PD01-11: Characterization of Wilms Tumor and Human Fetal Kidney Using Spatial Transcriptomics

Astgik Petrosyan, Valentina Villani, Paola Aguiari, Matthew Thornton, Zhou Shengmei, Brendan Grubbs, Marie Csete, Roger De Filippo, Kevin Lemley, Sargis Sedrakyan, Stefano Da Sacco, Laura Perin*, los angeles, CA

Introduction: Growing evidence links Wilms tumor (WT) to aberrant nephrogenesis. Studies highlighted the genetic complexity of WT, but little is known about the molecular mechanisms that regulate WT development. Using Spatial transcriptomics (ST), which allows analysis of the gene expression based on morphological context, we showed important differences between WT subtypes and defined the interactive gene networks involved in WT development using human fetal kidney (hFK) as reference.

Methods: Using Visium 10x Genomics, we generated spatial maps of gene expression in human fetal kidney (hFK) andfavorable (stage III) and unfavorable (stage I) WTs. Data were analyzed using Space Ranger software v1.0.0, Seurat v3.2, Panther V14, and Loupe Cell Browser and further analyzed against our previously generated bulk/sc-RNA seq data on the same samples.

Results: ST identified specific clusters in hFK that closely recapitulated the developmental stages of normal nephrogenesis (nephrogenic zone, glomeruli, tubules, and stroma). Unfavorable WT and favorable WT clusters showed heterogeneity of the tumor landscape (blastema, epithelium, and stroma and non-renal phenotypes). Blastema in WT favorable vs. WT unfavorable, though histologically identical, presented different transcriptomics profiles. WTs also showed gene expression typical of muscle tissue (or other non-renal phenotypes) rather than mature kidney structures, which correlated with the histologic absence of mature tubules and glomeruli. Comparative RNA-seq analysis identified cells expressing SIX2 and CITED1 as the root cells of the origin of the WT. Unfavorable WT expressed a higher level of CITED1 in blastema foci and higher expression of uncommitted genes and modulation of inductive nephrogenic signals like WNT and FGF. We also identified genes expressed specifically in WT subtypes and performed a preliminary characterization of the immune milieu of WT.

Conclusions: The spatiotemporal mapping combined with different transcriptomic data highlighted the heterogeneity of the WT subtypes confirming uncommitted nephron progenitors as driving the development of WT. We identified genes that may allow for better stratification of WT and potential therapeutic targets for distinct WT subtypes.

Source of Funding: GOFARR Fund