Hypoxia strongly affects the growth, invasion, and therapeutic response of solid tumors, including head and neck squamous cell carcinoma (HNSCC). Despite intensive research, only a few substances have progressed to clinical trials as radiosensitizers. Therefore, new clinically relevant tumor models are needed to identify agents that overcome radiation resistance in hypoxic tumors. To study radiosensitivity of hypoxic and normoxic cells, we developed a two-color spheroid model using two HNSCC cell lines, SAS and FaDu, with GFP-labeled inner cell layers and mCherry-labeled outer layers. Optimizing the ratios of fluorescent cells enabled formation of hypoxic and normoxic zones, confirmed by pimonidazole, HIF1α, and CA IX staining. A newly established fluorescence clonogenic survival assay demonstrated transferability of results from 2D normoxia and hypoxia assays to these 3D model. The inner GFP-labeled cells showed significantly lower plating efficiency and increased radiation resistance compared to outer mCherry-labeled cells, similar to 2D hypoxic cells. To improve the model by reducing normoxia-induced HIF1α expression in the outer layer, we added physiological concentrations of ascorbic acid. Ascorbic acid also increased spheroid growth, clonogenic survival, and radioresistance under normoxia, while hypoxic responses remained unchanged. These two-layer spheroid model with distinct fluorescent labels provide a simple, robust assay to distinguish hypoxic from normoxic tumor areas in radiotherapy research. Addition of ascorbic acid further refines the physiological relevance of 3D tumor models and modulates radiosensitivity of the outer mCherry-labeled cell layer in both HNSCC models.

Carcinomas of the pancreas and bile duct remain highly lethal malignancies, with surgical resection representing the only potentially curative treatment. Despite improvements in perioperative mortality, postoperative complications remain frequent and negatively affect long-term outcomes. Recent evidence suggests that the pancreas and bile ducts harbor distinct microbial communities, challenging the traditional concept of sterility in these environments. However, their composition and clinical relevance remain incompletely understood. This study aimed to characterize microbiome profiles across different anatomical sites in patients undergoing pancreatic surgery, evaluate the impact of preoperative biliary stenting, and assess associations between prevalent bacterial species and postoperative outcomes. A total of 224 samples (bile, pancreatic fluid, duodenal tissue, tumor tissue, and healthy pancreatic tissue) from 58 patients with pancreatic cancer, bile duct cancer, chronic pancreatitis, or healthy pancreas were analyzed using 16S rRNA gene sequencing. Microbial diversity was assessed using the Shannon index for alpha diversity and nMDS with PERMANOVA for beta diversity. Distinct microbial profiles were identified across body sites, with significant beta-diversity differences between duodenal, bile, and pancreatic fluid samples and between duodenal and pancreatic fluid samples from the same patient. Preoperative biliary stenting significantly influenced microbial composition. Enterococcus faecalis was associated with a reduced risk of severe postoperative complications (Clavien-Dindo ≥ III). Overall, microbial composition varies across anatomical sites and disease entities, and specific bacteria may influence surgical outcomes, warranting further investigation in larger cohorts.

RNA binding proteins (RBPs) are key post-transcriptional regulators controlling every aspect of the RNA life cycle from synthesis to decay. We extracted RBP target genes from publicly available enhanced cross-linking and immuno-precipitation followed by sequencing (eCLIP-seq) data of 168 RBPs and assembled a gene set collection that can be used to examine gene lists for enriched RBP targets via functional enrichment analysis methods like over-representation analysis (ORA), gene set enrichment analysis (GSEA) or gene set variation analysis (GSVA).

ABSTRACT Background Codon usage bias, the non-random usage of synonymous codons in coding sequences, represents a fundamental feature of genomic organization that has been largely understudied in cancer biology. Pancreatic ductal adenocarcinoma (PDAC), the predominant subtype of pancreatic cancer, is characterized by aggressive disease progression and limited therapeutic options, necessitating novel approaches to understand its molecular pathogenesis. Leveraging publicly available single-cell RNA sequencing data, we performed comprehensive codon usage analyses across different cellular populations in PDAC. Results Employing a variety of computational codon usage indices uncovered the connections between cancer-specific cellular state features and codon usage signatures. Our findings reveal that malignant pancreatic cells express genes with significantly higher GC content, demonstrate preferential usage of optimal codons through increased frequency of preferred synonymous codons, and exhibit a marked preference for more cost-effective amino acids. Analysis of transcript-level bulk RNA-seq data from PDAC tumors revealed that these codon optimization patterns extend to alternative isoform usage, with highly expressed isoforms displaying increased codon optimality and enhanced mRNA stability. Conclusion These codon usage-dependent adaptations operating at both gene expression and transcript isoform levels may enable malignant cells to enhance gene expression rates, potentially leading to increased translational efficiency and protein production. These insights into the codon usage landscape of PDAC may provide potential biomarkers for disease monitoring and treatment response prediction.

Background Chronic pancreatitis (CP) is a risk factor for pancreatic cancer, with inherited cases conferring a markedly increased risk. The underlying mechanisms driving malignant transformation by CP remain poorly understood. Objective Combining a recently developed mouse model of CP carrying the human carboxypeptidase A1 ( CPA1 ) p.N256K mutation with the established Kras G12D pancreatic cancer model, we characterised mechanisms linking chronic inflammation to early pancreatic carcinogenesis. Design We crossed Cpa1 N256K mice (Cpa1) with Ptf1a Cre ;Kras LSL-G12D (KC). In Cre, Cpa1, KC and KC-Cpa1 mice, we performed phenotypical characterisation at five early time points and in an ageing cohort. Assessment of histology combined with both RNA-sequencing and single-cell RNA-sequencing was performed to analyse metaplasia, preneoplastic lesions and cellular heterogeneity. Results KC-Cpa1 pancreata displayed a stark increase in remodelling, fibrosis and formation of metaplastic lesions as compared with KC. Cpa1 N256K induced extensive plasticity in both the acinar and ductal compartment, including an early acinar-to-ductal metaplasia state in acinar cells characterised by an upregulation of endoplasmic reticulum stress markers and an inflammatory ductal phenotype (iDucts). We characterised the complex cell-cell communication networks underlying both pancreatic inflammation and early carcinogenesis, revealing disease-specific signalling between ductal cells, granulocytes and fibroblasts. Conclusions The humanised KC-Cpa1 mouse model reveals the interplay of inflammation in hereditary CP and carcinogenesis. Cpa1 N256K -induced plasticity in acinar and ductal cells, inflammation and cell-cell interaction networks cooperate with Kras G12D in early pancreatic carcinogenesis.

Oral squamous cell carcinoma (OSCC), a major subgroup of head and neck squamous cell carcinoma (HNSCC), is an aggressive disease that preferentially spreads to cervical lymph nodes. Positive lymph node status is an important predictor of survival in OSCC [1-3]. Hence, a better understanding of the molecular mechanisms underlying oral cancer metastasis and the identification of therapeutic vulnerabilities are needed to prevent and treat metastatic disease. We collected 87 primary tumors and 21 lymph node metastasis (LNM) from 72 OSCC patients to conduct comprehensive transcriptome-wide expression and correlation analyses (Figure 1A). First, we performed expression-based clustering with all primary tumors and observed the best subdivision with k = 3 using protein-coding and non-coding genes (Figure 1B). Of note, we observed transcriptional heterogeneity among multiregional tumor samples in about 30% of the cases, leading to the assignment of these patients and their respective tumors to different clusters. Intriguingly, Kaplan-Meier analysis of patients whose tumors were unambiguously assigned to only one cluster revealed that cluster 3 (C3) had the worst outcome, with a median survival of 15.6 months compared to 28.57 and 36.53 months for clusters 1 (C1) and 2 (C2), respectively (Figure 1C). Importantly, prognostic factors known to negatively affect survival, such as high T, N, and G status, were not enriched in C3 tumors (Supplementary Figure S1A-C). However, gene expression analysis identified 244 genes that were significantly changed in C3 compared to C1/2 tumors (Supplementary Figure S1D, Supplementary Table S1). Of note, cell cycle-related gene sets, including Early region 2 binding factor (E2F) and Myelocytomatosis oncogene (MYC) target genes, along with other oncogenic signaling pathways, showed a positive normalized enrichment score (NES), potentially explaining the poor outcomes of C3 tumors (Supplementary Figure S1E). Deep transcriptome and isoform analysis reveals survival and metastasis biomarkers and targeting options for oral squamous cell carcinoma. (A) Overview of the Oral Squamous Cell Carcinoma (OSCC) patient and tissue cohort. The scheme was created in BioRender. Gutschner, T. (2025) https://BioRender.com/q66a024. (B) Consensus matrix of the primary tumor samples obtained via consensus clustering. Consensus values indicate the frequency with which samples clustered together (0 - never to 1 - always) during distinct permutations of the algorithm. (C) Kaplan-Meier overall survival (OS) analysis of patients with tumors unambiguously assigned to one of three clusters. Patients with cluster 3 tumors tend to have worse OS. (D) Hazard ratio (HR) values plotted against False Discovery Rate (FDR)-adjusted log-rank test-derived P-values of all protein-coding and non-coding genes. Samples were divided by tertile separation, and the gene list was filtered for at least 10 samples per group (genes with FDR ≤ 0.05 are highlighted). (E) Kaplan-Meier analysis of OS in the OSCC cohort (n = 70 patients) stratified by SHMT2 mRNA abundance. Samples were divided by median separation. (F) Representative images of OSCC cell lines after transfection with two independent SHMT2-targeting small interfering RNAs (siSHMT2a/b). Two-dimensional growth was strongly impaired after depletion of SHMT2 in all three cell lines. The white scale bar indicates 400 µm. (G) Venn diagram of significantly deregulated genes in metastases and LNMpos- tumors. Genes upregulated in both datasets and targeted by an approved drug, as well as gene ontology (GO) biological processes significantly overrepresented in downregulated genes found in both datasets, are indicated. The scheme was created in BioRender. Gutschner, T. (2025) https://BioRender.com/k79r687. (H) Plot of genes with significant isoform switches (q ≤ 0.05) in metastases, as obtained by the R package IsoformAnalyzeR. Genes with the highest combined differences in isoform fraction values (DIFs) and lowest q-values are highlighted. (I) Three-dimensional protein structure of the canonical WNT5A protein according to AlphaFold. The turquoise area marks the WNT5A-203 isoform, which is decreased in lymph node metastases. Letter codes indicate primer binding sites for the cloned overexpression constructs used for functional assays. Representative Western blot (n = 3) shows the overexpression of different WNT5A protein isoforms in SAS cells. Ribosomal Protein L7 (RPL7) was used as a loading control. (J-K) Representative pictures depicting spheroid growth (J) and matrigel-based invasion (K) upon overexpression of the respective WNT5A isoforms in SAS cells (scale bars = 800 µm). Bar graphs show the mean and standard deviation (n = 5). One-way ANOVA (Holm-Sidak corrected) was performed for statistical testing, with *P ≤ 0.05. Abbreviations: C1/2/3, cluster 1/2/3; DIF, differences in isoform fraction; GABRG3, Gamma-Aminobutyric Acid Type A Receptor Subunit Gamma 3; GO, gene ontology; FDR, False Discovery Rate; HR, Hazard ratio; HTR6, 5-Hydroxytryptamine Receptor 6; LNMpos, lymph node metastasis-positive; MAOB, Monoamine Oxidase B; MYL6, Myosin Light Chain 6; OS, overall survival; OSCC, oral squamous cell carcinoma; RORC, RAR Related Orphan Receptor C; Rpl7, Ribosomal Protein L7; SHMT2, Serine Hydroxymethyltransferase 2; siSHMT2a/b, small interfering RNAs a/b targeting SHMT2; STAC3, SH3 And Cysteine Rich Domain 3; TUBB4A, Tubulin Beta 4A Class Iva; WNT5A, Wnt family member 5A; ZNF443, Zinc Finger Protein 443. Next, we performed a gene expression-based overall survival (OS) analysis and identified two significant genes, namely Zinc Finger Protein 443 (ZNF443) and Serine Hydroxymethyltransferase 2 (SHMT2) (Figure 1D, Supplementary Table S2). Specifically, ZNF443 expression was associated with a reduced risk (Hazard ratio [HR] = 0.238), whereas expression of SHMT2 (HR = 4.028) suggested a higher risk of mortality. Thus, we further tested their prognostic relevance for OS and recurrence/disease-free survival (RFS/DFS) in our patient cohort (Figure 1E, Supplementary Figure S2A-C) as well as in The Cancer Genome Atlas (TCGA) HNSCC dataset (Supplementary Figure S2D-E) [4]. These analyses indicated that SHMT2, but not ZNF443, might serve as an OSCC-specific biomarker for OS. In line with this, SHMT2 expression was higher in HNSCC tissues compared to normal tissues, as well as in T4 versus T1 tumors of the OSCC subtype, and its expression level increased with higher tumor grade (Supplementary Figure S3). Furthermore, univariate Cox regression analysis demonstrated a significant association of age, T-stage, N-stage, and SHMT2 expression (Relative Risk [RR] = 1.548, P = 0.049; 95% confidence interval [CI] = 1.000-2.396]) with OS in OSCC patients. Multivariate analysis further confirmed the association of SHMT2 (RR = 1.616, P = 0.041; 95% CI = 1.020-2.559]) (Supplementary Table S3). Intriguingly, downregulating SHMT2 in SAS, Cal33, and XF354 cells or blocking its activity using an inhibitor [5] reduced proliferation and viability while inducing apoptosis (Figure 1F, Supplementary Figure S4). These data confirmed previous studies and underscored the therapeutic potential of SHMT2 in OSCC [6, 7]. Next, we aimed to characterize the metastasis-associated transcriptome in our OSCC cohort. First, we compared the transcriptome of primary tumors and matched LNM from each patient. This analysis identified 1,710 deregulated protein-coding and 990 long non-coding RNA (lncRNA) genes (false discovery rate [FDR] ≤ 0.05; | log2(Fold change) | ≥ 1). Subsequent gene set enrichment analysis revealed 32 gene sets, including Kirsten Rat Sarcoma Viral Oncogene Homology (KRAS) signaling and epithelial-to-mesenchymal transition (EMT)-promoting gene sets, among others, which showed positive enrichment in metastasis (Supplementary Figure S5A-C, Supplementary Tables S4-S5). In order to narrow down the list of putative metastasis-associated genes, we compared gene expression patterns between primary tumors with (n = 43) and without (n = 24) LNM. This complementary approach uncovered 482 deregulated protein-coding and 190 lncRNA genes (Supplementary Figure S5D-E, Supplementary Tables S6-S7). Intriguingly, a total of 31 gene sets were significantly enriched in LNM-positive (LNMpos) tumors, but only EMT-promoting genes showed consistent positive enrichment in both differential gene expression analysis (Supplementary Figure S5F). To identify individual genes driving and maintaining metastases, we intersected the lists of differentially expressed genes from both analyses (Supplementary Table S8). This revealed a common set of 41 upregulated and 40 downregulated genes. Gene ontology analysis suggested that differentiation-associated processes were impaired in both LNM and LNM-positive tumors (Figure 1G). Importantly, a database search using the canSAR knowledgebase [8] identified four consistently upregulated genes that are targetable with approved clinical drugs. However, their cellular and molecular functions as well as their contribution to OSCC metastasis needs to be established using appropriate in vitro and in vivo models. Finally, we extended our gene-level expression analysis and characterized gene isoform usage in primary tumors and their matched metastasis. We identified hundreds of alternative transcription events that were either enriched or diminished in metastasis (Supplementary Figure S6A). At the individual gene level, this analysis yielded a list of 114 genes with significant isoform switches (Figure 1H, Supplementary Table S9). The most significant isoform switches were observed for Wnt Family Member 5A (WNT5A) and Myosin Light Chain 6 (MYL6), whereas SH3 And Cysteine Rich Domain 3 (STAC3), and RAR Related Orphan Receptor C (RORC) showed the highest combined differences in isoform fraction (dIF) values (Figure 1H). We decided to investigate the isoform switch in WNT5A in greater detail. In metastases, the canonical isoform (WNT5-201) was more abundant, while the fractions of WNT5A-202 (encoding the same WNT5A protein) and WNT5-203 (encoding a shorter protein variant) were significantly reduced (Supplementary Figure S6B). We generated overexpression constructs, transduced SAS cells with these variants, and successfully detected all WNT5A proteins at their expected size (Figure 1I). Intriguingly, overexpression of the canonical WNT5A protein only slightly enhanced spheroid growth in SAS cells but strongly increased their invasive capacity (Supplementary Figure S6C-D). Moreover, comparison of the different WNT5A isoforms revealed no significant differences in growth (Figure 1J). However, cells overexpressing the WNT5A-203 (BC construct) isoform exhibited markedly reduced invasive potential in Matrigel compared to cells expressing other WNT5A protein variants (Figure 1K). These findings suggest that inhibiting the canonical WNT5A isoform may represent a therapeutic strategy to prevent metastasis in OSCC, consistent with previous reports [9]. In summary, our study contributes to OSCC profiling and target identification efforts [4, 10] in a unique manner. Our carefully selected sample collection included LNM-negative and LNMpos primary tumors as well as their matched metastases. Furthermore, RNA isolated in this study was subjected to total RNA sequencing upon ribosomal RNA depletion, providing a more unbiased view of the primary and metastatic oral cancer transcriptome. This approach enabled the identification of coding and non-coding genes, as well as isoforms, associated with OSCC metastasis. However, additional studies are needed (see Supplementary Discussion) to confirm the described associations and validate the clinical relevance of individual candidates in vitro and in vivo. The study was conceptualized by Jonas Pyko, Markus Glaß, and Tony Gutschner with input from Monika Hämmerle and Stefan Hüttelmaier. Experiments were performed by Jonas Pyko, Julia Rosemann, Jana Macho, and Sarah Qasem. Patient tumor samples were collected by Matthias Kappler and Alexander W. Eckert. Pathological sample evaluation was performed by Monika Hämmerle. Computational analysis was performed by Markus Glaß and Stefan Hüttelmaier. Analysis of experimental data was done by Jonas Pyko, Markus Glaß, and Tony Gutschner. The manuscript was written by Jonas Pyko, Markus Glaß, and Tony Gutschner with input from all authors. Figures were prepared by Jonas Pyko, Markus Glaß, and Tony Gutschner. All authors have read and agreed to the final version of the manuscript. The authors would like to thank all members of the Gutschner, Hämmerle & Hüttelmaier labs for helpful discussions and continuous support. Moreover, we thank Dr. Nadine Bley and members of the Core Facility Imaging for assistance with IncuCyte measurements. Importantly, we would like to thank all patients participating in this study. We would like to note that some results presented in this study are in whole or part based upon data generated by The Cancer Genome Atlas Program (TCGA, https://www.cancer.gov/ccg/research/genome-sequencing/tcga). The cBioPortal (https://cbioportal.org) was used to download gene expression data and clinical information of the TCGA-HNSCC cohort. The authors declare that they have no competing interests. This study was supported by intramural funding from the Medical Faculty (Wilhelm-Roux program, FKZ32/19). Ethical registry 210/19.08.09/10 was obtained from the Ethics Committee of the Medical Faculty of the University Halle. All patients gave written informed consent (Department of Oral and Maxillofacial Plastic Surgery, University of Halle-Wittenberg, Germany). The data underlying this article are available in the manuscript and in its online additional material. The raw sequencing data have been deposited at NCBI GEO (GSE275870; https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?&acc=GSE275870). Additional materials generated during the current study are available from the corresponding author upon reasonable request. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.

Cell lines are indispensable models for analyzing molecular mechanisms underlying human diseases. However, incorrect annotation and cross-contamination can introduce severe bias in respective studies. Accordingly, various publishers request authentication of cell lines before publication. Short tandem repeat profiling is commonly used to verify cell line identity and purity but does not guarantee that published results are based on the samples tested by this method. In this study, we demonstrate that RNA-seq-derived sequence variation information is eligible for unambiguous cell line-specific clustering. Based on this finding, we propose methods for reliable cell line identification from RNA-seq data using supervised machine learning methods. In addition, we demonstrate the ability to detect cross-contamination of human cell lines. The presented methods are insensitive to different data pre-processing steps and quality measures. The proposed topFracCCLE algorithm for cell line identification and detection of cross-contamination is available as R-script at https://github.com/HuettelmaierLab/topFracCCLE.

Pancreatic ductal adenocarcinoma (PDAC) remains a highly lethal cancer with hereditary chronic pancreatitis (CP) conferring a significantly increased risk. Hereditary CP is caused by mutations in genes, such as PRSS1, SPINK1, CTRC, or CPA1 and presents with variable onset. However, the underlying mechanisms through which CP contributes to oncogenic transformation remain poorly understood. To elucidate this, we generated a genetically engineered mouse model carrying CP-associated human p.N256K carboxypeptidase A1 (CPA1) mutation and oncogenic Kras mutation resulting in Ptf1a Cre ;Kras LSL-G12D ;Cpa1 N256K (KC-Cpa1) mouse strain. Histological analyses of the KC-Cpa1 pancreas revealed accelerated development of early steps of PDAC including acinar-to-ductal metaplasia (ADM), pancreatic intraepithelial neoplasia (PanIN) lesions, and extensive fibrosis. Ex vivo 3D acinar cultures from 8-week-old KC-Cpa1 mice pancreata demonstrated enhanced ADM formation compared to Ptf1a Cre (Cre), Cpa1 N256K/N256K (Cpa1), and Ptf1a Cre ;Kras LSL-G12D (KC). These findings suggest that inflammation induced by the Cpa1 N256K mutation synergizes with oncogenic Kras mutation to promote the early initiation of PDAC. To explore the cellular heterogeneity and transcriptional program of metaplastic cells, we performed single cell RNA sequencing of pancreata of KC-Cpa1, KC, Cpa1, and Cre which revealed Cpa1 N256K -induced exocrine plasticity marked by an early ADM state and inflammatory phenotype of ductal cells (iDucts). Furthermore, single cell RNA sequencing also suggested disease-specific signaling between ductal cells, granulocytes, and fibroblasts. These results support the utility of KC-Cpa1 mouse model for studying early stages of CP-induced PDAC. Tanvi V. Inamdar, Ferdinand Krannich, Nico Hesselbarth, Atul Verma, Teresa Vauti, Ghanem El Kassem, Jasmine Hillmer, Michael Boettcher, Ivonne Regel, Heidi Griesmann, Irene Esposito, Markus Glaß, Monika Hämmerle, Patrick Michl, Helmut Laumen, Jonas Rosendahl. Hereditary chronic pancreatitis induced plasticity cooperates with mutant Kras in early pancreatic carcinogenesis [abstract]. In: Proceedings of the AACR Special Conference in Cancer Research: Advances in Pancreatic Cancer Research—Emerging Science Driving Transformative Solutions; Boston, MA; 2025 Sep 28-Oct 1; Boston, MA. Philadelphia (PA): AACR; Cancer Res 2025;85(18_Suppl_3):Abstract nr A026.

RNA binding proteins (RBPs) are key post-transcriptional regulators controlling every aspect of the RNA life cycle from synthesis to decay. We extracted RBP target genes from publicly available enhanced cross-linking and immunoprecipitation followed by sequencing (eCLIP-seq) data of 168 RBPs and assembled a gene set collection that can be used to examine gene lists for enriched RBP targets via functional enrichment analysis methods like over-representation analysis (ORA), gene set enrichment analysis (GSEA) or gene set variation analysis (GSVA). ### Competing Interest Statement The authors have declared no competing interest. Deutsche Forschungsgemeinschaft, 468534282, 449501615

Pancreatic ductal adenocarcinoma (PDAC) is characterized by aggressive growth and metastasis, partly driven by fibroblast-mediated stromal interactions. Using RNA sequencing of fibroblasts from early-stage KPC mouse models, we identified significant upregulation of genes involved in adipogenesis, fatty acid metabolism, and the ROS pathway. ANGPTL4, a key adipogenesis regulator, was highly expressed in fibroblasts and promoted pancreatic cancer cell proliferation and migration through paracrine signaling. Notably, cancer cell-driven paracrine signals appear to regulate ANGPTL4 expression in fibroblasts, suggesting that ANGPTL4 may act as a reciprocal factor in a feedback loop that enhances tumor progression. LAMA2, an extracellular matrix gene with reduced expression, suppressed pancreatic cancer cell migration, proliferation, and invasion. This study provides the temporal transcriptional analysis of fibroblast subtypes during early PDAC, highlighting the roles of metabolic reprogramming and ECM remodeling in shaping the tumor microenvironment and identifying potential therapeutic targets.

Deciphering gene function via context-aware approaches is limited by various means. Especially, static gene sets used in enrichment analyses and the lack of single-gene resolution restrain flexible association of genes with specific contexts. Here, we introduce CENTRA (Centrality-Based Exploration of Network Topologies from Regulatory Assemblies), a framework that models gene contextuality through topic-specific gene co-occurrence networks derived from curated gene sets and associated literature. Using latent dirichlet allocation on 12 045 abstracts linked to Molecular Signatures Database C2 gene sets, we uncover 27 biological topics and construct corresponding topic-specific networks reflecting distinct biological states, perturbation conditions, and disease-related regulatory programs. Graph-topological metrics, including centrality, local fractality, and perturbation sensitivity, were computed for each gene to capture structural relevance within these topic-specific networks. We show that topological profiles distinguish well-characterized regulators, identify emerging functional candidates, and reveal context-specific roles. Our framework prioritizes understudied genes by assessing the robustness of their topological signatures across topic-specific networks. To support exploration of these results, we developed a publicly accessible interactive browser, CENTRA, enabling dynamic navigation of networks and functional annotations. CENTRA provides an interpretable, scalable framework for investigating context-dependent gene function and hypothesis generation, offering a novel entry point beyond traditional enrichment approaches.

The stroma of healthy pancreases contains various non-hematopoietic, non-endothelial mesenchymal cells. It is altered by chronic inflammation which in turn is a major contributor to the development of pancreatic adenocarcinoma (PDAC). In PDAC, the stroma plays a decisive and well-investigated role for tumor progression and therapy response. This review addresses the central role of stromal cells in the early inflammation-driven development of PDAC. It focuses on major subpopulations of pancreatic mesenchymal cells, i.e., fibroblasts, pancreatic stellate cells, and multipotent stroma cells, particularly their activation and functional alterations upon chronic inflammation including the development of different types of carcinoma-associated fibroblasts. In the second part, the current knowledge on the impact of activated stroma cells on acinar-to-ductal metaplasia and the transition to pancreatic intraepithelial neoplasia is summarized. Finally, putative strategies to target stroma cells and their signaling in early pancreatic carcinogenesis are reflected. In summary, the current data show that the activation of pancreatic stroma cells and the resulting fibrotic changes has pro- and anti-carcinogenetic effects but, overall, creates a carcinogenesis-promoting microenvironment. However, this is a dynamic process and the therapeutic targeting of specific pathways and cells requires in-depth knowledge of the molecular interplay of various cell types.

The Hippo/YAP1 signaling pathway regulates normal development by controlling contact inhibition of growth. In cancer, YAP1 activation is often dysregulated, leading to excessive tumor growth and metastasis. SRC kinase can cross talk to Hippo signaling by disrupting adherens junctions, repressing the Hippo cascade, or activating YAP1 to promote proliferation. Here, we demonstrate that the IGF2 messenger RNA-binding protein 1 (IGF2BP1) impedes the repression of YAP1 by Hippo signaling in carcinomas. IGF2BP1 stabilizes the YAP1 messenger RNA (mRNA) and enhances YAP1 protein synthesis through an m6A-dependent interaction with the 3 ' untranslated region of the YAP1 mRNA, thereby increasing YAP1/TAZ-driven transcription to bypass contact inhibition of tumor cell growth. Inhibiting IGF2BP1-mRNA binding using BTYNB reduces YAP1 levels and transcriptional activity, leading to significant growth inhibition in carcinoma cells and ovarian cancer organoids. In contrast, SRC inhibition with Saracatinib fails to inhibit YAP1/TAZ-driven transcription and cell growth in general. This is particularly significant in de-differentiated, rather mesenchymal carcinoma-derived cells, which exhibit high IGF2BP1 and YAP1 expression, rendering them less reliant on SRC-directed growth stimulation. In such invasive carcinoma models, the combined inhibition of SRC, IGF2BP1, and YAP1/TAZ proved superior over monotherapies. These findings highlight the therapeutic potential of targeting IGF2BP1, a key regulator of oncogenic transcription networks.

High-grade serous ovarian cancer (HGSC) accounts for more than 70% of ovarian cancer-related deaths, yet therapeutic progress remains stagnant. Among the four molecular subtypes reported for HGSC, the C5 subtype is distinguished by high proliferation and immune evasion with an unfavorable MHC-I/PD-L1 ratio. However, the molecular drivers of this immune desert state remain largely undefined. Here, we identify RNA-binding proteins (RBPs) as key regulators of immune evasion in C5-HGSC through integrated single-cell and bulk RNA sequencing. We perform a targeted loss-of-function screen in C5-like cell models and find IGF2BP1 as a central mediator of immune evasion in vitro and in vivo. Mechanistically, IGF2BP1 abrogates interferon-gamma signaling by accelerating IRF1 protein degradation, thereby suppressing MHC-I presentation. We also discover that IGF2BP1 decouples PD-L1 expression from IRF1-dependent transcription and reshapes the immune receptor landscape to limit immune cell infiltration and T cell activation. Therapeutically, the small-molecule BTYNB effectively inhibits IGF2BP1 and synergizes with PD-1 blockade to overcome immune evasion in vivo. Multi-spectral imaging confirms these findings in human HGSC tissues and highlights the role of oncofetal RBPs as molecular drivers of the C5-HGSC subtype. This subtype-wide survey uncovers a previously unrecognized RBP-interferon regulatory axis and establishes RBP inhibition as a therapeutic strategy to enhance immune checkpoint therapy in immunologically cold ovarian tumors.

The regulator of the canonical Wnt pathway, leucine-rich repeat-containing G protein-coupled receptor 5 (LGR5), is expressed in the stem cell compartment of several tissues and overexpressed in different human carcinomas. The isoform of the stem cell marker LGR5, named LGR5Δ5 and first described by our group, is associated with prognosis and metastasis in oral squamous cell carcinoma (OSCC) and soft tissue sarcoma (STS). In a proof-of-principle analysis, the function of LGR5Δ5 was investigated in HEK293T cells, a model cell line of the Wnt pathway, compared to full-length LGR5 (FL) expression. The CRISPR/CAS knockout of LGR5 and LGR4 (thereby avoiding the side effects of LGR4) resulted in a loss of Wnt activity that cannot be restored by LGR5Δ5 but by LGR5FL rescue. The ability to migrate was not affected by LGR5Δ5, but was reduced by LGR5FL overexpression. The CRISPR/CAS of LGR4 and 5 induced radiosensitization, which was enhanced by the overexpression of LGR5FL or LGR5Δ5. RNA sequencing analysis revealed a significant increase in the ligand R-spondin 1 (RSPO1) level by LGR5Δ5. Furthermore, LGR5Δ5 appears to be involved in the regulation of genes related to the cytoskeleton, extracellular matrix stiffness, and angiogenesis, while LGR5FL is associated with the regulation of collagens and histone proteins.

The RAVER1 protein serves as a co-factor in guiding the polypyrimidine tract-binding protein (PTBP)-dependent control of alternative splicing (AS). Whether RAVER1 solely acts in concert with PTBPs and how it affects cancer cell fate remained elusive. Here, we provide the first comprehensive investigation of RAVER1-controlled AS in cancer cell models. This reveals a pro-oncogenic role of RAVER1 in modulating tumor growth and epithelial-mesenchymal-transition (EMT). Splicing analyses and protein-association studies indicate that RAVER1 guides AS in association with other splicing regulators, including PTBPs and SRSFs. In cancer cells, one major function of RAVER1 is the stimulation of proliferation and restriction of apoptosis. This involves the modulation of AS events within the miR/RISC pathway. Disturbance of RAVER1 impairs miR/RISC activity resulting in severely deregulated gene expression, which promotes lethal TGFB-driven EMT. Among others, RAVER1-modulated splicing events affect the insertion of protein interaction modules in factors guiding miR/RISC-dependent gene silencing. Most prominently, in all three human TNRC6 proteins, RAVER1 controls AS of GW-enriched motifs, which are essential for AGO2-binding and the formation of active miR/RISC complexes. We propose, that RAVER1 is a key modulator of AS events in the miR/RISC pathway ensuring proper abundance and composition of miR/RISC effectors. This ensures balanced expression of TGFB signaling effectors and limits TGFB induced lethal EMT.

IntroductionObesity is associated with chronic low-grade inflammation of adipose tissue (AT) and an increase of AT macrophages (ATMs) that is linked to the onset of type 2 diabetes. We have recently shown that neutralization of interleukin (IL)-6 in obese AT organ cultures inhibits proliferation of ATMs, which occurs preferentially in alternatively activated macrophage phenotype.MethodsIn this study, we investigated AT biology and the metabolic phenotype of mice with myeloid cell-specific IL-6Rα deficiency (Il6raΔmyel) after normal chow and 20 weeks of high-fat diet focusing on AT inflammation, ATM polarization and proliferation. Using organotypical AT culture and bone marrow derived macrophages (BMDMs) of IL-4Rα knockout mice (Il4ra-/-) we studied IL-6 signaling.ResultsObese Il6raΔmyel mice exhibited no differences in insulin sensitivity or histological markers of AT inflammation. Notably, we found a reduction of ATMs expressing the mannose receptor 1 (CD206), as well as a decrease of the proliferation marker Ki67 in ATMs of Il6raΔmyel mice. Importantly, organotypical AT culture and BMDM data of Il4ra-/- mice revealed that IL-6 mediates a shift towards the M2 phenotype independent from the IL-6/IL-4Rα axis.DiscussionOur results demonstrate IL-4Rα-independent anti-inflammatory effects of IL-6 on macrophages and the ability of IL-6 to maintain proliferation rates in obese AT.

Plakophilin 4 (PKP4) is a component of cell–cell junctions that regulates intercellular adhesion and Rho-signaling during cytokinesis with an unknown function during epidermal differentiation. Here we show that keratinocytes lacking PKP4 fail to develop a cortical actin ring, preventing adherens junction maturation and generation of tissue tension. Instead, PKP4-depleted cells display increased stress fibers. PKP4-dependent RhoA localization at AJs was required to activate a RhoA-ROCK2-MLCK-MLC2 axis and organize actin into a cortical ring. AJ-associated PKP4 provided a scaffold for the Rho activator ARHGEF2 and the RhoA effectors MLCK and MLC2, facilitating the spatio-temporal activation of RhoA signaling at cell junctions to allow cortical ring formation and actomyosin contraction. In contrast, association of PKP4 with the Rho suppressor ARHGAP23 reduced ARHGAP23 binding to RhoA which prevented RhoA activation in the cytoplasm and stress fiber formation. These data identify PKP4 as an AJ component that transduces mechanical signals into cytoskeletal organization.

The oncofetal RNA-binding protein IGF2BP1 has been reported to be a driver of tumor progression in a multitude of cancer entities. Its main function is the stabilization of target transcripts by shielding these from miRNA-mediated degradation. However, there is growing evidence that several virus species recruit IGF2BP1 to promote their propagation. In particular, tumor-promoting viruses, such as hepatitis B/C and human papillomaviruses, benefit from IGF2BP1. Moreover, recent evidence suggests that non-oncogenic viruses, such as SARS-CoV-2, also take advantage of IGF2BP1. The only virus inhibited by IGF2BP1 reported to date is HIV-1. This review summarizes the current knowledge about the interactions between IGF2BP1 and different virus species. It further recapitulates several findings by presenting analyses from publicly available high-throughput datasets.