eRNAs Identify Immune Microenvironment Patterns and Provide a Novel Prognostic Tool in Acute Myeloid Leukemia

Ziming Jiang,Junyu Long,Kaige Deng,Yongchang Zheng,Miao Chen

DOI: https://doi.org/10.3389/fmolb.2022.877117

2022-05-02

Abstract:Background: Enhancer RNAs (eRNAs) play an essential role in tumorigenesis as non-coding RNAs transcribed from enhancer regions. However, the landscape of eRNAs in acute myeloid leukemia (AML) and the potential roles of eRNAs in the tumor microenvironment (TME) remain unclear. Method: Gene expression data collected from The Cancer Genome Atlas (TCGA) project were combined with Histone ChIP-seq so as to reveal the comprehensive landscape of eRNAs. Single-sample gene set enrichment analysis algorithm (ssGSEA) and ESTIMATE were employed to enumerate immune cell infiltration and tumor purity. Results: Most prognostic eRNAs were enriched in immune-related pathways. Two distinct immune microenvironment patterns, the immune-active subtype and the immune-resistant subtype, were identified in AML. We further developed an eRNA-derived score (E-score) that could quantify immune microenvironment patterns and predict the response to immune checkpoint inhibitor (ICI) treatment. Finally, we established a prognostic nomogram combining E-score and other clinical features, which showed great discriminative power in both the training set [Harrell's concordance index (C index): 0.714 (0.651-0.777), p < 0.0001] and validation set [C index: 0.684 (0.614-0.755), p < 0.0001]. Calibration of the nomogram was also validated independently. Conclusion: In this study, we systematically understood the roles of eRNAs in regulating TME diversity and complexity. Moreover, our E-score model provided the first predictive model for ICI treatment in AML.