Abstract 6350: Exploring genetic and non-genetic evolution of glioblastoma by integrating single cell DNA and RNA sequencing

Abstract Glioblastoma (GBM), the most frequent and aggressive primary brain tumor, almost invariably recurs as an incurable disease after standard therapy. Molecular heterogeneity underlies the failure of current treatment and drives GBM progression. Although significant efforts have been directed at resolving the subclonal architecture, identifying and characterizing the genetic subpopulation in GBM has proven challenging. Recent advances in single-cell genomics (scWGS) have provided new opportunities for exploring the genomic complexity of tumors at unprecedented resolution. However, earlier methods of scWGS faced limitations due to poor amplification uniformity resulting in high number of artefacts. The introduction of primary template-directed amplification (PTA) has overcome this challenge allowing for more accurate detection of genomic alterations at single-cell resolution. In this study, we combined PTA scWGS of 5,689 single cells with single-nucleus RNA sequencing (snRNAseq) of 41,071 single nuclei from 25 GBM samples to explore the clonal architecture and the dynamics of genetic and non-genetic subclonal programs activated in GBM. PTA-based scWGS data showed optimal performance in terms of overdispersion and coverage breadth when compared to other single cell methods for measuring somatic copy number alterations (SCNA). The SCNA substructure of GBM was reconstructed by applying a graph-based clustering to scWGS data, which identified a median of 4 subclones per tumor. GBM samples showed a heterogenous distribution of genetic clones constituted by variable numbers of cells. For each sample we computed the consensus SCNA profiles to reconstruct the genetic phylogeny of GBM. The predicted evolutionary model showed a first trajectory from the diploid cell to the most recent common ancestor (MRCA) with the highest number of SCNA (median 14), including frequent GBM driver alterations (chr 7 gain, chr 10 loss, EGFR amp, CDKN2A del). Conversely, number and frequency of SCNA decreased in the clades diversifying the MRCA into the later ancestors. By aligning the scWGS clonal lineages with snRNAseq gene expression profiles from the same tumors, we found significant coordination of genetic and non-genetic mechanisms that diversified tumor subpopulations and led to divergent phenotypes recurring across the patients. The functional phylogeny of GBM identified two main branches of tumor evolution that emerged from the MRCA with the activation of neuronal functions and differentiation in one clade, and the acquisition of cell cycle programs, stem cell activities, and epithelial to mesenchymal transition in the opposite clade, respectively. This multi-omics approach enabled the reconstruction of the evolutionary dynamics of GBM at both genetic and transcriptional levels, revealing critical insights into clonal evolution and therapeutic resistance. Citation Format: Fulvio D'Angelo, Aram Ko, Luciano Garofano, Pedro Davila, Luigi Ferraro, Benjamin Currall, Sion Williams, Sakir Gultekin, Ashish Shah, Ricardo Komotar, Michael Ivan, Corneliu Sologon, Alice Laurenge, Alberto Picca, Amel Dridi-Aloulou, Bertrand Mathon, Franck Bielle, Stephan Zuchner, Macarena de la Fuente, Marica Eoli, Chul-Kee Park, Michele Ceccarelli, Marc Sanson, Anna Lasorella, Antonio Iavarone. Exploring genetic and non-genetic evolution of glioblastoma by integrating single cell DNA and RNA sequencing [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2025; Part 1 (Regular Abstracts); 2025 Apr 25-30; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2025;85(8_Suppl_1):Abstract nr 6350.