A System for Genome-Wide Histone Variant Dynamics In ES Cells Reveals Dynamic MacroH2A2 Replacement at Promoters
Figure 2
Steady-state mapping of MacroH2A2.
(A) Overview of MacroH2A2 localization across chromosome 1. Shown are HA data from HA-Macro ES cells in the absence of doxycycline (top panel), or after 12 hours of doxycycline addition (in pink). Data from anti-MacroH2A2 is shown in blue. Rectangular blue and red tracks underneath the HA-12 hrs track represent RefSeq genes and Ensemble genes, respectively. Black boxes show gene “deserts” with relatively infrequent genes, which also exhibit reduced MacroH2A2 levels. (B) MacroH2A2 broadly localizes to gene-rich regions. The mouse genome was segmented into 1 Megabase tiles, and scatterplot shows strong correlation between gene number in a given tile (x axis) and overall MacroH2A2 signal in that tile (y axis). (C) TSS-aligned heatmaps for two related MacroH2A2 datasets. Note that left MacroH2A2 was generated from HA-Macro ES cells prior to doxycycline induction, while right panel represents t = 12 hours of doxycycline induction. All named genes are shown. Data was subject to k means clustering (k = 4), but given the strong similarities between two of the clusters we grouped them together into Cluster 1 leaving three clusters in this image. (D) Boxplot of MacroH2A2 enrichment for genes ranked by expression level [42]. Five categories on x axis show genes grouped from bottom 20% expression level in ES cells (left) to top 20% (right). For each quintile, MacroH2A2 enrichment (average enrichment level for 1.2 kb surrounding TSS) is shown as a boxplot, with box showing median and 1st and 3rd quartiles, and whiskers showing mean plus/minus standard deviation.