Biology subjects

Schlotawa, L.

Publications and source records attributed to Schlotawa, L..

4 recordsLinked to original sources

bioRxiv2026

Human-supervised Agentic AI for Hypothesis Generation and Experimental Assistance in Drug Repurposing

Computational drug repurposing has largely been focused on rapid hypothesis generation, yet real-world applications span a far broader lifecycle, from drug candidate suggestion to designing experiments, analyzing assay data, and iteratively refining candidates. Here, we demonstrate that agentic AI can operate throughout this lifecycle. To this end, we developed RepurAgent, a hierarchical multi-agent AI system comprising a supervisor agent and a planning agent that coordinate four specialized sub-agents (research, prediction, data, and report), through a human-in-the-loop design, with episodic memory and retrieval-augmented generation. The system is grounded in data, tools, and standard operating procedures specific for drug repurposing, developed within the REMEDi4ALL consortium. We validated the agentic system across three scenarios spanning the various stages within the repurposing lifecycle: in Acute Myeloid Leukemia, a blinded expert evaluation indicated that RepurAgent produced substantially more novel and mechanistically credible candidates compared to a vanilla LLM baseline; in a retrospective COVID-19 antiviral screen, RepurAgent acted as an adaptive experimental collaborator, prioritizing compounds with AUC-ROC up to 0.99 without predefined thresholds and flagging confounders missed in manual review; and for Multiple Sulfatase Deficiency, it prioritized 81 high-confidence candidates from 5000 compounds, which were further corroborated by domain experts. These results demonstrate that agentic AI can support across the drug repurposing lifecycle, from hypothesis generation to experimental analysis. RepurAgent is open source and deployed at https://repuragent.serve.scilifelab.se/.

bioinformatics↗

bioRxiv2024

Effective Gene Therapy for Metachromatic Leukodystrophy Achieved with Minimal Lentiviral Genomic Integrations

Metachromatic leukodystrophy (MLD) is a fatal lysosomal storage disease (LSD) characterized by the deficient enzymatic activity of arylsulfatase A (ARSA). Combined autologous hematopoietic stem cell transplant (HSCT) with lentiviral (LV) based gene therapy has great potential to treat MLD. However, if enzyme production is inadequate, this could result in continued loss of motor function, implying a high vector copy number (VCN) requirement for optimal enzymatic output. This may place children at increased risk for genomic toxicity due to higher VCN. We increased the expression of ARSA cDNA at single integration by generating novel LVs, optimizing ARSA expression, and enhancing safety. In addition, our vectors achieved optimal transduction in mouse and human HSC with minimal multiplicity of infection (MOI). Our top-performing vector (EA1) showed at least 4X more ARSA activity than the currently EU-approved vector and a superior ability to secrete vesicle-associated ARSA, a critical modality to transfer functional enzymes from microglia to oligodendrocytes. Three-month-old Arsa-KO MLD mice transplanted with Arsa-KO BM cells transduced with 0.6 VCN of EA1 demonstrated behavior and CNS histology matching WT mice. Our novel vector boosts efficacy while improving safety as a robust approach for treating early symptomatic MLD patients.

molecular biology↗

bioRxiv2024

Hematopoietic stem cell gene therapy improves outcomes in a clinically relevant mousemodel of Multiple Sulfatase Deficiency

Multiple sulfatase deficiency (MSD) is a severe, lysosomal storage disorder caused by pathogenic variants in the gene SUMF1, encoding the sulfatase modifying factor formylglycine-generating enzyme. Patients with MSD exhibit functional deficiencies in all cellular sulfatases. The inability of sulfatases to break down their substrates leads to progressive and multi-systemic complications in patients, similar to those seen in single-sulfatase disorders such as metachromatic leukodystrophy and mucopolysaccharidoses IIIA. Here, we aimed to determine if hematopoietic stem cell transplant with ex vivo SUMF1 lentiviral gene therapy could improve outcomes in a clinically relevant mouse model of MSD. We first tested our approach in MSD patient-derived cells and found that our SUMF1 lentiviral vector improved protein expression, sulfatase activities, and glycosaminoglycan accumulation. In vivo, we found that our gene therapy approach rescued biochemical deficits, including sulfatase activity and glycosaminoglycan accumulation, in affected organs of MSD mice treated post-symptom onset. In addition, treated mice demonstrated improved neuroinflammation and neurocognitive function. Together, these findings suggest that SUMF1 HSCT-GT can improve both biochemical and functional disease markers in the MSD mouse.

neuroscience↗

bioRxiv2023

A novel iPSC model reveals selective vulnerability of neurons in Multiple Sulfatase Deficiency

Multiple sulfatase deficiency (MSD) is an ultra-rare, inherited lysosomal storage disease caused by mutations in the gene sulfatase modifying factor 1 (SUMF1). MSD is characterized by the functional deficiency of all sulfatase enzymes, leading to the storage of sulfated substrates including glycosaminoglycans (GAGs), sulfolipids, and steroid sulfates. Patients with MSD experience severe neurological impairment, hearing loss, organomegaly, corneal clouding, cardiac valve disease, dysostosis multiplex, contractures, and ichthyosis. Here, we generated a novel human model of MSD by reprogramming patient peripheral blood mononuclear cells to establish an MSD iPSC line (SUMF1 p.A279V). We also generated an isogenic control iPSC line by correcting the pathogenic variant with CRISPR/Cas9 gene editing. We successfully differentiated these iPSC lines into neural progenitor cells (NPCs) and NGN2-induced neurons (NGN2-iN) to model the neuropathology of MSD. Mature neuronal cells exhibited decreased SUMF1 gene expression, increased lysosomal stress, impaired neurite outgrowth and maturation, reduced sulfatase activities, and GAG accumulation. Interestingly, MSD iPSCs and NPCs did not exhibit as severe of phenotypes, suggesting that as neurons differentiate and mature, they become more vulnerable to loss of SUMF1. In summary, we demonstrate that this human iPSC-derived neuronal model recapitulates the cellular and biochemical features of MSD. These cell models can be used as tools to further elucidate the mechanisms of MSD pathology and for the development of therapeutics.

cell biology↗