EZHIP IS NOT EXPRESSED WITHIN THE DEVELOPING MOUSE OR HUMAN BRAIN

Abstract The protein EZHIP, first described as part of the genetic signature of group A posterior fossa ependymoma (PFA), functions in a manner analogous to the onco-histone H3K27M, itself first described in diffuse midline glioma (DMG). These two proteins modulate activity of the Polycomb repressive complex 2 (PRC2), resulting in global loss of H3K27 trimethylation. Thus, both PFA and DMG are characterized by global loss of the H3K27me3 mark, and PRC2 modulation is critical to oncogenesis of these tumors. However, the precise mechanism(s) of EZHIP and H3K27M-induced tumorigenesis is not known. One possibility is that EZHIP is expressed within the developing brain where it acts to modulate cellular development. In turn, EZHIP overexpression or H3K27M mutation could co-opt this developmental program to drive tumorigenesis. Relatedly, whether EZHIP is expressed in the cell of origin for PFA or DMG is not known. In order to evaluate these hypotheses, we defined the landscape of EZHIP expression in both murine and human developing brain. Leveraging single cell RNA sequencing databases, we show that EZHIP is not expressed within the developing mouse brain, a finding that is consistent across separate datasets. Similarly, EZHIP is not expressed in the developing murine cerebellum, where the cell of origin for PFA is located. Furthermore, we examined bulk RNA sequencing datasets of the developing human brain. Similar to our findings in the mouse brain, EZHIP is not expressed at any time point or spatial location within the developing human brain. Taken together, these data reject the above hypothesis, that EZHIP or H3K27M co-opt an endogenous PRC2 inhibitory developmental program. Similarly, these results show that EZHIP is not expressed within the cell of origin for PFA or DMG. Further studies will seek to understand the endogenous function of EZHIP by further defining its normal expression pattern and function.