Spt6-Spn1 interaction is required for RNA polymerase II association and precise nucleosome positioning along transcribed genes

Abstract

Spt6-Spn1 is an essential histone chaperone complex that associates with RNA Polymerase II (RNAPII) and reassembles nucleosomes during gene transcription. While the interaction between Spt6 and Spn1 is important for its histone deposition and transcription functions, a precise mechanistic understanding is still limited. Here, using temperature-sensitive alleles of spt6 and spn1 that disrupt their interaction in yeast, we show that the Spt6-Spn1 association is important for its stable interaction with the elongating RNAPII complex and nucleosomes. Using micrococcal nuclease (MNase)-based chromatin occupancy profiling, we further find that Spt6-Spn1 interaction is required to maintain a preferred nucleosome positioning at actively transcribed genes; in the absence of Spt6-Spn1 interaction, we observe a return to replication-dependent phasing. In addition to positioning defects, Spt6-Spn1 disrupting mutants also resulted in an overall shift of nucleosomes toward the 5' end of genes that were correlated with decreased RNAPII levels. As loss of Spt6-Spn1 association results in cryptic transcription at a subset of genes, we examined these genes for their nucleosome profiles. These findings revealed that the chromatin organization at these loci is similar to other active genes, thus underscoring the critical role of DNA sequence in mediating cryptic transcription when nucleosome positioning is altered. Taken together, these findings reveal that Spt6-Spn1 interaction is key to its association with elongating RNAPII and to its ability to precisely organize nucleosomes across transcription units.

Keywords: chromatin; histone chaperone; histones; nucleosomes; spn1; spt6; transcription; yeast.

Figure 1

Spt6-Spn1 interaction is required for histone and RNAPII association.A, schematic representation of Spt6-Spn1 association with RNAPII. B, domain structures of Spt6 and Spn1. Red outline highlights the Spt6-Spn1 dimerization domains and mutants previously shown to disrupt Spt6-Spn1 interaction that were used in this study. C–E, volcano plots of fold-change of Spt6-3XFLAG interactors relative to WT in (C) spn1-F267E, (D) spn1-R263D, and (E) spt6-F249K. X-axis is log₂ fold-change of mutant/WT; y-axis is -log₁₀p-value. 1963 data points are shown on each graph. Proteins of interest are highlighted and colored by complex, as shown in the key. Significance threshold was set at p-value < 0.05 and is indicated by a dotted line. A total of 259 proteins had significantly changed interactions with Spt6 in spn1-F267E (C); 62 proteins had significantly changed interactions with Spt6 in spn1-R263D (D); and 567 proteins had significantly changed interactions with Spt6 in spt6-F249K (E). F, heatmap and clustering analysis (ClustVis, ref (47)) of protein abundance in Spt6-3XFLAG WT (n = 3), spn1-F267E (n = 3), spn1-R263D (n = 4), and spt6-F249K (n = 4). Shown are ln (protein abundance) in each replicate for bait protein (Spt6, indicated by ∗), histone proteins, RNAPII subunits, FACT, Paf1C, Spn1, and Spt5.

Figure 2

Disruption of Spt6-Spn1 interaction results in increased nucleosome phasing in transcribed genes.A, heat maps of nucleosome profiles were generated by plotting the midpoints of the nucleosome-sized fragments surrounding the transcription start sites of 4649 non-overlapping genes with annotated transcription start sites (24). All panels are ordered by decreasing transcriptional activity. B, aggregated nucleosome profiles of genes with regularly phased nucleosomes grouped into non-transcribed genes (left panel) and transcribed genes (right panel). C, nucleosome phasing of non-transcribed genes (left panel) and transcribed genes (right panel) with regularly-phased nucleosomes. D and E, chromatin occupancy profiles for (D) YGR053G, a non-transcribed gene and (E) YGL202W, a transcribed gene. Fragment midpoints are plotted by fragment length and genomic position relative to ORF start. The shading of each point is determined by a 2D kernel density estimate of the surrounding points. Gene bodies are shown in gray on the top. Grey dotted lines denote the position of nucleosome dyads. Red lines between the WT plot and the spt6-249K mutant plot is to illustrate the relative changes to nucleosome positioning found in the spt6 and spn1 mutants when using the WT dyad midpoints as an anchor.

Figure 3

Transcription-dependent increase in nucleosome phasing is associated with a 10 bp rotational shift of the nucleosomes.A, schematic describing the data processing steps for generating transformed nucleosome pile-ups. Nucleosome-sized reads (140–180 bp) around dyads of a group of nucleosomes were stacked up to generate the aggregate nucleosome profile. In the nucleosome profile, read midpoints form diagonal lines corresponding to changes in fragment length altering the midpoint of the protected fragment. The diagonal lines are spaced by multiples of 10 bp, reflecting alternative nucleosome dyad positions that can be attributed to the A/T dinucleotide repeat. To better visualize the pattern of nucleosome dyad positioning, the aggregate nucleosome profile was rotated 27 degrees clockwise, after which only the left half was retained. Similarly, the aggregate nucleosome profile was rotated 27 degrees counterclockwise, retaining the right half afterward. The two halves were joined to form the transformed nucleosome pile-up plots where alternative nucleosome dyad positions now align vertically. B, transformed nucleosome pile-up plots for the first three nucleosomes of non-transcribed genes (left panels) and transcribed genes (right panels). Nucleosomes are orientated relative to TSSs. C, density distribution of signals from the transformed nucleosome pile-up plots along the horizontals.

Figure 4

Disruption of Spt6-Spn1 interaction causes a shift in the distribution of nucleosome and RNAPII occupancy across gene bodies.A, chromatin occupancy profile at a representative locus showing a shift in nucleosome density across the gene body. B, mean nucleosome occupancy across gene bodies of non-transcribed genes and transcribed genes. C, mean RNA polymerase II occupancy across gene bodies of non-transcribed genes and transcribed genes.