Scientific Symposium 2026

How the developmental niche gates oncogenic competence

Some childhood cancers can arise in utero and can subsequently regress at birth, yet the cues that permit malignant proliferation in utero as opposed to postnatal life are often unclear. Transient abnormal myelopoiesis (TAM) is a fetal liver leukemia driven by GATA1s mutations and a rare exemplar of a spontaneously resolving cancer when blood formation shifts from liver to bone marrow (BM) during development. I discovered that high thrombopoietin signalling through a previously unrecognised heteromeric CSF2RB-MPL receptor complex promotes the expansion of GATA1s-mutant cells in the fetal liver. TAM can transform into myeloid leukemia of Down syndrome (ML-DS) upon acquisition of additional mutations. I also discovered that the ML-DS-associated CSF2RB A455D variant promotes MPL dimerization and constitutive JAK-STAT activation, bypassing THPO dependence and enabling GATA1s-mutant cells to proliferate in the postnatal BM. Together, these findings reveal a cytokine gate that developmentally restricts GATA1s oncogenic competence, reconciling why TAM expands in the fetal liver yet resolves after birth, and establishing a framework for understanding how developmental niches shape childhood cancer evolution.  

David Cruz Hernandez is currently a Postdoctoral Scientist at the MRC Weatherall Institute of Molecular Medicine. His research explores how genetic, cellular, and environmental constraints shape cellular fitness and evolution in blood cancer. He combines genome engineering, functional genomics, and primary haematopoietic cell models to investigate the mechanisms underlying leukaemia development.

MBL-domain RNA surveillance as a driver of transcriptional crosstalk 

RNA processing is closely coordinated with transcription, yet how defects in RNA maturation influence the broader transcriptional state of the cell remains poorly understood. My work addresses this question through MBLAC2, a metallo-?-lactamase-domain (MBL) endoribonuclease required for human ribosomal RNA maturation. Using nascent RNA profiling and long-read sequencing, we show that mammalian pre-rRNA processing occurs co-transcriptionally and that MBLAC2 depletion disrupts this process, alters nucleolar organisation and perturbs RNA polymerase I transcription. These perturbations also reduce global RNA polymerase II transcription and delay cell-cycle progression, revealing a connection between nucleolar RNA processing and transcriptional regulation beyond the nucleolus. These findings motivate the hypothesis that MBL-domain proteins participate in a compartment-specific RNA-surveillance network that communicates RNA-processing defects to the transcriptional machinery. Together, this work establishes the basis for an independent programme to define the RNA substrates and molecular mechanisms of MBL-domain proteins, uncover crosstalk between RNA polymerases I, II and III, and determine how RNA-processing failures are sensed and converted into transcriptional states that shape cancer cell identity, adaptation and vulnerability.

Luciano Marasco Gutierrez is a Postdoctoral Research Associate at the Sir William Dunn School of Pathology, University of Oxford, working with Professor Nicholas Proudfoot. He obtained his PhD in Molecular Biology from the University of Buenos Aires under Dr Alberto Kornblihtt, with research conducted in collaboration with Dr Adrian Krainer at Cold Spring Harbor Laboratory on antisense oligonucleotide therapy for spinal muscular atrophy. His current research explores how RNA-processing mechanisms coordinate transcription, DNA damage responses, nucleolar function and cellular homeostasis. Outside the laboratory, he enjoys graphic arts, particularly printmaking and screen printing.

Predicting CRISPR repair outcomes to identify in vivo dependencies in rare and paediatric cancers 

Rare and paediatric cancers often have a clearly defined genetic cause but few targeted therapeutic options. Predictable CRISPR repair outcomes can be exploited to identify the genes these tumours depend on for survival in vivo. During my doctoral research, I established Xenopus tropicalis as a rapid in vivo platform for modelling these cancers, in which genetically defined tumours arise in F0 animals within months. To make genome engineering predictable, I developed Pythia, a deep learning-guided framework that predicts CRISPR-Cas9 repair outcomes and enables precise genome integrations to be computationally designed, including at endogenous loci in vivo across species. Using CRISPR-SID, I demonstrated how selection acting on CRISPR-induced mutations can reveal therapeutic vulnerabilities, identifying EZH2 as a druggable dependency in desmoid tumours whose inhibition drove tumour regression in vivo. I will conclude by outlining how these approaches can be further combined to enable scalable in vivo dependency mapping in rare and paediatric cancers.

Thomas Naert is an FWO-funded senior postdoctoral researcher at Ghent University, Belgium, embedded in the VIB-UGent Center for Inflammation Research, where his research focuses on CRISPR-based genome engineering, genetically engineered cancer models, and AI-driven approaches to biological imaging and analysis. He developed Pythia, a deep learning-guided platform for precise and predictable genome editing, published in Nature Biotechnology (first author) and Nature Protocols (senior author, in press), and CRISPR-SID, a method for identifying cancer dependencies in vivo (PNAS). His work integrates genome engineering, computational methods, and advanced microscopy to identify therapeutic vulnerabilities in cancer and other human disease.

When tumor metabolism rewrites the neural niche: Schwann-cell reprogramming in pancreatic cancer

Perineural invasion (PNI) is a hallmark of pancreatic ductal adenocarcinoma (PDAC) and reflects reciprocal crosstalk between cancer cells and the neural microenvironment. Schwann cells (SCs), the principal glial cells of peripheral nerves, can acquire a repair-like phenotype and organize into tumor-activated Schwann-cell tracks that facilitate PNI. However, whether PDAC-derived metabolic cues reprogram SCs metabolism and how these adaptations influence tumor invasion remains poorly understood. During my talk, I will present findings demonstrating that metabolites released by glycolytic pancreatic cancer cells reprogram SCs toward a metabolically adaptive, repair-like phenotype that enhances pancreatic cancer cell invasion, supporting a tumor-supportive role for SCs metabolic adaptation.

Maria Daniela Santi is a Marie Skłodowska-Curie Fellow leading the biological research within Prof. Stefan Glöggler’s NMR Signal Enhancement Group at the Max Planck Institute for Multidisciplinary Sciences, Göttingen, Germany. Building on her research at the Uruguayan Center of Molecular Imaging, and at New York University’s Translational Research Center, her research integrates cancer neuroscience, metabolic biology, and hyperpolarized metabolic imaging to investigate complex tumor-microenvironment interactions.

PRIME-loading: The mechanistic basis for high-demand translation of mTOR-dependent transcripts

The cellular proteome shapes cellular function in health and disease, with mRNA translation initiation representing a key point of regulation. In the canonical model, initiation factors assemble on the 40S ribosomal subunit to form the 43S pre-initiation complex, including the ternary complex (eIF2–GTP– Met-tRNAᵢ), before recruitment to mRNAs through the 5′ cap. We find that highly translated housekeeping mRNAs, including those encoding ribosomal proteins and metabolic enzymes, can instead recruit 40S ribosomes before assembly of most initiation factors, enabling ribosome recruitment without extensive 43S pre-assembly. We termed this mechanism PRIME-loading. Biochemical analysis revealed an unexpected link between PRIME-loading and the S6K1–eEF2K axis of the mTOR pathway. eEF2, classically an elongation factor, is a component of PRIME-loaded complexes and remains associated with scanning ribosomes before recruitment of a full set of initiation factors and AUG recognition, with its association regulated by phosphorylation. Together, these findings reveal PRIME- loading as a mechanism that links cellular signaling to the selective control of protein synthesis and provides a unified framework for mTOR-dependent preferential translation and integrated stress response-associated selective translation.

Ramona Weber is a Postdoctoral Fellow at the Institute for Regenerative Medicine, University of Zurich, in the laboratory of Prof. Ataman Sendoel. Her research focuses on the molecular mechanisms that regulate mRNA translation and ribosome recruitment, with particular interest in how cellular signaling controls protein synthesis.