SRA

Our work and that of others (Zovkic et al 2014 Nature, Maze et al 2015 Neuron) recently established histone variants as novel regulators of memory, whereby enrichment of H2A.Z, a variant of histone H2A, impairs memory formation. However, H2A has several structurally and functionally distinct members whose function in memory has never been studied. macroH2A (mH2A) is of special interest because it has been implicated as a risk factor for neurodegenerative diseases, particularly Huntington's (Hu et al 2011 PNAS), but its functional role in the brain is not known. Here, we conducted ChIP sequencing against macroH2A1 (mH2A1) and macroH2A2 (mH2A2) in the mouse hippocampus and showed that both proteins are enriched on transcriptionally silent genes, suggesting that mH2A has a repressive effect on transcription in the brain. However, only mH2A1 binding was dynamically reduced during learning, such that mH2A1 removal was associated with increased gene expression in response to fear conditioning. Additionally, viral vector mediated depletion of mH2A1 protein resulted in extensive dysregulation of gene transcription, whereas mH2A2 depletion had minimal transcriptional consequences. To determine if these differential downstream transcriptomes are functionally relevant, we showed that mH2A1, but not mH2A2 depletion resulted in impaired fear memory. These results suggest that mH2A1 is required for memory and that its dynamic removal during learning promotes training-induced gene expression. In concert with our prior evidence for a suppressive effect of H2A.Z on memory, these data suggest that unique H2A variants have distinct effects on memory formation and as such, implicate altered nucleosome composition as a novel regulator of neuroplasticity. Overall design: 3 Sequencing data sets: 1) ChIP sequencing against macroH2A1 or macroH2A2, along with inputs with and without fear conditioning; 2) RNA seq in control mice and 1h after contextual fear conditioning; 3) RNA seq in mice with macroH2A1 or macroH2A2 knockdown in area CA1, along with mice receiving scramble control virus; Additional Sequencing data sets: 1) ChIP sequencing against macroH2A1 and macroH2A2, along with inputs with and without fear conditioning at 1 hour time-point; 2) ChIP sequencing against macroH2A1, along with inputs with and without fear conditioning at 6 hour time-point; NEW data set: RNA seq in mice with macroH2A1 or macroH2A1 knockdown in CA1, half untrained and half fear conditioned. Brains were collected 1h after training.