(P 286) EPIGENETIC MECHANISMS AND SIGNATURES OF SARCOMA PROGRESSION AND METASTASIS

Objective: Metastasis is a leading cause of death in sarcoma, but the molecular mechanisms driving metastasis are poorly understood. The overall objective of this study is to identify epigenetic mechanisms and signatures of sarcoma metastasis using preclinical platforms and patient datasets. The use of epigenetic modifications to predict tumor progression has not been well-explored in sarcoma due to the complexity of subtype-specific biology and limited number of samples. In our initial study, we determined that loss-of-function mutations in the histone methyltransferase PRC2 can drive MPNST metastasis and remodel the extracellular matrix. We extended these findings to analyses of multi-sarcoma epigenomic patient datasets to demonstrate that epigenetic cross-talk between histone modifications and DNA methyltransferases results in DNA methylation signatures that predict metastasis and tumor progression of multiple sarcoma subtypes.

Methods: To define mechanisms of metastasis in MPNSTs, we used a combination of in vitro metastasis assays, 3D collagen studies and orthotopic mouse models. These approaches were combined with time-lapse microscopy and analysis of patient samples to identify a critical role for PRC2 loss in MPNST metastasis and remodeling of the extracellular matrix. To extend these findings and identify epigenetic signatures of sarcoma metastasis in patient datasets, we used DNA methylation and RNAseq data from patient neurofibromas and MPNSTs with known PRC2 status. Following identification of 8 differentially methylated CpGs in MPNST datasets, we examined these sites in other sarcoma subtypes to determine a DNA methylation signature of metastasis that can be applied to multiple sarcomas.

Results: Loss of PRC2 drives MPNST metastatic phenotypes including collagen-dependent invasion, upregulation of matrix remodeling enzymes, and elevated lung metastases in orthotopic mouse models. Furthermore, clinical sample analysis demonstrates that PRC2 loss correlates with metastatic disease, increased fibrosis, and decreased survival in MPNST patients. Analysis of epigenomic patient data sets determines that altered histone modifications in MPNST are associated with a global DNA hypermethylation signature. We found over 900 CpGs are hypermethylated in MPNSTs with PRC2 loss. Furthermore, we identified 8 differentially methylated CpGs in Interleukin 17D and Interleukin 17 Receptor (IL17/R) family that correlate with the progression and metastasis of MPNSTs. Changes in IL17D/RD DNA methylation and RNA expression also correlate with progression and metastasis of other sarcoma subtypes including osteosarcoma (OS), rhabdomyosarcoma (RMS), and synovial sarcoma (SS), suggesting these CpGs can predict metastatic outcome in multiple sarcoma subtypes. Analysis of scRNAseq datasets shows that IL17D/RD is expressed in both tumor and stromal cell populations in MPNST, OS and RMS patients.

Conclusion: This study highlights the translational utility of preclinical mouse models to uncover mechanistic drivers of sarcoma biology that inform patient-relevant signatures of sarcoma progression and metastasis. To our knowledge, this is the first analysis of IL17D/RD expression and methylation patterns in sarcoma datasets. These findings have broad implications for disease monitoring and surveillance of multiple sarcoma subtypes. In addition, these data provide a strong rationale for potential targeting of ECM remodeling enzymes and tumor fibrosis to improve outcomes in MPNST patients.