Simultaneous Accessibility and DNA Methylation Sequencing (SAM-seq)

Background: Epigenetic modifications, including chromatin accessibility, nucleosome positioning, and DNA methylation intricately shape genome function. However, current chromatin profiling techniques relying on short-read sequencing technologies face limitations in adequately characterising repetitive genomic regions and detecting multiple chromatin features simultaneously. We introduce Simultaneous Accessibility and DNA Methylation Sequencing (SAM-seq), a robust method leveraging bacterial adenine methyltransferases (m6A-MTases) to label accessible regions in purified plant nuclei. Coupled with Oxford Nanopore Technology sequencing, SAM-seq enables high-resolution profiling of m6A-tagged chromatin accessibility together with cytosine methylation along chromatin fibres in plants. Analysis of naked genomic DNA revealed significant sequence preference biases of m6A-MTases, controllable through a normalisation step. SAM-seq applied to Arabidopsis and maize nuclei provide fine-grained accessibility and DNA methylation landscapes at genome-wide as well as local scales. We characterised crosstalks between chromatin accessibility and DNA methylation, notably at subnucleosomal regions over genes, TEs, and centromeric repeats. SAM-seq also facilitated the identification of DNA footprints over cis-regulatory regions. Furthermore, using the single-molecule information provided by SAM-seq we unveiled extensive cellular heterogeneity at chromatin domains harbouring antagonistic chromatin marks, suggesting that bivalency reflects cell-specific regulations of gene activity. SAM-seq provides a robust method for acquiring high-resolution accessibility and DNA methylation landscapes across entire plant genomes. Our results underscore the importance of considering the intrinsic substrate preferences of m6A-MTases for reliable chromatin profiling. SAM-seq opens new opportunities to simultaneously study multiple epigenetic features at unprecedented scale, even in non-model species with limited genomic and epigenomic information.