The Cancer Proteome Atlas (TCPA) were harmonized with outcome and therapy information, and thirteen supervised classifiers were systematically evaluated using stratified 5-fold cross-validation. Therapeutic outcome labels were operationally defined by integrating available treatment response annotations with complementary clinical outcome information. Across both cohorts, ensemble-based models consistently achieved the most stable and highest discriminative performance, supported by learning-curve analyses and consistent behavior across independent datasets. To enhance interpretability, we implemented a two-step feature selection strategy combining model-specific importance measures with a global consensus ranking, enabling the identification of a compact set of robust proteomic biomarkers associated with therapeutic outcome. Top-ranked features mapped to molecular programs relevant to breast cancer progression and treatment sensitivity, including regulators of cell survival, DNA damage response, PI3K/AKT/mTOR signaling, and invasion-related processes. Re-evaluation using only the top 30 globally ranked features preserved high predictive performance across both independent breast cancer cohorts, indicating that a parsimonious proteomic signature captures core molecular determinants of outcome. Overall, B.R.E.A.S.T. provides a robust and generalizable proteomics-driven framework for modeling outcome-associated therapeutic response patterns and supporting biologically informed biomarker discovery in breast cancer.

To investigate the biological attributes of core syndromes in colorectal cancer, namely, the damp-heat and stasis-toxin syndrome (SRYD). Between October 2021 and October 2022, a cohort comprising 40 patients with colorectal cancer (CRC) diagnosed with damp-heat and stasis-toxin syndrome (SRYD group), 40 patients with CRC without this syndrome (non-SRYD group), and 40 healthy controls (Normal group) was recruited at Jiangsu Province Hospital of Chinese Medicine. Untargeted metabolomics analysis was conducted on plasma samples from all 120 participants, while differential protein analysis using four-dimensional data-independent acquisition proteomics was performed on 20 randomly selected samples per group. A combined analysis of proteomics and metabolomics data followed, and the identified potential diagnostic biomarkers were subsequently used to train and validate multiple machine learning models. Proteomic analysis revealed 130 differential proteins in the colorectal cancer with damp-heat and stasis-toxin syndrome (CRC-SRYD) group, enriched in pathways including complement and coagulation cascades, as well as nuclear factor kappa-B (NF-κB) signaling. Metabolomic analysis identified 584 differential metabolites within the same group, showing enrichment in pathways such as primary bile acid biosynthesis, central carbon metabolism in cancer, and glucagon signaling. Integrated pathway analysis indicated heightened activity of the NF-κB signaling pathway in the CRC-SRYD group. A biomarker panel, comprising 6 proteins and 9 metabolites selected through the ReliefF algorithm, was used to construct a diagnostic model with random forest, achieving an accuracy of 93.33%, sensitivity of 80.00%, and specificity of 100%. This study systematically elucidates plasma metabolomic and proteomic alterations in patients with CRC, establishing a robust diagnostic model for CRC syndrome (CRC-SRYD). Further investigation is warranted to clarify the underlying molecular mechanisms and biological foundations.

MAPK13 (p38δ) is a frequently identified but often ignored MAP kinase in epithelial gene signatures. This study aimed to characterize p38δ expression in normal and cancerous epithelial tissues, investigate its possible role in epithelial identity, epithelial-mesenchymal transition (EMT), cancer progression, drug resistance, and evaluate its prognostic value. Quantitative analysis assessed p38δ expression across diverse normal human epithelial tissues and correlated it with epithelial markers in cancer cell lines. A p38δ gene signature was analyzed for pathway enrichment in epithelial processes. p38δ expression was compared between epithelial and mesenchymal lung cancer models and in EGFR inhibitor-resistant cells. Prognostic data on cancer patient survival was collected for various cancer types. Functional studies included p38δ knockdown to evaluate EMT markers, E-cadherin/EpCAM, Vimentin and proliferation, and exogenous p38δ expression to assess cell growth/migration. p38δ mRNA and protein levels were measured in Osimertinib-resistant cell lines. p38δ was predominantly expressed across diverse normal epithelial tissues and correlated strongly with epithelial markers in cancer cells. Its gene signature enriched pathways vital for epithelial architecture, adhesion, and differentiation. p38δ was consistently higher in epithelial-enriched lung cancer models and suppressed in EGFR inhibitor-resistant cells. Prognostic impact was context-dependent: high expression correlated with favorable overall survival in ovarian, lung, and rectum cancers, but inversely with outcomes in liver, breast, and pancreatic cancers. Functionally, p38δ knockdown induced EMT and increased proliferation. Conversely, exogenous p38δ suppressed cell growth/migration. Crucially, acquired Osimertinib resistance consistently corresponded with significant reductions in p38δ levels. p38δ is a critical marker and regulator of epithelial identity. Its dysregulation and loss are systematically linked to EMT and drug resistance in cancer. Given its complex, context-dependent prognostic value and functional significance, p38δ holds significant potential as a therapeutic target.

MEX3D, a member of the MEX3 RNA-binding protein family, has emerged as a potential regulatory molecule in cancer. However, its role across different tumor types remains largely unexplored. We conducted a pan-cancer analysis of MEX3D using transcriptomic and proteomic data from the Cancer Genome Atlas (TCGA), Genotype-Tissue Expression (GTEx), and Clinical Proteomic Tumor Analysis Consortium (CPTAC). Expression patterns, clinical correlations, survival outcomes, genetic alterations, RNA modification associations, immune infiltration, and functional enrichment were systematically evaluated. MEX3D was significantly dysregulated in numerous cancers at both mRNA and protein levels. Its expression correlated with tumor stage in ACC, LIHC, OV, SKCM, and THCA. Elevated MEX3D expression was associated with poor overall survival (OS) and disease-specific survival (DSS) in multiple malignancies, including ACC, LGG, LUAD, and MESO. Genetic alteration analysis revealed frequent amplifications and mutations, particularly in SARC and OV. MEX3D was positively correlated with RNA modification-related genes (m1A, m5C, m6A) and immune regulatory genes such as CD276, TGFB1, VEGFA, and ICOSLG. Additionally, MEX3D expression showed significant associations with tumor mutational burden (TMB), microsatellite instability (MSI), and cancer-associated fibroblast infiltration. Functional enrichment analyses indicated that MEX3D-related genes are involved in reproductive cellular processes, RNA binding, the Hippo signaling pathway, and microRNA-related oncogenic pathways. This pan-cancer analysis highlights the heterogeneous expression and cancer-specific prognostic significance of MEX3D. MEX3D is associated with immune infiltration, immune regulatory genes, RNA modification-related genes, TMB/MSI, and pathways involved in gene regulation and tumor progression. These findings suggest that MEX3D may participate in cancer-specific post-transcriptional and microenvironmental regulatory networks.

ObjectiveLuteolin, a natural flavonoid, exhibits anticancer effects; however, its mechanisms of action are unclear.MethodsThis retrospective computational study combined network pharmacology, molecular docking, and molecular dynamics simulation to explore luteolin's anticancer mechanisms in five types of cancers, including breast, lung, colorectal, gastric, and liver cancers. Network pharmacology helps uncover multi-target interactions in complex diseases. Public databases were used to build compound-target, target-disease, and protein-protein interaction networks. Molecular docking and dynamics simulations validated luteolin's binding to AKT1. The Clinical Proteomic Tumor Analysis Consortium dataset and Human Protein Atlas database were used to assess AKT1 expression in cancer versus normal tissues.ResultsTopological analysis identified AKT1 (a PI3 K/AKT pathway kinase) as a hub gene linked to apoptosis, cell cycle, and tumor suppression. Molecular studies confirmed strong luteolin-AKT1 binding. Clinical Proteomic Tumor Analysis Consortium and Human Protein Atlas data showed that compared with normal tissues, higher AKT1 expression was present in lung, gastric, and breast cancers and lower expression was observed in liver and colorectal cancers.ConclusionThe anticancer effects exerted by luteolin may involve AKT1 signaling modulation. Network pharmacology aids in revealing multi-target mechanisms of natural compounds, supporting further research on the therapeutic potential of luteolin.

Notch signaling exerts context-dependent effects in cancer; however, the prognostic relevance of inherited variation in Notch-related genes in prostate cancer remains unclear. We investigated whether germline single-nucleotide polymorphisms (SNPs) in Notch pathway genes are associated with clinical outcomes in men receiving androgen deprivation therapy (ADT). We genotyped 222 SNPs across 24 Notch pathway-related genes in 630 patients with advanced prostate cancer. Associations with cancer-specific survival (CSS) and overall survival (OS) were assessed using Cox proportional hazards models. Integrative bioinformatics analyses, including pooled transcriptomic dataset analysis, expression quantitative trait locus analysis, and pathway enrichment, were performed to assess functional relevance. Multiple SNPs were significantly associated with CSS and OS, with Yes1-associated transcriptional regulator (YAP1) rs1894116 showing the strongest prognostic signal. The minor G allele was associated with a 26% reduction in prostate cancer-specific mortality and a 22% reduction in all-cause mortality, independent of established clinical predictors. Functional annotation suggested that rs1894116 may be linked to increased YAP1 expression. Pooled transcriptomic analyses showed lower YAP1 expression in tumors than in normal tissue, while higher YAP1 expression correlated with more favorable outcomes. YAP1-correlated genes were enriched in the focal adhesion pathway, suggesting a tumor-suppressive YAP1-focal adhesion axis, particularly involving vinculin. Germline variation in the Notch-YAP axis, notably YAP1 rs1894116, predicts survival in patients with prostate cancer treated with ADT. Elevated YAP1 expression and coordinated activation of focal adhesion components may be associated with less aggressive disease. These findings provide mechanistic insight into Notch signaling in prostate cancer and highlight YAP1 and its variants as potential biomarkers for risk stratification and personalized therapy.

Sialyltransferase ST6GAL1 is known to be upregulated in various cancer types, including rectal cancer, and is linked to poor prognosis. This enzyme catalyzes the addition of sialic acids to various proteins, altering their activity and function. Our lab has previously shown that ST6GAL1 causes resistance to chemoradiation therapy (CRT) in rectal cancer, via sialylation of tumor necrosis factor receptor 1 (TNFR1). This leads to decreased TNFR1 mediated cell death/apoptosis. We aimed to elucidate whether beta-site amyloid precursor protein cleaving enzyme 1 (BACE1), which cleaves ST6GAL1, has a role in treatment response to CRT and whether its overexpression (OE) could improve response to CRT. We first assessed BACE1 and ST6GAL1 in pre-treatment biopsies from rectal cancer patients. We then evaluated the effects of BACE1 modulation in CRC cell lines. The high BACE1-expressing CRC cell line SW1463 was utilized for inhibition studies, and SW620 CRC cells lines were additionally lentivirally transfected for OE of BACE1 (validated by flow cytometry, qPCR, Western Blotting, and activity assay). BACE1 OE cells were also utilized for response to CRT. Analysis of patient biopsies showed that BACE1 mRNA was significantly increased in patients with a complete response to CRT. SW1463 colorectal cancer cells treated with a BACE1 inhibitor had increased cell-surface sialylation and increased survival after CRT compared to vehicle-treated cells. BACE1 overexpressing SW620 CRC clones exhibited protein expression and activity of BACE1, with reduced ST6GAL1 protein. BACE1 OE colorectal cancer cells also had significantly increased apoptosis and decreased survival after CRT. BACE1 OE clones had decreased sialylation of the apoptosis receptor, TNFR1. Collectively, we found that overexpressing BACE1 altered ST6GAL1-mediated CRT resistance in CRC cells. BACE1 enzymatic cleavage of ST6GAL1 may be a potential target for improving therapeutic response to CRT in rectal cancer.

The acquisition of stem-like properties and increased cellular plasticity is thought to drive tumor progression and therapeutic resistance, but grade-specific molecular trajectories across the epigenetic landscape in bladder cancer remain poorly defined. This study examined the dynamic rewiring of signaling networks and proteomic landscapes during reprogramming of low-grade (HTB-2) and high-grade (HTB-5) bladder cancer cells, and during their subsequent differentiation into embryoid bodies. Six experimental models were analyzed, including the parental HTB-2 and HTB-5 cells, their Sendai virus-reprogrammed counterparts (rep HTB-2 and rep HTB-5) and embryoid bodies generated from reprogrammed derivatives (rep HTB-2 EB and rep HTB-5 EB), with SV-HUC-1 uroepithelial cells used as a control. Phosphoproteomic and integrated proteomic analyses were performed to define grade-specific signaling architectures and shared plasticity-associated signatures, followed by clinical validation using pan-cancer datasets. Phosphoproteomic reconstruction revealed grade-dependent kinase network architectures associated with stem-like induction. These findings indicate that reprogramming induced cell-line-specific signaling rewiring, with HTB-2 cells showing enhanced MAPK/Src-family-associated phosphorylation and HTB-5 cells showing increased AKT/PRAS40 and STAT1/STAT3 phosphorylation together with reduced ERK1/2-MSK1/2 signaling. During differentiation, low-grade cells underwent metabolic reprogramming, while high-grade cells favored cytoskeletal remodeling and extracellular matrix organization. Integrated proteomics defined a shared plasticity signature, with reprogrammed models recapitulated key bladder cancer features and clinically relevant outcomes. These findings support a hierarchical model of bladder cancer progression where reprogramming induces a transient intermediate state that enables invasive features upon re-differentiation. The study reveals grade-specific signaling and proteomic adaptations, identifying differentiation as a critical window for uncovering prognostic biomarkers and therapeutic targets. Together, these results suggest that reprogrammed bladder cancer models provide a biologically relevant platform to study tumor plasticity, progression and therapeutic vulnerability.

Prostate cancer (PCa) remains a leading cause of cancer-related morbidity and mortality among men worldwide, with advanced stages often exhibiting resistance to standard therapies. Drug resistance in advanced prostate cancer is a multifactorial process influenced by genetic mutations, hereditary factors, epigenetic alterations, lifestyle choices, and dietary habits. Understanding this complex interplay is crucial for developing effective predictive and therapeutic strategies. In this context, a multiomics approach integrating genomics, transcriptomics, proteomics, metabolomics, and epigenomics offers a comprehensive framework to dissect the molecular mechanisms driving drug resistance.Current challenges include tumor heterogeneity, limited access to longitudinal patient data, and insufficient representation of diverse populations in omics datasets. This study highlights the novel application of multiomics integration to stratify patient subgroups based on molecular signatures, enabling early prediction of therapy resistance. We also emphasize key biomarkers and signaling pathways associated with treatment failure, including alterations in androgen receptor signaling, PI3K-AKT-motor pathways, and metabolic rewiring. The integration of these datasets not only enhances diagnostic precision but also aids in identifying actionable therapeutic targets. It also opens up new avenues for developing targeted therapies by integrative profiling and quantifying a broad spectrum of biomolecular features across distinct subtypes of malignant cells.

Endometrial cancer (EC) is the most common gynaecological malignancy worldwide, yet the prognosis for advanced and recurrent disease remains poor, highlighting the need for improved diagnostic, prognostic, and therapeutic decision-making frameworks. Conventional approaches, including histopathology, imaging, and single-layer molecular profiling, provide essential clinical information but may not fully capture EC's biological heterogeneity, especially within clinically challenging No Specific Molecular Profile (NSMP) and mismatch repair-deficient (MMRd) subgroups. Artificial intelligence (AI) and machine learning (ML) provide powerful approaches to analyse complex, high-dimensional datasets generated by multi-omics profiling, histopathology, imaging, and clinical records.In this review, we synthesize the latest evidence on AI-driven multi-omics research in EC, encompassing genomics, transcriptomics, proteomics, metabolomics, epigenomics, single-cell profiling, and spatial transcriptomics. Unlike other reviews that focus solely on AI, omics, or imaging, we integrate molecular, imaging, histopathological, and computational perspectives to underscore their collective impact on precision oncology in EC. We subsequently explore applications in molecular subtyping, predicting survival and recurrence, modelling treatment responses, discovering immunotherapy biomarkers, and identifying drug targets. Public resources such as The Cancer Genome Atlas (TCGA), Clinical Proteomic Tumour Analysis Consortium (CPTAC), Gene Expression Omnibus (GEO), cBioPortal, Human Protein Atlas, GTEx, and UCSC Xena have enabled large-scale reproducible analyses. However, challenges such as cohort heterogeneity, batch effects, ethnic underrepresentation, missing annotations, and the need for external validation remain significant hurdles.We then discuss the progression from conventional ML methods to deep learning architectures, including convolutional neural networks, transformers, graph neural networks, and multimodal fusion models applied to histopathological, radiological, and multi-omics data. Landmark models such as EndoNet, EndoRisk, and HECTOR illustrate the potential of AI-enabled approaches to support EC grading, molecular inference, lymph node metastasis prediction, and recurrence-risk stratification. Finally, we examine key translational barriers, including class imbalance, interpretability, data harmonization, regulatory requirements, and the implementation gap between high-performing retrospective models and routine clinical deployment. Ultimately, this review underscores how bridging these multi-modal computational approaches paves the way for precision oncology in EC.

Platinum (Pt)-based chemotherapeutics are widely used for cancer treatment in clinical trials. It has been reported that protein O-linked β-N-acetylglucosamine (O-GlcNAc) modification occurs on several important proteins implicated in regulating cancer cell response to Pt drugs. However, the O-GlcNAc proteomic landscape during this process remains poorly characterized. Herein, we report quantitative profiling of O-GlcNAcylation sites with carboplatin exposure, by using a chemoenzymatic labeling-assisted chemoproteomic approach. A total of 244 O-GlcNAcylated sites, many of which occur on essential genome stability regulators, are quantified. Furthermore, we discover that the cellular O-GlcNAc level is elevated upon carboplatin treatment and suppression of O-GlcNAcylation renders cancer cells more sensitive to carboplatin. These results establish a valuable resource for elucidating the functional role of O-GlcNAcylation in carboplatin response and suggest potential insights into cancer chemotherapy. The chemoproteomic strategy should be generally applicable for monitoring O-GlcNAc dynamic changes in various pharmacological processes.

Apolipoprotein C1 (APOC1) has been implicated in several malignancies, yet its expression patterns, clinical significance, and immunomodulatory roles across cancer types remain poorly characterized. We performed a comprehensive multi-omic analysis of APOC1 across 33 cancer types integrating transcriptomic, proteomic, genomic, epigenomic, and pharmacogenomic data from TCGA, GTEx, CPTAC, and multiple independent external cohorts. Immune infiltration was assessed using seven complementary algorithms. Spatial transcriptomics and single-cell RNA sequencing were employed to determine the cellular source of APOC1 expression. APOC1 upregulation in most cancers was associated with cancer type-specific prognosis. After adjustment for clinical covariates and macrophage infiltration, high APOC1 remained an independent adverse factor in KIRC, LGG, and STAD. APOC1 expression positively correlated with genomic instability hallmarks, including homologous recombination deficiency and aneuploidy, with these associations largely independent of immune infiltration; in contrast, associations with tumor mutational burden were substantially confounded by macrophage abundance. Immune infiltration analysis revealed a pattern consistent with adaptive immune resistance: APOC1 correlated positively with immune-activating signatures (STAT1, MHC-II, TCR signaling) and immunosuppressive M2 macrophages and Tregs, yet negatively with anti-tumor effectors (activated NK cells, dendritic cells). Spatial transcriptomics and single-cell RNA sequencing identified tumor-associated macrophages (TAMs) as the primary cellular source of APOC1, with transcripts co-localizing with CD68 in tissue sections. APOC1 expression correlated with multiple immune checkpoint molecules and was elevated in responders to immune checkpoint blockade, consistent with an inflamed yet regulated tumor microenvironment. Pharmacogenomic analyses revealed that APOC1-high tumors display distinct drug response profiles, characterized by resistance to MAPK pathway inhibitors and potential sensitivity to the HDAC inhibitor Entinostat. This pan-cancer analysis establishes APOC1 as a context-dependent biomarker and a TAM-derived modulator of adaptive immune resistance, with prognostic and therapeutic implications across malignancies. APOC1-expressing TAMs represent a potential target for combination immunotherapy strategies.

Female patients with ER+HER2- breast cancer have a favourable prognosis for 5-10 years. Later relapses are, however, common, yet predictions of late recurrence risk are suboptimal, particularly for patients with intermediate risk determined by the Oncotype Dx Recurrence Score (RS, 16-25). Here, we analyse tissue samples from patients with ER+HER2- breast cancer using spatial proteomics (multiplex immunofluorescence with 5 markers, n = 440) and spatial transcriptomics (n = 359), and find decoupled immune states between stroma and epithelia. Moreover, inflamed stroma express genes linked to tissue remodelling, immune exhaustion, and inhibitory/checkpoint receptors (CTLA4, TIGIT, CD96); inflamed epithelia similarly express genes associated with checkpoints (CTLA4) and exhaustion (CXCL13), but also genes attributed to antigen presentation. In our randomised, Intermediate RS cohort treated with chemotherapy we observe an association between higher stromal tumour-infiltrating CD8+ lymphocyte (sTIL CD8+) density and poor outcome (ΔLR-χ2: 6.79, p = 0.009), which we validate using data from whole-resection specimens (ΔLR-χ2: 8.90, p = 0.003). Our data thus provide insights into the immune states in ER+HER2- breast cancer, and propose sTIL CD8+ density as candidate biomarker for treatment decisions.

Acute myeloid leukemia (AML) is a heterogeneous and aggressive malignancy with limited therapeutic options and high relapse rates. Despite advances in genomic profiling, many genetic aberrations remain untargetable, and current risk stratification models often fail to predict treatment responses. Proteomics offers a complementary approach by directly measuring protein abundance, post-translational modifications, and protein-protein interactions, providing mechanistic insights into drug resistance, disease progression, and therapeutic vulnerabilities. In this review, we explore the emerging role of proteomics in AML, focusing on its application in biomarker discovery, prediction of drug responses, and identification of novel therapeutic targets. Special attention is given to antigen discovery for immunotherapy, where surface and immunopeptidomics enable the identification of AML-specific antigens and neoepitopes. These insights are critical for the development of antigen-targeted therapies, including chimeric antigen receptor (CAR) and T cell receptor (TCR)-based immunotherapies. Integrating proteomics into a multiomics framework could provide actionable insights for guiding precision medicine and improving AML outcomes.

Background/Objectives: The tumour immune microenvironment (TIME) critically influences colorectal cancer (CRC) progression and therapeutic response, yet mechanisms shaping immune phenotypes remain unclear. Mucin-type O-glycosylation regulates tumour-immune interactions at the cell surface. Methods: We analysed O-glycosylation activity in 988 colorectal cancer (CRC) tumours derived from three independent cohorts: The Cancer Genome Atlas (TCGA-CRC, n = 534), the Clinical Proteomic Tumour Analysis Consortium (CPTAC2-CRC, n = 106), and the Sidra-Leiden University Medical Center (Sidra-LUMC, n = 348). O-glycosylation activity was quantified using a transcriptomic gene signature and single-sample gene set enrichment analysis (ssGSEA). Tumours were stratified into high and low O-glycosylation groups based on the median score, and associations with immune phenotypes, genomic alterations, and tumour functional states were assessed. Results: High O-glycosylation tumours exhibited an immune-desert phenotype with reduced immune-inflamed (p = 3.65 × 10-10) and immune-excluded (p = 0.0070) signatures alongside increased immune-desert scores (p = 0.0049) and reduced Siglec signalling (p = 8.14 × 10-5). O-glycosylation was associated with genomic stability, including lower TP53 mutation frequency (p = 0.0056), reduced aneuploidy (p = 0.0116), and decreased fraction of genome altered (p = 0.0309). High O-glycosylation tumours also showed upregulation of multidrug resistance programmes and reduced epithelial-mesenchymal transition (p = 0.0141) and proliferation (p = 0.0294). Conclusions: O-glycosylation defines a CRC subtype characterised by immune exclusion, genomic stability, and multidrug resistance, highlighting its potential as a biomarker and therapeutic target.

Cancer metabolism is often viewed as a cooperative reliance on glucose and glutamine; however, whether these nutrients can enforce discrete, non-overlapping metabolic states remains unclear. This study aimed to isolate nutrient-specific regulatory programs. MDA-MB-231 human breast cancer cells were cultured under four distinct metabolic environments: glucose/glutamine nutrient-repleted (fed), dual glucose/glutamine deficiency, and isolated repletion of either glucose or glutamine. Groups were evaluated for integrated transcriptomic, metabolomic, and lipidomic profiles to identify only the non-redundant, nutrient-enforced architectures. The data show a mutually restrictive mechanistic state. Glutamine functions as a metabolic architect, restoring glycolytic enzyme transcripts (without lactate production), while inducing PDK1/3 which would decouple glycolysis from the TCA cycle. These changes are concomitant with a glutamine flux toward reductive TCA-driven lipogenesis, citric acid overflow, sterol synthesis (SREBF1/2), structural membrane expansion (phospholipids/sphingolipids) and the unique production of alanine as a nitrogen pool, independent of glycolytic flux. Conversely, glucose alone acts as the executor, licensing chromatin engagement, DNA replication, and mitotic progression. Glucose alone resolved ER stress, restored hexose-phosphate-derived glycosylation (mannose-6-phosphate), enabled lactic acid production, and diverted excess carbon into a triglyceride storage pool (>40% of lipids). Notably, each nutrient suppressed core elements of the other's program, revealing a reciprocal activation-braking system. Interestingly, ATP yield from glucose or glutamine alone were comparable, but not arbitrary; instead, aligned with the functional state of the cell. Glucose alone supported glycolytic phosphorylation and proliferative execution, as marked by lactate accumulation, whereas glutamine alone supported Krebs cycle-related phosphorylation, characterized by citrate accumulation and the maintenance of cellular structure and membrane infrastructure. Glucose and glutamine enforce a balance of two independent, reciprocally regulated metabolic states. This data provides a systems-level explanation for metabolic resilience in cancer and may lead to the identification of nutrient-specific targets for combination therapy.

Current liquid-biopsy technologies predominantly rely on genomic and proteomic analyses to detect tumor-derived signals in blood. However, the extremely low abundance of tumor secretome components remains a major limitation for reliable detection. Here, we introduce an AI-driven imaging framework that leverages chromatin-based cellular fingerprints as functional biosensors to detect cancer-associated blood-derived secretome signals. As a proof of concept, we demonstrate that human stromal fibroblasts exhibit a higher signal-to-noise ratio for detecting prostate cancer-derived secretomes isolated from whole blood compared with immune cells. Notably, aged fibroblasts display enhanced sensitivity relative to their young counterparts and T cells. To achieve a fully integrated sample-to-sensing workflow, we combined this imaging approach with a micropillar-guided secretome/plasma isolation platform coupled to an on-chip cell-culture chamber, enabling rapid exposure of sensor cells to tumor-derived secretome factors. Together, our results establish a proof-of-concept exploratory study demonstrating the feasibility of chromatin-based functional biosensing of cancer-associated blood-derived secretome signals.

Challenges in the precise diagnosis and treatment of nasopharyngeal carcinoma (NPC) remain, mainly due to the absence of a multi-omics-based molecular classification and effective targeted therapies. In this study, we performed proteomic and phosphoproteomic analyses of NPC and non-cancerous nasopharyngeal tissues to identify key dysregulated proteins and phosphorylation networks. Based on these profiles, we classified NPC into two distinct molecular subtypes: S1 and S2, which exhibit significant clinical heterogeneity. Notably, the S2 subtype displayed stronger immune suppressive characteristics. By leveraging proteomic data from both cancerous and non-cancerous tissues, as well as from the two molecular subtypes, we developed robust diagnostic and prognostic models. Through computational drug repurposing and experimental validation, we identified Panobinostat, a pan-histone deacetylase inhibitor, as a potent anti-tumor agent for NPC, demonstrating efficacy in both in vitro and in vivo models. Mechanistically, Panobinostat inhibits MYC expression, thereby suppressing the transcriptional activation of key components in the homologous recombination (HR) DNA repair pathway. This reduction in transcriptional activation impairs HR repair efficiency and leads to the accumulation of DNA double-strand breaks (DSBs). Furthermore, combination therapy with Panobinostat and radiotherapy produced a synergistic effect, significantly enhancing NPC suppression. Additionally, we predicted and validated potential drugs for targeting the S2 subtype of NPC. In conclusion, we identified molecular subtypes of NPC, constructed preliminary diagnostic and prognostic marker panels, and observed the therapeutic potential of Panobinostat, as a monotherapy and in combination with radiotherapy. These findings provide a solid foundation for precision diagnosis, prognostic stratification, and personalized treatment strategies for NPC.

Colorectal cancer liver metastases (CRLM) remain a leading cause of cancer-related mortality. Although hepatic resection is the only established curative option, recurrence exceeds 60%, underscoring substantial biologic heterogeneity. Liver transplantation (LT) has re-emerged for highly selected patients with unresectable CRLM, but optimal biologic selection criteria remain undefined. This study integrates genomic and clinical data to develop a biologically grounded framework for surgical and transplant decision-making. The Memorial Sloan Kettering 2017 metastatic colorectal cancer cohort was analyzed using cBioPortal-formatted clinical, genomic, and survival data. The study included patients with liver-only metastatic presentation. Genomic variables comprised KRAS, NRAS, BRAF, TP53, APC, PIK3CA, SMAD4 alterations, tumor mutational burden, microsatellite instability, and copy-number alteration burden. Survival was assessed using Kaplan-Meier and Cox models. A predefined molecular-risk classification stratified patients into low (RAS/RAF wild-type, SMAD4-intact), intermediate (isolated KRAS mutation), and high risk (NRAS, BRAF, or SMAD4 alterations). Machine-learning models predicted 24-month mortality. Among 503 patients, molecular-risk groups demonstrated distinct survival (5-year OS: 74.5%, 56.6%, and 49.8%; p<0.001). Intermediate- and high-risk groups were independently associated with worse survival (HR=1.71 and 2.60, respectively). Integrated clinicogenomic models modestly improved predictive performance (AUROC 0.697), with BRAF mutation, tumor sidedness, and age as key contributors. Inclusion of metastasectomy status increased AUROC but reflected post-treatment bias. Molecular-risk stratification and integrated modeling identify clinically meaningful prognostic groups in CRLM. These findings support incorporation of genomic profiling into precision surgical and transplant evaluation, while emphasizing the need for prospective validation.