What happens to gene expression during DNA replication?
As the genome replicates, and before the cell divides, the copy number of the replicated portions of the genome doubles. The DNA serves as a template for gene transcription, raising the question – what happens to gene expression during DNA replication?
This question was asked many years ago in classical studies and showed that in eukaryotes gene dosage is buffered during S-phase by measuring macromolecule synthesis along the cell cycle.
This has also been shown more recently using live cell tracking (Yaron Shav-Tal) and smFISH (Arjun Raj). They showed that transcription sites were clearly duplicated as cells entered S phase, yet transcription efficiency from either site was reduced upon doublet formation. Therefore, transcription was attenuated from both copies of the DNA, again to maintain stable transcription rates, compensating for the increased gene copy.
We set out to
(1) Establish buffering of gene expression during S-phase in a genome wide manner
(2) Find the molecular mechanism of expression homeostasis
We will examine what happens to gene expression during DNA replication by examining genes with different time of replication. In the cartoon, the temporal increase in the copy number of early (red) or late (green) replicated genes is shown (top panel) for DNA content. In the absence of regulation, an increase of relative mRNA levels is expected (brown, lower panel), as observed in bacteria. In contrast, in the presence of a buffering mechanism that inhibits transcription from replicated DNA (blue), little fluctuations in relative expression are expected, as suggested in different eukaryotic cells.
Expression during S phase: Average increase in DNA content (black) and expression levels (blue) of early-replicating genes relative to late-replicating ones, for cells progressing synchronously through S phase, after release from α-factor (left) or HU (right) arrest. Notice there are only little fluctuations in the relative expression along the cell cycle.
(Cell cycle, stress… etc genes were excluded from the analysis).
We employed 4TU-Seq to measure mRNA synthesis: we gave a short pulse of a labeled nucleotide to the cells, which incorporates into nascent mRNAs, and then isolated these transcripts and sequenced them. We see that mRNA synthesis also shows buffering during DNA replication.
We hypothesized that chromatin regulators may suppress transcription from replicated DNA.
In order to find a regulator of expression homeostasis (inhibitor of gene expression during S-phase), it is possible to do many time course experiments in different mutants. However, this would be very expensive and labor intensive.
Instead, we used published profiles (by Frank Holstege’s lab) of gene expression in asynchronous cultures. Each dot represents a single mutant. Correlation between replication timing and change in gene expression is shown on the y axis, and the difference in the expression of early-and late-replicating genes is shown on the x axis. The two measures are expected to be correlated. We are looking for mutants that have a large change in the expression of early replicating genes compare to late ones, and that have a negative correlation on the y-axis (earlier the gene is replicated, larger change in expression).
Of the three mutants showing the strongest correlation between gene expression and replication timing, two were involved in H3 acetylation: the acetyltransferase Rtt109 and its histone chaperone cofactor Asf1. We deleted Rtt109 and repeated the same time course experiments as before. We now see that mRNA synthesis follows DNA dosage, and that buffering is lost. Therefore, Rtt109 (and Asf1) govern expression homeostasis.
Rtt109 acetylates histone H3 on two residues, K56 and K9, and Asf1 is required for both functions. To differentiate which of these residues is responsible for the reduced transcription efficiency of replicated DNA, we considered mutants in which K56 or K9 were replaced by residues that mimic constant nonacetylation (lysine to alanine, K→A) or constant acetylation (lysine to glutamine, K→Q). We saw that only when mutating H3K56, the relative expression increases (bar graph are in HU-synchronized cultures of the mutants).
We also overexpressed the deacetylases that are specific for H3K56ac, and observed a similar loss of buffering.
In summary, we find that Rtt109/Asf1-dependent H3K56ac suppresses transcription from newly replicated DNA during S phase, thereby maintaining expression homeostasis during this time when the DNA dosage of different genes transiently differs.