Biology subjects

Blondel, A.

Publications and source records attributed to Blondel, A..

4 recordsLinked to original sources

bioRxiv2025

Deep Learning Bridges Histology and Transcriptomics to Predict Molecular Subtypes and Outcomes in Muscle-Invasive Bladder Cancer

Muscle-Invasive Bladder Cancer (MIBC) is a heterogeneous disease with distinct molecular subtypes influencing prognosis and therapeutic response. However, molecular profiling through RNA sequencing remains costly, time-consuming and complicated by intratumoral heterogeneity. We developed a Deep Learning (DL) approach to infer molecular subtypes from routine histopathological slides and to evaluate its prognostic value in patients treated with neoadjuvant chemotherapy (NAC). We developed an DL-model predicting the expression of 848 subtype-associated genes from histological images of transurethral resection of bladder tumor, enabling spatial molecular subtyping at tile level. The model was trained on 297 NAC-treated patients from the VESPER clinical trial and evaluated on three independent cohorts (COBLAnCE, n=224; Saint-Louis, n=30 and TCGA, n=315), covering diverse staining protocols and scanner types. Spatial transcriptomics from six VESPER patients confirmed the spatial consistency of the inferred expression profiles. Our approach achieved a ROC AUC of 0.94 for molecular subtype prediction, with 95% of genes significantly predicted, demonstrating its ability to capture transcriptomic dysregulations from histological morphology. Predicted expression maps revealed spatially coherent patterns and intratumoral molecular heterogeneity. Importantly, tumors predicted with basal/squamous features (pure or mixed), were associated with significantly worse progression-free and overall survival after NAC (log-rank p=0.014 and 0.037, respectively). This DL-based framework enables accurate and spatially resolved inference of gene expression and molecular subtypes in MIBC without sequencing. These findings could improve patient stratification in clinical practice and support the design of more targeted clinical trials. Further validation in larger cohorts is needed before routine clinical implementation.

cancer biology↗

bioRxiv2025

RNA2seg: a generalist model for cell segmentation in image-based spatial transcriptomics

Imaging-based spatial transcriptomics (IST) enables high-resolution spatial mapping of RNA species. A key challenge in IST is accurate cell segmentation to assign each RNA molecule to the right cell. Here, we present RNA2seg, a novel segmentation algorithm trained on over 4 million cells from MERFISH and CosMx datasets across seven organs using a teacher-student training scheme. RNA2seg integrates RNA point clouds and all available membrane and nuclear stainings. Validation on manually annotated data shows superior performance including in zero-shot and few-shot settings. The method is available as a documented pip package: https://github.com/fish-quant/rna2seg.

bioinformatics↗

bioRxiv2021

A highly sensitive cell-based luciferase assay for high-throughput automated screening of SARS-CoV-2 nsp5/3CLpro inhibitors

Effective drugs against SARS-CoV-2 are urgently needed to treat severe cases of infection and for prophylactic use. The main viral protease (nsp5 or 3CLpro) represents an attractive and possibly broad-spectrum target for drug development as it is essential to the virus life cycle and highly conserved among betacoronaviruses. Sensitive and efficient high-throughput screening methods are key for drug discovery. Here we report the development of a gain-of-signal, highly sensitive cell-based luciferase assay to monitor SARS-CoV-2 nsp5 activity and show that it is suitable for high-throughput screening of compounds in a 384-well format. A benefit of miniaturisation and automation is that screening can be performed in parallel on a wild-type and a catalytically inactive nsp5, which improves the selectivity of the assay. We performed molecular docking-based screening on a set of 14,468 compounds from an in-house chemical database, selected 359 candidate nsp5 inhibitors and tested them experimentally. We identified four molecules, including the broad-spectrum antiviral merimepodib/VX-497, which show anti-nsp5 activity and inhibit SARS-CoV-2 replication in A549-ACE2 cells with IC50 values in the 4-21 {micro}M range. The here described assay will allow the screening of large-scale compound libraries for SARS-CoV-2 nsp5 inhibitors. Moreover, we provide evidence that this assay can be adapted to other coronaviruses and viruses which rely on a viral protease.

microbiology↗

bioRxiv2020

Stress-induced tyrosine phosphorylation of RtcB modulates IRE1 activity and signaling outputs.

Endoplasmic Reticulum (ER) stress is a hallmark of various diseases, which is dealt with through the activation of an adaptive signaling pathway named the Unfolded Protein Response (UPR). This response is mediated by three ER-resident sensors and the most evolutionary conserved, IRE1 signals through its cytosolic kinase and endoribonuclease (RNase) activities. IRE1 RNase activity can either catalyze the initial step of XBP1 mRNA unconventional splicing or degrade a number of RNAs through Regulated IRE1- Dependent Decay (RIDD). The balance between these two activities plays an instrumental role in cells life and death decisions upon ER stress. Until now, the biochemical and biological outputs of IRE1 RNase activity have been well documented, however, the precise mechanisms controlling whether IRE1 signaling is adaptive or pro-death (terminal) remain unclear. This prompted us to further investigate those mechanisms and we hypothesized that XBP1 mRNA splicing and RIDD activity could be co-regulated by the IRE1 RNase regulatory network. We showed that a key nexus in this pathway is the tRNA ligase RtcB which, together with IRE1, is responsible for XBP1 mRNA splicing. We demonstrated that RtcB is tyrosine phosphorylated by c-Abl and dephosphorylated by PTP1B. Moreover, we identified RtcB Y306 as a key residue which, when phosphorylated, perturbs RtcB interaction with IRE1, thereby attenuating XBP1 mRNA splicing and favoring RIDD. Our results demonstrate that the IRE1 RNase regulatory network is dynamically fine-tuned by tyrosine kinases and phosphatases upon various stresses and that the nature of the stress determines cell adaptive or death outputs.

biochemistry↗