Przybyl, J. | Search bioRxiv

bioRxiv2026

Machine Learning Analysis to Define Cell Lineage in Leiomyosarcoma

IntroductionCellular differentiation and lineage commitment are known to be associated with differences in DNA methylation. Leiomyosarcoma (LMS) is a tumor thought to originate from smooth muscle cells in the walls of vessels in the soft tissue (STLMS) or from the uterine myometrium (ULMS). Here, we identify the methylation signatures of normal smooth muscle cells from blood vessels and the uterine wall and compare these with those found in STLMS and ULMS. We hypothesized that these methylation signatures could be used to assign a smooth muscle subtype of origin to individual leiomyosarcomas, and that tumors of different origin would show biological differences with potential therapeutic relevance. MethodsTo define methylation profiles for smooth muscle from vessel walls versus those found in myometrium, EPIC methylation profiling was performed on DNA from 49 formalin-fixed paraffin-embedded (FFPE) normal smooth muscle samples. A supervised machine learning algorithm (Random Forest) was used to distinguish the methylation patterns of normal smooth muscle cells in vessel walls from those in the myometrium. The resulting classifier was applied to methylation data on 67 cases of LMS with corresponding bulk RNAseq data to identify which tumors showed a methylation signature most consistent with either blood vessel wall (LMSvessel) or myometrial smooth muscle (LMSwall). A custom signature matrix derived from scRNAseq data from 6 samples of LMS was used in CIBERSORTx analysis to compare the cellular composition of LMS cases with a vessel or uterine wall methylation signature. ResultsA high degree of correlation was found between the known site of origin for LMS (STLMS vs ULMS) and the methylation signature derived from different types of normal smooth muscle. LMSwall tumors compared to LMSvessel tumors had significantly higher activation of the PD-1 checkpoint pathway in RNAseq analysis. Digital flow cytometry by CIBERSORTx analysis showed an increased expression of transcriptomic signatures of several immune cell subtypes in LMSvessel tumors. ConclusionUsing a supervised machine learning approach we classified LMS samples as either showing a high similarity in methylation patterns to normal smooth muscle cells of either the vessel wall or the myometrium. We found a correlation between LMS showing either a "vessel" or "muscle wall" methylation signature and their site of origin, but notably we also identified some exceptions. When classified based on their methylation signature LMSwall and LMSvessel differed in their PD-1 pathway activation and in their predicted immune cell populations, suggesting potential implications for immunotherapeutic approaches.

cancer biology↗

bioRxiv2019

CRISPR-NSID: an in vivo CRISPR/Cas9 negative selection screen reveals EZH2 as a druggable dependency factor in a genetic desmoid tumor model

Cancer precision medicine implies identification of tumor-specific vulnerabilities associated with defined oncogenic pathways. Desmoid tumors are soft-tissue neoplasms strictly driven by Wnt signaling network hyperactivation. Despite this clearly defined genetic etiology and the strict and unique implication of the Wnt/{beta}-catenin pathway, no specific molecular targets for these tumors have been identified. To address this caveat, we developed fast and semi-high throughput genetic Xenopus tropicalis desmoid tumor models to identify and characterize novel drug targets. We used multiplexed CRISPR/Cas9 genome editing in these models to simultaneously target a tumor suppressor gene (apc) and candidate dependency genes. Our methodology CRISPR/Cas9 Selection mediated Identification of Dependencies (CRISPR-SID) uses calculated deviations between experimentally observed gene editing outcomes and deep-learning-predicted double strand break repair patterns, to identify genes under negative selection during tumorigenesis. This revealed EZH2 and SUZ12, both encoding polycomb repressive complex 2 components, and the transcription factor CREB3L1, as genetic dependencies for desmoid tumors. In vivo EZH2 inhibition by Tazemetostat induced partial regression of established autochthonous tumors. In vitro models of patient desmoid tumor cells revealed a direct effect of Tazemetostat on Wnt pathway activity. CRISPR-SID represents a potent novel approach for in vivo mapping of tumor vulnerabilities and drug target identification. Significance StatementCRISPR-SID was established in the diploid frog Xenopus tropicalis for in vivo elucidation of cancer cell vulnerabilities. CRISPR-SID uses deep learning predictions and binomial theory to identify genes under positive or negative selection during autochthonous tumor development. Using CRISPR-SID in a genetic model for desmoid tumors, treatment-recalcitrant mesenchymal tumors driven by hyper-activation of the Wnt signaling pathway, we identified EZH2 and SUZ12, both encoding critical components of the polycomb repressive complex 2, as dependency genes for desmoid. Finally, we demonstrate the promise of EZH2 inhibition as a novel therapeutic strategy for desmoid tumors. With the simplicity of CRISPR sgRNA multiplexing in Xenopus embryos the CRISPR-SID method may be applicable to reveal vulnerabilities in other tumor types.

cancer biology↗