The Histone Chaperones FACT and Spt6 Restrict H2A.Z from Intragenic Locations

Abstract

H2A.Z is a highly conserved histone variant involved in several key nuclear processes. It is incorporated into promoters by SWR-C-related chromatin remodeling complexes, but whether it is also actively excluded from non-promoter regions is not clear. Here we provide genomic and biochemical evidence that the RNA polymerase II (RNA Pol II) elongation-associated histone chaperones FACT and Spt6 both contribute to restricting H2A.Z from intragenic regions. In the absence of FACT or Spt6, the lack of efficient nucleosome reassembly coupled to pervasive incorporation of H2A.Z by mislocalized SWR-C alters chromatin composition and contributes to cryptic initiation. Therefore, chaperone-mediated H2A.Z confinement is crucial for restricting the chromatin signature of gene promoters that otherwise may license or promote cryptic transcription.

Figure 1. A Survey of Chromatin Regulator Mutants Identified Histone Chaperones FACT and Spt6 as Important Regulators of H2A.Z Occupancy

(A–B) Aggregate profiles of H2A.Z/H2B log₂ enrichment ratios over all yeast genes longer than 1 kb (n=3439 genes) in various ATP-dependent chromatin remodeler mutants. ChIP experiments were either performed using epitope tagged H2A.Z (Myc) and H2B (HA) (A) or rabbit polyclonal antibodies against H2A.Z and H2B (B). (C) Aggregate profiles of H2A.Z/H2B log₂ enrichment ratios over all yeast genes longer than 1 kb in various histone chaperone and chromatin assembly factor mutants. The experiments for the spt16-197 and spt6-1004 strains, as well as their respective wild type, were performed after an 80 minute switch to non-permissive temperature (37°C). See also Figure S1.

Figure 2. Inappropriate Accumulation of H2A.Z in Gene Bodies in spt16 and spt6 Mutants

(A) A schematic representation of the spike-in strategy used to rescale H2A.Z ChIP-chip data in spt16-197 and spt6-1004 cells. (B) H2A.Z/H2B log₂ enrichment ratio along a 90 kb fragment of chromosome III is shown for WT (grey), spt16-197 (blue) and spt6-1004 (red) cells. The data from mutant cells are shown prior to (“raw”) and after (“rescaled”) rescaling using phiX174 exogenous spike-in controls. A zoom in around the ATG15 locus is shown at the bottom. (C) Aggregate profiles of H2A.Z/H2B log₂ enrichment ratios over all yeast genes longer than 1 kb in spt16-197 (blue) and spt6-1004 (red) cells, together with their respective WT (grey). The data from mutant cells are shown prior to (dashed trace) and after (solid trace) rescaling using phiX174 exogenous spike-in controls. (D) Absolute H2A.Z occupancy (expressed in % of Input) over the promoter (grey) and coding region (white, ORF) of selected genes, as determined by ChIP-qPCR. Control experiments using IgG antibodies are also shown. (E) Western blot showing bulk levels of H2A.Z in chromatin extracts prepared from spt16-197 and spt6-1004 cells, together with their respective WT. Loading was normalized using histone H4. The right panel shows bulk H2A.Z and H4 levels from chromatin extracts prepared from WT and htz1Δ cells, demonstrating the specificity of the H2A.Z antibody. (F) Aggregate profiles of H4 log₂ enrichment ratios over all yeast genes longer than 1 kb in WT (grey), spt16-197 (blue) and spt6-1004 (red) cells, which were shifted to 37 °C for 80 min. All traces were normalized by setting the minima (representing the NDR) to “0”. Reads density from an MNase-Seq experiment (black) from WT cells (Jiang and Pugh, 2009) is shown as a guide for the position of nucleosomes. See also Figure S2.

Figure 3. FACT and Spt6 Can Discriminate Between H2A and H2A.Z Dimers In Vitro

(A) Silver stained SDS-PAGE of the FACT (Spt16-TAP) and Spt6 (Spt6-TAP) protein complexes used. Traces of TEV protease remaining after purification are indicated. (B) Western blot showing the amount of HA-H2A.Z/H2B dimers incorporated within canonical nucleosomes (H2A Nuc) by purified SWR-C in the absence (−) or presence (+) of ATP. Histone H3 is shown as a loading control. (C) A scheme of the in vitro histone incorporation assay (left) and Western blots showing the amount of Flag-H2A (middle) or HA-H2A.Z (right) dimers incorporated within canonical nucleosomes in the absence (−) or presence (+) of purified FACT or Spt6 complexes. Inputs (50%) were loaded as controls. Blotting beads with an anti-H3 antibody shows that equivalent amount of mononucleosomes were used in both assays. (D) A scheme of the in vitro histone incorporation assay (left) and Western blots showing the amount of Flag-H2A (middle) or HA-H2A.Z (right) dimers incorporated within H2A.Z nucleosomes in the absence (−) or presence (+) of purified FACT or Spt6 complexes. The blots are also probed using an anti-H3 antibody. See also Figure S3.

Figure 4. FACT and Spt6 Prevent Pervasive SWR-C Recruitment In Vivo

(A) Left panel shows Aggregate profiles of Swr1-HA log₂ enrichment ratios (Tag vs No Tag) over all yeast genes longer than 1 kb in WT (grey), spt16-197 (blue) and spt6-1004 (red) cells. The right panel shows the difference in Swr1-HA levels between spt16-197 and WT cells (dashed blue) or spt6-1004 and WT cells (dashed red). (B) Absolute H2A.Z occupancy (expressed in % of Input) over the promoter (grey) and coding region (white, ORF) of selected genes, as determined by ChIP-qPCR in WT, swr1Δ, spt6-1004, spt6-1004/swr1Δ, spt16-197 and spt16-197/swr1Δ cells.

Figure 5. Deletion of HTZ1 Partially Suppresses Cryptic Transcription From spt16-197 and spt6-1004 Cells

(A) A schematic representation of the FLO8-HIS3 system used to detect cryptic transcription from the FLO8 gene (Cheung et al., 2008). (B) The indicated yeast strains were grown to saturation in YNB-complete medium, washed, resuspended at the same density in water, serial diluted (5 fold series) and spotted on YNB-complete (Complete) and YNB medium lacking histidine (−HIS). Plates were incubated at 33°C. (C) Levels of cryptic transcript transcribed from the FLO8-HIS3 locus, as determined by Northern blot, are shown after an 80 minute shift from 30°C to 33°C. SNR190 was used as a loading control. The experiments were performed four times. A representative example is shown on the left and quantification for four independent biological replicates (grey circles) together with the average (black bars) is shown on the right. Indicated P value is from T-test. (D) RT-qPCR was used to measure cryptic transcription at four genes in the indicated strains (four additional genes are shown in Figure S4C). Expression was measured in the 5− and 3− regions of each gene and the 3−/5− ratio was used as a measure of cryptic transcription. Values for four independent biological replicates are shown (grey circles) together with the average (black bars). Indicated P values are from T-tests. (E) Western blots showing the amount of H3K36me3 at permissive temperature (30 °C) in WT, spt6-1004 and spt6-1004/htz1Δ cells (left) and in WT, spt16-197 and spt16-197/htz1Δ cells (right). Histone H4 is shown as a loading control. (F) The indicated yeast strains were grown to saturation in YPD medium, washed, resuspended at the same density in water, serial diluted (10 fold series), spotted on YPD plates and incubated at 30 °C or 37 °C. (G) Aggregate profiles of H4 log₂ enrichment ratios over all yeast genes longer than 1 kb in WT (grey), spt16-197 (solid blue), spt16-197/htz1Δ (dashed blue), spt6-1004 (solid red) and spt6-1004/htz1Δ (dashed red) cells, which were shifted to 37 °C for 80 min. All traces were normalized by setting the minima (representing the NDR) to “0”. See also Figure S4.

Figure 6. A Schematic Model Describing the Activities of FACT and Spt6 in Preserving the Epigenetic Landscape and Guarding Against Cryptic Transcription

FACT and Spt6 prevent nucleosome loss and selectively reincorporate H2A within gene bodies during transcription elongation. This ensures proper chromatin structure over genes and prevents cryptic transcription. When either FACT or Spt6 is compromised, nucleosome loss occurs and H2A.Z is not efficiently removed from gene bodies. The paucity of nucleosomes in gene bodies leads to pervasive recruitment of SWR-C, which exacerbates H2A.Z accumulation in these regions. This nucleosome-poor/H2A.Z-rich chromatin promotes cryptic transcription.