Biology subjects

Sousa, A. A.

Publications and source records attributed to Sousa, A. A..

3 recordsLinked to original sources

bioRxiv2026

Creation of a High-Resolution, Continually Evolvable Lineage Tracer

Cellular diversity is required for the proper function of many tissues, including the retina, with its ~130 cell types.1-3 Previous lineage studies suggested that retinal progenitor cells (RPCs) generate cell diversity through biased cell divisions.4-9 However, the extent to which each progenitor cell contributes to diversity, and the overarching rules governing the lineages of the collection of progenitor cells, are unknown. While classic lineage tracing approaches including retroviral labeling4,6,10, dye injection11,12, and time-lapse microscopy13,14 were used to make the foundational insights regarding retinal lineages, they cannot be used to interrogate multigenerational lineages at scale. Here we describe the creation of an evolvable lineage tracing tool, "SCRIBBLE" for Sequential Combinatorial Recorder for Iterative Barcode-Based Lineage Evolution. SCRIBBLE uses prime editor to iteratively edit a barcode sequence while preserving the order of genesis of clonally related progeny. Through a series of rationally designed high-throughput screens and design optimizations, we identified novel sequences that support a high level of barcode editing and information content. Evaluation of SCRIBBLE in embryonic mouse retina recapitulated known lineages in retinal development, demonstrating that SCRIBBLE is capable of high-resolution lineage tracing in a complex tissue.

developmental biology↗

bioRxiv2026

In vivo HSPC gene therapy of hemoglobinopathies without drug selection of corrected cells

In vivo hematopoietic stem/progenitor cell (HSPC) gene therapy remains limited by low gene-editing efficiency and a lack of clinically applicable selection strategies to enrich therapeutically corrected progeny. We used in vivo base and prime editing to introduce a nonpathogenic EPOR variant into HSPCs, conferring erythropoietin hypersensitivity and promoting preferential expansion of gene-corrected erythroid cells. EPOR editing was combined with three therapeutic approaches for the correction of hemoglobinopathies: {gamma}-globin gene addition, {gamma}-globin reactivation, or correction of the sickle cell disease mutation. Tropism-modified helper-dependent adenoviral vectors (HDAd6/3+) targeting HSPCs were used to simultaneously deliver the EPOR-editing machinery and the corresponding therapeutic components. In vitro studies in an erythroid progenitor cell line and primary CD34+ cells demonstrated that the EPORW439* variant conferred a strong proliferative advantage to therapeutically modified erythroid progenitors. Mice humanized with CD34+ cells from a {beta}/{beta}-thalassemia patient were subjected to EPORW439*-mediated EPO hypersensitivity alongside a therapeutic {gamma}-globin transgene which resulted in >70% HbF-positive erythroid cells and substantial reversion of the disease-associated phenotype, including reduced oxidative stress, near-complete elimination of splenic iron deposition, and reduced splenomegaly. Importantly, these effects were achieved after simple intravenous administration of the vectors following HSPC mobilization and cytokine prophylaxis, without subsequent pharmacologic selection. Together, these findings establish a strategy to amplify the therapeutic benefit of otherwise limited in vivo HSPC gene editing for hemoglobinopathies.

bioengineering↗

bioRxiv2026

Mechanistic machine learning enables interpretable and generalizable prediction of prime editing outcomes

Although prime editing (PE) can effect virtually any specified local change to genomic DNA in living systems, its efficient application currently requires extensive optimization of prime editing guide RNA (pegRNA) sequences. We present OptiPrime, a machine learning model of PE efficiency based on our current understanding of the mechanism of prime editing. OptiPrime achieves state-of-the-art accuracy on PE efficiency prediction and also enables prediction of nicking guide RNA (PE3) and dual pegRNA (twinPE) outcomes. We validated that OptiPrime has learned the determinants of mammalian mismatch repair (MMR), and is therefore well suited for nominating MMR-evasive silent edits that improve PE efficiency. We demonstrate the utility of OptiPrime in a variety of prospective therapeutic contexts, including in primary human and mouse cells. Finally, we show how OptiPrime can be used to achieve highly streamlined and efficient in vivo correction of a pathogenic mutation in the brain of a mouse model of KIF1A-associated neurological disorder.

biochemistry↗