The TLR3 L412F polymorphism prevents TLR3-mediated tumor cell death induction in pediatric sarcomas

<strong>Abstract</strong> Toll-like receptor 3 (TLR3) is a pattern recognition receptor mainly known for its role in innate immune response to infection. Indeed, binding of double-stranded RNA (dsRNA) to TLR3 triggers a pro-inflammatory cascade leading to cytokine release and immune cell activation. Its anti-tumoral potential has emerged progressively, associated with a direct impact on tumor cell death induction and with an indirect action on immune system reactivation. Accordingly, TLR3 agonists are currently being tested in clinical trials for several adult cancers. Meanwhile, TLR3 variants have been linked to auto-immune disorders, and as risk factors of viral infection and cancers. However, aside from neuroblastoma, TLR3 role in childhood cancers has not been evaluated. Here, by integrating public transcriptomic data of pediatric tumors, we unveil that high TLR3 expression is largely associated with a better prognosis in childhood sarcomas. Using osteosarcomas and rhabdomyosarcomas as models, we show that TLR3 efficiently drives tumor cell death <em>in vitro </em>and induces tumor regression <em>in vivo</em>. Interestingly, this anti-tumoral effect was lost in cells expressing the homozygous TLR3 L412F polymorphism, which is enriched in a rhabdomyosarcoma cohort. Thus, our results demonstrate the therapeutic potential associated with the targeting of TLR3 in pediatric sarcomas, but also the need to stratify patients eligible for this clinical approach with respect to the TLR3 variants expressed. <strong>Data availability</strong> Schafer-Welle (RMS and normal muscle), Kuijjer (Osteosarcoma), Dirksen (Ewing sarcoma), TCGA 2022-v32 Sarcoma (Leiomyosarcoma) are available online on the R2 website (https://hgserver1.amc.nl/cgi- bin/r2/main.cgi). The Missiaglia (RMS) dataset is available on ArrayExpress under the identifier E-TABM-1202 (https://www.ebi.ac.uk/arrayexpress/experiments/E-TABM-1202/). The Khan (RMS) dataset can be requested by contacting M.D. Javed Khan (khanjav@mail.nih.gov). <strong>Repository information</strong> This repository contains the R scripts used to generate the figures of the paper.