Cytosine methylation and histone modification are epigenomic marks with effects on transposable elements (TE), transcription of genes and heterochromatin formation. While occurring mainly in a CG-dinucleotide context, DNA methylation in brain cells contains nearly an equal amount of non-CG methylation (mCH). mCH accumulates in neurons and correlates with transcriptional repression at a period coinciding with synaptogenesis and neuronal maturation. Embryonic CG-methylation patterns also change dramatically during the period between birth and the second postnatal week. The DNA methyltransferase Dnmt3a is highly expressed in brain during this period. Preliminary data in this application suggests that this enzyme is responsible for the accumulation of mC in neurons during the perinatal period. A conditional knockout mouse was created, in which deletion of Dnmt3a in pyramidal neurons occurs during the late embryonic period (~E15, driven by Neurod6-Cre). Contrary to results showing a shortened lifespan in animals with earlier embryonic deletion (driven by Nestin-Cre), or lack of phenotype when the deletion occurs past the second postnatal week (driven by CamK2a-Cre), NeuroD6-driven Dnmt3a-KO (pyrDnmt3a-KO) animals show no postnatal mC accumulation, have significantly altered gene expression, and develop pronounced changes in behavior without changes in lifespan. These results support the hypothesis that mC accumulation and patterning in neurons requires precise regulation of Dnmt3a activity during neuronal development. Based on these findings, it is proposed that mC accumulation during the perinatal period may be essential for the spatial and temporal gene regulation required for proper synapse development and circuit formation. This hypothesis will be tested by delineating the dynamics of Dnmt3a-dependent mC accumulation during brain development, by characterizing the disruptions in methylation patterns, transcriptional dysregulation and histone modifications in animals carrying a deletion of Dnmt3a in pyramidal and inhibitory neurons from cortex and hippocampus (Aim 1). To understand the mechanisms of activation of Dnmt3a during postnatal cortical development, this proposal will identify its binding-partners during the developmental transition between the first and second postnatal week in neurons using mass spectrometry of Dnmt3a immunocomplexes. It will also assess the requirement of these binding partners for Dnmt3a function by transcriptional knockdown experiments in cultured cells and animals using a viral deliver system (Aim2). Finally, this proposal will characterize the effects of Dnmt3a deletion on neuron development and synaptogenesis (Aim3).
Brain development shows dynamic changes in DNA methylation (mC) patterns, with one form of methylation appearing in neurons during the period of synaptogenesis. Our studies show that the DNA methyltransferase Dnmt3a is responsible for this form of methylation, and that its deletion from excitatory neurons leads to alterations in mC patterns, altered transcription and disrupted behavior. In this project we study how methylation patterns of post-mitotic neurons are laid, and how they affect transcription and synaptogenesis during neuronal development, as a fundamental step when trying to integrate physiological, behavioral, neurochemical and molecular data on brain development and disease.