Abstract Cyanobacteria are important primary producers and are used in biotechnology as microbial cell factories due to their ability to use solar light for oxygenic photosynthesis. Synechocystis sp. PCC 6803 is a popular model cyanobacterium, yet there are ambiguities in the precise coding regions of many genes, and numerous genes encoding small proteins have remained undetected. Here we present the results of a ribosome profiling (Ribo-seq) analysis involving inhibitors that stall ribosomes at translation initiation and termination sites (TIS- and TTS-Ribo-seq), combined with a proteogenomic reevaluation and reannotation of its entire genome. We report evidence for the translation of 3,050 annotated genes based on proteogenomics (83%), of 3,492 based on Ribo-seq (95.2%), and of 3,009 supported by both methods (82%). The data suggested both novel protein-coding genes and corrections for annotated ones. We validated 15 novel small proteins translated from antisense RNAs, from intergenic and intragenic regions and identified 69 novel, mostly small proteins based on proteogenomics. With slr0489, slr1079 and slr1082 we identified three genes with ~300 nt long intragenic out-of-frame coding regions and show that both the internal and host reading frames are translated. The resulting proteins interact with each other, resembling certain defense or toxin-antitoxin systems. Our data illustrate the enormous value of consolidating genome annotations in the context of integrated experimental data and suggest that genome annotations in general need to be extended and revised. All of our data can be accessed via an intuitive and interactive genome browser platform at https://www.bioinf.uni-freiburg.de/~ribobase/.

Chimeric antigen receptor (CAR)-T cell therapy has transformed the management of hematologic malignancies but faces obstacles, including severe treatment-related toxicities, highly suppressive tumor microenvironment (TME), inadequate long-term persistence, and poor trafficking/infiltration into solid tumor. This review summarizes recent genetic engineering strategies designed to overcome these barriers and to improve the safety, durability, and spatial effectiveness of CAR-T cell therapy. To mitigate cytokine release syndrome and neurotoxicity, approaches such as affinity-tuned and humanized scFvs, hinge/TM optimization, ITAM calibration have been developed alongside programmable “switch-off” and “switch-on” systems incorporating suicide genes, antibody-bridging switches, and optogenetic or hypoxia-gated circuits. TME remodelling strategies leverage nanomaterials for localized cytokine delivery, cell-surface “backpack” systems, and oncolytic viruses engineered to release cytokines or checkpoint-blocking biologics. Enhancing durability and resistance to exhaustion increasingly relies on precision genome engineering, including CRISPR-based editing and multiplexed shRNA platforms targeting inhibitory receptors and exhaustion-driving transcriptional programs. Finally, chemokine-receptor engineering and local biomaterial-based delivery systems are discussed as routes to improve CAR-T trafficking and intratumoral persistence. We also highlight the remaining translational challenges including checkpoint redundancy, in vivo payload dilution, vector capacity limits, and the safety of multiplex genome editing. Collectively, these interdisciplinary innovations point towards integrated, patient-tailored CAR-T platforms that combine safety control, metabolic and transcriptional resilience, and improved TME navigation to enable broader clinical application.

Epilepsy affects more than 50 million individuals worldwide, and approximately one-third of patients remain refractory to existing antiseizure medications. Advances in gene therapy and genome editing have opened new possibilities for disease-modifying interventions that directly target the molecular and circuit-level mechanisms underlying epileptogenesis. Recent progress in central nervous system tropic viral vectors, non-viral delivery systems, and programmable genome-editing technologies has enabled precise manipulation of neuronal and glial function in preclinical epilepsy models. Strategies range from restoration of haploinsufficient genes implicated in monogenic epilepsies, such as SCN1A in Dravet syndrome, to modulation of neuronal excitability through engineered ion channels, neuropeptides, and astrocyte-based approaches. In parallel, CRISPR-derived platforms, including transcriptional activation and repression systems, base editing, and prime editing, offer new avenues for regulating gene expression in post-mitotic neurons without introducing double-strand DNA breaks. Despite these advances, significant translational challenges remain, including efficient and cell-type-specific delivery, long-term safety, and the risk of network-level side effects in the epileptic brain. This review critically examines recent gene therapy and genome-editing approaches for epilepsy, highlights key technological and biological barriers to clinical translation, and discusses emerging strategies that may enable durable and targeted treatments for drug-resistant epilepsies.

Abstract Tagging a gene endogenously can identify when the gene is expressed and where the protein is localized. CRISPR is the primary tool for generating tags of endogenous genes, but it is error-prone and requires unique reagents for each gene and tag. Recombinases can insert DNA in an error-free and modular manner. Here, we tested eight recombinases for germline function in the nematode C. elegans, and introduce PhIT, a recombinase-based method for protein tagging. First, a short 39bp PhiC31 attB landing pad is inserted into the locus by CRISPR. This strain is a resource which can be used to insert a variety of modular tags. Second, tags are inserted by the integrase PhiC31, and in tandem, extraneous backbone sequences are removed by a tyrosine recombinase. Current modular tags include seven different fluorescent proteins, FLP-regulated cell-specific expression constructs, and degron tags. Importantly, tags can be inserted by genetic crosses instead of by microinjection.

Abstract Bacillus subtilis is a pivotal model organism in both industrial biotechnology and scientific research, where the efficiency of its genetic engineering is very important. However, achieving highly efficient gene insertion in this bacterium remains a significant technical challenge. To address this, we aimed to develop a novel gene insertion tool in B. subtilis . Building upon the Vibrio cholerae -derived Vch CAST system, we systematically optimized and successfully established a high-performance VchCAST system. The core components of this system include the TniQ-Cas678 complex, a guide RNA for precise targeting, and the TnsABC transposase complex responsible for DNA integration. Under antibiotic selection, screening and employing a strong promoter to drive crRNA expression increased the single-locus transposition efficiency to 41%. Subsequent genomic integration of the transposase operational unit further enhanced the efficiency to 80%. Moreover, we demonstrated that overexpressing the auxiliary factor BmrR enables simultaneous integration at two distinct genomic loci. Through protein engineering of the key transposase TnsB, we obtained optimized variants V178F and V178L with significantly enhanced activity, which improved the overall transposition efficiency by 232.6% and 178.07%, respectively. We then conducted transposition validation with the optimized system, achieving a site-specific gene insertion efficiency of approximately 95.25%. In conclusion, this study not only provides a robust gene insertion platform for B. subtilis microbial cell factory engineering, but also stands as a valuable reference for the construction of gene insertion tool in other microbial.

Abstract CRISPR-Cas12a (Cpf1) offers distinct advantages for genome editing due to its flexible, T-rich PAM recognition. However, variable cleavage efficiency—modulated by sequence context and epigenetic features—remains a challenge, with existing predictors limited in accuracy and interpretability. Here, we present DeepCas12a, a hybrid deep learning framework integrating Convolutional Neural Networks (CNNs) and a Vision Transformer (ViT) encoder to capture both local sequence motifs and long-range dependencies. The model fuses DNA sequence data with epigenetic profiles (DNA methylation and chromatin accessibility) in an end-to-end architecture. Benchmarked on an independent test set, DeepCas12a outperformed state-of-the-art predictors, achieving an Average Precision of 0.783, an AUC of 0.868, and a Spearman correlation of 0.630. Furthermore, interpretability analysis via saliency maps confirms the model captures biologically relevant features, including PAM specificity and seed region sensitivity, facilitating rational guide RNA design.

Abstract The U6 promoter plays a pivotal role in the CRISPR/Cas9 system by driving the transcription of single guide RNA (sgRNA), which directs Cas9 to achieve precise genome editing. Endogenous U6 promoters typically exhibit superior transcriptional activation efficiency compared to exogenous counterparts, thereby enhancing the efficacy of genome editing. However, the endogenous U6 promoter in kenaf ( Hibiscus cannabinus L.) remains uncharacterized. In this study, we conducted a homologous search of the kenaf genome using the Arabidopsis U6 (AtU6-26) RNA sequence as a reference, identifying two candidate promoters, HcU6-1 and HcU6-14. Promoter fragments were amplified from the genomic DNA of kenaf cultivar 'Fuhong 952' and subsequently cloned into a GUS fusion expression vector. Histochemical staining revealed transcriptional activity for both promoters, with HcU6-14 demonstrating significantly stronger activity. To evaluate editing efficiency, we constructed a CRISPR/Cas9 vector containing HcALS sgRNA, driven by either the kenaf U6-14P promoter or the cotton U6-9P (GbU6-9P) promoter. Kenaf hairy roots were regenerated via Agrobacterium rhizogenes K599-mediated transformation. Sequencing analysis of ALS gene fragments from these hairy roots confirmed successful targeted editing when using the kenaf U6-14P promoter, whereas no base mutations were detected with the cotton U6 promoter. These findings highlight the superior editing efficiency of the kenaf U6 promoter and provide a critical foundation for advancing functional genomics research in kenaf.

Hepatocellular carcinoma (HCC) remains a leading cause of cancer-related death, yet adequate in vitro models mimicking the tumor immune microenvironment (TIME) are rare. Specifically, the role of the transcription factor nuclear factor erythroid 2-related factor 2 (Nrf2) in modulating interactions between tumor cells and tumor-associated macrophages (TAMs) is not fully understood. We established a 3D multicellular tumor spheroid (MCT) model using murine N-HCC25 cells with CRISPR/Cas9-mediated knockouts of Nrf2 and its negative regulator Kelch-like ECH-associated protein 1 (Keap1), the latter mimicking constitutive activation. N-HCC25 cells were co-cultured with bone marrow-derived macrophages (BMDMs) isolated from wild-type and Nrf2-knockout C57BL/6J mice. We compared co-culture setups (conditioned media, transwell systems, direct contact) using RT-qPCR, flow cytometry, and invasion assays. 3D spheroid systems better preserved stemness than 2D cultures and revealed functional Nrf2-dependent effects such as increased Vegf-α secretion in Keap1-deficient spheroids. Among the different co-cultivation models, most profound effects were observed in the MCT model. Macrophages successfully integrated into the spheroids and triggered invasive outgrowth, whereas MCTs containing Nrf2-deficient macrophages displayed markedly reduced invasion and lower programmed cell death ligand-1 expression. These findings demonstrate that Nrf2 signaling in macrophages fosters an immunosuppressive and pro-invasive microenvironment. The established MCT model provides a suitable platform to further unravel Nrf2-dependent mechanisms in the HCC TIME.

Abstract Apoptosis is a highly conserved form of programmed cell death controlled by a core molecular pathway that was first defined in Caenorhabditis elegans and is conserved in mammals. This pathway is composed of egl-1/ BH3-only, ced-9 /Bcl-2, ced-4 /Apaf-1, and ced-3/ Caspase. Despite being discovered more than 20 years ago, tissue-specific apoptosis induction as well as endogenous expression pattern and dynamic subcellular localization of apoptosis proteins remain incompletely defined. Here, we generated a complete set of CRISPR/Cas9-engineered transcriptional and translational reporters for all four apoptosis genes and systematically analyzed their expression and subcellular localization in the C. elegans germline and embryo. We show that somatic apoptosis is driven by precise, lineage-specific activation of egl-1 , whereas ced-9 , ced-4 , and ced-3 are ubiquitously expressed. In contrast, DNA-damage triggers a robust CEP-1/p53-dependent-induction of egl-1 throughout the germline, yet apoptosis occurs only in late pachytene cells. We also identify intron1 of egl-1 as essential for CEP-1–dependent transcriptional activation. Analysis of brc-1 and syp-2 mutants demonstrates that distinct meiotic surveillance pathways converge on egl-1 induction. Analysis of the subcellular localization of the downstream regulators CED-9, CED-4, and CED-3 reveals dynamic, tissue-specific localizations that refine the classical apoptosis model. CED-4 transitions from a perinuclear distribution in the germline and early embryos to a predominantly mitochondrial localization later in embryogenesis, while CED-3 changes its subcellular localization depending on developmental stage and apoptotic status. CED-9 localizes to distinct mitochondrial foci in both embryo and germline. Together, these reporters reveal that C. elegans apoptosis is governed by two mechanistically distinct programs: (1) lineage-specific egl-1 activation in embryos and (2) checkpoint-mediated activation of egl-1 in the germline, where additional, yet unidentified pathways restrict apoptotic execution. These reporters also provide a comprehensive toolbox for dissecting apoptotic and non-apoptotic functions of the conserved apoptotic machinery in vivo .

The integration of 3D bioprinting technology and CRISPR-Cas9 genome editing has become a game-changing method for creating complex organotypic cancer models. This integrated platform overcomes the drawbacks of traditional 2D culture systems by enabling precise genetic modifications within physiologically relevant, biomimetic tumor microenvironments. Researchers can more precisely recreate tumor progression, oncogenic mutations, cellular heterogeneity, and drug resistance mechanisms by utilizing the structural complexity provided by 3D bioprinting and the specificity of CRISPR-Cas9-mediated gene editing. CRISPR-Cas9 enables specific gene modifications, including oncogene knockout (e.g., MYC, KRAS) or immune checkpoint genes (e.g., PD-1, PD-L1), in 3D-bioprinted structures made from tumorigenic or patient-specific cell populations. It has been demonstrated that these modified models maintain important histopathological and molecular characteristics of original tumors, allowing for accurate high-throughput screening of immunotherapeutics and anticancer drugs. Significantly speeding up the modeling of tumorigenesis, studies using prostate cancer organoids showed gene correction efficiencies ranging from 50 to 90 %. Additionally, in 3D cultures, combinatorial CRISPR-Cas9 editing has demonstrated synergistic drug responses in models of lung and breast cancer, underscoring the platform's potential for discovering new therapeutic targets. Biomaterial-based vectors, like hydrogels and nanocarriers, are being improved to reduce off-target effects and increase intracellular uptake to increase the accuracy and safety of CRISPR delivery. However, issues with scalability, reproducibility, and standardization still exist, requiring ongoing interdisciplinary cooperation to improve downstream validation procedures, gene-editing tactics, and bioink formulations. The potential of CRISPR-Cas9-integrated 3D bioprinting as a state-of-the-art technique for drug discovery and cancer modeling is highlighted in this review. It emphasizes how the platform can speed up translational research in oncology, lessen dependency on animal models, and customize treatment plans. The goal of this review is to present a thorough summary of current developments, technical difficulties, and potential paths forward in this quickly developing field.

Abstract Background To investigate the distribution of CRISPR-Cas systems in Escherichia coli ( E. coli ) isolates and evaluate their associations with multilocus sequence types (MLST), antimicrobial resistance genes (ARGs), and plasmid features. Results ST1193 (16/65, 24.6%) and ST95 (11/65, 16.9%) were the predominant lineages. ST1193 showed a higher resistance gene burden than ST95, and bla TEM−1 was detected in 87.5% of ST1193 isolates. CRISPR-Cas systems were detected in 22 isolates (33.8%), including 11 with type I-F (50.0%), 10 with type I-E (45.5%), and one with both types. Spacer sequences were primarily directed against plasmid DNA. Plasmid replicons were frequently detected, and plasmid burden varied across lineages. All the ST1193 isolates lacked detectable CRISPR-Cas systems, whereas 90.9% (10/11) of ST95 isolates harbored type I-F systems. Conclusions CRISPR-Cas carriage was strongly lineage-dependent and showed an inverse association with predicted resistance gene burden in this cohort; this pattern should be interpreted as a lineage-structured correlation rather than mechanistic evidence of CRISPR-mediated restriction of ARG acquisition.

Abstract A growing body of evidence supports that targeting the tumour extracellular matrix (ECM) in solid cancers holds great promises to reactivate T cell migration in immune-excluded patients. By matrisome profiling of triple-negative breast cancer (TNBC) patients, we identified two core ECM proteins enriched in fibrotic immune-excluded tumours and associated with reduced CD8+ T cell stromal infiltration, versican (VCAN) and fibronectin (FN1). Both cancer-associated fibroblasts and aggressive cancer cell lines were found to deposit these two proteins in vitro, conferring resistance to T-cell-mediating cytotoxicity. Characterisation of in vivo murine breast cancer models 4T1 and EMT6 revealed significant differences in tumour ECM deposition and immune cell composition. Accordingly, targeting Fn1 and Vcan in both models induced opposite effects on tumour growth. While it appeared unfavourable in inflamed non fibrotic tumours, deletion of Fn1 in cancer cells was beneficial in immune-excluded tumours by promoting TCF7+ T cells and restoring anti-PD1 response.

Abstract Methods Based on two corolla tube types(L-T,C-T), 256 F1 individuals from a complete diallel cross involving, 5000 F2 individuals and 100 S6 (selfed for six generations) populations. Phenotypic traits, cytology, qRT-PCR, and CRISPR/Cas9 assays were conducted to examine genetic differential expression between the two corolla types. Results The order of dominance in hybrids was labiate > campanulate; For shared typical traits, the dominance order was angled > pubescent > veined. Dorsal and ventral cells of the L-T differed significantly, exhibiting thin walls, fewer stomata, and more trichomes; Whereas the C-T displayed uniformly arranged cells, thicker walls, and more stomata. Gene analysis revealed that 20 genes from five classes (A, B, C, E, and AGL6 subfamily) of the MIKCC-type MADS-box family, 30–32 genes from the TCP family, 8 genes from the YABBY family, and 12 genes from the WOX family transcription factors were involved in regulating typical corolla tube traits. Continuous selfing for six generations did not cause significant changes in the related regulatory genes, with the L-T exhibiting higher stability than the C-T. Both corolla tube types underwent six distinct developmental stages. The L-T was primarily activated by CYC/DICH genes, while the C-T was activated by RAD genes, leading to global expression of MADS-box identity transcription factors. The regulatory patterns, genes involved, and formation mechanisms differed significantly Conclusion The corolla tube of Sinningia speciosa exhibits high genetic stability, and the formation of its typical traits demonstrates significant representativeness and classification characteristics.

Abstract Transparent model organisms are invaluable for live imaging, yet generating them remains challenging. Here, we present a robust strategy to produce translucent Xenopus laevis , enabling non-invasive, deep-tissue imaging in intact organisms. Using CRISPR/Cas9 technology, we generated quadruple knockouts of slc2a7 , hps6 , and the two tyrosinase homeologs. Instead of in vitro -transcribed single guide RNAs, we employed chemically modified, commercially synthesized two-part guide RNAs, which enabled efficient multiplex genome editing. We produced a high proportion of translucent frogs directly in F0 founder tadpoles, eliminating the need for multi-generational breeding. We validated in vivo live imaging using two transgenic reporter lines with GFP expression in the eye, brain, and heart. Loss of both eumelanin and iridescent pigments in tyr ; hps6 knockouts markedly improved optical clarity and fluorescence visibility. Overall, this multiplexed genome-editing strategy enables the rapid generation of translucent transgenic X. laevis suitable for live imaging, while also providing a simple and efficient protocol for simultaneous multi-gene targeting in X. laevis .

Abstract Colorectal cancer (CRC) is the third most spread cancer globally, commonly arising from activation of the Wnt/β-catenin signaling pathway and microsatellite instability (MSI). Long non-coding RNAs (lncRNAs) are transcripts longer than 200 nucleotides and serve as crucial regulators in CRC progression. Depending on their subcellular localization, lncRNAs can be nuclear (e.g., MALAT1), cytoplasmic (e.g., H19), or exosomal, and are genomically classified as sense, antisense, divergent, intergenic, and intronic. They modify gene expression through cis- and trans-regulation functions, regulate mRNA stability through RNA-binding proteins, and function as competitive endogenous RNAs (ceRNAs). LncRNAs control the function of proteins using scaffolding or decoy mechanisms. While exosomal RPPH1 and CAF-derived H19 enhance metastasis by activating β-catenin signaling, oncogenic lncRNAs like HOTAIR and CCAT1-L promote the growth of colorectal cancer. New RNA-based therapies, including RNA interference (RNAi), antisense oligonucleotides (ASOs), and CRISPR technologies (CRISPRi, CRISPR-Cas13, and CRISPR/Cas9), show promising methods for downregulating oncogenic miRNAs (miR-21), restoring tumor-suppressive miRNAs (miR-145), and suppressing carcinogenic lncRNAs like EPIC1, CCAT1, and CCAT2.

Abstract The efficiency of CRISPR interference (CRISPRi) depends on functional dCas9 activity, yet practical and reproducible validation of dCas9-expressing cell lines remains limited. Here, we describe a simple and reproducible assay to assess dCas9 functionality using a single sgRNA targeting the ubiquitously expressed surface protein CD81. We evaluated this approach in multiple hematological and solid tumor cell lines expressing the dCas9-KRAB-MeCP2 repressor complex. In all tested models, CD81 targeting resulted in a consistent reduction of surface protein levels, quantified by flow cytometry. This assay provides a rapid and quantitative functional readout of dCas9 activity without the need for reporter constructs or transcriptional assays. The CD81-targeting sgRNA and validated cell lines are made available to support reproducibility and technical standardization in CRISPRi experiments. This strategy can be readily implemented in any laboratory using CRISPRi-based approaches.

The Solanaceae family includes some of the most economically and agronomically im-portant crops, such as tomato, potato, pepper and eggplant. Recently, CRISPR/Cas-based genome editing has emerged as a powerful tool for functional genomics and crop improvement, enabling precise and efficient genetic modifications. This review provides an overview of CRISPR/Cas-mediated genome editing technologies and their applications in the major cultivated Solanaceae crops. The use of systems for targeted gene knockout and knock-in approaches is described, together with advances in precision editing strategies such as base editing and prime editing, which allow precise nucleotide substitutions and small sequence changes. The expanding CRISPR toolbox is further explored through alternative Cas proteins, such as Cas12a and Cas13 with distinct targeting features and potential applications. Emerging delivery strategies, including ribonucleoprotein-mediated editing in protoplasts, virus-induced gene editing (VIGE), and de novo induction of meristems, represent promising approaches to generate transgene-free edited plants. In addition, the current status of field trials involving genome-edited Solanaceae crops in Europe is outlined, considering the regulatory landscape and legislative requirements for their release in the environment. Despite regulatory constraints, some ge-nomeedited crops have reached the market, highlighting their potential to contribute to sustainable agriculture and crop improvement.

Abstract Phages have evolved various anti-CRISPR(Acr) proteins evade hosts’ immunity by direct CRISPR interference. However, whether other counter-CRISPR mechanisms exist remains unexplored. Here, we report a phage-encoded two-protein system, Healer, which neutralizes CRISPR immunity via a post-cleavage DNA repair mechanism. This phage replication-associated system comprises two proteins: Gp63 (a DUF669 domain-containing protein), and Gp64 (a AAA domain). Mechanistic investigations elucidate that Gp63 acts as a rapid-response effector of CRISPR-induced DNA break, following the binding of ssDNA allows the Gp64 to mediate homologous recombination, repairing CRISPR-induced phage DNA break and enabling phage survival. Notably, co-expression Healer system with CRISPR-Cas9/Cas12 in E. coli, P. aeruginosa, and A. baumannii, demonstrated higher phage genome-editing efficiency. Conclusively, our findings represent a vital anti-CRISPR complementary strategy, providing a promising tool for genome manipulation.

Abstract Background. Bladder cancer (BC) can be characterized clinically as either non-muscle-invasive (NMIBC) or muscle-invasive (MIBC). While NMIBC generally has a favorable prognosis, MIBC is characterized by high morbidity and mortality. Understanding the molecular determinants of tumor invasion is critical, yet research is hampered by the limitations of current experimental models. Standard assays like the Boyden chamber lack physiological complexity, while porcine bladder models suffer from tissue contamination and genetic variability. There is an urgent need for reliable models that mimic the intact tissue architecture. Methods. We established a unique Ex vivo Tissue Model (EXTIM) to evaluate the invasive capacity of BC cells within a largely intact tissue context, using freshly prepared bladders from mice. The invasiveness of human BC cells (RT4, T24, UMUC3) and the immortal urothelial cell strain (Y235T) was comparably evaluated using EXTIM, the Boyden chamber and porcine models. Gene knockdown or ectopic expression of GJB3 or ORP3 indicated suitability of EXTIM to investigate the impact of specific factors on tumor cell invasion. To identify novel genetic regulators of cell invasion, we combined EXTIM with a genome-wide CRISPR-Cas9 knockout screen. Additionally, we utilized the EXTIM to perform a pharmacological screen of a small molecule library comprising 90 substances to identify compounds capable of suppressing BC cell dissemination. Results. Importantly, by combining EXTIM with genome-wide CRISPR-Cas9 screening, we identified several candidate genes involved in BC progression. Notably, discoidin domain receptor tyrosine kinase 1 (DDR1) was identified as a functional inhibitor of tumor cell invasion. Furthermore, the small molecule screen revealed that PD-156707, a selective antagonist of the endothelin receptor A (ETA), significantly suppresses cancer cell invasion within the EXTIM environment. Conclusions. EXTIM serves as a robust and physiologically relevant tool for assessing tumor cell invasion and migration under ex vivo conditions. EXTIM can be used to identify factors involved in the progression of invasive BC by high-troughput genetic screenings in an ex post organ culture system, by culturing cells after transmigration through the bladder tissue. Moreover, the impact of specific genetic factors in the process of tumor cell dissemination can be assessed by placing bladders from genetically modified mice into the EXTIM.

Abstract The critical need for highly sensitive, rapid, low-cost and equipment-free pathogens diagnostics is acutely felt in resource-limited settings, such as in low-income region and at home, where existing methods force a compromise among these key metrics. We break this trade-off with a nucleic acid chip that leverages a novel “dam-break drainage” readout model and a CRISPR-Cas13a-mediated interfacial wettability switching sensing mechanism. This design unlocks both single-plex and duplex visual detection, delivering attomolar sensitivity (10 and 100 aM), rapid results (2 and 5 min) and low cost (0.20$ and 0.30$ per test), respectively. In a 40 clinical-samples validation for SARS-CoV-2, Influenza A virus and Influenza B virus, our platform demonstrated robust clinical concordance and successfully detected samples with ambiguous RT-PCR results (Ct = 35). This work provides a powerful and accessible solution for next-generation point-of-care molecular testing.

Abstract Tumor-associated macrophages (TAMs) are central drivers of resistance to conventional and immune-based therapies. Recent single-cell studies have uncovered an unexpected heterogeneity of macrophage states in human cancers, moving beyond the simplistic “M1–M2” paradigm. However, progress in understanding human TAM biology has been hampered by the lack of physiologically relevant in vivo models. Here, we leverage next-generation humanized mouse models implanted with cell line xenografts (CDX) to dissect the transcriptional and epigenetic programs that shape monocyte-to-macrophage differentiation within the tumor microenvironment. We show that these immuno-CDX models faithfully reproduce the phenotypic and molecular hallmarks of TAMs observed in patient tumors. By integrating single-cell transcriptomic and chromatin accessibility profiling with high-dimensional co-expression and enhancer-driven regulatory network analyses, we reconstruct a comprehensive framework of human TAM states. Furthermore, CRISPR-mediated gene editing of hematopoietic stem cells prior to humanization identifies MAFB as a key regulator of TAM plasticity. Our findings establish humanized models as a powerful platform to study TAM biology in vivo, uncover fundamental regulators of macrophage plasticity, and highlight new therapeutic targets for next-generation cancer immunotherapies.

Abstract Alternative splicing is a fundamental mechanism underlying protein diversity. The microtubule-associated protein tau (MAPT) undergoes age-associated alternative splicing of exon 10 to generate 3R and 4R isoforms, and disruption of the 4R:3R ratio is a central feature of tauopathies. However, the molecular mechanisms regulating tau exon 10 splicing remain incompletely understood. Here, we identify the RNA-binding protein CELF2 as a key promoter of tau exon 10 inclusion. Loss of CELF2 in the mouse brain reduces exon 10 inclusion, resulting in a decreased 4R:3R ratio. We show that an intrinsically disordered region (IDR) within the CELF2 hinge domain drives protein condensation and is essential for its splicing activity. This IDR can be functionally substituted by those of FUS or TAF15. CRISPR-based imaging reveals colocalization of CELF2 condensates with tau RNA. Proteomic analyses identify NOVA2 and SFPQ as CELF2 interactors, which co-condense with CELF2 to cooperatively regulate tau exon 10 splicing. A conserved negatively charged residue (D388) within the IDR is critical for condensate formation, protein interactions, and splicing function. Finally, CELF2 condensation capacity correlates with 4R tau expression in vivo and influences locomotor and cognitive performance. These findings uncover a condensate-based mechanism for tau splicing regulation with implications for tau-related neurodegeneration.

Abstract Background Computational prediction of CRISPR-Cas9 off-target activity is essential for safe guide-RNA design, yet models trained on large proxy datasets often fail to generalize to experimentally validated sites. Methods We present a modular two-stage deep learning framework that separates sequence representation learning from off-target classification. In Stage 1, guide RNA sequences are encoded using frozen, pretrained DNABERT embeddings learned from large genomic corpora. In Stage 2, these embeddings are integrated with mismatch-level and pairwise sequence features within a hybrid CNN-Transformer classifier trained exclusively on a high-throughput proxy dataset. Results On the external TrueOT benchmark, a curated collection of low-throughput, experimentally confirmed off-target sites, the full model achieved a mean ROC-AUC of 0 . 70   ±   0 . 03 and a PR-AUC of 0 . 30   ±   0 . 03 , markedly surpassing the proxy-only baseline (ROC-AUC = 0.64, PR-AUC = 0.22). Ablation studies confirmed that the performance gain arises from the pretrained sequence representations rather than architectural complexity. Conclusions Decoupling representation learning from downstream classification and leveraging frozen transformer-based embeddings substantially improves generalization to biologically relevant off-target predictions. The proposed framework provides a reproducible baseline for the assessment of CRISPR-Cas9 risk and underscores the importance of transfer learning in the integration of proxy test data and experimental results in the real-world.

Abstract Base editing, an extension of CRISPR-Cas9 gene editing, presents a promising strategy for correcting genetic mutations underlying inherited bone marrow failure syndromes. This approach enables ex vivo editing of haematopoietic stem cells, which can be reintroduced into patients without the risk of immune rejection. We have identified pathogenic single nucleotide mutations associated with bone marrow failure syndromes that have a nearby protospacer adjacent motif site and were amenable to adenine base editing, and designed guideRNAs to target. We performed a pooled CRISPR screen to simultaneously target these mutations in haematopoietic cell lines. This pooled strategy addresses the limitation of focusing solely on more prevalent mutations, thereby enabling the inclusion of rarer variants. Our comprehensive analysis revealed correction at 48 loci, with editing efficiencies of up to 26.22% in Jurkat cells and 19.75% in K562 cells. These findings highlight the potential of adenine base editing to correct a broad range of pathogenic variants and accelerate the development of targeted therapies for bone marrow failure syndromes.

Abstract Sex chromosome evolution is traditionally viewed as a unidirectional process of recombination suppression followed by progressive Y degeneration, yet many vertebrates deviate from this canonical trajectory. Frogs represent a striking exception: they retain homomorphic sex chromosomes, exhibit extreme heterochiasmy, and undergo X-Y recombination in sex-reversed females, providing a powerful system to explore alternative evolutionary pathways. Here, we investigate a Swiss Alpine population of the European common frog ( Rana temporaria ) with homomorphic XY chromosomes and three coexisting male genotypes, using whole genome sequencing of pooled samples carrying distinct Y haplotypes, whole genome sequencing of doubled haploid YY individuals, and transcriptomic (RNA-seq) data. We uncover extreme heterogeneity in X–Y divergence: two fully differentiated Y haplotypes span nearly 90% of the sex chromosome (~618–625 Mb), forming the largest non-recombining region (large NRR) described in a vertebrate to date, whereas a semi-differentiated Y harbors only a NRR of 4.64 Mb (small NRR), and XX males display no detectable X-Y divergence. Despite extensive recombination suppression, Y degeneration is minimal even within the large NRR: we detect elevated transposable element insertions but no gene loss, Y-linked gene copy decay, faster-X effects, or enrichment of sex-biased genes. All Y haplotypes share a small NRR encompassing the candidate master sex-determining gene Dmrt1 , defining a common sex-determining locus. Our data support a dynamic model in which extreme heterochiasmy and recurrent X–Y recombination via sex reversal repeatedly reset X-Y divergence, generating and potentially maintaining multiple Y haplotypes with distinct evolutionary histories. The long-term coexistence of these haplotypes across the species range is compatible with a dynamic equilibrium between drift-driven loss and mutation- and recombination-driven renewal of Y chromosome variation. Together, our findings reveal a non-canonical, reversible pathway of sex chromosome evolution shaped by sex-specific recombination patterns and sex reversal, challenging the universality of classical models and highlighting the value of non-model vertebrates for understanding sex chromosome diversity.