Histone code dynamics during nitrogen response in tomato

In plants, environmental stimuli trigger rapid transcriptional reprogramming of relevant gene suites. Regulation occurs down to the local chromatin landscape of the genes, and while epigenetic modifications proceed in a manner dependent on specific organ and tissue contexts, detailed investigations at this level remain limited. To better understand the organ specificity of dynamic chromatin modifications in response to external signals, we treated nitrogen-limited tomato seedlings with a supply of nitrate and measured the genome-wide changes of four histone marks, the permissive histone marks H3K27ac, H3K4me3, and H3K36me3 and repressive mark H3K27me3, in shoots and roots separately. We observed dynamic histone acetylation and methylation events which are largely organ-specific in scope at functionally relevant gene loci. Integration of transcriptomic and epigenomic datasets generated from the same tissue samples revealed largely syngenetic relations between changes in transcript levels and histone modifications, with the exception of H3K27me3 where an increased level at genes up-regulated in response to nitrate supply is observed in only the shoots. This non-canonical pattern of H3K27me3 deposition could possibly function to prevent over-expression of certain activated genes. To study the determinant roles of histone code in predicting gene regulation at the genome-wide level we applied a machine learning approach. While gene regulation could be best predicted using all four histone marks together, we observed different rules regarding the importance of individual histone marks between shoots, where H3K36me3 is the most successful mark in predicting gene activation and repression events, and the roots, where H3K4me3 is the strongest individual predictor. In summary, our integrated study substantiates a view that during plant environmental responses, the histone code dynamics that govern relationships between chromatin modification and gene regulation are highly dependent on tissue specific contexts. Three replicates of ChIP-Seq from root and shoot in nitrogen supplied (6 hours) or nitrogen replete control, with input and immunopreciptation for H3K4me3, H3K27ac, H3K27me3, and H3K36me3, for a total of 60 samples. Three replicates of RNA-Seq in the root and shoot tissue in the two nitrogen conditions (total of 12 samples) were used for expression analysis. For H3K9me2 analyses, three replicates from input or from immunoprecipitation for each of the conditions were used for a total of 12 samples.